Oral contraceptives (OCs) are a major class of prescription drug, used by a large proportion of women starting from early adolescence. Much research has been conducted to investigate the physiological changes that occur in women who take OCs. These include changes in general health as well as in nutritional needs. In terms of nutrition, several studies investigated whether women on OCs need different amounts of some vitamins and minerals. In particular, a report from the World Health Organization (WHO) points out that the influence of OCs on nutrient requirements is a topic of high clinical relevance and should, therefore, receive great attention. It has been shown that the key nutrient depletions concern folic acid, vitamins B2, B6, B12, vitamin C and E and the minerals magnesium, selenium and zinc. Most research has focused on the levels of these vitamins and minerals in the blood of women who take OCs compared to women who do not. Since women who take OCs not always have adequate diet, may have unhealthy life style or may suffer from pathologies of malabsorption, the possibility to prevent vitamin and mineral deficiencies by taking appropriate dietary supplements should be considered a first-line approach by clinicians.

Introduction: Major depressive disorder is a serious mental disorder in which treatment with antidepressant medication is associated with incidence of adverse events, such as constipation, diarrhea, dry mouth, headache, insomnia, and sexual dysfunction (SDys). Escitalopram (ESC), an effective and safe selective serotonin reuptake inhibitor with good tolerability, was used in this study. In this study, we investigated the prospective effect of Pycnogenol (PYC), an antioxidant, anti-inflammatory, and vasodilator agent, on ESC-induced SDys.

Methods: This was a randomized, parallel, open-label study. Seventy-two outpatients of both genders with depression were randomized into two groups as follows: 37 patients from the ESC + PYC group took 50 mg of PYC per day for 4 months in ESC co-treatment, and 35 subjects from the ESC group took ESC only. Five patients dropped out and were excluded from the analysis. The participants were examined every month (visits 1-4).

Results: ESC use led to improvement of depressive symptoms and severity scored by standardized psychiatric tests. PYC co-treatment resulted in attenuation of SDys beginning at 1 month of treatment and continuing for two consecutive months. Furthermore, an increase in heart rate in the PYC group was registered.

Conclusions: We propose that PYC-mediated SDys attenuation is based on its ability to improve endothelial functions by its antioxidant, anti-inflammatory, vasodilatory, and anticoagulant action. We assume that the action of PYC on heart rate is in accordance with the aforementioned vasodilatory action of PYC and consequent baroreflex-mediated heart rate response. PYC co-treatment reduced ESC-induced SDys and elevated heart rate.

We review the findings in major depression of a low plasma and particularly red cell folate, but also of low vitamin B12 status. Both low folate and low vitamin B12 status have been found in studies of depressive patients, and an association between depression and low levels of the two vitamins is found in studies of the general population. Low plasma or serum folate has also been found in patients with recurrent mood disorders treated by lithium. A link between depression and low folate has similarly been found in patients with alcoholism. It is interesting to note that Hong Kong and Taiwan populations with traditional Chinese diets (rich in folate), including patients with major depression, have high serum folate concentrations. However, these countries have very low life time rates of major depression. Low folate levels are furthermore linked to a poor response to antidepressants, and treatment with folic acid is shown to improve response to antidepressants. A recent study also suggests that high vitamin B12 status may be associated with better treatment outcome. Folate and vitamin B12 are major determinants of one-carbon metabolism, in which S-adenosylmethionine (SAM) is formed. SAM donates methyl groups that are crucial for neurological function. Increased plasma homocysteine is a functional marker of both folate and vitamin B12 deficiency. Increased homocysteine levels are found in depressive patients. In a large population study from Norway increased plasma homocysteine was associated with increased risk of depression but not anxiety. There is now substantial evidence of a common decrease in serum/red blood cell folate, serum vitamin B12 and an increase in plasma homocysteine in depression.

Furthermore, the MTHFR C677T polymorphism that impairs the homocysteine metabolism is shown to be overrepresented among depressive patients, which strengthens the association. On the basis of current data, we suggest that oral doses of both folic acid (800 microg daily) and vitamin B12 (1 mg daily) should be tried to improve treatment outcome in depression.

Major Depressive Disorder is an important global public health problem associated with the significant burden and is projected to be the second leading cause of disability by 2020 [1]. Prevalence figures worldwide range between 4.2 – 17% and a systematic review of studies from Pakistan gave the estimates as high as 34% for anxiety and depressive disorders [2]. Clinical guidelines recommend the use of SSRIs as first-line treatment [3]. Remission rates in the acute phase of treatment are 30-40% and an overall 30% show poor response to anti-depressant monotherapy [4-6]. In such cases, addressing other co-morbid conditions is suggested in addition to upgrading or augmenting anti-depressant doses [3].

Pakistan, a developing country of 17 million people, is fraught with poverty, inadequate nutrition and underprivileged health systems. Several small-case studies from Pakistan show vitamin deficiencies to be common [7,8]. A small study from Islamabad on a clinical population of patients with megaloblastic anemia reported figures of 76% with vitamin B12 deficiency [9]. Although no large scale studies on general population are available here, studies from neighboring countries that share the culture and climate show vitamin deficiencies to be common [10-12]. A study on a healthy Indian population showed the prevalence of B12 deficiency to be 47% [13].

Vitamin B12 plays an important role in DNA synthesis and neurological function. Its deficiency is associated with hematological, neurological and psychiatric manifestations of which the latter includes irritability, personality change, depression, dementia and rarely, psychosis [14]. Recent literature has identified the links between this vitamin deficiency and depression. High B12 levels in serum are associated with good treatment response, high homocysteine levels which are common in folate / B12 deficiency and in those suffering from depression are associated with poor response to anti depressant treatment [15-17]. Hyperhomocysteinemia may have direct effects on neurotransmitters implicated in depression [18]. 

Randomized trials have shown that folate and other nutritional supplementations have a significant effect in treating the treatment-resistant depression [19,20]. Folate deficiency has also been linked with the delay in treatment response as well as relapse [21,22]. 

To date no adequately designed trial compares anti-depressant monotherapy with B12 augmentation. We aimed to compare the clinical response of antidepressants monotherapy with that of B12-augmentation in a sample of depressed patients with low normalB12 levels (190 pg/ml to 300 pg/ml).

METHODS

The study was designed as an open label randomized controlled trial (clinical trials registration number (Clinical Trials.gov ID NCT00939718). Patients were enrolled from outpatient clinics of the department of Psychiatry at Aga Khan University Hospital, Karachi Pakistan from December 2009 - June 2010. Depression was defined as Patients scoring ≥ 16 on the 20-item Hamilton Rating Scale for Depression-Urdu version (HAM-D) [23,24].

Low normal B12 level was defined as B12 level ranging between 190 and 300 pg/ml. Patients with B12 deficiency (level below190 pg/ml) were not enrolled due to ethical reasons (patients with established B12 deficiency must be treated and not subjected to randomization).

Ethics Statement

A study protocol was approved by the Ethics Review Committee of Aga Khan University hospital. All patients signed an informed consent form. 

Randomization and Masking

Patients with depression and low normal B12 levels were randomized by a computer program into the control arm (antidepressants only) or treatment arm (antidepressants and injectable vitamin B12 supplementation). The randomization was carried out using a Microsoft Excel spreadsheet using the function “RAND” to generate uniform random numbers. 

The randomized assignment was determined by partitioning the range of the random number. Randomization was not concealed for investigators. Antidepressants that were prescribed included either a tricyclic antidepressant (TCA) with a dose equivalent of imipramine 100 mg to 250 mg/ day and the SSRIs dose equivalent of Fluoxetine 20 – 40 mg/day. Patients with concurrent unstable medical illness, history of manic episodes or psychotic illness, psychotic symptoms in a depressive episode, concurrent substance misuse and patients with suicidal ideation were excluded. Participants in the control arm only received the antidepressants. Those in the treatment arm received B12 intramuscular injectable as 1000 mcg every week during the 6 weeks in addition to the antidepressants.

Any adverse effects of oral medications and injections (B12) were monitored and documented on an adverse effects sheet in the patient folder. 

A decline in HAM-D score of 20% or more from the baseline, indicating an improvement in depression, was defined as the primary outcome. The total change in HAM-D score from baseline to follow up was defined as a secondary outcome. Additional secondary outcomes included the follow up HAM-D score and a reduction in HAM-D score of 50% or more from the baseline. Follow up of the HAM-D score was recorded during 12 weeks by a research officer who was unaware of the patient’s randomization.

Statistical Considerations

A total of 248 patients equally allocated to the two groups provides at least 90 percent power, with a 5 percent type I error rate, using a two-sided hypothesis test. This calculation assumes an anticipated difference in the response rate of 20 percent between the two groups with a 30 percent response rate in the SSRI group and at least a 50 percent response rate in the combination treatment group. Considering a 15 percent dropout rate in each group, an additional sample of 44 patients equally divided into the two groups would be needed to yield a total required sample size of 292 patients. Therefore, the target sample size for each arm was 146.

Descriptive statistics were prepared for all characteristics including means and standard deviations for continuous measures and frequencies and percentages for categorical measures. Chi-square or Fisher’s exact tests were used to compare categorical baseline characteristics. Two-sample t tests were used to compare continuous baseline characteristics. Chi-square tests were used to examine the association between treatment and the 20% and 50% reductions in HAM-D outcomes. These analyses were extended to logistic regression to adjust for the baseline HAM-D score. Two-sample t tests were used to examine the association between treatment and the follow up HAM-D score or change in HAM-D score from baseline to follow up. These analyses were extended to analysis of covariance to adjust for the baseline HAM-D score. All hypotheses tests were two-sided and p-values less than 0.05 were considered statistically significant. Analyses were conducted using SAS Version 9.2 (SAS Institute, Inc., Cary, NC, USA).

Results

A total of 199 depressed patients were screened for the B12 levels. Vitamin B12 deficiency was present in 44 (22%) patients; 73 (36%) had low normal B12 levels and 82 (42%) had normal B12 levels. Patients with low B12 levels were given the prescriptions for B12 replacement therapy in addition to the antidepressants. These patients were not randomized. Out of 73 patients with low normal B12 levels 34 (47%) were randomized to the treatment group while 39 (53%) were randomized to the control arm. There were no significant differences between the two groups at baseline except for the higher depression scores in the treatment group (Table ​11). No adverse effects or complications were noted in either group. For the primary outcome of 20% reduction in HAM-D score, significantly more subjects from the treatment group showed a 20% reduction unadjusted for baseline HAM-D score (100% vs. 69%; p < 0.001). Examining a50% reduction from baseline, this effect remained significant (44% vs. 5%; p < 0.001). We also adjusted the analyses of reduction for the baseline HAM-D score and our findings remained significant (Table ​22). Additionally, the change in HAM-D score was significantly greater for the treatment group, unadjusted and adjusted for the baseline HAM-D score (Table ​22).

Table 1.

Baseline Characteristics Shown as Mean ± Standard Deviation or Frequency (Percent)

Treatment (n=34)Control (n=39)p-valueAge, in years37.68 ± 13.3836.56 ± 12.280.71Total HAM-D 23.21 ± 5.8519.38 ± 5.700.006Vitamin B12 level238.49 ± 33.21245.16 ± 27.820.36GenderMale18 (52.9)20 (51.3)0.89Female16 (47.1)19 (48.7)Marital statusSingle never married9 (26.5) 13 (33.3)0.84Married21 (61.8) 21 (53.9)Divorced/separated/widowed4 (11.8)5 (12.8)Table 2.

Outcomes for Follow Up After 3 Months Shown as Mean ± Standard Deviation or Frequency (Percent)

Treatment (n=34)Control (n=39)Unadjusted P-valueAdjusted* P-value20% reduction in HAM-D score34 (100)27 (69.2)<0.0010.00150% reduction in HAM-D score15 (44.1)2 (5.1)<0.001<0.001Follow up HAM-D score12.12 ± 5.1214.38 ± 4.730.053<0.001Change in HAM-D score11.09 ± 4.585.00 ± 3.38<0.001<0.001Go to:DISCUSSION

Coexistence of depression and vitamin B12 deficiency are not uncommon. These patients are routinely treated with SSRI and B12 supplementation, however it is not well established whether the people with low normal B12 and co-occurring depression should also receive B12 supplementation. Our study tried to address this issue. Despite not attaining the targeted sample size, the findings appear significant. Vitamin B12 deficiency was present in 22% of our depressed population. This frequency is high in a non-vegetarian, relatively young, middle to upper income population of our patients. A recent study of healthy adults from Karachi showed a population based prevalence of vitamin B12 deficiency (less than 200 pg/ml) in 9.74% people [25]. These findings indicate a substantial co-morbidity of B12 deficiency (22%) in our depressed patient population. 

To our knowledge, this is the first randomized control trial in Pakistani population which has looked at clinically depressed patients with low B12 levels. We did not randomize B12 deficient patients due to ethical reasons. Low normal B12 levels were present in 36% of patients. These patients are often not treated with B12 supplementation. The findings of our study clearly indicate that these patients demonstrated significant improvement with B12 supplementation in addition to SSRI as compared to the control group even after adjustment for baseline HAM-D score. We think that these patients represent sub group within the clinically depressed population and a supplementation with B12 along with the conventional antidepressants may be a useful strategy in the treatment of depression in such cases. 

We did not give sham injections to the control group. A placebo response due to injections may be responsible for the significant improvement in the treatment group. We also did not look at differences in response at specific doses of antidepressants nor did we stratify our groups on the basis of type of antidepressants prescribed. We also could not reach our sample size because we were not able to get another extension in our grant. Due to financial constraints we were not able to obtain final B12 levels. These are some of the limitations of our study. However, these findings have important clinical implications. B12 deficiency and low normal B12 levels are common and may be associated with depression and the inadequate response to antidepressant treatment in patients with depression. Vitamin B12 supplementation with antidepressants has significantly improved depressive symptoms in our group. Larger, multi-center studies are required to extend and replicate our findings.

CONFLICT OF INTEREST

Authors have no conflicts of interests to declare

ACKNOWLEDGEMENT

Authors would like to acknowledge Michele Shaffer PhD from School of Public Health, Penn State University for her valuable input in statistical analysis and manuscript revision.

FUNDING

Study funding was supported by University research council grant of Aga Khan University (G&C code; 70210)

REFERENCES

1. Murray C, Lopez A. The global burden of diseases a comprehensive assessment of Mortality and disability from diseases injuries and risk factors in 1990 and projected to 2020. Boston Harvard School of Public Health. WHO and World Bank. 1996 [Google Scholar]

2. Mirza I, Jenkins R. Risk factors prevalence and treatment of anxiety and depressive disorders in Pakistan systematic review. BMJ. 2004;328:794–8. [PMC free article] [PubMed] [Google Scholar]

3. Sohail Z, editor. Practice Guideline for the Treatment of Patients With Major Depressive Disorder DOI10.1176/appi.books.9780890423387.654001. 3rd ed. American Psychiatric Association. [Google Scholar]

4. Thase ME, Entsuah AR, Rudolph RL. Remission rates during treatment with Venlafaxine or selective serotonin reuptake inhibitors. Br J Psychiatry. 2001;178:234 –41. [PubMed] [Google Scholar]

5. Puech A, Montgomery SA, Prost JF, Solles A, Briley M. Milnacipran,a new serotonin and Noradrenaline reuptake inhibitor an overview of its antidepressant activity and clinical tolerability. Int Clin Psychopharmacol. 1997;12(2):99–108. [PubMed] [Google Scholar]

6. Antai-Otong D. Monotherapy Antidepressant.A Thing of the Past Implications for the Treatment of Major Depressive Disorder. Perspect Psych Care. 2007;43(3):142–5. [PubMed] [Google Scholar]

7. Shah RA, Qureshi MB, Khan AA. Magnitude of vitamin A deficiency in poor communities of the four selected districts of Punjab using (rapid assessment technique). Ann King Edward Med Coll. 2005;11(3):314–8. [Google Scholar]

8. Khan RM, Iqbal P. Deficiency of vitamin C in South Asia. Pak J Med Sci. 2006;22(3):347–55.[Google Scholar]

9. Hashim H, Tahir F. Frequency of vitamin B12 and folic acid deficiencies among patients with megaloblastic anemia. Ann Pak Inst Med Sci. 2006;2(3):192–4. [Google Scholar]

10. Yajnik CS, Deshpande SS, Lubree HG, et al. Vitamin B12 deficiency and hyperhomocysteinemia in rural and urban Indians. J Assoc Physicians India. 2006;54:775–82. [PubMed] [Google Scholar]

11. Fakhrzadeh H, Ghotbi S, Pourebrahim R, et al. Total plasma homocysteine folate and vitamin b12 status in healthy Iranian adults the Tehran homocysteine survey 2003–2004)/a cross sectional population based study. BMC Public Health. 2006;6:29. [PMC free article] [PubMed] [Google Scholar]

12. Lim HS, Heo YR. Plasma total homocysteine folate and vitamin B12 status in Korean adults. J Nutr Sci Vitaminol (Tokyo). 2002;48(4):290–7 . [PubMed] [Google Scholar]

13. Refsum H, Yajnik CS, Gadkari M, et al. Hyperhomocysteinemia and elevated methylmalonic acid indicate a high prevalence of cobalamin deficiency in Asian Indians. Am J Clin Nut. 2001;74(2):233–241.[PubMed] [Google Scholar]

14. Robert C, Brown DL. Vitamin B12 Deficiency. Am Fam Physician. 2003;67:979–86. [PubMed] [Google Scholar]

15. Hintikka J, Tolmunen T, Tanskanen A, Viinamäki H. High vitamin B12 level and good treatment outcome may be associated in major depressive disorder. BMC Psychiatry. 2003;3:17–22.[PMC free article] [PubMed] [Google Scholar]

16. Tiemeier H, vanTuijl HR, Hofman A, Meijer J, Kiliaan AJ, Breteler MM. Vitamin B12 folate and homocysteine in depression the Rotterdam Study. Am J Psychiatry. 2002;159(12):2099–01. [PubMed] [Google Scholar]

17. Sachdev PS, Parslow RA, Lux O, et al. Relationship of homocysteine folic acid and vitamin B12 with depression in a middle aged community sample. Psychol Med. 2005;35(4):529–38. [PubMed] [Google Scholar]

18. Bottiglieri T, Laundy M, Crellin R, Toone BK, Carney MW, Reynolds EH. Homocysteine folate methylation and monoamine metabolism in depression. J Neurol Neurosurg Psychiatry. 2000;69:228–232.[PMC free article] [PubMed] [Google Scholar]

19. Alpert JE, Mischoulon D. Folinic acid (Leucovorin) as an adjunctive treatment for SSRI-refractory depression. Ann Clin Psychiatry. 2002;14(1):33–8. [PubMed] [Google Scholar]

20. Fava M, Borus JS. Folate vitamin B12 and homocysteine in major depressive disorder. Am J Psychiatry. 1997;154(3):426–8. [PubMed] [Google Scholar]

21. Papakostas GI, Peterson T. Serum folate vitamin B12 and homocysteine in major depressive disorder Part I.predictors of clinical response in fluoxetine-resistant depression. J Clin Psychiatry. 2004;65(8):1090–5. [PubMed] [Google Scholar]

22. Papakostas GI, Peterson T. Serum folate vitamin B12 and homocysteine in major depressive disorder Part II predictors of clinical response in fluoxetine-resistant depression. J Clin Psychiatry. 2004;65(8):1096–8. [PubMed] [Google Scholar]

23. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960;23:56–62.[PMC free article] [PubMed] [Google Scholar]

24. Chaudhry HR, Qureshi Z, Tareen IA, Yazdani I. Efficacy and tolerability of Paroxetine 20 mg daily in the treatment of depression and depression associated anxiety. J Pak Med Assoc. 2002;52(11):518–25.[PubMed] [Google Scholar]

25. Yakub M, Iqbal MP, Kakepoto GN, et al. High prevalence of mild hyperhomocystienemia and folate B12 and B6 deficiencies in an urban population in Karachi Pakistan. Pak J Med Sci. 2010;26(4):923–929.[Google Scholar]

INTRODUCTION

Antidepressants are among the most widely used medications in clinical practice for the treatment of depressive disorder, panic disorder, generalized anxiety disorder, and numerous other psychiatric diseases[1,2]. Among the many classes of antidepressants, selective serotonin reuptake inhibitors (SSRIs) are the most commonly prescribed because of their effectiveness in treating many psychiatric disorders. SSRIs are safer and have a more favorable side-effect profile than the previous generations of antidepressants[3]. However, numerous studies have indicated that short-term use of SSRIs can cause many adverse side effects, including nausea, diarrhea and unstable mood swings. Additionally, suicide attempts, gastrointestinal bleeding, sexual dysfunction, and hyponatremia may occur during long-term use[4-6].

Irritable bowel syndrome (IBS) is part of a large group of functional gastrointestinal disorders that are characterized by recurrent abdominal discomfort or pain and disturbed defecation in the absence of organic disease[7,8]. IBS is one of the most commonly treated diseases by primary care physicians as well as gastroenterologists. The prevalence of IBS is estimated to be 7.5%-21% worldwide[9,10]. IBS is a functional disorder and has no contribution towards mortality[11]; however, it is chronic and significantly reduces patients’ quality of life[9,12,13]. The suggested treatments for IBS include antispasmodics, antidiarrheal agents, laxatives, prokinetics, probiotics, anxiolytics, SSRIs, tricyclic antidepressants (TCAs), 5-HT3 antagonists, cGMA agonists, and antibiotics according to each patient’s clinical symptoms[14,15]. However, there is no universally accepted or recommended therapy that effectively cures this disease.

The pathophysiology of IBS is believed to be associated with abnormal gastrointestinal motility, visceral hypersensitivity, low-grade inflammation, stress and brain-gut interactions[9,16]. Additionally, a high prevalence of psychiatric disorders in patients with IBS, in particular anxiety and depressive disorders, has been reported in previous studies[16,17]. Antidepressants are often used to treat a variety of functional bowel disorders. Tricyclic antidepressants have been proven to offer statistically significant control of IBS symptoms in a previous meta-analysis[18,19]. Additionally, several randomized controlled trials have evaluated the safety and efficacy of fluoxetine, citalopram and paroxetine for the treatment of IBS[20-23]. However, the evidence regarding the effectiveness of SSRIs in providing symptom relief in IBS is inconsistent. The American Gastroenterological Association Institute guidelines advise against using SSRIs for patients with IBS, based on the lack of improvement in global relief of symptoms identified in pooled estimates of five randomized control trials[24].

The present study aimed to explore the relationship of SSRIs with the subsequent development of IBS using the Taiwan National Health Insurance Research Database (NHIRD).

MATERIALS AND METHODS

Data source

The database used in our study was the NHIRD of Taiwan. The National Health Insurance (NHI) program in Taiwan was initiated in 1995 and enrolled over 24 million people by the end of 2014, representing 99% of the population in Taiwan. In the present study, data from the Longitudinal Health Insurance Database (LHID) 2000 were analyzed. The LHID 2000 is a subset of the NHIRD that contains data from 1000000 randomly sampled patients, which is approximately 5% of the Taiwan’s general population. We conducted a retrospective observational study on the correlation of SSRIs with and their possible influence on IBS. This study was approved by the Institutional Review Board (IRB) of Taichung Veterans General Hospital (IRB number: CE13152B-3), and because the data were obtained from the LHID 2000, informed consent from the participants was not obtained. The IRB specifically waived the requirement for consent.

Study population

We extracted data from the LHID 2000 for this retrospective study. Patients with a more than two-month medical prescription for an SSRI during one year between January 1, 2000 and December 31, 2010 were selected. The prescribed SSRIs included fluoxetine, citalopram, paroxetine, sertraline, fluvoxamine and escitalopram. We excluded any patients who were diagnosed with IBS (ICD-9-CM code: 564.1) prior to medical treatment with SSRIs. The comparison cohort included patients without any medical history of SSRI use who were frequency-matched with the SSRI cohort by age, sex, index date and year at a ratio of 1:4 (Figure ​(Figure1).1). To ensure the validity of the diagnosis, we included only patients who were diagnosed with IBS (ICD-9-CM code: 564.1) in more than three outpatient visits or more than one inpatient hospitalization. The included patients taking SSRIs were regularly followed at the database for at least three outpatient visits. Thus, we supposed that these patients needed these medication, and had compliance in medication.

In addition, patients with inflammatory bowel disease (ICD-9:555, 556) were excluded because some of these patients exhibit symptoms similar to those of IBS during the inactive or remission stage of inflammatory bowel syndrome[25-29]. Additionally, patients with a diagnosis of infectious enterocolitis (ICD 9: 0078,0079, 0080-0088, 0090-0093, 558, 0030, 0062, 11285) within three months prior to the diagnosis of IBS were excluded due to the potentially increased risk of post-infectious IBS associated with bacterial, protozoan, helminth, or viral infections, all of which have been reported[30-34]. The main outcome was the incidence of newly diagnosed IBS during the follow-up period, which was estimated as the duration from the index date until IBS, withdrawal from the insurance system, or the end of study in 2011.

Furthermore, common comorbidities diagnosed before enrollment in this study, including hypertension, diabetes mellitus, dyslipidemia, colorectal cancer, major depressive disorder, anxiety disorder, bipolar disorder, and posttraumatic disorder, were compared between the SSRI and comparison cohorts.

Statistical analysis

All analyses were performed using SAS software, version 9.3 (SAS Institute, Cary, NC, United States). The distributions of SSRI exposure based on subject’s age, gender and clinical comorbidities were examined using χ2 tests for categorical variables.

In the cohort study, multivariable Cox proportional hazard models were used to explore the relationship between exposure to SSRI and the diagnosis of IBS, after adjusting for age, gender and medical comorbidities. All statistical tests were two-sided, conducted at a significance level of 0.05, and reported using P values and/or 95%CIs.

Go to:RESULTSDemographic characteristics of study subjects

The eligible study participants included 19653 patients in the SSRI cohort and 78612 persons in the comparison cohort, with a similar age and sex distribution (Figure ​(Figure11 and Table ​Table1).1). Males represented 41.5% and females 58.5% of the entire study population. The mean follow-up time in the present study was 5.9 ± 3.0 years in the non-SSRI cohort and 5.5 ± 3.2 years in the SSRI cohort. The mean follow-up period of SSRI exposure to IBS diagnosis was 2.05 years. The majority of psychiatric disorders leading to a prescription of SSRI included anxiety (48.2%) and major depressive disorders (21.8%). There was more concomitant anti-psychotic usage in the SSRI cohort than in the non-SSRI group. However, most participants in both groups did not use anti-psychotics.

At the baseline, comorbid diabetes, hypertension, hyperlipidemia, colorectal cancer, major depressive disorder and anxiety disorder were more prevalent in the SSRI cohort than in the comparison cohort.

Risk of IBS in SSRI users

A total of 236 patients in the SSRI cohort (incidence, 2.17/1000 person-years) and 478 patients in the comparison cohort (incidence, 1.04/1000 person-years) had a new diagnosis of IBS during the follow-up period (Table ​(Table2).2). The incidence of IBS increased with advancing age. Comorbidities such as diabetes mellitus, hypertension, hyperlipidemia, colorectal cancer, and major depressive disorder did not influence the HR of IBS. However, patients with anxiety disorders had a significantly increased HR of IBS (HR = 1.33, 95%CI: 1.11-1.59, P = 0.002). The use of anti-psychotics did not affect the incidence of IBS, whereas the use of SSRIs was associated with an increased HR of IBS.

Incident rate and HR of IBS associated with SSRI use in Cox regression analyses

After adjusting for sex, age and other comorbidities including diabetes, hypertension, hyperlipidemia and colorectal cancer, the overall adjusted HR (aHR) in the SSRI cohort compared with the comparison cohort was 1.74 (95%CI: 1.44-2.10; P < 0.001) using Cox regression analysis. The subgroup analysis showed that the aHR was higher in SSRI users than in non-SSRI users among females (aHR = 1.65; 95%CI: 1.29-2.11), males (aHR = 1.85; 95%CI: 1.38-2.48) and individuals aged between 20 and 60 years (Table ​(Table33).

The cumulative incidence of IBS was higher in the SSRI cohort than in the non-SSRI cohort (log-rank test, P < 0.001) (Figure ​(Figure22).

HR of IBS associated with the duration of SSRI exposure

Table ​Table44 shows the association between SSRI exposure days over one year and the HR of a subsequent diagnosis of IBS. The aHR was highest in individuals with a one-year SSRI exposure of less than 90 d (aHR = 3.27, 95%CI: 2.61-4.08). The aHR remained significantly higher in patients with longer durations of SSRI exposure.

Hazard ratio of IBS among SSRI and antidepressant users

A subgroup analysis of SSRI and non-SSRI users showed a significantly increased aHR of IBS in SSRI only users (aHR = 1.82, 95%CI: 1.50-2.21, P < 0.001) but not in the users of other antidepressants only (aHR = 1.33, 95%CI: 0.75-2.36, P = 0.338) or the combined SSRI and other antidepressant users (aHR = 1.30, 95%CI: 0.84-2.01, P = 0.235) (Table ​(Table55).

DISCUSSION

Based on our review of the literature, this is the first study using nationwide population database to investigate the relationship between SSRI prescriptions and IBS in Taiwan. Our results revealed that patients exposed to SSRIs had an increased risk of developing IBS (aHR = 1.74) after adjusting for sex, age, and major comorbidities.

In this study, we demonstrated that IBS in SSRI users tended to occur in older patients. The incidence of IBS became higher as age increased in both SSRI users and non-users, and this finding conflict with both the global data and a previous questionnaire survey conducted in Taiwan[10,35]. A previous global meta-analysis and questionnaire study in Taiwan showed that the prevalence of IBS decreased with advancing age. However, the participants in the previous questionnaire study in Taiwan were healthy volunteers, and thus the prevalence of IBS in the Taiwan’s general population may have been underestimated. Additionally, IBS symptoms may occur at early ages with relatively benign symptoms. Most patients may tolerate these symptoms and not seek medical advice; this tendency may have also led to an underestimation of the incidence of IBS in young individuals.

To evaluate the dose effect of SSRIs on the subsequent development of IBS, we analyzed the days of SSRI exposure within one year and determined the HR of subsequent IBS diagnosis. The aHR was highest in individuals with SSRI exposure for less than 90 d but remained significantly higher for patients with longer exposure times. Although SSRIs include widely used antidepressants that are more tolerable and have relatively benign adverse effects compared to TCAs or monoamine oxidase inhibitors[3], they continue to have some early onset adverse effects and problems associated with long-term treatment. The early onset adverse effects include gastrointestinal discomfort, nausea, dyspepsia and diarrhea and disappear within two to three weeks[5]. The higher HR of IBS in individuals with lower SSRI exposure times may be due to these early onset gastrointestinal side effects that lead to misdiagnosis of IBS by clinical physicians. However, Table ​Table44 shows that patients with SSRI exposure for more than 90 d had a significantly higher HR of IBS, and this finding cannot be explained by the early gastrointestinal adverse effects of SSRIs. In addition, the mean follow-up time from SSRI exposure to IBS diagnosis was 2.05 years, which is long enough to exclude transient adverse effects of SSRIs. Therefore, it can be hypothesized that the long duration of psychiatric disorders leads to an increased risk of subsequent IBS.

Many studies have identified a relationship between IBS and psychiatric disorders[17,36]. In patients with IBS who seek treatment, the rates of comorbid psychiatric disorders range from 54% to 94%[9,12]. Anxiety and depression disorders are associated with gastrointestinal symptoms in accordance with brain-gut interactions[37-40]. Additionally, the communication between the central nervous system and enteric nervous system appears to be bidirectional[36]. The biopsychological model of IBS suggests that deterioration of gastrointestinal symptoms could exacerbate anxiety and depression (bottom-up model) and that psychological factors themselves similarly influenced physiological factors (top-down model)[41]. Fond et al[17] demonstrated in a systematic review and meta-analysis that patients with IBS have significantly higher levels of anxiety and depression than healthy controls. To exclude the influence of other antidepressants, we performed a subgroup analysis of SSRI and non-SSRI users. Table ​Table55 shows that the aHR was significantly higher in the users of SSRIs only compared with the users of non-SSRIs or users of combined SSRIs and other antidepressants. The use of other antidepressants only or in combination with SSRIs was not associated with an increased risk of IBS.

One possible reason for the higher incidence of IBS in the SSRI cohort is that the patients’ underlying psychiatric disorders, particularly anxiety, may have deteriorated and thus their subclinical gastrointestinal symptoms became overt. Under these conditions, clinical physicians may provide new prescriptions for SSRIs, and patients may seek medical advice for IBS symptoms. This process is consistent with our study results, which showed that patients with anxiety disorders had a significantly higher HR of IBS. Poorly controlled anxiety disorders and unstable mood may exacerbate the symptoms of IBS. The increased HR of IBS in the SSRI cohort, compared with the comparison cohort in our study, may be due to the increased severity of anxiety disorders, poor compliance to SSRIs due to adverse medication effects or personal reasons.

There were some limitations to our study. First, there is only one code for IBS (ICD-9-CM code: 564.1) in the ICD-9 system, and further subgroup analyses were therefore not feasible. Second, data on lifestyle factors, such as smoking and alcohol use, were not available in the NHIRD. Third, information on drug compliance was not obtained from this health care database. Fourth, the severity of any psychiatric disorder upon enrollment in the study was also not available in the NHIRD. Fifth, there are always coding issues in database studies. There may be registration differences among the physicians and psychiatrists who did not code for IBS and the gastroenterologists who did not code for depression/anxiety. However, most physicians should adhere to the proper coding standards due to NHI payment rules.

In conclusion, SSRI use is associated with subsequently diagnosed IBS. In clinical practice, it is important to pay attention to the gastrointestinal symptoms of patients with psychiatric disorders who use SSRIs.

Go to:ACKNOWLEDGMENTS

We thank the National Health Research Institute and the Bureau of National Health Insurance for providing the data for this study.

Go to:COMMENTSBackground

Irritable bowel syndrome (IBS) is part of a large group of functional gastrointestinal disorders that significantly reduce patients’ quality of life. Antidepressants, selective serotonin reuptake inhibitors (SSRIs) in particular, have been used to treat refractory IBS. The influence of SSRIs on the subsequent development of IBS remains unknown.

Research frontiers

The results regarding the clinical efficacy of SSRI in treating IBS are inconsistent. This study presents the first attempt to elucidate the relationship between SSRI use and subsequent diagnosis of IBS.

Innovations and breakthroughs

The overall adjusted hazard ratio was higher in the SSRI cohort than in the comparison cohort. The incidence of IBS in SSRI users increased with advancing age.

Applications

Physicians in clinical practice should pay attention to the gastrointestinal symptoms of patients with psychiatric disorders and SSRI use.

Terminology

IBS is characterized by recurrent abdominal discomfort or pain and disturbed defecation in the absence of an organic disease.

Peer-review

This study by Lin et al probes the relationship between SSRI use and the subsequent diagnosis of IBS over a 10-year span, using a national health insurance research database. The data indicate an adjusted increase in HR of 1.74 (P = 0.002) for the diagnosis of IBS in patients treated with SSRIs.

REFERENCES

1. Ables AZ, Baughman OL. Antidepressants: update on new agents and indications. Am Fam Physician. 2003;67:547–554. [PubMed] [Google Scholar]

2. Masand PS, Gupta S. Selective serotonin-reuptake inhibitors: an update. Harv Rev Psychiatry. 1999;7:69–84. [PubMed] [Google Scholar]

3. Brambilla P, Cipriani A, Hotopf M, Barbui C. Side-effect profile of fluoxetine in comparison with other SSRIs, tricyclic and newer antidepressants: a meta-analysis of clinical trial data. Pharmacopsychiatry. 2005;38:69–77. [PubMed] [Google Scholar]

4. Nierenberg AA, Ostacher MJ, Huffman JC, Ametrano RM, Fava M, Perlis RH. A brief review of antidepressant efficacy, effectiveness, indications, and usage for major depressive disorder. J Occup Environ Med. 2008;50:428–436. [PubMed] [Google Scholar]

5. Moret C, Isaac M, Briley M. Problems associated with long-term treatment with selective serotonin reuptake inhibitors. J Psychopharmacol. 2009;23:967–974. [PubMed] [Google Scholar]

6. Spigset O. Adverse reactions of selective serotonin reuptake inhibitors: reports from a spontaneous reporting system. Drug Saf. 1999;20:277–287. [PubMed] [Google Scholar]

7. Longstreth G, Thompson W, Chey W, Houghton L, Mearin F, Spiller R. Rome III: the functional gastrointestinal disorders. 3rd ed. Functional bowel disorders 2006: 487-555 [Google Scholar]

8. Ford AC, Moayyedi P, Lacy BE, Lembo AJ, Saito YA, Schiller LR, Soffer EE, Spiegel BM, Quigley EM. American College of Gastroenterology monograph on the management of irritable bowel syndrome and chronic idiopathic constipation. Am J Gastroenterol. 2014;109 Suppl 1:S2–S26; quiz S27. [PubMed] [Google Scholar]

9. Drossman DA, Camilleri M, Mayer EA, Whitehead WE. AGA technical review on irritable bowel syndrome. Gastroenterology. 2002;123:2108–2131. [PubMed] [Google Scholar]

10. Lovell RM, Ford AC. Global prevalence of and risk factors for irritable bowel syndrome: a meta-analysis. Clin Gastroenterol Hepatol. 2012;10:712–721.e4. [PubMed] [Google Scholar]

11. Canavan C, West J, Card T. The epidemiology of irritable bowel syndrome. Clin Epidemiol. 2014;6:71–80. [PMC free article] [PubMed] [Google Scholar]

12. Whitehead WE, Palsson O, Jones KR. Systematic review of the comorbidity of irritable bowel syndrome with other disorders: what are the causes and implications? Gastroenterology. 2002;122:1140–1156. [PubMed] [Google Scholar]

13. Canavan C, West J, Card T. Review article: the economic impact of the irritable bowel syndrome. Aliment Pharmacol Ther. 2014;40:1023–1034. [PubMed] [Google Scholar]

14. Chang L, Lembo A, Sultan S. American Gastroenterological Association Institute Technical Review on the pharmacological management of irritable bowel syndrome. Gastroenterology. 2014;147:1149–1172.e2.[PubMed] [Google Scholar]

15. Gwee KA, Bak YT, Ghoshal UC, Gonlachanvit S, Lee OY, Fock KM, Chua AS, Lu CL, Goh KL, Kositchaiwat C, et al. Asian consensus on irritable bowel syndrome. J Gastroenterol Hepatol. 2010;25:1189–1205. [PubMed] [Google Scholar]

16. Talley NJ, Spiller R. Irritable bowel syndrome: a little understood organic bowel disease? Lancet. 2002;360:555–564. [PubMed] [Google Scholar]

17. Fond G, Loundou A, Hamdani N, Boukouaci W, Dargel A, Oliveira J, Roger M, Tamouza R, Leboyer M, Boyer L. Anxiety and depression comorbidities in irritable bowel syndrome (IBS): a systematic review and meta-analysis. Eur Arch Psychiatry Clin Neurosci. 2014;264:651–660. [PubMed] [Google Scholar]

18. Rahimi R, Nikfar S, Rezaie A, Abdollahi M. Efficacy of tricyclic antidepressants in irritable bowel syndrome: a meta-analysis. World J Gastroenterol. 2009;15:1548–1553. [PMC free article] [PubMed] [Google Scholar]

19. Xie C, Tang Y, Wang Y, Yu T, Wang Y, Jiang L, Lin L. Efficacy and Safety of Antidepressants for the Treatment of Irritable Bowel Syndrome: A Meta-Analysis. PLoS One. 2015;10:e0127815.[PMC free article] [PubMed] [Google Scholar]

20. Creed F, Fernandes L, Guthrie E, Palmer S, Ratcliffe J, Read N, Rigby C, Thompson D, Tomenson B. The cost-effectiveness of psychotherapy and paroxetine for severe irritable bowel syndrome. Gastroenterology. 2003;124:303–317. [PubMed] [Google Scholar]

21. Ladabaum U, Sharabidze A, Levin TR, Zhao WK, Chung E, Bacchetti P, Jin C, Grimes B, Pepin CJ. Citalopram provides little or no benefit in nondepressed patients with irritable bowel syndrome. Clin Gastroenterol Hepatol. 2010;8:42–48.e1. [PMC free article] [PubMed] [Google Scholar]

22. Kuiken SD, Tytgat GN, Boeckxstaens GE. The selective serotonin reuptake inhibitor fluoxetine does not change rectal sensitivity and symptoms in patients with irritable bowel syndrome: a double blind, randomized, placebo-controlled study. Clin Gastroenterol Hepatol. 2003;1:219–228. [PubMed] [Google Scholar]

23. Tack J, Broekaert D, Fischler B, Van Oudenhove L, Gevers AM, Janssens J. A controlled crossover study of the selective serotonin reuptake inhibitor citalopram in irritable bowel syndrome. Gut. 2006;55:1095–1103. [PMC free article] [PubMed] [Google Scholar]

24. Weinberg DS, Smalley W, Heidelbaugh JJ, Sultan S. American Gastroenterological Association Institute Guideline on the pharmacological management of irritable bowel syndrome. Gastroenterology. 2014;147:1146–1148. [PubMed] [Google Scholar]

25. Bayless TM, Harris ML. Inflammatory bowel disease and irritable bowel syndrome. Med Clin North Am. 1990;74:21–28. [PubMed] [Google Scholar]

26. Annese V, Bassotti G, Napolitano G, Usai P, Andriulli A, Vantrappen G. Gastrointestinal motility disorders in patients with inactive Crohn’s disease. Scand J Gastroenterol. 1997;32:1107–1117. [PubMed] [Google Scholar]

27. Pezzone MA, Wald A. Functional bowel disorders in inflammatory bowel disease. Gastroenterol Clin North Am. 2002;31:347–357. [PubMed] [Google Scholar]

28. Ginsburg PM, Bayless TM. Managing Functional Disturbances in Patients with Inflammatory Bowel Disease. Curr Treat Options Gastroenterol. 2005;8:211–221. [PubMed] [Google Scholar]

29. Isgar B, Harman M, Kaye MD, Whorwell PJ. Symptoms of irritable bowel syndrome in ulcerative colitis in remission. Gut. 1983;24:190–192. [PMC free article] [PubMed] [Google Scholar]

30. Wang LH, Fang XC, Pan GZ. Bacillary dysentery as a causative factor of irritable bowel syndrome and its pathogenesis. Gut. 2004;53:1096–1101. [PMC free article] [PubMed] [Google Scholar]

31. Marshall JK, Thabane M, Garg AX, Clark WF, Salvadori M, Collins SM. Incidence and epidemiology of irritable bowel syndrome after a large waterborne outbreak of bacterial dysentery. Gastroenterology. 2006;131:445–450; quiz 660. [PubMed] [Google Scholar]

32. Zanini B, Ricci C, Bandera F, Caselani F, Magni A, Laronga AM, Lanzini A. Incidence of post-infectious irritable bowel syndrome and functional intestinal disorders following a water-borne viral gastroenteritis outbreak. Am J Gastroenterol. 2012;107:891–899. [PubMed] [Google Scholar]

33. Dizdar V, Gilja OH, Hausken T. Increased visceral sensitivity in Giardia-induced postinfectious irritable bowel syndrome and functional dyspepsia. Effect of the 5HT3-antagonist ondansetron. Neurogastroenterol Motil. 2007;19:977–982. [PubMed] [Google Scholar]

34. Hanevik K, Wensaas KA, Rortveit G, Eide GE, Mørch K, Langeland N. Irritable bowel syndrome and chronic fatigue 6 years after giardia infection: a controlled prospective cohort study. Clin Infect Dis. 2014;59:1394–1400. [PMC free article] [PubMed] [Google Scholar]

35. Lu CL, Chen CY, Lang HC, Luo JC, Wang SS, Chang FY, Lee SD. Current patterns of irritable bowel syndrome in Taiwan: the Rome II questionnaire on a Chinese population. Aliment Pharmacol Ther. 2003;18:1159–1169. [PubMed] [Google Scholar]

36. Fadgyas-Stanculete M, Buga AM, Popa-Wagner A, Dumitrascu DL. The relationship between irritable bowel syndrome and psychiatric disorders: from molecular changes to clinical manifestations. J Mol Psychiatry. 2014;2:4. [PMC free article] [PubMed] [Google Scholar]

37. Fioramonti J, Bueno L. Centrally acting agents and visceral sensitivity. Gut. 2002;51 Suppl 1:i91–i95.[PMC free article] [PubMed] [Google Scholar]

38. Kennedy PJ, Cryan JF, Dinan TG, Clarke G. Irritable bowel syndrome: a microbiome-gut-brain axis disorder? World J Gastroenterol. 2014;20:14105–14125. [PMC free article] [PubMed] [Google Scholar]

39. Gunter WD, Shepard JD, Foreman RD, Myers DA, Greenwood-Van Meerveld B. Evidence for visceral hypersensitivity in high-anxiety rats. Physiol Behav. 2000;69:379–382. [PubMed] [Google Scholar]

40. Accarino AM, Azpiroz F, Malagelada JR. Attention and distraction: effects on gut perception. Gastroenterology. 1997;113:415–422. [PubMed] [Google Scholar]

41. Stasi C, Rosselli M, Bellini M, Laffi G, Milani S. Altered neuro-endocrine-immune pathways in the irritable bowel syndrome: the top-down and the bottom-up model. J Gastroenterol. 2012;47:1177–1185.[PubMed] [Google Scholar]

Introduction

Depression is a severe and common mental disorder with 12-month prevalence as high as 3.2% in subjects without comorbid physical disease and 9.3 to 23.0% in subjects with chronic medical conditions [1]. Despite its frequent occurrence, high likelihood of a chronic course, negative impact on quality of life and ability to work, and strong association with an increased suicide risk, the available treatment options for depression are still not satisfactory for many patients. The most neglected pharmacological needs in the treatment of depression are the lack of early-onset response to the treatment, the moderate response and low remission rate to the first antidepressant trial, and side effects which frequently cause treatment non-compliance [2]. Among the most common side effects of antidepressants and residual symptoms leading to incomplete remission from depression are those related to sleep. The aim of this review article is to summarize the literature published in recent years on how antidepressants affect sleep, as an addition to our [3] and previous reviews on this topic [e.g. 4–8]. We also summarize recent data which has shaped our personal view on the use of antidepressants in treating insomnia in depressed and non-depressed subjects.

Go to:Polysomnographic Sleep Studies in Depression

The most detailed information on sleep in depression was provided by studies using polysomnography (PSG), that is considered the gold standard for sleep assessment. Based on registration of the three physiological parameters such as the brain (EEG), muscle (EMG) and eye movements (EOG) bioelectric activity, PSG allows human sleep to be scored into sleep stages. Subsequently, it is possible to calculate several sleep parameters (Table ​(Table1)1) that express sleep continuity, sleep depth, and distribution of sleep stages.

Patients with depression show abnormalities of sleep parameters across all three groups. Disrupted sleep continuity manifests as prolongation of sleep latency, increased number, and duration of awakenings from sleep expressed as increased wake after sleep onset (WASO) time, decreased sleep efficiency, and early morning awakenings. Early morning, awakening together with altered distribution of REM sleep is considered a biological marker of circadian rhythm disturbances in depression and is a characteristic biological marker of depression with melancholic features [9]. Sleep depth is substantially reduced in depressed patients. Furthermore, the distribution of deep sleep scored in PSG as sleep stage N3, also called delta or slow wave sleep (SWS), is altered in depressed patients. In healthy subjects, the highest delta wave activity in EEG can be observed in the first sleep cycle, whereas in depressed patients there is a frequent shift of delta activity from the first to the second sleep cycle. It is expressed in sleep parameters as a reduced delta ratio (ratio between delta wave activity in the first and second sleep cycle). Alterations of REM sleep are the most prominent feature of sleep architecture in depressed subjects. They include shortened REM sleep latency, increased REM sleep time (especially in the first sleep cycle that is usually very short in healthy subjects), and increased REM sleep density. The recent meta-analysis summarizing the current evidence from studies using PSG about sleep architecture in mental disorders has confirmed that disturbed sleep is a core symptom of depression. It was found that affective disorders and major depression were associated with alterations in most variables compared to healthy controls [10••]. Although none of the sleep parameters was specific to depression, the high prevalence and severity of sleep abnormalities in depressed patients are of a great clinical importance. The complaints of insomnia are present in 60–90% of patients with major depression, depending on the episode’s severity. When it comes to bipolar disorder, insomnia is present during a depressive episode in 60% of patients, while 20–30% suffer from prolonged sleep (hypersomnia) and increased daytime sleepiness [11, 12]. Fortunately, in most patients, sleep disturbances diminish with the improvement of depressive symptoms, especially if the clinical improvement is related to the recurrence of interest and pleasure in everyday activities. Because it is usually related to the substantially increased physical daytime activity, it increases homeostatic sleep need, which improves sleep depth and duration. However in many patients, difficulties with sleep persist. The Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study, which included a large sample of outpatients with non-psychotic MDD who responded without remitting after up to 12 weeks treatment with citalopram, found that midnocturnal insomnia was the most commonly observed residual symptom of depression, as it was still present in 79% patients [13].

The observations on the high prevalence of subjective insomnia complaints and objective worsening of sleep architecture in PSG studies in depressed patients are important for the choice of pharmacological treatment. Antidepressant drugs substantially differ in their acute effects on sleep. Some of them alleviate sleep disturbances, but other may disrupt sleep, which is related to poor treatment compliance. Persistent insomnia symptoms may also result in unfavorable clinical outcome, e.g., increased suicide risk [14]. Therefore, it is important to know what is the preferred pharmacological treatment in a depressed patient with clinically relevant insomnia symptoms.

Go to:Effects on Antidepressants on Sleep

It is well known that some classes of antidepressant drugs may deteriorate sleep quality mainly due to activation of serotonergic 5-HT2 receptors and increased noradrenergic and dopaminergic neurotransmission (Table ​(Table2).2). Among them, most prominent are serotonin and norepinephrine reuptake inhibitors (SNRI), norepinephrine reuptake inhibitors (NRI), monoamine oxidase inhibitors (MAOI), selective serotonin reuptake inhibitors (SSRI), and activating tricyclic antidepressants (TCA).

On the contrary, antidepressants with antihistaminergic action, like sedating TCA, mirtazapine, mianserine, or strong antagonistic action at serotonergic 5-HT2 receptors, like trazodone and nefazodone quickly improve sleep. Some patients show improvement of sleep quality already after the first drug dose [15], which was specially discussed for mirtazapine as related to the faster onset of antidepressant action [16].

In a recent review article on the prevalence of treatment emergent insomnia and somnolence in depressed patients, it was shown that subjective complaints of insomnia or daytime somnolence were frequent in patients suffering from depression or anxiety disorders treated with SSRI and SNRI [16].

Based on data from the US Food and Drug Administration (FDA) study register [16], the average prevalence of treatment-emergent insomnia in clinical trials with SSRI was 17% as compared to 9% out of patients randomized to the placebo arm. The average rate of treatment emergent somnolence in patients being treated with SSRI amounted to 16% as compared to 8% out of patients receiving placebo. The lowest rate of treatment emergent insomnia complaints (below 2%) was reported in the study with citalopram. The highest rate of treatment-emergent insomnia and somnolence was found in patients suffering from obsessive-compulsive disorder (OCD) being treated with high-dose fluvoxamine, 31 and 27%, respectively [17••]. In clinical trials with SNRI, treatment-emergent insomnia was reported on average in 13% out of SNRI-treated patients as compared to 7% out of the placebo arm and treatment-emergent somnolence in 10% of SNRI-treated patients in comparison to 5% out of patients receiving placebo. Both treatment-emergent insomnia and somnolence were the most frequent (both equal to 24%) in patients with generalized anxiety disorders treated with venlafaxine. The lowest rate of treatment emergent insomnia and somnolence (both below 2%) was reported for levomilnacipran. On the contrary to the treatment with SSRI and SNRI, in clinical studies with sedating antidepressants, e.g., mirtazapine and trazodone, the reported prevalence of treatment-emergent insomnia complaints in patients with major depressive disorder (MDD) was very low (below 2%). However, the rate of treatment-emergent somnolence was very high, 54% in patients being treated with mirtazapine as compared to 18% out of patients in the placebo arm and 46% in patients being treated with trazodone as compared to 19% out of patients receiving placebo. It is important to note that acute effects of antidepressants on sleep are reflected not only in the patients’ subjective complaints but they can also be demonstrated in studies with PSG (Table ​(Table2).2). While SSRI, SNRI, and activating TCA increase REM latency, suppress REM sleep, and may impair sleep continuity, sedating antidepressants decrease sleep latency, improve sleep efficiency, increase SWS, and usually have little or no effect on REM sleep [3, 6, 7, 17••]. Although both the sleep-disrupting and sleep-promoting effects of the antidepressants are the strongest only in the first few weeks of treatment, in some patients they may persist, aggravating insomnia complaints or causing daytime somnolence [18]. Therefore, for the depressed patients with clinically significant insomnia, a treatment with a sedative antidepressant is usually more recommended [19]. It was recently shown that such treatment significantly reduces the need to use benzodiazepines in patients with MDD [20•]. Such an approach, combination treatment with benzodiazepines and SSRI/SNRI is often necessary to reduce anxiety and insomnia as early as in the first week of treatment. However, there is a related risk that the patient suffering from depression and insomnia will not be able to stop such treatment after 14–28 days of therapy and will develop a dependence [21, 22]. On the other hand, because the sleep complaints usually improve after a few weeks of effective treatment of depression with SSRI/SNRI, it is important to consider whether the use of hypnotics is not a better short-term treatment option for a patient than risking oversedation during treatment with a sedative antidepressant. The sedating effect of those antidepressants is usually an increasing problem in long-term maintenance treatment, frequently resulting in a need to reduce the drug dose. It may substantially diminish the efficacy of the maintenance treatment. The sedative antidepressants may also induce a weight gain, what is particularly shown for mirtazapine but not for trazodone [23].

Agomelatine should be considered as an alternative approach to the treatment of depressed patients with marked insomnia symptoms. Agomelatine is a non-sedative antidepressant drug exerting agonistic action at melatonergic M1 and M2 receptors, and antagonistic action at serotonergic 5-HT2c receptors [24]. Such pharmacodynamic profile is related to sleep-promoting action without the risk of sedation and weight gain. In comparison to escitalopram, agomelatine is known to improve sleep latency after both short (after 2 weeks) and long (after 24 weeks) treatment. Moreover, both drugs differ significantly in their effect on sleep continuity. In the second week, agomelatine slightly improves sleep continuity (increased total sleep time and sleep efficiency) and escitalopram worsens it [25]. Moreover, treatment with agomelatine is not related to the suppression of REM sleep: it restores cyclical sleep profile, may increase the amount of SWS, and most importantly leads to the improvement of daytime alertness [26].

Effects on sleep has recently been also reported for a vortioxetine, with clinical action mediated mainly by selective blockade of serotonin reuptake and direct modulation of serotonergic receptors activity (such as 5-HT3, 5-HT7, 5-HT1D, and 5-HT1B) [27]. In a study which has compared the effects of vortioxetine and paroxetine to the placebo in a group of 24 healthy male volunteers, it has been shown that the vortioxetine dose of 20 and 40 mg similarly to the paroxetine dose of 20 mg suppresses REM sleep by increasing REM sleep latency and diminishing duration of REM sleep. Both drugs also decrease total sleep time and increase duration of sleep stage N1. These negative effects of vortioxetine on sleep continuity are significant only for the higher dose [28•]. According to the FDA clinical trial register, the rate of treatment-emergent insomnia complaints or somnolence during the therapy with vortioxetine is lower when compared to SSRI and SNRI drugs [17••].

The use of antidepressants, also those with sedative properties, may impair sleep due to the induction of sleep disorders or worsening already existing ones. Mianserin and mirtazapine may induce restless legs syndrome in as many as 28% of patients [29]. Treatment-emergent RLS has also been described for SSRI and venlafaxine [30]. SSRI, SNRI, and TCA are known to induce or exacerbate sleep bruxism and disturb regulation of muscle tone during REM sleep, causing REM sleep without atonia, which may induce or worsen REM Sleep Behavior Disorder [3, 6]. Moreover, although antidepressants are recommended for the treatment of post-traumatic sleep disorder, they can induce nightmares. We observe this side effect most frequently during the treatment with mirtazapine, just as it was recently reported [31]. Finally, antidepressants inducing weight gain are contraindicated in patients with sleep apnea, that is an overlooked but frequent sleep disorder in people suffering from mental illness [32].

Go to:Treatment of Insomnia Disorder with Low-Dose Antidepressants

Insomnia belongs to the most frequent disorders of the brain [33]. In industrialized countries, approximately 6% of the adults suffer from insomnia as a disorder [34] and as many as 50% may suffer from transient insomnia symptoms [35•]. Although insomnia is not regarded as a severe mental disorder, it shares many features with depression. In order to offer a patient an effective treatment of insomnia, there is a need for a broader perspective, one that reaches far beyond the prescription of hypnotics. Current treatment guidelines [36••] strongly recommend the use of cognitive-behavioral therapy (CBT-I) as the initial treatment for chronic insomnia disorder and only short-term use of the sleep-promoting drugs in patients with insufficient response to CBI-I. However, in daily clinical practice, the use of pharmacotherapy for insomnia is very common. The most frequently used drugs to treat insomnia aside from benzodiazepines and non-benzodiazepine (eszopiclone/zopiclone, zaleplon, zolpidem) hypnotics are sedative antidepressants. However, due to the lack of methodologically sound randomized clinical trials in insomnia, only one of them, doxepin, is approved by FDA for the treatment of sleep maintenance insomnia. Furthermore, recent recommendations discourage the use of other drugs from this class than doxepin for the insomnia treatment [37••]. In our opinion, sedative antidepressants are a valuable treatment option of insomnia in a situation in which despite being in CBT-I therapy, the patient still requires sleep-promoting drugs more than 3–4 times per week. The use of sedative antidepressants should be also considered when there is a comorbid mood or anxiety disorder because such patients are at increased risk of developing hypnotic dependency. Moreover, in many insomnia patients, physiological parameters, e.g., hormone secretion, whole body, and brain metabolic rate, are altered in a similar fashion to the depressed ones what is called a hyperarousal [38, 39], supporting the use of sedative antidepressants to treat such patients. The pros and cons of using sedative antidepressants in insomnia patients were discussed extensively in the earlier papers. This is especially true for trazodone that is very often used as a sleep-promoting drug [3, 39–42]. Frequently expressed concern with the usage of sedative antidepressants in insomnia is that their side effect profile and interactions with other drugs may be underrated [40]. Indeed, although there is evidence for efficacy of sedative antidepressants to promote sleep, for example for TCA in a form of a recent meta-analytic study [43•], it is important to remember that these drugs should be used in insomnia patients only in a very low dose, e.g., for doxepin as low as 3 to 6 mg or 25–50 mg for trazodone. Many psychiatrists are astonished that a sedative antidepressant can promote sleep in such a low dose. Firstly, it should be noted that such low doses are appropriate only for patients with primary insomnia. In the presence of a comorbid mood disorder, the antidepressants have to be used in a recommended therapeutic dose [42]. Secondly, such treatment should be used only when combined with behavioral interventions from CBT-I protocol. When a patient restricts time in bed and uses stimulus control technique, even low-dosage pharmacological treatment starts to work. Thirdly, to be effective in treating sleep-onset insomnia, sedative antidepressants have to be taken much earlier than hypnotics in regard to their pharmacokinetics, especially the time they take to reach the maximum serum concentration (Cmax). It usually means at least 2 hours before sleep (in the case of more rapid drug action the patient should be encouraged to shorten this time). In our opinion, sedative antidepressants are a safe class of drugs when given in low doses. We use them in many patient groups where hypnotics are contraindicated, e.g., in the elderly patients, in patients with sleep apnea and in patients with a history of alcohol and substance abuse. Despite the fact that the use of atypical antipsychotics, mostly quetiapine [44], is increasing for treatment of insomnia accompanying bipolar disorder and schizophrenia, we hold the conviction that sedative antidepressants are a valuable treatment option for such patients as well. Based on our clinical experience and review of published case reports, we believe that the use of sedative antidepressants in a low dose is not related to the increased risk of phase shift in bipolar disorder [45]. Moreover, we have observed that for the treatment of insomnia low doses (5–10 mg) of citalopram administered in the morning can be an alternative to sedative antidepressants with good treatment effects [46].

Go to:Conclusions

Disturbed sleep is a core symptom of depression and its normalization is necessary to achieve remission from the illness. In the long term, all antidepressants which show clinical efficacy improve sleep secondary to improvement of mood and daytime activity. However, in the short term, while some of them may impair sleep due to the activating effects, other may improve sleep due to the sedative properties. Although sleep-promoting action is desired in depressed patients with coexisting anxiety or insomnia, it may be problematic during the maintenance treatment after recovery from depression due to oversedation. Thus, it is necessary to understand the effects of these drugs on the sleep and daytime alertness. It is particularly noteworthy that for sleep-promoting effect, it is sufficient to use a sedative antidepressant in a low dose. In such dose, these drugs can be also combined with other antidepressants as an alternative to hypnotic drugs, especially if there is a clinical necessity to promote sleep for longer than 2–4 weeks with a frequency higher than 3–4 times per week.

REFERENCES

1. Moussavi S, Chatterji S, Verdes E, Tandon A, Patel V, Ustun B. Depression, chronic diseases, and decrements in health: results from the world health surveys. Lancet. 2007;370:851–858. doi: 10.1016/S0140-6736(07)61415-9. [PubMed] [CrossRef] [Google Scholar]

2. Greden JF. Unmet need: what justifies the search for a new antidepressant? J Clin Psychiatry. 2002;63(Suppl 2):3–7. [PubMed] [Google Scholar]

3. Wichniak A, Wierzbicka A, Jernajczyk W. Sleep and antidepressant treatment. Curr Pharm Des. 2012;18:5802–5817. doi: 10.2174/138161212803523608. [PubMed] [CrossRef] [Google Scholar]

4. Le Bon O. Contribution of sleep research to the development of new antidepressants. Dialogues Clin Neurosci. 2005;7:305–313. [PMC free article] [PubMed] [Google Scholar]

5. Mendlewicz J. Sleep disturbances: core symptoms of major depressive disorder rather than associated or comorbid disorders. World J Biol Psychiatry. 2009;10:269–275. doi: 10.3109/15622970802503086.[PubMed] [CrossRef] [Google Scholar]

6. Wilson S, Argyropoulos S. Antidepressants and sleep: a qualitative review of the literature. Drugs. 2005;65:927–947. doi: 10.2165/00003495-200565070-00003. [PubMed] [CrossRef] [Google Scholar]

7. Gursky JT, Krahn LE. The effects of antidepressants on sleep: a review. Harv Rev Psychiatry. 2000;8:298–306. doi: 10.1080/hrp.8.6.298. [PubMed] [CrossRef] [Google Scholar]

8. Thase ME. Depression, sleep, and antidepressants. J Clin Psychiatry. 1998;59(Suppl 4):55–65.[PubMed] [Google Scholar]

9. Wichniak A, Wierzbicka A, Jernajczyk W. Sleep as a biomarker for depression. Int Rev Psychiatry. 2013;25:632–645. doi: 10.3109/09540261.2013.812067. [PubMed] [CrossRef] [Google Scholar]

10. Baglioni C, Nanovska S, Regen W, Spiegelhalder K, Feige B, Nissen C, Reynolds CF, Riemann D. Sleep and mental disorders: a meta-analysis of polysomnographic research. Psychol Bull. 2016;142:969–990. doi: 10.1037/bul0000053. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

11. Abad VC, Guilleminault C. Sleep and psychiatry. Dialogues Clin Neurosci. 2005;7:291–303.[PMC free article] [PubMed] [Google Scholar]

12. Kaplan KA, Harvey AG. Hypersomnia across mood disorders: a review and synthesis. Sleep Med Rev. 2009;13:275–285. doi: 10.1016/j.smrv.2008.09.001. [PubMed] [CrossRef] [Google Scholar]

13. McClintock SM, Husain MM, Wisniewski SR, Nierenberg AA, Stewart JW, Trivedi MH, Cook I, Morris D, Warden D, Rush AJ. Residual symptoms in depressed outpatients who respond by 50% but do not remit to antidepressant medication. J Clin Psychopharmacol. 2011;31:180–186. doi: 10.1097/JCP.0b013e31820ebd2c. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

14. Kay DB, Dombrovski AY, Buysse DJ, Reynolds CF, Begley A, Szanto K. Insomnia is associated with suicide attempt in middle-aged and older adults with depression. Int Psychogeriatr. 2016;28:613–619. doi: 10.1017/S104161021500174X. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

15. Schmid DA, Wichniak A, Uhr M, Ising M, Brunner H, Held K, Weikel JC, Sonntag A, Steiger A. Changes of sleep architecture, spectral composition of sleep EEG, the nocturnal secretion of cortisol, ACTH, GH, prolactin, melatonin, ghrelin, and leptin, and the DEX-CRH test in depressed patients during treatment with mirtazapine. Neuropsychopharmacology. 2006;31:832–844. doi: 10.1038/sj.npp.1300923.[PubMed] [CrossRef] [Google Scholar]

16. Thompson C. Onset of action of antidepressants: results of different analyses. Hum Psychopharmacol. 2002;17(Suppl 1):S27–S32. doi: 10.1002/hup.386. [PubMed] [CrossRef] [Google Scholar]

17. Doghramji K, Jangro WC. Adverse effects of psychotropic medications on sleep. Psychiatr Clin North Am. 2016;39:487–502. doi: 10.1016/j.psc.2016.04.009. [PubMed] [CrossRef] [Google Scholar]

18. Fava M. Daytime sleepiness and insomnia as correlates of depression. J Clin Psychiatry. 2004;65(Suppl 16):27–32. [PubMed] [Google Scholar]

19. Jindal RD, Thase ME. Treatment of insomnia associated with clinical depression. Sleep Med Rev. 2004;8:19–30. doi: 10.1016/S1087-0792(03)00025-X. [PubMed] [CrossRef] [Google Scholar]

20. Hashimoto T, Shiina A, Hasegawa T, Kimura H, Oda Y, Niitsu T, Ishikawa M, Tachibana M, Muneoka K, Matsuki S, Nakazato M, Iyo M. Effect of mirtazapine versus selective serotonin reuptake inhibitors on benzodiazepine use in patients with major depressive disorder: a pragmatic, multicenter, open-label, randomized, active-controlled, 24-week trial. Ann General Psychiatry. 2016;15:27. doi: 10.1186/s12991-016-0115-1. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

21. Jindal RD, Friedman ES, Berman SR, Fasiczka AL, Howland RH, Thase ME. Effects of sertraline on sleep architecture in patients with depression. J Clin Psychopharmacol. 2003;23:540–548. doi: 10.1097/01.jcp.0000095345.32154.9a. [PubMed] [CrossRef] [Google Scholar]

22. Furukawa TA, Streiner DL, Young LT. Antidepressant and benzodiazepine for major depression. Cochrane Database Syst Rev. 2002; CD001026. [PubMed]

23. Serretti A, Mandelli L. Antidepressants and body weight: a comprehensive review and meta-analysis. J Clin Psychiatry. 2010;71:1259–1272. doi: 10.4088/JCP.09r05346blu. [PubMed] [CrossRef] [Google Scholar]

24. Hickie IB, Rogers NL. Novel melatonin-based therapies: potential advances in the treatment of major depression. Lancet. 2011;378:621–631. doi: 10.1016/S0140-6736(11)60095-0. [PubMed] [CrossRef] [Google Scholar]

25. Quera-Salva MA, Hajak G, Philip P, Montplaisir J, Keufer-Le GS, Laredo J, Guilleminault C. Comparison of agomelatine and escitalopram on nighttime sleep and daytime condition and efficacy in major depressive disorder patients. Int Clin Psychopharmacol. 2011;26:252–262. doi: 10.1097/YIC.0b013e328349b117. [PubMed] [CrossRef] [Google Scholar]

26. Quera-Salva MA, Lemoine P, Guilleminault C. Impact of the novel antidepressant agomelatine on disturbed sleep-wake cycles in depressed patients. Hum Psychopharmacol. 2010;25:222–229. doi: 10.1002/hup.1112. [PubMed] [CrossRef] [Google Scholar]

27. Sowa-Kucma M, Panczyszyn-Trzewik P, Misztak P, Jaeschke RR, Sendek K, Styczen K, Datka W, Koperny M. Vortioxetine: a review of the pharmacology and clinical profile of the novel antidepressant. Pharmacol Rep. 2017;69:595–601. doi: 10.1016/j.pharep.2017.01.030. [PubMed] [CrossRef] [Google Scholar]

28. Wilson S, Hojer AM, Buchberg J, Areberg J, Nutt DJ. Differentiated effects of the multimodal antidepressant vortioxetine on sleep architecture: part 1, a pharmacokinetic/pharmacodynamic comparison with paroxetine in healthy men. J Psychopharmacol. 2015;29:1085–1091. doi: 10.1177/0269881115599387. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

29. Rottach KG, Schaner BM, Kirch MH, Zivotofsky AZ, Teufel LM, Gallwitz T, Messer T. Restless legs syndrome as side effect of second generation antidepressants. J Psychiatr Res. 2008;43:70–75. doi: 10.1016/j.jpsychires.2008.02.006. [PubMed] [CrossRef] [Google Scholar]

30. Narowska D, Bozek M, Krysiak K, Antczak J, Holka-Pokorska J, Jernajczyk W, Wichniak A. Frequent difficulties in the treatment of restless legs syndrome—case report and literature review. Psychiatr Pol. 2015;49:921–930. doi: 10.12740/PP/35395. [PubMed] [CrossRef] [Google Scholar]

31. Buschkamp JA, Frohn C, Juckel G. Mirtazapine induces nightmares in depressed patients. Pharmacopsychiatry. 2017. [PubMed]

32. Szaulińska K, Pływaczewski R, Sikorska O, Holka-Pokorska J, Wierzbicka A, Wichniak A, Sliwinski P. Obstructive sleep apnea in severe mental disorders. Psychiatr Pol. 2015;49:883–895. doi: 10.12740/PP/32566. [PubMed] [CrossRef] [Google Scholar]

33. Wittchen HU, Jacobi F, Rehm J, Gustavsson A, Svensson M, Jonsson B, Olesen J, Allgulander C, Alonso J, Faravelli C, Fratiglioni L, Jennum P, Lieb R, Maercker A, van OJ PM, Salvador-Carulla L, Simon R, Steinhausen HC. The size and burden of mental disorders and other disorders of the brain in Europe 2010. Eur Neuropsychopharmacol. 2011;21:655–679. doi: 10.1016/j.euroneuro.2011.07.018.[PubMed] [CrossRef] [Google Scholar]

34. Ohayon MM. Epidemiology of insomnia: what we know and what we still need to learn. Sleep Med Rev. 2002;6:97–111. doi: 10.1053/smrv.2002.0186. [PubMed] [CrossRef] [Google Scholar]

35. Nowicki Z, Grabowski K, Cubala WJ, Nowicka-Sauer K, Zdrojewski T, Rutkowski M, Bandosz P. Prevalence of self-reported insomnia in general population of Poland. Psychiatr Pol. 2016;50:165–173. doi: 10.12740/PP/58771. [PubMed] [CrossRef] [Google Scholar]

36. Qaseem A, Kansagara D, Forciea MA, Cooke M, Denberg TD. Management of chronic insomnia disorder in adults: a clinical practice guideline from the American College of Physicians. Ann Intern Med. 2016;165:125–133. doi: 10.7326/M15-2175. [PubMed] [CrossRef] [Google Scholar]

37. Sateia MJ, Buysse DJ, Krystal AD, Neubauer DN, Heald JL. Clinical practice guideline for the pharmacologic treatment of chronic insomnia in adults: an American Academy of sleep medicine clinical practice guideline. J Clin Sleep Med. 2017;13:307–349. doi: 10.5664/jcsm.6470. [PMC free article][PubMed] [CrossRef] [Google Scholar]

38. Bonnet MH, Arand DL. Hyperarousal and insomnia: state of the science. Sleep Med Rev. 2010;14:9–15. doi: 10.1016/j.smrv.2009.05.002. [PubMed] [CrossRef] [Google Scholar]

39. Riemann D, Spiegelhalder K, Feige B, Voderholzer U, Berger M, Perlis M, Nissen C. The hyperarousal model of insomnia: a review of the concept and its evidence. Sleep Med Rev. 2010;14:19–31. doi: 10.1016/j.smrv.2009.04.002. [PubMed] [CrossRef] [Google Scholar]

40. Mendelson WB, Roth T, Cassella J, Roehrs T, Walsh JK, Woods JH, Buysse DJ, Meyer RE. The treatment of chronic insomnia: drug indications, chronic use and abuse liability. Summary of a 2001 new clinical drug evaluation unit meeting symposium. Sleep Med Rev. 2004;8:7–17. doi: 10.1016/S1087-0792(03)00042-X. [PubMed] [CrossRef] [Google Scholar]

41. Wiegand MH. Antidepressants for the treatment of insomnia : a suitable approach? Drugs. 2008;68:2411–2417. doi: 10.2165/0003495-200868170-00001. [PubMed] [CrossRef] [Google Scholar]

42. Wilson SJ, Nutt DJ, Alford C, Argyropoulos SV, Baldwin DS, Bateson A, Britton T, Crowe C, Dijk DJ, Espie C, Gringras P, Hajak G, Idzikowski C, Krystal A, Nash JR, Selsick H, Sharpley AL, Wade AG. British Association for Psychopharmacology consensus statement on evidence-based treatment of insomnia, parasomnias and circadian rhythm disorders. J Psychopharmacol. 2010;24:1577–1601. doi: 10.1177/0269881110379307. [PubMed] [CrossRef] [Google Scholar]

43. Liu Y, Xu X, Dong M, Jia S, Wei Y. Treatment of insomnia with tricyclic antidepressants: a meta-analysis of polysomnographic randomized controlled trials. Sleep Med. 2017;34:126–133. doi: 10.1016/j.sleep.2017.03.007. [PubMed] [CrossRef] [Google Scholar]

44. Thompson W, Quay TA, Rojas-Fernandez C, Farrell B, Bjerre LM. Atypical antipsychotics for insomnia: a systematic review. Sleep Med. 2016;22:13–17. doi: 10.1016/j.sleep.2016.04.003. [PubMed] [CrossRef] [Google Scholar]

45. Wichniak A, Jarkiewicz M, Okruszek L, Wierzbicka A, Holka-Pokorska J, Rybakowski JK. Low risk for switch to mania during treatment with sleep promoting antidepressants. Pharmacopsychiatry. 2015;48:83–88. doi: 10.1055/s-0034-1396802. [PubMed] [CrossRef] [Google Scholar]

46. Jernajczyk W, Sobańska A, Wierzbicka A, Wichniak A, Szatkowska E. The therapy of primary insomnia with citalopram. J Sleep Res. 2006;15(Suppl. 1):154. [Google Scholar]

Introduction

The prevalence of depression in community settings has been estimated to be between 0.4% and 2.5% in children and 0.4% and 8.3% in adolescents [1]. However, in a more recent community study of children without depression who were initially assessed between the ages of 9 and 13 years, more than 7% of boys and almost 12% of girls developed a depressive disorder by the age of 16 [2].

While the diagnosis of major depressive disorder (MDD) in younger patients generally follows the criteria set forth in the Diagnostic and Statistical Manual of Mental Disorders, 4th edition, text revision (DSM-IV-TR) [3], the treatment of pediatric depression presents many challenges. Not only do children with MDD have multiple co-morbid disorders [4,5], psychosocial and academic problems, and are at increased risk for suicide attempts, self-harm, and substance abuse [1,6-10], treatment options are often limited and ineffective, poorly tolerated, and generally present long delays in delivering a therapeutic benefit [11-14].

One of the few antidepressants approved for use in children is the selective serotonin reuptake inhibitor (SSRI) fluoxetine [15]. The first study that demonstrated the positive effects of using fluoxetine to treat depression in child and adolescent patients was published in 1997 [16], following a small trial in which no difference was observed between fluoxetine treatment and placebo [17]. Overall, five clinical trials have been conducted which show the positive effects of using fluoxetine to treat pediatric depression [16,18-21]. In addition, the improvement response rate on the Clinical Global Impressions (CGI) for antidepressant use was found to be between 52% and 61% for fluoxetine patients versus 33% to 37% for patients treated with placebo [22]. The CGI measures whether or not depressive symptoms have improved after treatment.

Despite being one of the most popular treatments for pediatric patients, in 2004 the use of prescription medication such as fluoxetine as well as other antidepressant medications declined by approximately 20% in the United States [23]. This shift in prescription patterns is likely due to warnings issued by regulatory agencies, initially in the United Kingdom [24] and later in the United States [25], against the use of SSRIs to treat depression in pediatric populations due to the possible link between antidepressant usage and an increased incidence of suicidal ideations or attempts. Subsequently, there is a compelling need for better understanding of the pathophysiology of MDD as well as the development of novel treatment methods that can be used to improve the current clinical management of pediatric depression.

Nutrients like vitamin C (ascorbic acid) have become of interest in adjuvant therapy settings for the management of depressive symptoms due to the fact that psychological abnormalities are among the characteristics of vitamin C deficiency [26-29]. A recent population-based survey revealed that 60% of the patients in the acute medical wards of a Montreal teaching hospital were vitamin C deficient, while this deficiency was only detected in 16% of people attending the hospital’s outpatient center [30]. There is also preliminary evidence that the administration of vitamin C may be able to reduce the severity of MDD in both children [31] and adults [32], as well as improve mood in healthy individuals [33-35]. In addition, a recent study reported a 35% reduction in average mood disturbance in hospitalized patients following treatment with vitamin C (1000 mg/day) [36]. In one particular study that investigated mood, patients who were acutely hospitalized were either treated with vitamin C or vitamin D as a deficiency in both of these vitamins has been associated with psychological abnormalities [32]. The results showed that only vitamin C led to an improved mood. More specifically, treating the vitamin C deficiency led to a decrease in mood disturbance while vitamin D supplementation had no effect on mood. Similar findings were observed in non-critically ill hospitalized patients who were treated with vitamin C for hypovitaminosis C [36]. Moreover, an animal study showed that the co-administration of vitamin C was found to potentiate the action of subeffective doses of fluoxetine (1 mg/kg) [37]. This synergistic antidepressant effect of vitamin C and fluoxetine suggests that this vitamin could be helpful in improving conventional pharmacotherapy for pediatric MDD and potentially reduce side effects.

This study would be the first to examine the efficacy of vitamin C as an adjunct to SSRIs in the treatment of pediatric depression. In addition, the low potential toxicity, inexpensiveness, and over-the-counter availability, we sought to investigate whether oral supplementation of vitamin C would improve clinical depressive symptoms. Therefore, the present study was designed to measure the effect of vitamin C on the Children’s Depression Rating Scale (CDRS), the Children’s Depression Inventory (CDI), and the CGI scores in pediatric patients with depression taking fluoxetine.

Go to:MethodsTrial design

The study was a prospective, double-blind, placebo-controlled, six-month clinical trial. Two parallel groups of outpatient pediatric patients with depression in Mansoura University Hospital, Egypt participated in the study from October 2009 to September 2011. The study was approved by the institution’s review board.

Participants

The authors screened pediatric patients (less than 18 years of age) who were referred to the outpatient psychiatry clinic for MDD based on a semi-structured interview and DSM-IV-TR criteria [3]. Exclusion criteria included clinically significant organic or neurological disorder, psychotic disorder or depression with psychotic features, a history of substance abuse or dependence, or prior use of psychotropic medication. Young patients with bipolar disorder may experience adverse psychological effects such as mania and hypomania due to antidepressants and were therefore, excluded from the study. It has been shown that patients who are young in age at the onset of bipolar disorder demonstrate an illness progression that is characterized by high rates of switching into mania or hypomania in response to antidepressant treatment [38]. Among the 32 patients screened during this period, five were excluded (two had depression with psychotic features, two had a history of hypomania, and one had a substance abuse disorder). The remaining 27 patients agreed to participate in this study after informed consent from at least one parent was obtained. The patients did not receive any other treatment such as cognitive behavioral therapy during the trial period. This trial was performed in accordance with the Declaration of Helsinki and subsequent revisions [39]. Written consent was obtained from each patient’s parent or guardian before entering the study.

Intervention

Vitamin C and placebo were formulated into capsules by the Mansoura University Hospital. The patients were randomly allocated to either the treatment or control group using a computer-generated list of random numbers. Fourteen patients were assigned to the treatment group and were given fluoxetine (10–20 mg/day) plus vitamin C (1000 mg/day; 500 mg BID). Thirteen patients were assigned to the control group and were given fluoxetine (10–20 mg/day) plus placebo. Patients less than eight years of age received fluoxetine (10 mg/day), whereas patients eight years of age or older were given 10 mg/day of fluoxetine for one week and 20 mg/day all subsequent weeks as per the prescribing information [40]. There are several published studies which support the administration of 20 mg/day of fluoxetine for children at least eight years of age [16,18,20,21], and it is within FDA indication. The use of fluoxetine for children under the age of eight is off-label. A dose of 1000 mg/day of vitamin C (500 mg BID) was chosen based on human studies suggest that psychiatric patients generally require higher levels of vitamin C to improve symptoms than the doses that are recommended for healthy individuals [32,41]. The recommended dose of vitamin C for healthy individuals is 70 mg/day, while a dose of 1000 mg/day needs to be consumed before symptoms begin to improve in psychiatric patients [41].

Patients in the placebo group received two identical capsules (morning and evening). No other psychotropic medications were prescribed. Three subjects were removed from the trial due to noncompliance (two patients from the vitamin C group and one from the placebo group). Patients were assessed using CDRS, an Arabic version of CDI, and CGI at the baseline as well as 3 and 6 months after the start of treatment. The scores for the CDRS were based on parent ratings, CDI on children ratings, and CGI on clinician ratings. Examinations of patients during the treatment period were performed by a psychiatrist trained in the use of these instruments.

Instruments

The Children’s Depression Rating Scale (CDRS) is a 16-item measure used to determine the severity of depression in children and adolescents aged 6 to 12 [42]. The CDRS is derived from the Hamilton Rating Scale for Depression (HAM-D) [43] and is based on parent, child, and schoolteacher interviews. CDRS scores show good concordance with research diagnostic criteria for depression [44] and correlate highly with other interview and self-report measures of depression severity [45].

The Children’s Depression Inventory (CDI) is a 27-item, self-rated, symptom-oriented scale suitable for children and adolescents aged 7 to 17 [46]. The CDI is sensitive to changes in depressive symptoms over time, making it a useful index for the severity of MDD. The CDI is reported to have high internal consistency and test-retest reliability [47]. The CDI assessment utilized in this study was based on the previously developed instrument [46] and was translated and normalized for Arab children by Gharib (1988) [48]. Reliability and validity data for the Arabic version are comparable to those provided for the original instrument.

The Clinical Global Impressions Scale (CGI) is a 3-item, observer-rated scale that measures illness severity, global improvement or change, and therapeutic response [49]. The CGI is rated on a 7-point scale with each component being rated separately; the instrument does not yield a global score. Over the past 30 years, the CGI has been shown to correlate well with standard, well-known research drug efficacy scales, including the Hamilton Rating Scale for Depression, the Brief Psychiatric Rating Scale, and the Scale for the Assessment of Negative Symptoms across a wide range of psychiatric indications [50].

Statistical analysis

Student’s t-tests and chi-squared tests were used to evaluate possible differences in baseline demographics. Two-way repeated measures analysis of variance (ANOVA) were used to assess the effects of treatment (treatment versus placebo), time (months of visit), and an interaction between the treatment and time. Significant differences in the mean scores for each visit were assessed through unpaired Student’s t-tests. Quantitative variables were tested for normal distributions by the Kolmogorov-Smirnov test. The variables were presented as means ± standard deviations (SD). Statistical significance was set at the 5% level. SPSS for Windows version 13 was used for the statistical analysis of the data obtained from the study.

Go to:ResultsDemographic characteristics and attrition

Thirty-two patients were initially examined, but five patients did not satisfy the inclusion criteria. Therefore 27 patients enrolled in the study; 14 assigned to the vitamin C group and 13 to the placebo group. Twenty-four patients aged between 7 and 14 completed the six-month trial. Two patients from the vitamin C group and one patient from the placebo group were removed from the trial due to noncompliance (Figure 1).

Clinical complications and adverse effects

No major adverse effects were observed.

Go to:Discussion

These results show that orally administered vitamin C as an adjunct to fluoxetine treatment leads to significantly greater decreases in depressive symptoms in comparison to fluoxetine treatment alone. This was demonstrated by the decrease in depressive symptoms, which was observed in the improved CDRS and CDI scores. A significant effect was not observed for the CGI, but this may be related to the response items for this instrument. For instance, symptoms were scored according to whether “much improvement” or “very much improvement” was observed [49]. Although there may have been a slight increase in CGI scores, response items such as these may have made it difficult to detect a significant improvement of symptoms. The differences between the scores may have also been related to the individuals who supplied the ratings for each instrument. More specifically, the scores for the CDRS were based on parent ratings, the scores for the CDI were based on children ratings, and the scores for the CGI were based on clinician ratings. The scores from the CGI were computed based on clinical criteria such as that which is listed in the DSM-IV-TR as well as semi-structured interviews. Therefore, the clinician’s rating and score interpretations adhered to strict guidelines and training, whereas the ratings from parents and children may have been more subjective leading to significantly different scores. Nonetheless, these preliminary findings, including the results of ANOVA suggest that vitamin C may be an effective adjuvant agent for the treatment of depression in pediatric patients. Furthermore, the results support the notion that vitamin C has antidepressant-like properties and are in accordance with previous animal research that demonstrated vitamin C’s ability to potentiate the action of conventional antidepressants [37].

Despite the lack of research investigating the effects of vitamin C in pediatric patients with MDD, previous studies have suggested that vitamin C improves clinical symptoms in other psychiatric disorders [51-53], and that vitamin C supplementation can be used to positively modulate mood [33-35]. Furthermore, Khanzode et al., (2003) showed that plasma levels of vitamin C were decreased in depressive patients [54]. In a more recent study, Chang et al., (2007) described a case in which a patient with depression developed scurvy, suggesting that reduced plasma levels of vitamin C due to inadequate vitamin C intake could be associated with the pathophysiology of depression [29]. Other studies have also shown that depressive symptoms are associated with scurvy [55-57].

While the exact role of vitamin C in the etiology of MDD is not well understood, a growing body of evidence suggests that oxidative stress, characterized by an accumulation of free radicals due to an organism’s inhibited antioxidant capacity, may play a primary or secondary role in the pathogenesis of neurological and psychiatric diseases like MDD [58,59]. The brain is much more vulnerable to oxidative free radicals than other tissues since it utilizes 20% of the oxygen consumed by the body, contains large amounts of polyunsaturated fatty acids and iron, and typically has low concentrations of antioxidant enzymes [60]. Previous studies have shown that MDD may be accompanied by disturbances in the balance between pro- and anti-oxidative processes, demonstrated by decreased blood plasma levels of the antioxidants enzymes superoxide dismutase, catalase, and glutathione peroxidase and an increased level of lipid peroxidation by-products in patients with depression versus healthy controls [54,61,62].

While antidepressant drugs may affect the oxidative or antioxidative systems [54], partly due to their effects on the immune [63] and P450 systems [64], adjunctive therapy with vitamin C may provide additional protection as it is the brain’s most abundant antioxidant and plays an important role in preventing free radical-induced damage [65,66]. In addition to its neuroprotective properties, vitamin C has also been identified as a neuromodulator in the brain, modulating both dopamine- and glutamate-mediated neurotransmission [67-69]. As there is a considerable amount of pharmacological evidence demonstrating the efficacy of antidepressants with dopaminergic effects in the treatment of depression [70], vitamin C’s complex interaction with the dopaminergic system may be another potential mechanism of action. However this effect appears to be dose-dependent. Wambebe and Sokomba (1986) showed that administering 50–200 mg/kg of vitamin C to rats enhanced dopamine-mediated behavioral effects [71], while higher dosages have been shown to antagonize such effects [68].

There are a number of other potential biological substrates that underlie vitamin C’s effects on depression and mood. For example, Binfaré et al. [37] identified the involvement of 5-HT1A receptors in the antidepressant-like effect of vitamin C. Additionally, adjuvant administration of vitamin C may also prove useful in decreasing the risk of suicidal thoughts and behaviors linked to antidepressant therapy in pediatric patients [72]. Meta-analyses of placebo-controlled studies have indicated that antidepressants may cause a significant, although small and short-term, risk of self-harm or suicide-related events in children and adolescents with MDD, no completed suicides were reported in any trial included in the analysis [73,74]. Li et al. [75] reported that a history of attempted suicide was shown to be associated with a low level of antioxidant vitamins and carotenoids. Therefore, increasing plasma vitamin C levels in children and adolescents who are being treated with antidepressants may help mitigate some of this risk. However, as suicidal thoughts and behaviors were not measured in the present study, future clinical research is needed to test this hypothesis.

Limitations

The present study has several limitations, one being its small sample size. While pilot clinical trials can play an important role in the early assessment of novel treatment methods when they are well designed and evaluated [76], further studies with larger sample sizes are needed to substantiate the results of this study. Secondly, drawing conclusions from a combined sample of children and adolescents with regard to the response to medication should be done with precaution as there is reason to believe that children respond differently than adolescents to antidepressants [77]. Also, due to the low potential for adverse drug reactions related to vitamin C in this study, the effect of doses higher than 1000 mg/day should be considered in future studies.

Measuring plasma vitamin C levels pre- and post-treatment may also be of interest, but although these levels were not measured, previous studies have demonstrated the association between hypovitaminosis C (vitamin C deficiency) and psychological abnormalities and this deficiency is highly prevalent in acutely hospitalized patients [32,36]. Furthermore, the increase in plasma and mononuclear leukocyte vitamin C from subnormal to normal concentrations after the administration of vitamin C administration implicate that the metabolic properties of hypovitaminosis C are consistent with deficiency as opposed to different mechanisms such as tissue redistribution [36]. These findings also indicate that patients with depression, such as those who participated in this study, may experience vitamin C deficiency and that the decrease in depressive symptoms that was observed may be directly attributed to the synergistic antidepressant effect of vitamin C and fluoxetine. Future studies that involve measuring plasma vitamin C levels may further support these findings. Finally, in the current study, participants were only treated and assessed for a short period of time (six months). The most striking effects were observed for the interaction between treatment and time and this finding suggests that longer trials are needed to better assess the efficacy of vitamin C as an adjunct to fluoxetine therapy.

Go to:Conclusion

Treatment with 1000 mg/day of vitamin C potentiated the efficacy of fluoxetine in pediatric patients being treated for MDD. Furthermore, vitamin C was shown to be a particularly attractive therapeutic adjuvant due to the absence of substantial side effects and its inexpensive cost. The observed improvements in CDRS and CDI scores also imply that this type of treatment effectively increases blood plasma levels of vitamin C as it has been shown that ascorbic acid deficiency is associated with psychological abnormalities [26-29]. Future, large-scale clinical trials are warranted to evaluate the therapeutic efficacy of vitamin C for the treatment of depression in pediatric patients as well as its effectiveness as an adjuvant treatment to antidepressants.

Abstract

Accumulating evidence demonstrates that the gut microbiota affects brain function and behavior, including depressive behavior. Antidepressants are the main drugs used for treatment of depression. We hypothesized that antidepressant treatment could modify gut microbiota which can partially mediate their antidepressant effects. Mice were chronically treated with one of five antidepressants (fluoxetine, escitalopram, venlafaxine, duloxetine or desipramine), and gut microbiota was analyzed, using 16s rRNA gene sequencing. After characterization of differences in the microbiota, chosen bacterial species were supplemented to vehicle and antidepressant-treated mice, and depressive-like behavior was assessed to determine bacterial effects. RNA-seq analysis was performed to determine effects of bacterial treatment in the brain. Antidepressants reduced richness and increased beta diversity of gut bacteria, compared to controls. At the genus level, antidepressants reduced abundances of Ruminococcus, Adlercreutzia, and an unclassified Alphaproteobacteria. To examine implications of the dysregulated bacteria, we chose one of antidepressants (duloxetine) and investigated if its antidepressive effects can be attenuated by simultaneous treatment with Ruminococcus flavefaciens or Adlercreutzia equolifaciens. Supplementation with R. flavefaciens diminished duloxetine-induced decrease in depressive-like behavior, while A. equolifaciens had no such effect. R. flavefaciens treatment induced changes in cortical gene expression, up-regulating genes involved in mitochondrial oxidative phosphorylation, while down-regulating genes involved in neuronal plasticity. Our results demonstrate that various types of antidepressants alter gut microbiota composition, and further implicate a role for R. flavefaciens in alleviating depressive-like behavior. Moreover, R. flavefaciens affects gene networks in the brain, suggesting a mechanism for microbial regulation of antidepressant treatment efficiency.

Introduction

During the past decade, there has been an increase in understanding of how the gut microbiota affects various aspects of brain development and function, as well as behavior. For example, studies on germ free mice revealed that gut bacteria influence development of stress response, appropriate maturation and function of microglia, affect anxiety, social and depressive-like behaviors, along with alterations in gene expression and neurochemistry of different brain regions1,2,3,4,5. Regarding the etiology of depression, it has been shown that germ free mice exhibit less behavioral despair, along with higher brain serotonin levels, in comparison to their conventional counterparts2,6. The importance of gut bacteria in development of mood disorders was further confirmed by several recent studies showing that patients with depression had altered diversity and composition of gut microbiota, and these changes were causally related to depressive-like behavior in rodent models7,8.

Antidepressants are major drugs used for treatment of depression9,10. Some of the most effective antidepressants act as inhibitors of serotonin and/or norepinephrine reuptake which leads to increased synaptic concentrations of these neurotransmitters11,12,13,14. However, even though they are in use for >50 years, the precise molecular mechanisms of their therapeutic action are still not completely understood. Of particular importance, it is not clear what is the biological mechanism behind the variability of efficacy of antidepressants between different individuals. It is presumed that their therapeutic effects are achieved through slow onset auto-receptor down-regulation, and subsequent adaptation of downstream neural signaling pathways, including promotion of neural plasticity15,16,17,18. Besides this, although the antidepressants are considered to be efficient, still the relative risk reduction of relapse by the continuous treatments was estimated to be 50–60%19,20. All that point out further need for better understanding of antidepressant actions, along with searching for new treatments, or complementary ways to improve the efficiency of current antidepressant medication.

Considering the evidence for a role of microbiota in depressive behavior, we hypothesize that antidepressants also change gut microbiota composition, and through modulation of the microbiota, at least partly, exert their antidepressant effects. Indeed, some antidepressants were shown to have antimicrobial effects in vitro against several groups of microorganisms, and inhibit number of processes in microorganisms, such as slime production and bacterial motility21,22,23. On the other hand, serotonin and noradrenaline, which are found in high amounts in the gut, can promote growth and virulence in certain bacteria, acting as interkingdom signaling molecules24,25,26,27. Also, recently it was found that knockout of rat serotonin transporter disrupted gut bacteria homeostasis, augmentating early life stress effects as well28. Considering that there is considerable variation in the microbiome between individuals29, we may also consider that microbiome variation may be partly a mechanism for the variability of antidepressant efficacy among different individuals.

In addition to the effects of antidepressants on microbiota, microbiota may also effect depressive-like behaviors through modulation of neurotransmitters and other key molecules. Microbiota can produce neuroactive compounds, including neurotransmitters, that may influence host physiology and behavior30,31. In addition, host microbiota can influence the production of serotonin by enterochromaffin cells in the host gut32. This is interesting, considering that the gut is the main source of serotonin. This can provide a further mechanism to support the hypothesis that antidepressants may partially mediate their effects through regulation of microbiota.

To explore the aforementioned hypothesis, we treated BALB/c mice with one of five antidepressants, commonly used in clinical practice and different in their mode of action. The choice of BALB/c strain for the study was based on their natural characteristics of exhibiting higher depressive-like behavior33,34,35,36,37 and anxiety36,38,39 compared to other strains. Furthermore, they were shown to be responsive to chronic antidepressant treatments that reduce their immobility in test of behavioral despair37,40,41. All these, make BALB/c strain as a suitable model to study antidepressant responses relevant for depressive disorder. Indeed, our results demonstrated that antidepressants change diversity and composition of gut bacterial communities and one of the identified bacterial species, affected by their treatment, is able to mediate alleviation of depressive behavior.

Methods and materialsAnimals

Male BALB/c OlaHsd mice, purchased from Harlan (Israel), were used in the study. The mice were housed under reverse 12 h light/dark cycle conditions, with water and food available ad libitum. All animals were group housed with 3–5 animals per cage. When studying antidepressant effects on microbiota, animals were divided into 3 cages, in order to minimize cage effect. Animals that developed illnesses during the experiments were excluded from the study. The experiments started when mice were 8–10 weeks of age. All experimental protocols were approved by the Animal Care and Use Committee of Bar-Ilan University. Animals were randomly assigned to each experimental group.

Antidepressant and bacterial treatments

All five antidepressants used in the study were purchased from Sigma-Aldrich. Antidepressants were diluted in PBS, in the following doses: fluoxetine 10 mg/kg, escitalopram 10 mg/kg, venlafaxine 10 mg/kg, duloxetine 10 mg/kg and desipramine 20 mg/kg, and delivered by daily i.p. in volume of 8 ml/kg (Fig. 1a). In all experiments, mice were treated with antidepressants for 21 days before stool collection or beginning of behavioral tests. The effective doses were selected according to the available literature regarding their therapeutic concentrations in mice40,41,42,43,44. The control group received corresponding volume of PBS. In the initial experiment (Figs. 1 and 2), the number of animals used per experimental group were n = 9 (control), n = 11 (flu), n = 12 (esc), n = 12 (ven), n = 11 (dul), n = 12 (des).

Behavioral testing

Depressive-like behavior of mice was assessed after 21st day of treatment, using tail suspension test (TST), forced swim test (FST) and sucrose preference test (SPT) (for details see Supplement). Locomotor activity was assessed in a separate group of animals, after 21stday of treatment as well, using open field test and rotarod (for details see Supplement). All tests were done during the dark phase, between 10 a.m. and 3 p.m., and animals were acclimated to the behavior room for 1 h before testing (except for sucrose preference test that was performed in home cages). Between each of behavioral tests, mice had at least one day of rest. On the test day, the i.p. injections of an antidepressant or pbs was done 1 h before testing.

Stool collection, DNA extraction and sequencing of 16 s rRNA gene

Mice fecal samples were collected under aseptic conditions, 1 h after i.p. injection of antidepressants or PBS, and stored at −80 °C until further analyses. DNA was isolated using the PowerSoil DNA isolation kit (MoBio Laboratories) according to the manufacturer’s instructions following an initial 2 minute beadbeating step (BioSpec). (For details about 16s rRNA gene amplification see Supplement.)

Bioinformatic analyses of 16S rRNA gene sequences

Obtained 16s rRNA gene sequencing data were analyzed by QIIME 1 pipeline47 (for details see Supplement). Alpha diversity (within community diversity) was estimated by Faith’s phylogenetic diversity (PD)48 and Chao149, as a measure of community richness, and by Gini coefficient, as a measure of community evenness50. Beta diversity (between communities diversity) was calculated using unweighted and weighted UniFrac distances51. The diversity parameters were compared between groups using a nonparametric t-test with Monte Carlo permutations (999) to calculate p values, and Benjamini and Hochberg FDR method was used afterwards to correct p values for multiple comparisons between different pairs of groups.

Differences in relative abundances of bacterial taxa between groups were identified using the linear discriminant analysis (LDA) effect size (LEfSe) method (version 1.0)52 as well as permutational multivariate analysis of variance (PERMANOVA). LEfSe uses the nonparametric Kruskal–Wallis rank-sum test to detect features which have significantly different abundances between groups. Then, it performs LDA to estimate the effect size of each feature (alpha significance level was set at 0.05 and an effect-size threshold was set at 2).

Bacterial detection by PCR

Bacterial abundances of the stool samples used in 16S rRNA gene sequencing were verified by quantitative real-time PCR (qRT-PCR). It was performed using Fast Start Universal SYBR Green Master (Rox) (Roche) and ViiA™7 Real-Time PCR System (Life Technologies). PCR consisted of 40 cycles, using melting temperature of 95 °C for ten seconds per cycle, and an annealing/elongation temperature of 60 °C or 57 °C, as appropriate, of thirty seconds per cycle. Relative quantification by ddCt method was used to verify bacterial abundances in the gut. The primer sequences used in the reactions are indicated in the Supplementary Table S1. In order to detect bacterial species in bacteria treated mice, DNA from stool samples were amplified by PCR as described above, and then ran on a 2% agarose gel to visualize PCR bands.

Brain tissue isolation and RNA extraction

Mice were sacrificed by rapid decapitation and brains were quickly removed. The medial prefrontal cortices (mPFC) were isolated using brain matrix and gauge 13 (from the slice between 3 mm and 1 mm anterior to bregma), and immediately frozen on dry ice. Total mPFC RNA was extracted from six samples per experimental group using RNeasy Mini Kit (Qiagen) according to the manufacture protocol. NanoDrop 1000 (Thermo Scientific) and Qubit were used to determine the purity and concentration of RNA, respectively. Bioanalyzer 2100 (Aligent Technologies) was used to verify RNA integrity number (RIN), and all samples displayed RIN greater than 7.90.

mRNA sequencing

From 100 ng of total RNA of each sample, mRNA enrichment was done by NEBNext Poly(A) mRNA Magnetic isolation Module (NEB # E7490), followed by preparation of RNA libraries using NEBNext Ultra RNA Prep kit (NEB #E7530), according to manufacturer’s protocols. Libraries’ concentrations were determined by Qubit, while quality and size distribution was analyzed using Bioanalyzer 2100. The sequencing was performed at the Technion Genome Center, Haifa with the Illumina HiSeq 2500. Fastq files are available at GEO under the accession number GSE129359.

Bioinformatic analyses of mRNA sequencing

The quality of the sequenced data, as well as read length distributions after trimming, was evaluated by FASTQCT (0.11.5.). The reads were mapped to the Mus musculus reference genome, GRCm38.p5, using the Tophat2 software. (For details regarding differential expression analyses see Supplement). The weighted gene correlation network analysis (WGCNA) R software package was applied to the entire set of normalized gene counts with the aim of identifying gene modules affected by the treatments53 (for details see Supplement). Enrichment analyses for the Gene ontology (GO) terms (molecular function, biological process and cellular component) were performed using online ToppGene Suite software. GO terms were considered to be significant when the Benjamini and Hochberg FDR adjusted p value was below 0.05. For protein-protein interaction (PPI) network analysis, the STRING database followed by Cytoscape (version 3.2.1) network construction was used (for details see Supplement).

Serotonin and noradrenalin brain levels

mPFC was extracted as previously described for RNA collection. After weighting, brain tissue was homogenized in 0.01 N HCl with 1 mM EDTA and 4 mM sodium metabisulfite on ice. Levels of serotonin and noradrenaline were measured using Serotonin Research ELISA and Noradrenaline Research ELISA kits (LDN, Nordhorn, DE), respectively, according to manufacturer’s protocols.

Statistical analyses

Statistical analyses were done using SPSS. When the assumptions of normality and homogeneity of variances were met, the data were analyzed by ANOVA. When these assumptions were violated, non-parametric tests were used. Namely, the effects of antidepressants and bacteria on behavior and neurotransmitter brain levels were analyzed by one-way or two-way ANOVA, as appropriate. Comparison between groups was performed by Dunnett or Tukey post hoc test, as appropriate. The effects of antidepressants on bacterial levels in gut were analyzed by Kruskal–Wallis test, followed by pairwise Mann-Whitney tests. Level of significance was set at p < 0.05.

ResultsAntidepressants affect gut microbiota composition

To investigate whether antidepressants may alter gut microbiota, we chose five different antidepressants common in clinical practice and different in their mode of action. We used two selective serotonin reuptake inhibitors (SSRIs) - fluoxetine and escitalopram, two serotonin norepinephrine reuptake inhibitors (SNRIs) - venlafaxine and duloxetine, and desipramine that acts as norepinephrine reuptake inhibitor. Since antidepressants require at least three weeks to show their therapeutic effects, we characterized the gut microbial community after 21 day of daily antidepressant treatment (Fig. 1a; OTU table with all detected bacteria can be found in Supplementary Table 1). Antidepressants were injected i.p. in order to provide a specific concentration of antidepressants directly to the gut. Treatment by drinking water would cause potential cofounders in the analysis of specific drug effects on microbiota, due to potential variability among experimental groups in drinking water volumes and differential kinetics of breakdown of antidepressants in drinking water. Nonetheless, one important limitation of our approach is that daily i.p. administration adds an element of stress to the experimental outline which can affect experimental findings.

Analyses of alpha diversity revealed that all antidepressants, except desipramine, reduced the richness of microbial communities (Fig. 1b; Supplemental Figure S1a, b), but did not affect their evenness (Fig. 1c). Beta diversity measures of gut microbiota were also affected by all studied antidepressants, but more pronounced effects were observed in unweighted UniFrac analyses (Fig. 1d–f) than in weighted UniFrac (Supplemental Figure S1c–e). Namely, beta diversity of fecal microbial communities from mice receiving antidepressants was higher than beta diversity of control samples (Fig. 1d, f). Further, microbial communities of control group were more similar to each other than when they were compared to samples of any of antidepressant treated groups (Fig. 1e, f).

In order to identify bacterial taxa which differed in relative abundances in antidepressant treated mice in comparison to controls, we analyzed the 16S rRNA gene sequencing results using linear discriminant analysis (LDA) effect size (LEfSe) algorithm. A comparison between control mice and all antidepressant groups together revealed that Ruminococcus, Adlercreutzia and an undefined genus in the order RF32, class Alphaproteobacteria were less abundant in antidepressant treated mice (Fig. 2a, b). When analyzing the effects of each individual antidepressant in comparison to control using pairwise comparisons, the same genera were shown to be less abundant in escitalopram, venlafaxine, duloxetine and desipramine groups, but not in fluoxetine group (Supplemental Figure S2). Further, in both the overall effects of antidepressants and in the pairwise analysis, the differences in genus Adlercreutzia contributed to observed decreased abundance of family Coriobacteriaceae, order Coriobacteriales, class Coriobacteriia, phylum Actinobacteria in antidepressant groups (Fig. 2a, b; Supplemental Figure S2). In addition to the LEfSe analysis, we further analyzed the 16S rRNA gene sequencing data with PERMANOVA (Supplementary Table 2). At the genus level, both Ruminococcus and Adlercreutzia were effected by antidepressant treatment (p < 0.05), although only Adlercreutzia remained significant after corrections for multiple comparisons. In direct pairwise analysis between control and each of the antidepressant groups, some antidepressants significantly decreased levels of Ruminococcus, although they were no longer significant after corrections for multiple comparisons. Therefore, the findings of decrease in Ruminococcus and Adlercreutzia were more significant in the LEfSe analysis, and we further explored these findings using qRT-PCR.

Validation of antidepressant-induced decrease in Ruminococcus and Adlercreutzia levels was done by qRT-PCR (undefined genus in order RF32 could not be analyzed by qRT-PCR because of the shortage of knowledge about its sequence). In order to determine which species of Ruminococcus was most affected, examination of the OTU table revealed that the abundance of the OTU assigned to Ruminococcus flavefaciens was altered by most of the antidepressant treatments (Kruskal–Wallis test H = 27.88, p < 0.001) (Fig. 2c). The qRT-PCR verified that R. flavefaciens was indeed less abundant in all antidepressant groups compared to controls (Kruskal–Wallis test H = 22.33, p < 0.001) (Fig. 2d). In the genus Adlercreutzia, Adlercreutzia equolifaciens is the only characterized species of this genus. The OTU that had the highest (98%) similarity to Adlercreutzia equolifaciens and was the most abundant in mice gut, was also decreased by most of antidepressants (Kruskal–Wallis test H = 16.06, p < 0.01) (Fig. 2e). Likewise, the qRT-PCR results showed that all antidepressants except desipramine, reduced levels of A. equolifaciens in the mice gut (Kruskal–Wallis test H = 26.60, p < 0.001) (Fig. 2f).

R. flavefaciens but not A. equolifaciens reduces antidepressive effects of duloxetine

In our next group of experiments, we wanted to explore the hypothesis that R. flavefaciens or A. equolifacien may mediate the effects of antidepressant treatment on depressive-like behavior. With that aim, we treated BALB/c mice with antidepressant, bacteria (R. flavefaciens or A. equolifaciens), or both, and then performed behavioral testing (Fig. 3a). For this group of experiments, we chose antidepressant duloxetine because it decreased both R. flavefaciens and A. equolifaciens, as well as it showed the biggest effect in TST in BALB/c mice (Supplemental Figure S3). Presence of gavaged bacteria in the gut was confirmed by detecting R. flavefaciens or A. equolifaciens in stool samples of treated mice (Supplemental Figure S4).

First, we examined effects of R. flavefaciens in the TST. As expected, duloxetine induced a significant effect (two-way ANOVA: Fdul = 30.70, p < 0.001), and duloxetine treated animals displayed significantly less immobility, compared to control animals (Fig. 3b). Interestingly, there was also a significant effect of R. flavefaciens (two-way ANOVA: FRum = 5.90, p < 0.05), and animals treated with both R. flavefaciens and duloxetine displayed significantly more immobility compared to those treated with duloxetine alone (Fig. 3b). Therefore, R. flavefaciens treatment was able attenuate the duloxetine effect in the TST. In the FST, significant effects were also obtained by both duloxetine (two-way ANOVA: Fdul = 5.70, p < 0.05) and R. flavefaciens (two-way ANOVA: FRum = 8.74, p < 0.01) (Fig. 3c). Specifically, animals concomitantly treated with R. flavefaciens and duloxetine were more immobile than animals treated with duloxetine alone, and displayed behavior comparable to control animals (Fig. 3c). Therefore, R. flavefaciens treatment was able to abolish the effect of duloxetine treatment in the FST. The effects of R. flavefaciens in TST and FST were not confounded by locomotor deficits (Supplemental Figure S5a, b). Together, these results demonstrate that R. flavefaciens can attenuate duloxetine effects in behavior despair paradigms.

Further, we examined the effect of R. flavefaciens on anhedonia, by SPT. Both the antidepressant and the bacteria showed significant effects (two-way ANOVA: Fdul = 6.42, p < 0.05; FRum = 13.73, p = 0.001) (Fig. 3d). Namely, duloxetine increased mice preference for 2% sucrose, while it did not change the sucrose preference in the group receiving R. flavefacienstogether with the drug (Fig. 3d). Also, sucrose preference was higher in group receiving duloxetine than in group receiving both duloxetine and R. flavefaciens (Fig. 3d). Moreover, the interaction of duloxetine and the bacteria treatment was significant in SPT (two-way ANOVA: Fdul*Rum = 8.34, p < 0.01) (Fig. 3d). In conclusion, R. flavefaciens supplementation reduced or abolished antidepressive properties of duloxetine.

In addition, the effect of R. flavefaciens on general gastrointestinal health was evaluated by number of fecal pellets in open field test (Supplemental Figure S5c). R. flavefaciensabolished constipation induced by duloxetine. The decreased defecation in a new environment induced by duloxetine can be attributed to its anxiolytic effect as well, yet we note that there were no changes in time spent in center of the open field arena by any of treatments, which is also taken as sign of anxiety behavior54 (data not shown).

Additionally, we determined whether antidepressant effects can be modulated by A. equolifaciens. As in the previous experiment, duloxetine had significant effects on immobility time in TST (two-way ANOVA: Fdul = 55.63, p < 0.001) and to lesser degree in FST (two-way ANOVA: Fdul = 6.52, p < 0.05) (Fig. 3e, f). However, when the drug was given together with A. equolifaciens, it still exhibited its antidepressant properties (Fig. 3e, f). Therefore, contrary to R. flavefaciens, A. equolifaciens did not abolish antidepressive effects of duloxetine.

R. flavefaciens up-regulates mitochondrial genes while down-regulates neural genes in mPFC

In order to reveal mechanisms by which R. flavefaciens exerts its effects on the brain, we did whole transcriptome analyses on RNA extracted from the mPFC from each of the four experimental groups: control, duloxetine treated, R. flavefaciens treated, and those treated with both R. flavefaciens and duloxetine. The mPFC is known to be a significant center of behavior regulation, receiving inputs from different limbic structures, and has been implicated in both depression and antidepressant treatment effects55,56.

Differential expression analyses

RNA-seq data were firstly analyzed by differential expression analyses (DEA) (Supplementary Table 3). Duloxetine treatment alone, compared to control, changed expression of only one gene, Adrb1, one of the norepinephrine receptors (Fig. 4a). In contrast, R. flavefacienstreatment resulted in 324 differentially expressed genes (DEGs), in comparison to controls (Fig. 4a). GO analyses revealed that genes up-regulated by R. flavefaciens treatment were enriched for mitochondrial processes (Fig. 4b), especially oxidative phosphorylation, while the down-regulated genes were enriched with genes involved in neural plasticity (Fig. 4c). When the group treated by both R. flavefaciens and duloxetine was compared to controls, there were 185 DEGs (Fig. 4a). Interestingly, concomitant R. flavefaciens and duloxetine treatments also resulted in the up-regulation of genes enriched for mitochondrial energy metabolism (Fig. 4d), and down-regulation of genes enriched for neural plasticity (Fig. 4e). When the group receiving both the bacteria and duloxetine was compared to the group receiving duloxetine alone, only one gene, Dpysl2, was shown to be differentially expressed, i.e., down-regulated. The list of all DEGs can be found in Supplemental Table 3. Therefore, treatment with R. flavefaciens induces expression changes of genes related to mitochondrial and neuronal process in the mPFC.

https://www.nature.com/articles/s41398-019-0466-x

Abstract

Given that antidepressants (ADs) work slowly, there is interest in means to accelerate their therapeutic effect and to reduce side effects. In this regard, thiamine (vitamin B1) is attracting growing interest. Thiamine is an essential nutrient, while thiamine deficiency leads to a broad variety of disorders including irritability and symptoms of depression. Here, we tested the hypothesis that adjuvant thiamine would reduce depression, compared to placebo. A total of 51 inpatients (mean age: 35.2 years; 53 % females) with MDD (Hamilton Depression Rating Scale score (HDRS) at baseline: >24) took part in the study. A standardized treatment with SSRI was introduced and kept at therapeutic levels throughout the study. Patients were randomly assigned either to the thiamine or the placebo condition. Experts rated (HDRS) symptoms of depression at baseline, and after 3, 6, and 12 weeks (end of the study). Between baseline and the end of the study, depression had reduced in both groups. Compared to placebo, adjuvant thiamine improved symptoms of depression after 6 week of treatment, and improvements remained fairly stable until the end of the study, though mean differences at week 12 were not statistically significant anymore. No adverse side effects were reported in either group. Results suggest that among younger patients with MDD adjuvant thiamine alleviated symptoms of depression faster compared to placebo. Importantly, improvements were observed within 6 weeks of initiation of treatment. Thus, thiamine might have the potential to counteract the time lag in the antidepressant effects of ADs.

1. Background

Surprisingly, given their pivotal physiological significance, our understanding of the role of the B group of vitamins (thiamine (B1), riboflavin (B2), niacin (B3), pantothenic acid (B5), vitamin B6, folate (B9) and vitamin B12) in health and brain function is limited in several respects. As an example, the major human epidemiological and controlled trial research effort in this area has concentrated almost exclusively on that small sub-set of B vitamins (folate, vitamin B12 and, to a lesser extent vitamin B6) that play the most obvious roles in homocysteine metabolism. The multifarious inter-related roles of the remaining five B vitamins have been largely overlooked. Possibly as a result of this, the many intervention studies that have involved administering just folic acid ± vitamins B12 and/or B6, have generated equivocal results. Similarly, whilst we have some knowledge of the minimum levels of each B vitamin required in order to prevent explicit deficiency related diseases, we have a poor understanding of the negative effects of levels of consumption that lie above the minimum, but under the optimal level of consumption for these vitamins. Indeed, we have no clear idea of where the optimal level of consumption may lie. The following review will therefore describe some of the closely inter-related cellular functions of the entire group of B vitamins in catabolic and anabolic metabolism; examine evidence from human studies suggesting widespread sub-optimal consumption of a number of these vitamins in developed societies, and the related case for consumption of these vitamins well in excess of governmental minimum recommendations. It will also marshal evidence from the largely equivocal human literature describing intervention with a small sub-set of B vitamins, and the more promising literature describing the effects of “multi-vitamin” treatments. Taken together, these strands of evidence suggest that supplementation with the entire B group of vitamins is a more rational approach than selecting one, two or three compounds from this sub-group of vitamins.

Brain Specific Roles of B Vitamins

The brain is by far the most metabolically active organ in the body, representing only 2% of body weight but accounting for over 20% of the body’s total energy expenditure [38]. The B vitamins’ general metabolic functions, alongside their roles in neurochemical synthesis, may therefore be conceived as having a particular impact on brain function. Indeed, the importance of the B vitamins for brain function is illustrated by the fact that each vitamin is actively transported across the blood brain barrier and/or choroid plexus by dedicated transport mechanisms. Once in the brain, specific cellular uptake mechanisms dictate distribution, and, whilst the B vitamins all have high turnovers, ranging from 8% to 100% per day, their levels are tightly regulated by multiple homeostatic mechanisms in the brain [39,40]. This guarantees that brain concentrations remain comparatively high. For example, the concentration of methyltetrahydrofolate (the principal circulating form of folate) in the brain is four times that seen in plasma [39], whereas biotin and pantothenic acid exist in the brain at concentrations of up to 50 times that seen in plasma [41].

2.1.1. Thiamine (Vitamin B1)

Thiamine is a coenzyme in the pentose phosphate pathway, which is a necessary step in the synthesis of fatty acids, steroids, nucleic acids and the aromatic amino acid precursors to a range of neurotransmitters and other bioactive compounds essential for brain function [9]. Thiamine plays a neuro-modulatory role in the acetylcholine neurotransmitter system, distinct from its actions as a cofactor during metabolic processes [42] and contributes to the structure and function of cellular membranes, including neurons and neuroglia [35].

2.1.2. Riboflavin (Vitamin B2)

The two flavoprotein coenzymes derived from riboflavin, FMN and FAD are crucial rate limiting factors in most cellular enzymatic processes. As an example, they are crucial for the synthesis, conversion and recycling of niacin, folate and vitamin B6, and for the synthesis of all heme proteins, including hemeglobin, nitric oxide synthases, P450 enzymes, and proteins involved in electron transfer and oxygen transport and storage [11]. The flavoproteins are also co-factors in the metabolism of essential fatty acids in brain lipids [12], the absorption and utilisation of iron [43], and the regulation of thyroid hormones [11]. Dysregulation of any of these processes by riboflavin deficiency would be associated with its own broad negative consequences for brain function. Riboflavin derivatives also have direct antioxidant properties and increase endogenous antioxidant status as essential cofactors in the glutathione redox cycle [44].

2.1.3. Niacin (Vitamin B3)

A vast array of processes and enzymes involved in every aspect of peripheral and brain cell function are dependent on niacin derived nucleotides such as nicotinamide adenine dinucleotide (NAD) and NAD phosphate (NADP). Beyond energy production, these include oxidative reactions, antioxidant protection, DNA metabolism and repair, cellular signalling events (via intracellular calcium), and the conversion of folate to its tetrahydrofolate derivative [45]. Niacin also binds agonistically at two G protein receptors, the high affinity Niacin receptor 1 (NIACR1), responsible for the skin flush associated with high intake of niacin, and the low affinity NIACR2. Niacin receptors are distributed both peripherally in immune cells and adipose tissue, and throughout the brain. Currently established roles include modulation of inflammatory cascades [46,47] and anti-atherogenic lipolysis in adipose tissue [48,49]. NIACR1 receptor populations have been shown to be down-regulated in the anterior cingulate cortex of schizophrenia sufferers [46] and upregulated in the substantia nigra of Parkinson’s disease sufferers, (a group that have low niacin levels generally) with levels correlating with poorer sleep architecture in this group [50]. A recent case study demonstrated that 250 mg niacin administration modulated peripheral immune cell NIACR1 expression and attenuated the disturbed sleep architecture associated with Parkinson’s disease [51].

2.1.4. Pantothenic Acid (Vitamin B5)

This vitamin is a substrate for the synthesis of the ubiquitous coenzyme A (CoA). Beyond its role in oxidative metabolism, CoA contributes to the structure and function of brain cells via its involvement in the synthesis of cholesterol, amino acids, phospholipids, and fatty acids. Of particular relevance, pantothenic acid, via CoA, is also involved in the synthesis of multiple neurotransmitters and steroid hormones [14].

2.1.5. Vitamin B6 (Pyridoxine, Pyridoxal, Pyridoxamine)

Beyond its role as a necessary cofactor in the folate cycle (see above and folate section below), the role of vitamin B6 in amino acid metabolism makes it a rate limiting cofactor in the synthesis of neurotransmitters such as dopamine, serotonin, γ-aminobutyric acid (GABA), noradrenaline and the hormone melatonin. The synthesis of these neurotransmitters is differentially sensitive to vitamin B6 levels, with even mild deficiency resulting in preferential down-regulation of GABA and serotonin synthesis, leading to the removal of inhibition of neural activity by GABA and disordered sleep, behaviour, and cardiovascular function and a loss of hypothalamus-pituitary control of hormone excretion. Vitamin B6 also has a direct effect on immune function and gene transcription/expression [15] and plays a role in brain glucose regulation [52]. More broadly, levels of pyridoxal-5′-phosphate are associated with increased functional indices and biomarkers of inflammation, and levels of pyridoxal-5′-phosphate are down-regulated as a function of more severe inflammation [53,54], potentially as a consequence of pyridoxal-5′-phosphate’s role either in the metabolism of tryptophan or in one-carbon metabolism [53]. This role is particularly pertinent as inflammatory processes contribute to the aetiology of numerous pathological states including dementia and cognitive decline [55].

2.1.6. Biotin (Vitamin B7)

The brain is particularly sensitive to the delivery and metabolism of glucose. Biotin plays a key role in glucose metabolism and haemostasis, including regulation of hepatic glucose uptake, gluconeogenesis (and lipogenesis), insulin receptor transcription and pancreatic β-cell function [18]. Frank deficiency in biotin is rarely reported, although lower circulating levels of biotin have been reported in those suffering gluco-regulatory dysfunction, for instance Type II diabetes, alongside an inverse relationship between fasting plasma glucose and biotin levels [18].

2.1.7. Folate (Vitamin B9) and Vitamin B12 (Cobolamin)

The functions of these two vitamins are inextricably linked due to their complementary roles in the “folate” and “methionine” cycles. Indeed, a deficiency in vitamin B12 results in a functional folate deficiency, as folate becomes trapped in the form of methyltetrahydrofolate [11,19]. An actual or functional folate deficiency, with an attendant reduction in purine/pyrimidine synthesis and genomic and non-genomic methylation reactions in brain tissue, leads to decreased DNA stability and repair and gene expression/transcription, which could hamper neuronal differentiation and repair, promote hippocampal atrophy, demyelination and compromise the integrity of membrane phospholipids impairing the propagation of action potentials [45]. Folate related downregulation of the synthesis of proteins and the nucleotides required for DNA/RNA synthesis, has ramifications for rapidly dividing tissue in particular, and therefore underlies the foetal developmental disorders and megaloblastic anaemia (alongside aspects of neuronal dysfunction), associated with either folate or vitamin B12 deficiency [11,19,45]. The efficient functioning of the folate cycle is also necessary for the synthesis and regeneration of tetrahydrobiopterin, an essential cofactor for the enzymes that convert amino acids to both monoamine neurotransmitters (serotonin, melatonin, dopamine, noradrenaline, adrenaline), and nitric oxide [56,57] (see Figure 2).

The importance of all of the B vitamins to brain function is illustrated by the neurological and psychiatric symptoms commonly associated with deficiency in any one of these eight vitamins [11,45,58,59] (see Table 1). For example, the primary symptoms of vitamin B6 deficiency are neurological, including depression, cognitive decline, dementia, and autonomic dysfunction [15] and vitamin B12 deficiency is often manifested in the form of neurological symptoms prior to the appearance of more typical haematological changes [20]. Notably, whilst about a third of those suffering folate or vitamin B12deficiency present only with anaemia, a similar proportion present only with neuropsychiatric symptoms. Indeed, more than a third of psychiatric admissions have been found to be suffering deficiencies in folate or vitamin B12 [19].

Go to:3. The Homocysteine Hypothesis

No description of the mechanisms of action of the B vitamins would be complete without some consideration of the predominant mechanistic theory that has driven much of the human research in this area. The “homocysteine hypothesis” originally stemmed from the observation that increased fasting plasma levels of the potentially toxic amino acid homocysteine were an independent predictor of cardiovascular disease [60,61],with this observation subsequently extended to cognitive function [62], Alzheimer’s disease and dementia [63]. In essence, the hypothesis attributed mild to moderate increases in homocysteine levels with being a causal contributor to these disease states. Insufficiencies in several of the key vitamins involved in effectively recycling homocysteine in the methionine cycle, in particular folate, but also vitamins B12 and B6, were then implicated as the underlying cause [61]. The mechanisms by which homocysteine has been hypothesised to have these detrimental effects on brain function include its theoretical roles in increasing oxidative stress, the inhibition of methylation reactions, increased damage to DNA and dysregulation of its repair, and direct and indirect neurotoxicity leading to cell death and apoptosis. These processes are suggested to then lead to general effects such as the accumulation of beta-amyloid, hyper-phosphorylation of tau, brain tissue atrophy and compromised cerebrovascular circulation [64].

This hypothesis has been the driver not only for the majority of observational studies investigating epidemiological relationships between vitamins and cardiovascular or brain function, but also for a huge research effort that has seen a flood of clinical trials that have involved the administration of folic acid, either alone or in combination with vitamin B12, and less frequently, vitamin B6. These studies have been conducted on the basis that increasing the levels of these vitamins will reliably reduce homocysteine levels. However, the results of the intervention trials have been entirely equivocal. As an example, meta-analyses of the data from 17 trials, involving 39,107 participants [65] and 12 trials involving 47,429 participants [66] found that whereas administering folic acid ± vitamins B12/B6 reliably reduced homocysteine levels, these vitamins had no protective effect against cardiovascular or cerebrovascular disease events or all-cause mortality. The findings with regards to brain function, reviewed below, are equally equivocal. In addition, studies investigating the relationship between a common genetic polymorphism associated with higher homocysteine levels (methylenetetrahydrofolate reductase (MTHFR) 677TT) and cardiovascular disease [61] and cognitive function [67] have also been equivocal. These findings suggest that homocysteine is likely to be a simple biomarker or epiphenomenon related either to the circulating levels of the relevant vitamins or a disease related mechanism or process [61,68,69,70].

One unfortunate consequence of the “homocysteine hypothesis” is that it has effectively funneled the majority of clinical trial research in this area towards elucidating the effects of folic acid, and to a decreasing extent vitamin B12 followed by vitamin B6. The potential effects and roles of the other five B vitamins have been almost entirely ignored, despite the fact that the entire palette of B vitamins work intricately in concert. As an example, staying with the homocysteine theme, the status of folate and vitamin B6/B12 are themselves dependent on levels of riboflavin derived flavoproteins. Riboflavin is also essential for the metabolism of homocysteine as a cofactor for methylenetetrahydrofolate reductase (MTHFR) and methionine synthase reductase (MTRR) [11,12,22]. In line with this, homocysteine levels have been shown to correlate negatively with plasma riboflavin and dietary riboflavin intake [71,72], and supplementation with riboflavin has been shown to attenuate both increased homocysteine levels and blood pressure in individuals with the MTHFR 677TT polymorphism [73,74]. Although it has received even less attention than riboflavin, it is notable that niacin is also a necessary cofactor for the enzymes dihydrofolate reductase and S-adenosylhomocysteine synthase in the folate/tetrahydrobiopterin and methionine cycles, respectively, and that all of the remaining B vitamins play roles in the interlinked folate/methionine and citric acid cycles [8,11,14] (see Figure 1 and Figure 2).

The potential limitations of administering a restricted range of B vitamins are illustrated by evidence showing that approximately a third of supplementation studies to date have involved the administration of folic acid alone [65,66]. As noted above, folate and vitamin B12 are intimately interlinked within the folate/methionine cycles, and increasing the level of folate can mask the accrual of permanent neurological damage associated with a specific vitamin B12 deficiency [20]. A striking illustration of this was provided by an epidemiological study by Morris et al. [75] who reported that high folate status was associated with protected cognitive function, but only in those with normal vitamin B12 status, with this relationship reversed in participants with low vitamin B12 status. For this group, high folate status exacerbated the detrimental effect of vitamin B12 deficiency, increasing the risk of cognitive impairment and anaemia by a factor of five, compared to those with normal vitamin status. A further study also demonstrated that low vitamin B12 status was associated with a significantly increased decline in cognitive performance over the subsequent eight years, with this effect exacerbated in those having high levels of folate, or those taking folic acid supplements [76]. Alongside these observations it is interesting to note that in one study supplementation with folic acid also significantly increased the proportion of participants with riboflavin deficiency [72].

It is also notable, firstly, that supplementation with folic acid may not be effective in terms of regulating homocysteine: a recent study showed that folic acid supplementation reduced plasma homocysteine levels as expected, but left the more important cellular levels of homocysteine untouched, with evidence suggesting that cellular one-carbon metabolism was also disturbed [77]. Secondly, folate may affect physiological functioning via an alternative mechanism, for instance via the role the folate cycle plays in the synthesis and regeneration of tetrahydrobiopterin [57], a folate-dependent rate limiting cofactor in the enzymatic pathways to both nitric oxide and monoamine neurotransmitter synthesis [37,78,79]. This mechanism would accommodate the observation that folate increases endothelial vasodilation via a mechanism entirely unrelated to homocysteine [57,79] and would accommodate epidemiological observations of a relationship between reduced folate status and depression and disturbed cognitive function [56,78,80].

It seems reasonable to conclude, from the above and the following, that concentrating solely on one potential hypothesis as to the mechanisms of action of a small group of vitamins with multifarious complex cellular functions, at the expense of elucidating the mechanisms and effects of a broader group of inter-related vitamins, in hindsight, may not be a rational approach to research in this area.

Abstract

Background

Oral contraceptives (OCs) may affect oxidative stress status. We aimed to assess whether supplementation with vitamins E and C reduced this OC effect.

Study Design

One hundred twenty healthy female individuals were divided into three groups: A, control; B, untreated OCU (OC users); and C, treated OCU (OC users with vitamin E and C supplementation). In all cases, plasma glutathione peroxidase (GPx) and glutathione reductase (GR) activities and malondialdehyde (MDA) level were determined.ResultsSignificant increases were found in the plasma MDA level, and activities of GPx and GR in plasma were decreased in Group B compared to the control group. Supplementation with vitamin C and E significantly increased the activity of GPx and GR activity, and reduced plasma MDA levels in Group C (p<.05).

Conclusions

These data suggest that low-dose OCs, by enhancing the stress oxidative and lipid peroxidation, may represent a potential cardiovascular risk factor, and the use of vitamins E and C may be beneficial in ameliorating this side effect of OCs.

Abstract

Oral contraceptives agents (OCA) have been in use for more than two decades, and at the present time, 150 to 200 million women are using the preparations. Apart from their gynecologic influence, the hormones have been shown to affect a number of metabolic and nutritional processes, some advantageously and others disadvantageously. Concern over the nutritional status of females consuming OCA prompted this review. Eight vitamins and three minerals were investigated. Contraceptive steroid ingestion was shown to depress the physiologic levels of six nutrients (riboflavin, pyridoxine, folacin, vitamin B12, ascorbic acid and zinc), elevate the levels of three others (vitamin K, iron and copper) and provide little or no change in one (alpha tocopherol) and questionable increases in another (vitamin A). It was concluded that females consuming OCA should pay particular attention to vitamin and mineral intake and, if warranted, consume physiologic supplements of needed nutrients.

PIP: The state of knowledge concerning the effects of OCs (oral contraceptives) and mineral metabolism is assessed. A review of the literature indicates that OCs depress the levels of Vitamin B2, or riboflavin, Vitamin B6, or pyridoxine, folacin, Vitamin B12, Vitamin C, or ascorbic acid, zinc and elevate levels of Vitamin K, copper, and iron. The ingestion of OCs produces little effect on Vitamin E, or alpha tocopherol. Findings on the effects of OC ingestion on Vitamin A are ambiguous. OC users have 50%-80% higher serum levels of Vitamin A than nonusers; however, OC users may have a greater need for Vitamin A than nonusers. The need for riboflavin may also be higher for OC users. OC users need more pyridoxine and riboflavin is needed to oxidize pyridoxine phosphate to pyridoxal phosphate. Most studies support the contention that OC usage leads to a deficiency of Vitamin B6. Approximately 80% of all women using OCs for 6 or more months experience abnormal typtophan metabolism. In order to correct this problem, 25 mg daily, or 12 times the normal daily requirement, is needed. Some investigators recommend givng this dosage to women, who experience abnormal tryptophan metabolism, while others warn that the long-term effects of such high dosages are unknown. Most investigators recommend that OC users, with Vitamin B12 or Vitamin C deficiencies, should be given supplementary vitamins.

2.12. Oral Contraceptives(OC)

2.12.1. Vitamin B6

Tryptophan metabolism, an indirect measure of vitamin B6 status, is abnormal in OC users compared to controls and can be corrected with supplemental doses of vitamin B6 [358,359,360,361,362,363,364]. However, estrogens may influence tryptophan metabolism independently of B6 status [172,364], thus other markers of vitamin B6 levels, such as plasma 5′-phosphate (PLP), urinary 4-pyridoxic acid (4-PA), urinary B6, and erythrocyte aminotransferase or transaminase activity, should be used when studying the potential interaction between OCs and vitamin B6. OC users were observed to have significantly lower PLP in both fasting and non-fasting plasma compared to nonusers [365,366]. However, not all studies accounted for subjects’ dietary intake of vitamin B6, which can influence PLP levels.

The potential effects of OC use on vitamin B6 status have also been assessed in controlled, depletion–repletion feeding studies. In these studies, OC users and nonusers were fed a vitamin B6 deficient diet for one menstrual cycle, and then given a replete diet with 0.8–16.6 mg/d pyridoxine for another cycle [363,364]. A comprehensive array of markers was used to assess B6 status weekly. Among these markers, only tryptophan metabolites differed significantly between the OC users and nonusers. During the depletion period the decline in all parameters was similar between groups, and during repletion 1.8 mg pyridoxine was equally effective in increasing parameters of B6 status in both groups. Thus, vitamin B6 requirements were not higher in OC users, which was confirmed in a separate but similarly designed study [360,367].

When using erythrocyte transaminase as a biomarker of vitamin B6 status, one cross-sectional study of 233 women taking OCs found almost 50% of OC users, compared to 18% of nonusers, had marginal or deficient B6 status [368]. Conversely, in a study that analyzed both cross-sectional and longitudinal data of erythrocyte transaminase activity in young OC users vs. nonusers, no differences were reported [369].In summary, results are mixed regarding the relationship between OC use and vitamin B6 requirements. Depending on the biomarker used to measure B6 status, OC use may negatively impact or have no effect on B6 levels, thus, widespread supplementation is not currently recommended [370]. However, intervention studies with vitamin B6 supplementation have reported improvements in clinical symptoms of B6 deficiency and fewer side effects in OC users who may be deficient in this vitamin [371,372]. 

Abstract

Purpose: Elevation of homocysteine levels have been involved as a remarkable risk factor for cardiovascular disease. Decreased bioavailability of nitric oxide (NO) may result in abnormal reactions between the vessel wall and platelets and is thus involved in the initiation and progression of atherosclerosis. We aimed to assess the effect of a low dose oral contraceptive pills on homocysteine and NO levels which may influence the individual cardiovascular risk by regulation of endothelial function and development of atherosclerosis.

Methods: The study was conducted in 50 healthy women with normal menstrual cycles as control group and 50 healthy women receiving oral contraceptive pill for at least three menstrual cycles. Homocysteine concentration was assayed by an enzyme immunoassay method and a colorimetric assay was used for determination of NO levels.

Results: After 3 months of treatment, homocysteine levels were significantly increased (P = 0.027), and there was a significant and considerable decrease (P = 0.048) in NO concentration of oral contraceptive pill (OCP) consumers.

Conclusions: Our data indicates that OCPs increase the homocysteine (HCY) and reduce the NO levels. These findings warrant a possible implication of OCP in change of risk of development of coronary heart disease.

Abstract

This study addressed the associations between oral contraceptive (OC) use and obesity as measured by recording the body mass index (BMI) of premenopausal females, and possible interactions with micronutrient intake were considered. A group of 39,189 premenopausal females aged 35–59 were included in the analysis; they were in the Health Examinee cohort. Participant BMIs were calculated from anthropometric measurements, and females with a BMI≥25kg/m2 were considered obese. Individual OC use, age at first OC use, duration of OC use, nutrient intake, and other covariates were measured with a structured questionnaire. A multivariate logistic regression with an interaction term was applied to identify the odds ratio (OR) and 95% confidence intervals (CI) between OC use and obesity along with consideration of micronutrient intake interactions. OC use is associated with an increased risk of obesity (OR = 1.12, 95% CI = 1.04–1.20), and females who used OCs for more than 6 months over their lifetimes were more likely to be obese (OR = 1.15, 95% CI = 1.01–1.32) compared with those who used OCs for <6 months. There were interaction effects between phosphorus, potassium, vitamin A, vitamin B1, vitamin B2, niacin, vitamin C intake and total duration of OC use on being obesity (P-value<0.05). When stratified by micronutrient intake, the associations between total OC use duration and obesity were present only among those with calcium, phosphorus, potassium, vitamin A, B1, B2, C, niacin, and folate intakes below the recommended levels. Efforts to estimate nutrient intake and prevent micronutrient depletion with supplements or food should be considered by clinicians for females who take OC for a long period.

Introduction

Contraception use by females in their childbearing years and the choice of having safe, effective, affordable and acceptable contraceptive methods as part of family planning are important dimensions of reproductive health. Oral contraceptives (OCs) are currently the most commonly used contraception method in developed countries. In developing countries, female sterilization and intrauterine devices are the two most commonly used contraceptive methods, accounting for 58% of all contraceptive use [1]. However, in developed countries in Asia such as Japan and the Republic of Korea, the prevalence of OC use is lower than it is in other Western countries [1], suggesting that there are different preferred methods depending on the region.

In Western countries, adherence to OC use is poor, with a 6-month discontinuation rate of more than 50% [2, 3]; side effects are the most common reason for poor compliance [2, 4, 5]. Although previous longitudinal studies and randomized controlled trials have shown that OCs do not have an effect on weight change [6, 7], and a recent review by Cochrane suggested that OC use did not significantly affect weight [8], weight gain is frequently reported as a side effect of OC use [9]. The OC discontinuation rate among the females who gained weight from OCs was 40% higher compared to the females who did not gain weight [2].

Another potential effect of OC use is higher micronutrient requirements. Females who used OCs showed lower blood nutrient levels compared with non-users, suggesting that they may need higher amounts of vitamins and minerals. In addition, an association between micronutrient deficiencies and obesity has been suggested to occur via changes in leptin concentrations or increases in the inflammatory response [10–13]. Therefore, OC use and micronutrient deficiency and micronutrient deficiency and obesity are correlated, and interplay between OC use and micronutrient intake in the effect on obesity seems plausible.

In this study, we investigated possible associations between OC use and obesity as measured by recording the body mass indexes (BMI) of premenopausal females in health examinee-based samples in Korea. In particular, the possible interactions between OC use and micronutrient intake in association with obesity were examined.

Go to:Materials and MethodsData source and study population

Data from the Health Examinee (HEXA) cohort was part of the Korea Genome Epidemiology Study (KoGES), which is an ongoing cohort study that was started in 2001 to investigate the effects of various factors and their interactions on the development of chronic diseases in the Korean population, and it was used in this study. Health examinees from health examination centers in 14 urban areas around Korea were recruited from 2004–2013, and the baseline survey and measurement information was analyzed. The interviewers gave information about the HEXA cohort, and informed consents were obtained from health examinees who agreed to participate. All participants completed a questionnaire including their past medical history, sociodemographic and behavioral characteristics, reproductive factors, and a validated food frequency questionnaire, which assessed the intake of 106 food items, was administered by the interviewers. Daily nutrient intake was estimated using the sum of the nutrient intake of each food item [14, 15]. In addition, information from physical examinations including anthropometric measurements and biochemical measurements with a fasting blood sample that was part of a routine process was obtained from all participants. Details of the KoGES and HEXA cohort are described elsewhere [16] or at the Korea National Institute of Health website (http://www.nih.go.kr/NIH/eng/main.jsp).

Among the 111,592 female participants aged 35–73, 62,640 had stopped menstruating for more than 12 months, 2,716 had stopped menstruating for approximately 3 months, 9,440 were premenopausal with an irregular cycle, 30,004 were premenopausal with a regular cycle, and 6,792 had missing information about their menstruation. Among the 39,444 premenopausal females, 39,189 females under 60 were included in the analysis. The Institutional Review Board of the National Cancer Center approved this study protocol, which was in compliance with the Declaration of Helsinki (IRB No: NCC2014-0098).

Variables

For the obesity index, participant BMIs were calculated as weight (kg)/height2 (m2) by using anthropometric measurements, and females with a BMI ≥25 kg/m2 were considered obese according to the World Health Organization’s revised guidelines for Asians [17, 18]. Individual OC use, age at first OC use, and duration of OC use were assessed by a questionnaire. OC use was categorized as never use and use. The age at first OC use and the total duration of OC use were used to calculate the median (<28 years old and ≥28 years old) or quartile values (<3months, 3–6 months, 6–12 months, and >12 months) in the premenopausal females included in this study. The other covariates included age (in 1 year increments), smoking status (never, ex-smoker, or current smoker), drinking status (never, ex-drinker, or current drinker), regular exercise (none, < 150 minutes/week, or ≥ 150 minutes/week) based on current Europe [19], and World Health Organization physical activity guidelines [20], educational attainment (< high school, ≥ college), monthly household income (<$3000/month, ≥$3000/month), marital status (married, single/divorced/widowed), occupation (unemployed/housewife, white collar job, blue collar job), parity (nulliparous, 1–2, 3 or more), and total energy intake (kcal/day, <1400, 1400–1699, 1700–2049, ≥ 2050 as quartile in the included subjects).

Among the assessed nutrients, we included calcium, phosphorus, potassium, vitamin A, vitamin B1, vitamin B2, niacin, vitamin C and folate because the recommended daily intake level or adequate intake level of these nutrients have been identified for Koreans.

Statistical analysis

The sociodemographic and health behavioral characteristics of those who never used OCs and those who did were compared by using a chi-square test or a t-test. To identify the association between OC use and obesity, we conducted a multivariate logistic regression analysis to estimate the odds ratios (ORs) and 95% confidence intervals (CI) by OC use, age at first OC use, and duration of OC use, and the values were adjusted for age, smoking status, drinking status, regular exercise, educational attainment, monthly household income, marital status, occupation, parity, and total energy intake.

To identify the interaction effects between age of first OC use, duration of OC use and micronutrient intake in association with obesity, the interactions were assessed by using multivariate logistic regression with interaction terms. In addition, the association between the age at first OC use, the duration of OC use and obesity was stratified by micronutrient intake (less than recommended level, recommended level or more). All statistical analyses were performed with SAS software ver. 9.1 (SAS Institute, Cary, NC, USA).

Go to:Results

Of the 39,189 females, 5,508 (16.1%) had some experience in using OC including 5,291 past users and 217 current users. The sociodemographic and health behavioral characteristics of people who used OCs and those who never did are shown in Table 1. The females who took OCs were older and had lower rates of education beyond high school, increased tobacco or alcohol use and engaged in exercise less regularly. They reported lower household incomes, and being unmarried. They were more likely to be engaged in blue collar work. The proportion of obese females was higher in OC users (21.8% in never-users compared with 25.2% in users, P-value <0.001). There were no significant differences in energy intake distribution between the two groups. Among the OC users, the median age at first OC use was 28 years old and the quartile ranges for the duration of OC use were <3 months, 3–6 months, 6–12 months and >12 months.

Discussion

This study employed baseline information to show that premenopausal females who used OCs had 12% higher odds of being obese, and among OC users, OC use duration of more than 3 months had a positive association with obesity. However, when stratified by micronutrient intake, OC use of more than 3 months had a positive association with obesity among females whose micronutrient intake was not sufficient. For those who took in more than recommended, the duration of OC use was not associated with being obese. Although this study design did not allow for causal inference, the findings suggested that consuming adequate levels of micronutrients may be necessary for OC users, who may also face higher odds of obesity.

Although a direct comparison with previous clinical trials that showed OC use did not affect weight change [6–8] might be difficult, our results showed that OC users had an increased tendency to be obese after controlling for the effects of other factors. Although the distribution of energy intake was not different, OC users exhibited more unhealthy behaviors and lower socioeconomic characteristics, such as smoking, drinking, completing less education, earning lower incomes, being unmarried and working a blue collar job. Thus, differences in unmeasured covariates and health behaviors between OC users and non-users may increase the odds of obesity in OC users. Otherwise, the effect of OCs may be different according to regions or ethnicities when considering that all previous clinical trials included in the Cochrane review were conducted in Western countries, and there was no Asian study included. To identify the causal association, we suggest that future studies target females living in Asia where OC use is not a common contraceptive method.

Several review studies have shown that blood folate, vitamin B2, vitamin B6, vitamin B12, vitamin C, vitamin E, zinc, selenium, and magnesium was lower in OC users in comparison with non-users [13, 21]. It is well-known that nutrient deficiencies in antioxidants, vitamin A, vitamin B complex, vitamin D, calcium, iron, and zinc have been associated with obesity [10, 22]. These deficiencies included low serum levels or lower intake [10]. The increased intake of calcium prevents fatty acid synthesis and promotes lipolysis and lipid oxidation [10, 23]; phosphorus consumption from food compromises adenosine triphosphate production, which regulates energy metabolism and regulates insulin release [24]; vitamin A inhibits adipogenesis, increases fat cell apoptosis, and regulates leptin [10, 25–27]; and vitamin C, an antioxidant, regulates gene expression involved in adipogenesis and glucocorticoid metabolism [10]. Although a direct mechanism linking potassium, B vitamins and obesity is not evident, lower potassium consumption induces insulin resistance and increases metabolic syndrome [28], and several previous studies have shown lower levels of B vitamins in obese people [10]. Considering that OC use may increase micronutrient intake requirements [13], the relationships among total duration of OC use and micronutrient intake in association with obesity observed in this study may be related to unmet micronutrient needs, an intermediately associated factor in the relationship between longer OC use and obesity. In addition, we did not observe significant associations for longer duration OC use among those who ingested micronutrients above the recommended levels. No significant interplay between age at which OCs were started and micronutrient intake on the associations with obesity was observed, suggesting an interaction between cumulative exposure to OCs and higher micronutrient requirements, which may be associated with obesity. Although we adjusted for possible confounding variables, further prospective studies are necessary to understand whether these interactions are causal.

As the authors acknowledge, this is the first study to investigate possible interactions between OC use and micronutrient intake, and the major limitation of this study should be mentioned. This study employed a baseline survey along with measurements from premenopausal females who visited health examination centers, and it used a cross-sectional approach that has an important limitation in the interpretation of causal relations between exposure and outcome. However, it seems unlikely that obesity could be the determinant of OC use, and obesity was measured at the time of study recruitment; OC use reflected both past and current exposure, by including both current and past users. The number of current OC users was too small (217 of 39,189, 0.55%), and thus we could not observe the current OC use effect on obesity. The study subjects were health examinees who went for health check-ups and agreed to participate in the HEXA cohort study. Although the Korean government offers various health screenings to the whole population for free or at low cost to increase accessibility, the participants might be more concerned about their health status than populations who did not receive health examinations. Thus, the generalization of our results to all premenopausal females who ever used OCs should be treated with caution. The prevalence of current OC use among married women aged 15–44 was 2.3% in Korea [29], which was higher than our results (0.55%), but the marital status and age distribution of these two populations were not different, and therefore, a direct comparison might be not possible. Following the 1990s, most OCs in use have been a combination of estrogens and progestins, and a few might contain estrogen alone or progesterone only, but component information was not available. Information on heights and weights was taken by trained health center workers, but OC use and nutrient intake were obtained with questionnaires. Although there is a possibility of some recall bias, misclassification of OC use may not be systematic because we used the baseline information from the cohort study, which is less affected by recall bias [30], and its influence over the true effect would be minimal. The validated food frequency questionnaire, which covered food items that are generally consumed by Koreans [14], was applied for nutrient intake, and thus a bias caused by the nutrition intake measurement also would not be serious. We could not determine whether the source of the nutrients was from food or dietary supplements because daily nutrient intake was calculated using the sum of the nutrient intake for each food item and was not divided by the intake source. We tried to control for the effects of other covariates through multivariate adjustment, but residual confusion from uncontrolled variables would remain. For example, polycystic ovarian syndrome, which has a prevalence in Korean females of 4–6% [31, 32] could be a confounding factor by possibly increasing obesity and OC intake [33]; however, neither the questionnaire nor medical examination assessed each subject’s history of polycystic ovarian syndrome.

Despite these limitations, the strength of this study is that it includes a large number of subjects with a sufficient number of females who have used OCs. Considering the lower prevalence of OC use in Asian females [1], a large study population is needed to obtain enough OC users. Because all the participants were health examinees, we expect that they would have relatively homogenous characteristics compared with population-based samples, and the comparability of the study population would be increased.

In conclusion, females who have taken OCs had higher odds of obesity than never-users, and those who took OC for 3 months or more over their lifetimes had an increased association with obesity. However, the association between total duration of OC use of 3 months or more and obesity was modified by micronutrient intake, and it was only present among those whose micronutrient intakes were below the recommended levels. Considering that these micronutrients play a role not only in obesity but also in many functions used to regulate our multiple metabolic pathways, efforts to increase micronutrient intake should be considered for females taking OCs. Future clinical trials may confirm the effects of the interaction between cumulative OC use and micronutrient intake on obesity, and given the results of this study, dietary recommendations to ensure sufficient micronutrient intake should be considered by clinicians for those who take OCs for longer periods.

Abstract

Oral contraceptives agents (OCA) have been in use for more than two decades, and at the present time, 150 to 200 million women are using the preparations. Apart from their gynecologic influence, the hormones have been shown to affect a number of metabolic and nutritional processes, some advantageously and others disadvantageously. Concern over the nutritional status of females consuming OCA prompted this review. Eight vitamins and three minerals were investigated. Contraceptive steroid ingestion was shown to depress the physiologic levels of six nutrients (riboflavin, pyridoxine, folacin, vitamin B12, ascorbic acid and zinc), elevate the levels of three others (vitamin K, iron and copper) and provide little or no change in one (alpha tocopherol) and questionable increases in another (vitamin A). It was concluded that females consuming OCA should pay particular attention to vitamin and mineral intake and, if warranted, consume physiologic supplements of needed nutrients.

Abstract

Background: Women on different contraceptive methods have been linked with the development of various diseases and possible changes in serum trace elements and vitamins of women on contraceptives have been postulated. Therefore, the relationship between contraceptive use and trace elements needs to be investigated.

Methods: This is a cross-sectional randomized study. After informed consent was obtained, blood samples were collected from a total of 100 women of child-bearing age on different contraceptive methods: 50 on oral contraceptives, 25 on injectables and another 25 on intra-uterine device. Blood samples were also collected from another 50 age-matched non-contraceptive users to serve as control. Serum was analysed using atomic absorption spectrophotometer for zinc, copper manganese, iron, selenium, cadmium, lead and magnesium while colorimetric method was used for phosphorus and calcium. Body mass index (BMI) was calculated as weight in kilogram/height in meter squared. Results obtained from laboratory analysis and anthropometric measurements were analysed using computer SPSS package.

Results: The mean serum zinc, selenium, phosphorus and magnesium levels obtained from subjects on contraceptives were significantly lower (p < 0.01, p < 0.05, p < 0.05 and p < 0.05 respectively) than those of the control group. However, the mean serum copper iron, calcium and cadmium levels were significantly higher (p < 0.05) in participants on contraceptive when compared with the control group. Manganese and lead levels were similar in participants and control groups. Correlation analysis shows significant association between some trace elements and the duration of contraception and body mass index of the participants.

Conclusion: The study showed and confirmed reduced levels of trace elements in women on contraceptives. The reduction is proportional to the duration of contraceptive use.

Abstract

Objective:

Third generation oral contraceptives (OC) may increase the risk of cardiovascular events in healthy young women.

Design and methods:

We assessed the effect of second and third generation OC on CRP, lipids and apolipoproteins in 128 women.

Results:

CRP was significantly higher in third generation contraceptive users. The main determinant of CRP in OC users was triglycerides.

Conclusions:

Young women using oral contraceptives, especially third generation formulas, might not be free of cardiovascular risk having increased CRP concentration.

Introduction

Disturbances in lipoprotein levels play an important role in CVD risk. In women, hormonal effects on lipoproteins are complex and influenced by contraceptive administration. Oral contraceptives are the most commonly used birth control methods all over the world. Recent studies on mortality from CVD have shown that it has increased in midlife women [1]. The use of hormonal contraception causes increase in the prevalence of atherosclerosis and its complications in young, apparently healthy women. New formulations have been developed to decrease adverse effects, such as thrombotic and cardiovascular. However, recent evidence suggests that third generation contraceptives do not reduce, but even increase the risk of adverse effects compared to previous generation. Third generation OC, containing desogestrel, gestodene or norgestimate have been associated with elevated risk of venous thromboembolism and cardiovascular events compared to second generation containing only levonorgestrel [1], [2]. Recent studies demonstrated that use of contraceptives is associated with elevated serum CRP in women, which may partially explain increased risk of future cardiovascular events [2], [3]. CRP is a useful marker of low-grade inflammation and the strongest independent predictor of myocardial infarction, stroke and cardiovascular mortality in apparently healthy women [3]. Available evidence suggests that measurement of ApoB and AI, instead of LDL-C and HDL-C may add valuable information in the assessment of cardiovascular risk. In healthy people with optimal LDL-C concentration, higher ApoB indicates an increased number of small dense LDL that promotes an inflammatory state and the development of atherosclerotic plaques. It was demonstrated that ApoAI, the major apolipoprotein in HDL, manifests anti-inflammatory and anti-atherogenic properties [4]. Moreover, ApoB:ApoAI ratio is suggested as a strong predictor of cardiovascular risk [5]. We evaluated the effects of different generation contraceptives on novel and traditional biomarkers of inflammation and atherosclerosis in young non-obese apparently healthy women with normolipidemia.

One hundred twenty-eight non-obese non-smoking women, aged 25–40 years were recruited into the study. Thirty-four were current oral contraceptive users; 14 used second generation contraceptives containing levonorgestrel and 20 used third generation formula containing desogestrel, gestodene or norgestimate for over 6 months, but not longer than 3 years. All participants provided informed written consent. Study protocol was approved by Bioethics Committee at Collegium Medicum, Nicolaus Copernicus 

Results

In the whole group, hsCRP correlated directly with TG, ApoB concentration (R = 0.4; p < 0.01, p < 0.05), ApoB:ApoAI, Wc (R = 0.3; p < 0.01, p < 0.05) and inversely with HDL-C (R = −0.3; p < 0.05). OC raised ApoA (p < 0.03) and TG concentration (p < 0.0001) (Table 1). Compared with the non-users, OC users had over 3-fold higher median hsCRP (0.5 vs. 1.85, p < 0.001). hsCRP > 1 mg/L was found in 65% of the OC group and 30% of non-users whereas hsCRP > 3 mg/L was found in 35% of the OC group and in 18% of non-users.

Discussion

Widespread use of hormonal contraceptives has given rise to concerns about their effects on cardiovascular risk factors, including lipoproteins, glucose, vascular and inflammatory factors. We have found that healthy women using contraceptives had over 3-fold higher hsCRP concentration than non-users and hsCRP > 3 mg/L occurred twice more frequently in OC. Our results indicated that third generation contraceptives increased CRP to significantly higher extent.

Newer generations of contraceptives

Conclusions

We are aware of the limitation of this study that includes a relatively small group of contraceptive users; however, we believe that our data could have several important clinical implications. We conclude that healthy, young, non-smoking, non-obese, normolipidemic women using OC, especially third generation formula, might not be free of cardiovascular risk having increased CRP concentration.

2.12.6. Vitamin C and E 

While some studies indicate circulating vitamin C levels are lower in OC users compared to nonusers [383,392,393], others indicate little threat to individuals who live a healthy lifestyle and consume a diet adequate in vitamin C [175,394]. Then again, several studies show chronic OC use leads to increased oxidative stress, in particular lipid peroxidation, and lower circulating vitamin E [394,395,396,397,398,399,400]. Enhanced oxidative stress and lipid peroxidation may represent a potential risk for cardiovascular disease. One study reported an increase in catalase and glutathione peroxidase activities, i.e., endogenous antioxidant defenses, in 19 young, healthy, non-smoking women after nine cycles of OC use when compared to baseline levels [401]. In a larger, controlled study of 120 healthy women, age 18–40 years, OC users were randomized to receive either supplements with 150 mg vitamin C and 200 IU vitamin or no supplements, and compared with non-supplemented, non-OC users. After four weeks, increased plasma malondialdehyde levels, a marker of lipid peroxidation, and reduced glutathione peroxidase and reductase activities were reported in the non-supplemented OC users, while the directions of each were reversed in the supplemented OC users when compared with the control group [395].

Influence of oral contraceptives on ascorbic acid concentrations in healthy, sexually mature womenVera Joyce McLeroy, M.Ed., M.S., Ph.D.,  Harold Eugene Schendel, Ph.D.The American Journal of Clinical Nutrition, Volume 26, Issue 2, February 1973, Pages 191–196, https://doi.org/10.1093/ajcn/26.2.191

The concentration of ascorbic acid was measured in the leukocytes of 126 sexually mature and healthy female subjects. Sixty-three of these served as control subjects, whereas the other 63 had received a variety of oral contraceptives for at least 1 year (mean 34.6 months). The two groups were similar in age, weight, parity, and educational level. The mean dietary intake of vitamin C was similar and above that recommended for their age group.

The mean concentration of ascorbic acid in the leukocytes of women who had received oral contraceptive drugs (19.0 mg/100 g leukocytes) was significantly less (P < 0.01) than that of the control group (25.7 mg/100 g leukocytes). The mean concentration remained at 19 mg regardless of whether or not the experimental subjects were taking ascorbic acid supplements. The mean concentration of ascorbic acid in the leukocytes of the eight control subjects taking vitamin C supplements (35.2 mg/100 g leukocytes) was significantly greater (P < 0.05) than in the 54 control subjects who had not taken vitamin C supplements (24.1 mg/100 g leukocytes).

Furthermore, the duration of contraceptive treatment after 1 year and the type of preparation had no significant effect on the mean ascorbic acid concentrations of the experimental group. The phase of the menstrual cycle, likewise, had no significant effect on the values of either the experimental or control groups. However, compared with the control subjects, the concentration of ascorbic acid was significantly reduced during both the follicular and luteal phases of the menstrual cycle in healthy, sexually mature women using oral contraceptive agents longer than 1 year, regardless of the type.

2.12.5. Magnesium 

Most cross-sectional studies have shown serum magnesium levels are lower in OC users compared to both nonusers and women on other forms of contraception [194,195,196,197,198], with one exception [199]. An increase in the blood calcium to magnesium ratio due to low magnesium levels can influence blood coagulation processes [390,391]. Indeed, a systematic review and network meta-analysis of 26 observational studies that investigated the risk of venous thrombosis for different combined OCs concluded OC use increased the risk of venous thrombosis, and effect size depended on the combination used [200].

Further studies are needed to investigate whether the effects of dietary or supplemental magnesium can influence the calcium: magnesium ratio and the potential for blood clotting in OC users

Abstract

STUDY QUESTION

What is the effect of alternative administration routes of combined contraceptives (CCs) on androgen secretion, chronic inflammation, glucose tolerance and lipid profile?

SUMMARY ANSWER

The use of oral, transdermal and vaginal CCs impairs glucose tolerance and induces chronic inflammation.

WHAT IS KNOWN AND WHAT THIS PAPER ADDS

Oral CCs worsen insulin sensitivity and are associated with increased levels of circulating inflammatory markers, whereas the metabolic effects of transdermal and vaginal CCs have been reported to be minimal. This is the first study comparing three different administration routes of CCs on metabolic variables.

STUDY DESIGN, SIZE AND DURATION

This randomized (computer-generated) open-label 9-week follow-up study was conducted at the Oulu University Hospital, Finland. Fasting blood samples were collected at baseline and thereafter at 5 and 9 weeks of treatment, and serum levels of 17-hydroxyprogesterone, androstenedione, testosterone, C-reactive protein (CRP), sex hormone-binding globulin (SHBG), glucose, insulin, C-peptide, total, low-density lipoprotein and high-density lipoprotein cholesterol and triglycerides were measured. Oral glucose tolerance tests were performed and plasma levels of pentraxin 3 (PTX-3) were measured at 0 and 9 weeks. The randomization list, with an allocation ratio of 1:1:1 and block size of six, was computer generated and constructed by a pharmacist at the Oulu University Hospital. The research nurse controlled the randomization list and assigned participants to their groups at the first visit.

PARTICIPANTS AND SETTING

Forty-two of 54 healthy women who entered the study used oral contraceptive pills (n = 13), transdermal contraceptive patches (n = 15) or contraceptive vaginal rings (n = 14) continuously for 9 weeks. Inclusion criteria were regular menstrual cycles, at least a 2-month washout as regards hormonal contraceptives and no medication.

MAIN RESULTS AND THE ROLE OF CHANCE

Serum levels of SHBG increased and consequently the free androgen index (FAI) decreased in all study groups from baseline to 9 weeks of treatment [FAI, oral: 1.3 (95% confidence interval, CI: 0.94; 1.62) to 0.40 (0.25; 0.54); transdermal: 1.2 (0.96; 1.4) to 0.36 (0.30; 0.43); vaginal: 1.6 (1.1; 2.1) to 0.43 (0.29; 0.58), P < 0.001 in all groups]. Insulin sensitivity was reduced at 9 weeks in all three groups according to the Matsuda index [oral: 7.3 (5.5; 9.0) to 5.6 (3.9; 7.3); transdermal: 9.1 (6.7; 11.4) to 6.6 (4.5; 8.8); vaginal: 7.7 (5.9; 9.5) to 5.4 (3.9; 7.0), P= 0.004–0.024]. Levels of HDL cholesterol, triglycerides and CRP rose in all three groups [CRP, oral: 0.70 (0.38; 1.0) to 5.4 (1.0; 9.9) mg/l; transdermal: 0.77 (0.45; 1.1) to 2.9 (1.4;4.4) mg/l; vaginal: 0.98 (0.52; 1.4) to 3.7 (−0.25; 7.7, a negative value due to skewed distribution to right) mg/l, P≤ 0.002 in all groups] and PTX-3 levels increased in the oral and transdermal study groups (P = 0.007 and P = 0.002).

WIDER IMPLICATIONS OF THE FINDINGS

Although the long-term consequences of the present results remain undetermined, these findings emphasize the importance of monitoring glucose metabolism during the use of CCs, especially in women with known risks of type 2 diabetes or cardiovascular diseases.

BIAS, LIMITATIONS, GENERALIZABILITY

The number of subjects was relatively low. Moreover, the 9-week exposure to CCs is too short to draw conclusions about the long-term health consequences. However, as the subjects were healthy, normal-weight young women, the possible alterations in the glucose and inflammatory profiles among women with known metabolic risks might be even greater.

STUDY FUNDING/COMPETING INTERESTS

This work was supported by grants from the Academy of Finland, the Sigrid Jusélius Foundation, the Finnish Medical Foundation, the Research Foundation of Obstetrics and Gynecology, Oulu University Scholarship Foundation, the North Ostrobothnia Regional Fund of the Finnish Cultural Foundation, the Tyyni Tani Foundation of the University of Oulu and the Finnish-Norwegian Medical Foundation. No competing interests.

Abstract

A number of side effects have been linked to the use of hormonal contraceptives, among others, alterations in glucose levels. Hence, the objective of this mini-review is to show the main effects of hormonal contraceptive intake on glycemic regulation. First, the most relevant studies on this topic are described, then the mechanisms that might be accountable for this glycemic regulation impairment as exerted by hormonal contraceptives are discussed. Finally, we briefly discuss the ethical responsibility of health professionals to inform about the potential risks on glycemic homeostasis regarding hormonal contraceptive intake.

Introduction

Since the early 1950s, when Mexican chemist Luis. E. Miramontes and co-researchers carried out the synthesis of norethisterone (norethindrone), the first oral contraceptive (Miramontes, Rosenkranz and Djerassi 1951; Djerassi et al. 1954), the subsequent mass use of hormonal contraceptive methods has resulted in immense and significant changes to mankind, from artificial birth control, to the social phenomenon which came to be known as “Women's Liberation,” or the “Sexual Revolution” and related behaviors previously discussed in this journal (Norris 2013). Contraceptive use has radically affected population pyramids, especially in the more developed countries which, from having broad-based pyramids reflecting a high birthrate, have moved to narrow-based pyramids, evidence of an aging population and the associated burdens on their health, and retirement systems, as well as on their workforce. When the so-called “pill” went on the market in the 1960s, the flourishing pharmaceutical industry offered it as a universal and safe method, free from side effects which they already knew of or “suspected.” However, the arrival of synthetic oral contraceptives has not been without risks to health. As any other drug, they possess not only therapeutic effects but also side effects (Sitruk-Ware and Nath 2013), which contraindicate their use in some patients. In fact, following the conference on “Metabolic Effects of Gonadal Hormones and Contraceptive Steroids” held in Boston in 1968, it was stated that, according to available data, no organ was free from the effects of the pill (Salhanick, Kipnis, and Vande Wiele 1969).

Glycemia constitutes a fundamental homeostatic variable, and hence its alteration can lead to a number of pathophysiological conditions affecting the internal milieu of the human being. Since the early 1960s, the intake of oral contraceptives has been associated with an increased risk of developing disorders of glucose metabolism (Waine et al. 1963). For that reason, the objective of this article is to review the main effects of the use of hormonal contraceptives on glycemic regulation.

Go to:Search for Bibliographic Information

Articles were searched for in the following bibliographic databases: PubMed, ISI Web of Knowledge, SCOPUS Database, SciELO, ScienceDirect, Google Scholar, and Google Books. Search languages used were English and Spanish; among the words used when searching were “oral contraceptives” and “glycemia,” “oral contraceptives” and “insulin resistance,” “oral contraceptives” and “diabetes,” “anticonceptivos orales” and “glicemia,” “anticonceptivos orales” and “resistencia insulínica,” and “anticonceptivos orales” and “diabetes.” Finally, twenty-four references on these topics were reviewed depending on their availability. In addition, sixteen other references were added to be used in the rationale and concluding remarks.

Go to:Studies Relating the Use of Oral Hormonal Contraceptives and the Impairment of Glycemic Regulation

In an interesting article, Shawe and Lawrenson (2003) have discussed the recommendation for the best practice when prescribing hormonal contraceptives in women, especially those suffering from glycemic disorders such as diabetes mellitus. They argue that there is little evidence that any changes in glycemic control caused by combined oral contraceptives are of clinical relevance. However, there are several studies that show the opposite. In the late 1960s a classic article written by Spellacy (1969) suggested that an abnormal carbohydrate metabolism in oral contraceptive users was characterized by impaired glucose tolerance. In this regard, Wynn et al. (1979) argued that even though there is evidence that estrogen and progestin1 oral contraceptives modify carbohydrate metabolism, the results of related studies are non-conclusive due to the scarce consideration given to subject selection and estrogen doses, and to doses and type of progestin used. In an investigation involving 2,205 women (1,628 of whom used combined estrogen/progestin contraceptives, and 577 did not), these researchers performed glucose tolerance tests on both user and non-user subjects (women using oral contraceptives were separated in six groups based on contraceptive composition), finding altered glucose tolerance in all groups of subjects using estrone progestin (nortestosterone derived) and gonane d-norgestrel (levonorgestrel) oral contraceptives. No changes were observed regarding glucose tolerance among subjects using pregnane progestin (progesterone derived). This research team also reported that women using the oral contraceptive with the highest estrogen level (75 µg or higher) presented the greatest glucose tolerance alteration. They also noted increased insulin release in all the groups except among users of contraceptives containing pregnane progestin, which showed no change (Wynn et al. 1979). Later, Skouby et al. (1985) studied the metabolic effects of a low-dose triphasic oral contraceptive (ethinyl estradiol and levonorgestrel) on glucose tolerance and plasma insulin response among other metabolic variables, in sixteen women with previous gestational diabetes and in nineteen healthy women. Investigations were performed prior to the hormonal intake and after intake for 2 and 6 months, using the oral glucose tolerance test. Before treatment, the women with previous gestational diabetes had significantly elevated fasting glucose and impaired glucose tolerance when compared to those of the healthy control women. Following the intake period, the glucose and insulin responses to oral glucose remained unchanged. Using the euglycemic clamp, the same research group compared this variable between six non-diabetic and six women suffering from previous gestational diabetes pre- and post-intake of a low-dose triphasic oral contraceptive (ethinyl estradiol and levonorgestrel) over a 6-month period. As a result, an increased insulin resistance was observed which was not sufficient to impair glucose tolerance either in the previous gestational diabetic women nor in the non-diabetic women; however, given its reduced sample, this study is not conclusive enough (Skouby et al. 1987). Pérez et al. (1987) studied glucose tolerance in 200 women taking oral hormonal contraceptives (ethinyl estradiol and norgestrel), grouped in cohorts of patients with and without cardiovascular risk. Even though no differences were found between the two groups, both evidenced significant differences when comparing, following an oral glucose overload, their basal glycemia before taking the contraceptive, after 6 months of use, and after a year of intake. In a cross-sectional study, Simon et al. (1990) studied 1,290 consecutive, healthy, non-pregnant women of child-bearing age. Compared with non-users taking no progestagens, oral contraceptive users had higher 2-h plasma glucose and higher fasting plasma insulin. These authors argued that oral contraceptive intake appears to induce an increase of insulin-resistance markers. Godsland et al. (1990)studied 1,060 women taking oral contraceptives (different progestin formulations: levonorgestrel, norethindrone, and desogestrel). These women were subjected to an oral glucose overload and, when comparing their metabolic variables with a four hundred and eighteen woman control group, it was observed that depending on the dose and type of progestin, combination drugs were associated with glycemias 43–61 percent higher than in controls, insulin responses 12–40 percent higher, and C-peptide responses 18–45 percent higher. Conversely, progestin-only formulations had only minor metabolic effects. Watanabe et al. (1994) studied one hundred and eighty-six women, fifty six of whom constituted the control group (they had never used oral contraceptives or at least had not used them during the last 2 years), sixty eight used them in low doses (contraceptive 1, 30 µg ethinyl estradiol and 300 µg norgestrel; and contraceptive 2, 30 µg ethinyl estradiol and 150 µg levonogestrel) and sixty two used a contraceptive in high doses (high-dose contraceptive, 50 µg ethinyl estradiol and 500 µg desogestrel); the last two groups had been using contraceptives for at least 6 months. Oral glucose tolerance tests were performed on all participants, and the results confirmed the development of impaired glucose tolerance in both pill groups, allowing for an estimation of insulin sensitivity and glucose effectiveness, as well as for beta-cell function. Low-dose users had lower insulin sensitivity and glucose effectiveness compared to controls and inappropriately low beta-cell function in relation to the insulin. High-dose contraceptive users, on the other hand, had metabolic variables that did not differ from controls. These researchers concluded that low-dose contraceptive use results in insulin and glucose resistance, which is not compensated by increased beta-cell function. The reduced glucose tolerance would be primarily due to the defect in glucose effectiveness, and these oral contraceptive users may be at risk of contracting diabetes or cardiovascular disease. In 1995, Shamma et al. (1995) used the hyperglycemic–hyperinsulinemic clamp in seven healthy, normally cycling, non-obese, non-diabetic women before and after using an implant contraceptive (36 mg of levonorgestrel) over an 8-week treatment, observing decreased insulin sensitivity and increased pancreatic insulin release as compensatory response, which might constitute a problem for diabetic patients; this study, nonetheless, cannot be considered conclusive based on its reduced sample size. Mastorakos et al. (2006) compared the effects of combined oral contraceptives containing cyproterone acetate or desogestrel on insulin sensitivity in adolescent girls with polycystic ovary syndrome. For that purpose, they compared a group of eighteen patients who received 0.15 mg of desogestrel plus 0.030 mg of ethinyl estradiol daily, and a group of eighteen patients who received 2 mg of cyproterone acetate plus 0.035 mg of ethinyl estradiol daily, for 21 days followed by a 7-day rest, for 1 year. All patients performed an oral glucose tolerance test before and after the 12-month treatment. These researchers found that, following a 1-year treatment, the homeostasis model assessment index of insulin resistance had increased significantly in both groups, concluding that treatment of adolescent girls with polycystic ovary syndrome with the two combined oral contraceptives administered, resulted in unfavorable changes of insulin sensitivity. In addition, these investigators found that cyproterone acetate is associated with increased insulin secretion and hyperinsulinemia. In an interesting work, Friedrich et al. (2012) studied the effect of combined oral contraceptives on the responsiveness of growth hormone. The contraceptive contained ethynil estradiol as estrogen, and levonorgestrel, desogestrel, norgestigmate, dienogest, or chlomardinon acetate as the progestin. These researchers found an enhanced responsiveness of the growth hormone to hyper- and hypoglycemia in women using the oral contraceptives (n = 15) as compared with the control subjects (without contraceptive, n = 10). According to the results, these authors recognize the effect of an increase in glucose levels attributed to oral contraceptives and even propose them as candidates to revert deep hypoglycemic episodes and hypoglycemia unawareness in women with diabetes in the future. A recent study lead by Piltonen et al. (2012) studied forty-two women (thirteen used oral contraceptives, fifteen used transdermal contraceptive patches, and fourteen used contraceptive vaginal rings). After continuous use over 9 weeks, fasting serum levels of glucose remained unchanged but the area under the curve values of glucose in oral glucose tolerance test rose significantly in all three study groups. Fasting serum levels of insulin increased significantly from baseline during the use of oral and vaginal contraceptives, and a similar trend was seen in the transdermal patch group. The area under the curve of insulin rose significantly during the use of oral and transdermal contraceptives, and there was a tendency to increase in the vaginal ring group. In conclusion, the results obtained by these authors demonstrate that commonly used contraceptives have some unfavorable effects on glucose metabolism.

Go to:Mechanisms that Could Explain Impaired Glycemic Regulation due to the Use of Hormonal Contraceptives

What causes hormonal contraceptives to have an effect on glycemic homeostasis? This could result from estrogens, progestins, or the molar concentration ratio of the administered estrogen–progestin. According to Alonso, Llaneza, and González (2008) several clinical and experimental data show that the physiological action of sex steroids and insulin interacts in the target tissues for these hormones. For example, the existence of high concentrations of sex steroids in women seems to contribute to the development of insulin resistance (Sutter-Dub 2002; Alonso, Llaneza and González 2008). Likewise, low plasma levels of the mentioned steroids, or high testosterone appear to increase the risk of developing type 2 diabetes. Although the close link between insulin resistance and plasma steroid levels seems clear, the nature of this relationship has not yet been sufficiently elucidated, especially in humans (Sutter-Dub 2002; Alonso, Llaneza, and González 2008). Sitruk-Ware and Nath (2013) argue that the estrogenic component of contraceptives exerts a relevant role in the alteration of insulin sensitivity. In this regard, studies in rats carried out by Nadal, Díaz and Valverde (2001) show that, at the level of beta cells in pancreatic islets, estrogens can modulate insulin secretion. Particularly, these researchers have reported that in the presence of glucose, estradiol enhances insulin secretion (i.e., an insulinotropic effect of estradiol) (Nadal et al. 1998). The latter confirms that, somehow, the estrogens contained in hormonal contraceptives can alter the dynamics of insulin secretion of the users. Along this line, González et al. (2002) investigated the influence of estradiol on the insulin receptor of ovariectomized rats treated with different hormonal doses. Their results showed that high doses of estradiol cause the carbohydrate mechanism to deteriorate and decrease insulin sensitivity, evidencing the relevance of estrogen dose and concentration for the glycosidic metabolism of women using oral hormonal contraceptives or undergoing hormone replacement. On this topic, Patiño, Díaz-Toledo, and del Barrio (2008)suggest that, in general, the changes detected on carbohydrate metabolism are dependent on ethinyl estradiol doses and on the androgenic effect of progestins (those prepared with 50 µg ethinyl estradiol have been described to lead to decreased glucose tolerance, which is compensated with higher insulin levels following an oral glucose overload (Skouby Petersen and Jespersen 1996)), and therefore there would be no hyperglycemia in healthy women (2008), though this leads to controversy. As regards progestagens, it has been reported that progesterone accelerates the progression of diabetes in female db/db mice (Picard et al. 2002). Moreover, female, but not male, mice in which the progesterone receptors (PR) have been knocked out (PR–/–), showed lower fasting glycemia than PR + /+ (intact receptors) mice and had higher insulinemia following a glucose injection. It was also found that pancreatic islets from female PR–/– mice were larger and secreted more insulin, due to increased beta-cell mass due to stimulated pancreatic beta cells. This shows the importance of progesterone in the signaling triggering insulin secretion, and also leads one to think that its progestin derivatives would also alter insulin release from the pancreas, even though the specific mechanisms are not clear as yet. On this matter, it has been suggested (Godsland et al. 1992; Sitruk-Ware and Nath 2013) that most progestins could bind and transactivate the PR, modifying the half-life of insulin and increasing insulin response to increased glucose, a fact dependent both on dosage, progestin molecular structure, and on its combination with estrogen. In the late 1970s, Wynn et al. (1979)proved that progestins decreased insulin sensitivity, thus causing insulin resistance, with the content of levonorgestrel (d-norgestrel) in combined contraceptives the strongest progestin to stimulate insulin secretion. Patiño, Díaz-Toledo and del Barrio (2008) suggested that the action mechanism of progestins on glycemic regulation could be due to the direct action on the pancreatic beta cell, maybe by modifying the insulin release rate (Howell, Tyhurst, and Green 1977), to a decrease in the number of insulin receptors at peripheral level, or to an alteration in the post-receptor response mechanisms, a fact leading to compensatory hyperinsulinemia. In this way, progestins would be acting through an “anti-insulin” effect, increasing peripheral resistance to insulin, causing reduced glucose utilization in the muscle and adipose tissue, but producing increased glycogen storage in the liver. The latter is in agreement with early studies by Spellacy, Buhi, and Birk (1975) showing that the progestin norethindrone (norethisterone) could affect the peripheral action of insulin. A study by Cagnacci et al. (2009) has shown that, in comparison with oral contraceptives, the vaginal ring presents no negative effect on insulin sensitivity and that, seemingly, progestins do not show the same effect on insulin response when non-orally administered. According to Patiño, Díaz-Toledo and del Barrio (2008), progestin androgenic activity exerts an important role on the diabetogenic effect of hormonal contraceptives, hence the importance of determining the selectivity index (i.e., the ratio between the wanted progestational response and the unwanted androgenic response at a given dose) associated with each compound. Even though the hormone formulations currently used in contraceptives contain lower estrogen doses (e.g., ethinyl estradiol) and third-generation progestins (e.g., desogestrel, gestodene, and norgestimate) possess a very low androgenic profile (there is no imbalance towards progestins), and hence their effect on glucose and insulin levels would be minimal (Patiño, Díaz-Toledo, and del Barrio 2008), the available information is still insufficient to rule out possible long-term effects of hormonal contraceptives on glycemia.

Go to:Concluding Remarks

The evidence discussed in this mini-review is sufficient to state that hormonal contraceptives exert some degree of influence on the mechanisms modulating glycemia. Currently, all physicians, and especially endocrinologists, are recommended to provide counseling for their patients who may have contraceptive prescriptions (Christin-Maitre 2013); however, from our perspective, little is being done to warn patients of the potential health risks involved. Thus, we consider it the duty and ethical responsibility of these health professionals, in the light of the risks linked to the use of hormonal contraceptives, to offer advice and guidance to users, and to warn them of the secondary effects these drugs involve. In this regard, we think that one of the fundamental principles of health professionals and educators should always be the promotion of health maintenance and the prevention of risky behaviors. It should always be kept in mind that to care for a patient as an individual, physicians, and other health professionals must recognize the patient as a person. Every professional is morally obliged to properly inform the patient; in fact, patients are legally entitled to be informed (Parra 2013). Health professionals should maintain an open dialogue with the patient and family regarding the potential risks that certain treatment and drugs (such as hormonal contraceptives) pose to their health. Health professionals must put the well-being of patients above their own, i.e., should prioritize the well-being of the patient. This primacy of patient well-being should be the guiding principle of health professionals. The altruism of these professionals, to generate confidence in patients, should be immune to any political and economic pressures they and their patients may be facing (Goldman and Dennis 2004). Peck and Norris (2012) argue that prescribing hormonal contraceptives without proper warning of its risks to the user violates the Hippocratic Oath to “do no harm.” In addition, these authors argue that while physicians ethically feel they cannot “impose” their own Catholic morality they should rightly insist that their patients be given opportune, adequate, and complete informed consent about all the risks of oral contraceptives. A real alternative, free from side effects, is the use of natural family planning based on fertility awareness, which includes the acceptance of one's fertility, and the shared responsibility of man and woman to live their own fertility in mutual confidence and cooperation (Fehring, Klaus, and Williams 2012), accepting it as a gift. In addition, fertility awareness can be very useful in the assessment of a woman's health (Vigil, Blackwell, and Cortés 2012).

In the future, research in this area ought to be focused on studying the long-term effect of hormonal contraceptives on glycemic homeostasis, on the basis of large, sufficiently representative sample sizes, and on performing reliable clinical tests suitable to determine alterations in insulin sensitivity among women using the contraceptives. We recommend the use of the euglycemic–hyperinsulinemic clamp, considered the gold standard in the assessment of insulin sensitivity (Greenfield et al. 1981), or alternatively the insulin suppression test (Pei et al. 1994), which is highly correlated (r = 0.93) with the euglycemic–hyperinsulinemic clamp (Greenfield et al. 1981), and has proved useful even in the identification of women subpopulations regarding insulin sensitivity (Vigil et al. 2007). In case of determining a deleterious effect on glycemic homeostasis, research should aim at identifying the underlying specific mechanisms altered at molecular, cellular, and physiological level.

Objective: The purpose of this study is to provide insight on the continuing high rate of unin- tended pregnancy among adult women.

Study design: Contracepting women were recruited while they waited for primary care appoint- ments. A total of 369 completed the baseline questionnaire, and 145 oral contraceptive (OC) users were enrolled in a 5-week, diary-based study of adherence and sexual activity.

Results: Most women who reported having discontinued OCs did so because of medical side effects, and most had switched to less effective methods. Among OC users, 26.4% had sexual intercourse on days they missed pills just before or after their placebo week. Nonadherence did not differ by socioeconomic factors or obesity.

Conclusion: Clinicians may need to encourage their patients to discuss their reasons for wanting to discontinue the use of an effective contraceptive method and assist them with their concerns or to switch to other effective methods to protect themselves from unintended pregnancy.

The National Survey of Family Growth (NSFG) provides researchers with important information on factors affecting pregnancy and women’s health among women age 15 to 44 years. Since 1982, in-depth interviews conducted by NSFG staff have increased the knowledge on a number of issues, including the use of contracep- tion.1 Between 1995 and the latest NSFG in 2002, the per- centage of sexually active women not using contraception increased from 5.2% in 1995 to 7.4%. This represents an apparent increase of 1.43 million women at risk for unintended pregnancies.1

Oral contraceptives (OCs) are the most commonly used non-permanent method of contraception in the United States.1 Among OC users, adherence to the dos- age regimen and frequency of intercourse are major de- terminants of unintended pregnancy.2 Though integral to studies of fertility and contraceptive effectiveness, including recent studies of a possible obesity-oral contraceptive failure association,3-6 information on these determinants is limited. The purpose of the Contraceptive History, Initiation, and Choice (CHIC) Study is 2-fold: first, to provide more insight into the de- mographic and lifestyle characteristics associated with contraceptive use and discontinuation, and second, to determine whether body mass index (BMI) is associated with adherence or frequency of sexual intercourse among OC users.

Material and methods Study population and design

The CHIC Study was conducted at a suburban Family Medicine clinic in the Atlanta area from May to Novem- ber 2004. The clinic is affiliated with Emory University and serves as the primary training area for residents of the Family Medicine Residency Program. The CHIC Study protocol was approved by the Emory University Institutional Review Board on April 24, 2004. Women between the ages of 18 and 45 years who were using any method of birth control were approached while waiting for their appointment (n = 413), and those who agreed to participate signed an informed consent form and filled out a short, baseline questionnaire (n = 369 by Septem- ber 2004).

OC users were invited to participate in a longitudinal study, in which the women were requested to fill out five 1-week diaries, in which they recorded daily information on whether they took their pill, the exact time they took their pill, and if they engaged in sexual intercourse. Women who agreed to participate received a coupon for a free movie rental for each diary they completed and mailed back to the study personnel. After September 2004, recruitment focused on enrolling OC users in the longitudinal portion of the study to increase sample size. Thus, during the remainder of the study period, only OC users between the ages of 18 and 45 years were asked to fill out the baseline questionnaire. An additional 12 women were recruited for the longitudinal study during this time. Of the 158 OC users, 13 (8.2%) declined to take part in the longitudinal study. Of the 145 women who agreed to participate in the longitudinal portion of the study, 98 (67.6%) returned at least 1 diary. Nonre- spondents were contacted by telephone up to 3 times, and finally by mail. Of the 47 nonrespondents, 6 indicated discontinuation of OCs as the reason for not partici- pating, whereas no information was available for the remaining women. Nonrespondents and women who declined to participate were less educated (25.0% at- tended graduate school vs 40.8% of respondents) and more likely to be of black race/ethnicity as compared with respondents (36.7% vs 19.4%). Respondents and nonrespondents did not differ by age (65.3% vs 65.0% !30 years), weight (44.9% vs 42.3% !150 lbs), or marital status (54.1% vs 56.7% single).

Measurement of covariates

The baseline questionnaire collected information on age, race/ethnicity, marital status, education, dual method use (use of an additional contraceptive method), prior con- traceptive method use, and reason for discontinuation of a contraceptive method. Diaries collected information on the following variables: income, smoking, reason(s) for using OCs, length of time using OCs, parity, future birth intention, history of abortion, and whether the diary week was typical in terms of family, work, and social responsibilities. Information on height and weight were abstracted from medical records.

Definition of adherence and sexual intercourse outcomes

Adherence during a week was defined as missing no pills during an active week. Diary information was used to create a dichotomous sexual intercourse variable for each week.

Analysis

Summary statistics of the demographic and lifestyle characteristics of participants were calculated. For con- traceptive method categories with sufficient numbers, characteristics of these specific users were further inves- tigated. Unadjusted odds ratios (ORs) and 95% CIs were obtained to examine the association between BMI and the adherence with an OC regimen and frequency of intercourse outcomes. For the adherence analyses, weeks during which women took placebo pills were excluded. Because the data include repeated measures for each woman, a generalized estimating equations approach was used to carry out logistic regression for correlated responses for the weekly adherence and sexual inter- course outcomes.7 Multivariate logistic regression for correlated responses was used to further explore the rela- tionship between obesity and the adherence and sexual intercourse outcomes. Only those variables that changed point estimates by 10% or more were included in the final model as confounders.8 All analyses were performed with the SAS System for Windows Version 8.2 (SAS Institute, Cary, NC).

Results

Of the 369 women who were initially recruited into the CHIC Study, the majority were 26 to 35 years old (mean: 29.3), single, and highly educated (Table I). Most partic- ipants were of race/ethnicity white or black (48.8% and 40.9%, respectively) and nearly half had BMIs in the overweight or obese range (mean BMI = 27.2). The most popular method of contraception was OCs (39.6%), followed by male condoms (21.4%), and tubal ligation (7.6%). The majority of women were not dual...

https://www.ajog.org/article/S0002-9378(05)02601-3/pdf

Abstract

Objective: To conduct a systematic review and meta-analysis of the effect of oral contraceptive use on plasma and red blood cell (RBC) folate concentrations .

Methods: We searched Medline, EMBASE, Web of Science, and the Cochrane library for human studies published from inception to June 2013 evaluating oral contraceptive use and folate status . Case–control studies, cohort studies, and clinical trials were included . A random-effects model of outcomes was used for the meta-analysis .

Results: A total of 2831 women in 17 studies were included in the analysis . In those whose plasma folate concentrations were available, there was a significant folate-lowering effect of oral contraceptives observed (mean reduction 1 .27 μg/L; 95% CI 1 .85 to 0 .69, P < 0 .001) . Similarly, after analyzing data from 1389 women in 12 studies whose RBC folate concentrations were available, significantly lower folate status was observed among oral contraceptive users (mean reduction 59 .32 μg/L; 95% CI 58 .03 to 23 .04, P < 0 .001) .

Conclusion: Because of the reduction in blood folate concentrations associated with the use of oral contraceptives, it is critical for women of childbearing age to continue folate supplementation during oral contraceptive use .

Abstract

BACKGROUND

Combined oral contraceptives (COCs) reduce levels of androgen, especially testosterone (T), by inhibiting ovarian and adrenal androgen synthesis and by increasing levels of sex hormone-binding globulin (SHBG). Although this suppressive effect has been investigated by numerous studies over many years, to our knowledge no systematic review concerning this issue had been performed. This systematic review and meta-analysis was performed to evaluate the effect of COCs on concentrations of total T, free T and SHBG in healthy women and to evaluate differences between the various types of COCs (e.g. estrogen dose, type of progestin) and the assays used to assess total T and free T.

METHODS

A review of the literature was performed using database searches (MEDLINE, EMBASE and the Cochrane Central Register of Clinical Trials) and all publications (from inception date until July 2012) investigating the effect of COCs on androgen levels in healthy women were considered eligible for selection. Three reviewers were involved in study selection, data extraction and critical appraisal. For the meta-analysis, data on total T, free T and SHBG were extracted and combined using random effects analysis. Additional subgroup analyses were performed to evaluate differences between the various types of COCs (e.g. estrogen dose, type of progestin) and the assays used to assess total T or free T.

RESULTS

A total of 151 records were identified by systematic review and 42 studies with a total of 1495 healthy young women (age range: 18–40 years) were included in the meta-analysis. All included studies were experimental studies and 21 were non-comparative. Pooling of the results derived from all the included papers showed that total T levels significantly decreased during COC use [mean difference (MD) (95% confidence interval, CI) −0.49 nmol/l (−0.55, −0.42); P < 0.001]. Significantly lower levels of free T were also found [relative change (95% CI) 0.39 (0.35, 0.43); P < 0.001], with a mean decrease of 61%. On the contrary, SHBG concentrations significantly increased during all types of COC use [MD (95% CI) 99.08 nmol/l (86.43, 111.73); P < 0.001]. Subgroup analyses revealed that COCs containing 20–25 µg EE had similar effects on total and free T compared with COCs with 30–35 µg EE. In addition, suppressive effects on T levels were not different when comparing different types of progestins. However, subgroup analyses for the estrogen dose and the progestin type in relation to changes in SHBG levels did show significant differences: COCs containing second generation progestins and/or the lower estrogen doses (20–25 µg EE) were found to have less impact on SHBG concentrations.

CONCLUSIONS

The current literature review and meta-analysis demonstrates that COCs decrease circulating levels of total T and free T and increase SBHG concentrations. Due to the SHBG increase, free T levels decrease twice as much as total T. The estrogen dose and progestin type of the COC do not influence the decline of total and free T, but both affect SHBG. The clinical implications of suppressed androgen levels during COC use remain to be elucidated.

Abstract

Statins (3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors) are currently the most effective medications for reducing low-density lipoprotein cholesterol concentrations. Although generally safe, they have been associated with a variety of myopathic complaints. Statins block production of farnesyl pyrophosphate, an intermediate in the synthesis of ubiquinone or coenzyme Q10 (CoQ10). This fact, plus the role of CoQ10 in mitochondrial energy production, has prompted the hypothesis that statin-induced CoQ10 deficiency is involved in the pathogenesis of statin myopathy. We identified English language articles relating statin treatment and CoQ10 levels via a PubMed search through August 2006. Abstracts were reviewed and articles addressing the relationship between statin treatment and CoQ10 levels were examined in detail. Statin treatment reduces circulating levels of CoQ10. The effect of statin therapy on intramuscular levels of CoQ10 is not clear, and data on intramuscular CoQ10 levels in symptomatic patients with statin-associated myopathy are scarce. Mitochondrial function may be impaired by statin therapy, and this effect may be exacerbated by exercise. Supplementation can raise the circulating levels of CoQ10, but data on the effect of CoQ10 supplementation on myopathic symptoms are scarce and contradictory. We conclude that there is insufficient evidence to prove the etiologic role of CoQ10 deficiency in statin-associated myopathy and that large, well-designed clinical trials are required to address this issue. The routine use of CoQ10 cannot be recommended in statin-treated patients. Nevertheless, there are no known risks to this supplement and there is some anecdotal and preliminary trial evidence of its effectiveness. Consequently, CoQ10 can be tested in patients requiring statin treatment, who develop statin myalgia, and who cannot be satisfactorily treated with other agents. Some patients may respond, if only via a placebo effect.

Statins

Statins or 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors competitively inhibit HMG-CoA reductase thereby decreasing synthesis of mevalonate, a critical intermediary in the cholesterol synthesis pathway (Fig. 1).Their most serious and frequent side effects are a variety of myopathic complaints ranging from mild myalgia to fatal rhabdomyolysis. The incidence of statin-associated fatal rhabdomyolysis is only 1.5 deaths per 10 million prescriptions (1). The mechanism of statin myopathy is unknown, but possible mechanisms include decreased sarcolemmal cholesterol (2), reduction in small guanosine triphosphate-binding proteins (2), increased intracellular lipid producing a lipid myopathy (3, 4), increased myocellular phytosterols (5), and mitochondrial dysfunction possibly from reduced intramuscular coenzyme Q10(CoQ10) (2, 3). The present review examines the evidence that CoQ10 is an etiologic factor in statin myopathy.

CoQ10

Coenzyme Q10 was discovered by Crane et al. (6) in 1957. Coenzyme Q10 is a naturally occurring, fat-soluble quinone that is localized in hydrophobic portions of cellular membranes. Approximately half of the body’s CoQ10 is obtained through dietary fat ingestion, whereas the remainder results from endogenous synthesis (7). Coenzyme Q10 participates in electron transport during oxidative phosphorylationin mitochondria, protects against oxidative stress produced by free radicals (8), and regenerates active forms of the antioxidants ascorbic acid and tocopherol (vitamin E) (9, 10). Statins block production of farnesyl pyrophosphate, an intermediate in the production of CoQ10 (Fig. 1). This fact plus the role of CoQ10 in mitochondrial energy production and the importance of mitochondria in muscle function has prompted the hypothesis that statin-induced CoQ10 deficiency participates in statin-associated myopathy.

Methods

English language articles relating statin treatment and CoQ10 levels were identified via a PubMed search through August 2006 and from reference citations in other articles. The PubMed search was performed using various combinations of the terms myopathy, rhabdomyolysis, statin, HMG-CoA reductase inhibitor, CoQ10, ubiquinone, and skeletal muscle. Abstracts were reviewed, and articles addressing the relationship between statin treatment and CoQ10 levels were examined in detail.

Effect of Statins on Circulating CoQ10 Levels

Statins have been known to reduce circulating CoQ10 levels in animal models (11) and humans (12) since at least 1990. Since that time and with rare exceptions (13, 14) at least 9 observational studies (12, 13, 15, 16, 17, 18, 19, 20, 21, 22, 23) and 6 randomized controlled trials (7, 24, 25, 26, 27, 28) have demonstrated that statins reduce plasma/serum levels of CoQ10 16% to 54% (Table 1).The largest trial (25) included 1,049 patients and noted reductions in plasma CoQ10 levels of 38% and 27% after treatment with atorvastatin 10 mg/day to 20 mg/day or lovastatin 20 mg/day to 40 mg/day, respectively. The decrease in blood CoQ10 levels with statintreatment is probably due to reductions in lower-density lipoproteins. Coenzyme Q10 is transported in low-density lipoprotein (LDL) (58 ± 10% of serum CoQ10), high-density lipoprotein (26 ± 8%), and very low-density lipoprotein + intermediate-density lipoprotein (16 ± 8%) particles (29). Normalizing CoQ10 concentrations for the reduction in LDL cholesterol or total cholesterol suggests that there is no change in CoQ10 lipoprotein particle concentration. For example, treatment with simvastatin 20 mg/day for 4 weeks (16) reduced serum total cholesterol and CoQ10 levels by 26% and 31%, respectively (p < 0.001), but the ratio of serum CoQ10 to total cholesterol decreased only 9% (p = NS). Similar apparently parallel reductions in total cholesterol have been reported by at least 7 other authors using a variety of statins (17, 18, 19, 24, 25, 26, 27) although 2 trials have not found reductions in circulating CoQ10 levels with statin treatment (14, 30).

The hypothesis that a combination of ezetimibe with simvastatin would counteract the statin-induced decrease in circulating CoQ10 was tested in a recent trial (20). Ezetimibe inhibits intestinal absorption of dietary cholesterol and increases endogenous cholesterol synthesis. Seventy-two healthy men were randomized to receive either ezetimibe 10 mg/day, simvastatin 40 mg/day, or both for 2 weeks. Ezetimibe alone did not reduce plasma CoQ10 concentrations (+1.1 ± 21%). The combination of simvastatin and ezetimibe reduced plasma CoQ10 levels as did simvastatin alone (−28 ± 12% and −16 ± 16%, respectively). The decrease in LDL cholesterol correlated with the decrease in plasma CoQ10 levels in all 3 groups (R = 0.67, p < 0.0001). In addition, the ratios of plasma CoQ10 to LDL cholesterol concentrations were increased in all groups (p < 0.0001).

Few studies have examined the effect of statins on nonlipoprotein CoQ10 levels in the circulation, although at least 1 study (24) reported 12.5% reductions in platelet CoQ10 levels. The mechanism and significance of this finding are unclear, however, because platelets do not contain mitochondria.

Effect of Statins on Skeletal Muscle CoQ10 Levels

Animal studies document that statins can reduce ubiquinone levels in cardiac muscle and liver (11, 31, 32, 33). If CoQ10 deficiency contributes, at least in part, to statin-associated myopathy, CoQ10 levels in skeletal muscle should also be reduced, but data from animal models are inconsistent. Whereas simvastatin or pravastatindecreased skeletal muscle ubiquinone up to 72% (p < 0.005) when administered to rabbits, other studies using rabbits do not support these results (34). High-dose statin treatment (50 mg/kg of simvastatin or pravastatin per day for 14 days) did not reduce skeletal muscle concentrations of ubiquinone in animals treated with either drug. Severe lesions in skeletal muscles developed in the simvastatin-treated rabbits, despite the absence of decreases in muscle ubiquinone levels (32). A recent animal study treated rats with various doses of cerivastatin for 15 days and demonstrated no significant difference, in most cases, between skeletal muscle mean ubiquinone levels in statin-treated animals and nontreated control animals (35).

In humans, low-dose statin treatment does not appear to reduce intramuscular CoQ10 concentrations (Table 2).In the first human study to address this issue, skeletal muscle CoQ10 concentrations were actually 47% higher (p < 0.001) after 4 weeks of treatment with 20 mg of lovastatin daily.

Longer treatment, also with simvastatin 20 mg daily, yielded similar results (18). Coenzyme Q10 levels in muscle obtained from hypercholesterolemic men at baseline and after 6 months of statin therapy showed virtually identical CoQ10 concentrations. These values were similar to those measured in 15 normolipidemic untreated subjects before and after 6 months of observation (88 vs. 95 μmol/kg, respectively).

The effect of statins on muscle CoQ10 may be drug and dose dependent. In a recent trial using simvastatin 80 mg/day, atorvastatin 40 mg/day, or placebo for 8 weeks (5), mean muscle concentrations of CoQ10 in the simvastatin-treated patients decreased by 34%. To date, this is the only trial comparing intramuscular CoQ10 levels among the statins.

There are few published studies on intramuscular CoQ10 levels in subjects symptomatic from statin myopathy. Almost half (47% or 17 of 36 patients) of patients referred for myopathic complaints presumably due to statin-associated myopathy had intramuscular CoQ10 levels >2 standard deviations below the mean (G. Vladitu, personal communication, August 26, 2005). It is not clear, however, whether these decreased intramuscular CoQ10 levels produced statin myopathy or were simply associated with a reduction in mitochondrial volume associated with the statin myopathy itself (5) or associated with physical inactivity due to myopathic symptoms. A recent study (36) examined intramuscular CoQ10 levels in patients with statin-associated myopathy and found levels 2 standard deviations below the normal mean in 3 patients, below 1 standard deviation in 7 patients, normal levels in 4 patients, and increased levels in 4. No evidence of myocyte apoptosis was found, using terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick-end labeling assay and immunohistochemical studies with antibodies against Bax, Bcl-2, and caspase-3, although only 11 biopsies had enough tissue to perform the test.

Effect of Statins on Mitochondrial Function

If reduced CoQ10 levels mediate statin myopathy, not only should there be evidence of reduced intramuscular CoQ10 levels, but there should also be evidence of impaired mitochondrial function. One animal study (37) showed decreased phosphorylation potential of adenosine diphosphate in the cardiac mitochondria of guinea pigs treated with lovastatin 20 mg/kg twice daily. This correlated with a 37% and 31% decrease in CoQ10 concentration in cardiac mitochondria and myocardium, respectively. Curiously, these changes were observed in the older animals (2 years old), but not younger ones (2 to 4 months old), suggesting a possible age effect. Similarly, myocardium of dogs treated with simvastatin 2 mg/kg/day for 3 weeks contained lower concentrations of CoQ10 and decreased mitochondrial respiration during ischemia induced by left anterior descending coronary artery ligation, compared with untreated control patients or a pravastatin-treated group (38). In rats, pravastatin reduced mitochondrial complex I activity in the diaphragm and psoas major muscles in 35- to 55-week-old animals and decreased complex IV activity in 55-week-old animals (39). Various doses of cerivastatin (0.1, 0.5, or 1.0 mg/kg/day for 15 days) administered to rats did not produce morphologic abnormalities in skeletal muscle after 10 days of statin treatment (35). At 15 days, inflammation and necrosis with structurally altered mitochondria involving type II myofibers was observed in the mid/high dose groups. However, mitochondria in the unaffected adjacent myocytes appeared normal, prompting the authors to conclude that mitochondrial damage did not precede myofiber degeneration. No significant changes in mitochondrial function were observed, and mitochondrial function did not correlate with changes in ubiquinone concentration. This study supports an earlier trial in rats (40) in which simvastatin treatment did not result in a significant change in ATP production in muscle, heart, liver, or brain, despite lower blood ATP levels.

Few studies have directly or indirectly addressed this issue in humans. de Pinieux et al. (21) examined the ratio of serum lactate to pyruvate as an indirect measure of mitochondrial function. Statin-treated hypercholesterolemic patients had 16% higher fasting, nonexercise, lactate:pyruvate ratios than untreated hypercholesterolemic patients (p < 0.05). In contrast mean lactate:pyruvate ratios with fibrates were not significantly higher than those of the untreated hyperlipidemic patients. Curiously, regardless of statin, fibrate, or no treatment, mean lactate:pyruvate ratios were higher in the hypercholesterolemic patients than in the normocholesterolemic control patients.

Others have also suggested that mitochondrial function may be impaired by statin therapy. Muscle biopsies from 4 patients with statin-associated myopathy, despite normal creatine kinase (CK) levels, demonstrated findings consistent with mitochondrial dysfunction, including increased intramuscular lipid, diminished cytochrome oxidase staining, and ragged red fibers (3). Three of these patients had repeat biopsies performed after discontinuation of statin, which showed resolution of the pathologic abnormalities. These same investigators used respiratory exchange ratios (RER = V̇co2/V̇o2) during exercise as a measure of exercise aerobic metabolism and an indirect measure of mitochondrial function (41). Sixteen healthy subjects received atorvastatin 5 to 10 mg for 6 weeks. Exercise results in these subjects were compared with those from 9 men and 2 women who developed muscle complaints plus CK elevations during statin alone (n = 6), statin and fibrate (n = 3), or statin and niacin (n = 2) therapy. There was no effect of statin therapy on maximal oxygen uptake (V̇o2max), a measure of aerobic fitness, in the normal patients, but V̇o2max was significantly lower in the myopathic patients, a finding the authors attributed to these patients’ physical inactivity due to their myopathy. Interestingly, the resting RER increased with statin therapy in the normal subjects (0.75 ± 0.02 to 0.86 ± 0.06), and was also elevated in the myopathic patients off statin therapy (0.90 ± 0.07). Respiratory exchange ratio decreases with fat oxidation and increases as carbohydrate is used as fuel, but is a crude measure of these processes. The authors interpreted their findings as demonstrating that statins impair fat metabolism in healthy patients and that victims of statin myopathy have a pretreatment abnormality in fat metabolism that is exacerbated by statin therapy leading to the muscle complaints. There was no apparent change in the onset of lactate accumulation or “anaerobic threshold” during the exercise test, however, which argues against an alteration in exercise fat metabolism with statin treatment. One additional study of 195 noninsulin diabetic patients noted a 6% increase in resting RER (0.78 to 0.83), which the authors attributed to improved glucose metabolism with statin treatment, but exercise parameters were not measured in that study (42). Consequently, these exercise results raise the possibility of mitochondrial dysfunction during exercise, a problem that could relate to CoQ10 depletion.

In contrast to these studies suggesting a mitochondrial problem in statin myopathy (13, 21, 22, 23, 41, 42, 43, 44, 45, 46, 47, 48), muscle biopsies from 18 patients with statin-associated myopathy found only 2 patients with decreased intramuscular levels of CoQ10 and some morphologic evidence of mitochondrial dysfunction (36). Muscle biopsies from these patients revealed the presence of a few ragged red fibers or cytochrome coxidase-negative fibers. It is not clear whether these changes were produced by the depletion of CoQ10 or were related to aging, given that both patients were older than 60 years old. In addition, the activity of complex III of the mitochondrial respiratory chain, which is dependent on CoQ10 activity, was normal in all participants.

Others have directly measured concentrations of high energy phosphates, including adenosine triphosphate and creatine phosphate, in the skeletal muscle of statin-treated patients and found no changes despite 6 months of treatment with 20 mg/day of simvastatin, suggesting that energy supply to the muscle was not compromised (18). These studies were performed at rest, however, and more subtle differences may be obvious with exercise or during exercise recovery.

CoQ10 Supplementation

A number of studies (12, 13, 24, 30) have demonstrated that CoQ10 administration can increase CoQ10 blood levels in patients treated with statins. Coenzyme Q10supplementation may also reduce the symptoms of statin-induced myopathy in patients treated with massive statin doses (43, 44). Lovastatin was investigated as a potential cancer treatment in doses up to 45 mg/kg body weight. Among 56 cancer patients treated with monthly 7-day courses of high-dose (2 mg/kg/day to 45 mg/kg/day) lovastatin, over the course of 2 and a half years, CoQ10 supplementation (240 mg/day) did not reduce the frequency of lovastatin-induced myopathy compared with that in unsupplemented patients, but did decrease the severity of myopathic symptoms, judged by a grading scale that included duration of myalgiaand degree of CK elevations. Interestingly, patients receiving lovastatin at doses <25 mg/kg/day did not develop musculoskeletal toxicity, and at higher doses there was no correlation between the incidence of myotoxicity and the dose of lovastatin (p = 0.24) (43). Another trial (44) treated 16 patients with advanced gastric adenocarcinoma with lovastatin 35 mg/kg/day for 7 consecutive days. All subjects received 240 mg of CoQ10 daily. Only 2 patients developed increased CK levels, myalgia, and muscle weakness. The authors reported that symptoms were successfully managed with CoQ10 and symptomatic treatment. The dose of CoQ10 required for this treatment was not specified nor was a control group without supplementation included.

There are only 2 randomized trials, both in abstract form, that were designed to evaluate CoQ10 as a treatment for statin-associated myopathy (45, 46). The first treated 41 patients with statin-associated myalgia with either 400 IU of vitamin E or 100 mg of CoQ10 daily for 30 days. Preliminary results suggest a significant improvement in pain scores in patients treated with CoQ10, with 18 of 21 reporting improvement in symptom severity and a reduction in mean pain scores (6.2 ± 1.7 to 3.1 ± 2.2 at baseline using a 10-point scale (p < 0.001). By comparison, only 3 of 20 patients on vitamin E reported improvement in their symptoms, and there was no change in this group’s mean pain score (3.9 ± 2.2 vs. 3.1 ± 2.2, p = NS). These results require confirmation.

A more recent trial (46) randomized 44 dyslipidemic patients with prior statin-induced myalgia to 12 weeks of treatment with escalating doses of simvastatin (10 mg/day to 40 mg/day) and CoQ10 200 mg/day or placebo. Plasma CoQ10 levels increased with CoQ10 supplementation, but there were no differences in myalgia scores (p = 0.63) or in statin tolerance between the 2 treatment groups. Specifically, similar numbers of patients in both groups were able to tolerate simvastatin 40 mg (p = 0.34) and 10 mg (p = 0.35) daily.

Conclusions and Clinical Implications

Statins are the most effective medications for reducing LDL cholesterolconcentrations. They have been proven to decrease the incidence of adverse cardiac events in diverse patient populations. The primary adverse effect limiting their use is myopathy, ranging from benign myalgias to rare cases of fatal rhabdomyolysis. Statins’ interference with the production of CoQ10 prompted the hypothesis that CoQ10 deficiency may play a role in statin-associated myopathy. Although statins reduce circulating levels of CoQ10, this effect is nullified by normalizing CoQ10 concentrations for the reduction in LDL cholesterol or total cholesterol. Low-dose statin treatment does not appear to reduce intramuscular levels of CoQ10 in humans. Few data are available regarding intramuscular CoQ10 levels in symptomatic patients with statin-associated myopathy. Mitochondrial function may be impaired by statin therapy, and this effect may be exacerbated by exercise, but confirmatory data are needed. Animal models of statin myopathy demonstrate similar results in that decreases in skeletal muscle ubiquinone levels and mitochondrial function are not consistent and skeletal muscle injury can occur without decreases in muscle CoQ10 concentrations. Supplementation can raise the circulating levels of CoQ10, but whether or not this relieves myopathic symptoms is not clear. The 2 available double-blind studies report contrasting results and are only available in abstract form. Presently, insufficient evidence exists to implicate CoQ10 deficiency as the cause of statin-associated myopathy. Our evaluation of this data is that intramuscular CoQ10 depletion does not cause statin myopathy and that CoQ10 supplementation does not mitigate the symptoms of statin myopathy. Nevertheless, there are no known risks to this supplement. Also, reports from cancer trials using high-dose statins (44, 45), results from 1 of the 2 available clinical trials (45, 46), and clinical anecdotes suggest that some patients benefit from CoQ10 supplementation. This variable response could be explained by a placebo effect or by genetic and other variability among patients. For example, there are at least 2 reported cases of mitochondrial encephalopathy lactic acidosis and stroke-like episodes temporally related to statin therapy that appeared to benefit from CoQ10 therapy (47, 48). Consequently, although there is no definite evidence of its effectiveness, CoQ10 supplementation, 200 mg daily, can be trialed in patients requiring statin treatment, who develop statin myalgia, and who cannot be satisfactorily treated with other agents. Some patients may respond if only via a placebo effect.

Abstract

Background

Aggressive lipid lowering with high doses of statins increases the risk of statin-induced myopathy. However, the cellular mechanisms leading to muscle damage are not known and sensitive biomarkers are needed to identify patients at risk of developing statin-induced serious side effects.

Methodology

We performed bioinformatics analysis of whole genome expression profiling of muscle specimens and UPLC/MS based lipidomics analyses of plasma samples obtained in an earlier randomized trial from patients either on high dose simvastatin (80 mg), atorvastatin (40 mg), or placebo.

Principal Findings

High dose simvastatin treatment resulted in 111 differentially expressed genes (1.5-fold change and p-value<0.05), while expression of only one and five genes was altered in the placebo and atorvastatin groups, respectively. The Gene Set Enrichment Analysis identified several affected pathways (23 gene lists with False Discovery Rate q-value<0.1) in muscle following high dose simvastatin, including eicosanoid synthesis and Phospholipase C pathways. Using lipidomic analysis we identified previously uncharacterized drug-specific changes in the plasma lipid profile despite similar statin-induced changes in plasma LDL-cholesterol. We also found that the plasma lipidomic changes following simvastatin treatment correlate with the muscle expression of the arachidonate 5-lipoxygenase-activating protein.

Conclusions

High dose simvastatin affects multiple metabolic and signaling pathways in skeletal muscle, including the pro-inflammatory pathways. Thus, our results demonstrate that clinically used high statin dosages may lead to unexpected metabolic effects in non-hepatic tissues. The lipidomic profiles may serve as highly sensitive biomarkers of statin-induced metabolic alterations in muscle and may thus allow us to identify patients who should be treated with a lower dose to prevent a possible toxicity.

Introduction

Large-scale clinical trials have shown that statins are effective and safe cholesterol lowering drugs [1]–[3]. Recently more patients have been titrated to higher doses of statins in order to reach the new goals of LDL-cholesterol lowering and achieve even greater reductions of atherosclerotic complications. However, aggressive treatment with high dosages increases the risk of statin-induced myopathy [4]. Elucidation of myopathy mechanisms and identification of patients likely not to tolerate the treatment is therefore of great clinical interest. In addition, comparison of different statin drugs used for aggressive treatment is essential. The currently used statins do have clear differences for instance in their pharmacokinetic properties [5], therefore it is likely that some of the statins at high dosages are more prone to have unexpected and unwanted effects in non-hepatic tissues.

We do know that some diseases such as hypothyroidism, liver dysfunction and diabetes increase the risk of muscle complications due to statin treatment [6]. Furthermore, exercise, alcohol, infections or underlying metabolic diseases seems to exacerbate this risk [7]. Under these circumstances development of myopathy may be exacerbated by interactions with statins [8]. Mukhtar and Reckless listed four potential myopathy mechanisms in their recent review: Depletion of intracellular cholesterol leading to calcium influx; inhibited protein synthesis, signal transduction and metabolism due to decreased mevalonate acid and its metabolite concentrations; reduced ubiquinone (coenzyme Q10) concentrations; and enhanced apoptosis [8]. Muscle biopsies obtained from patients with statin-induced myopathy without creatine kinase (CK) elevations have shown evidence of mitochondrial dysfunction, including abnormally increased lipid stores in muscles [9]. We observed decreased mitochondrial function in patients on high dose simvastatin treatment, with no signs of myopathy [10]. Later, we confirmed that high dose (80 mg) simvastatin affects muscle mitochondria by assessing a significant decrease in the muscle mitochondrial DNA (mtDNA) content during treatment (Schink et al, submitted). Thus, statins are causing unwanted mitochondrial effects and defective mitochondrial metabolism may already be involved during the early development of statin-induced myopathy when the currently used plasma CK measurements are not sensitive enough to identify these patients at risk of developing muscle damage.

Given the existing gap in knowledge and understanding of statin-induced muscle damage, we embarked in a systems biology approach aiming to gain insight into the mechanism and potential biomarkers of myopathy in the clinical setting. Of particular relevance is to gain new knowledge of signaling and metabolic pathways in muscle involved in myopathy and early molecular markers applicable in clinical setting. To address both objectives, gene expression profiling of muscle tissue is an obvious strategy of choice. As changes in plasma lipid composition are of particular interest in the study of statins, plasma lipidomics is one possible option to address the latter. Recent advances in liquid chromatography and mass spectrometry have empowered us with ability to reliably measure hundreds of lipid molecular species from biological samples in parallel [11], [12].

In this paper we report the study of muscle gene expression profiles in combination with plasma lipidome analysis before and during high dose statin treatment. The specimens were from patients who participated in our earlier controlled and randomized study comparing placebo and high dosages of atorvastatin and simvastatin [10]. The simvastatin treated patients were particularly unique and suitable for sensitive early marker discovery due to changes in their mitochondrial function, mtDNA and ubiquinone concentration in muscle. The systems biology approach allowed us to compare two widely used statins in terms of their effects on muscle gene expression and plasma lipidome. In addition, we were able to illuminate relevant biological pathways and to identify biomarker candidates related to unwanted and potentially toxic statin-induced changes in muscle metabolism.

MethodsPatients

Plasma samples from 37 subjects of an earlier study [10], focusing on the effect of high dose statin treatment on skeletal muscle metabolism, were used for plasma lipidome analysis; placebo (N = 11), simvastatin (N = 13), and atorvastatin (N = 14). The subjects aged between 45 and 69 years and their average serum total cholesterol concentration was 5.8±0.9 mmol/L and serum triglycerides below 4.5 mmol/L. Muscle specimens from eighteen age matched men being treated either with atorvastatin (n = 6), simvastatin (n = 6) or placebo (n = 6) were selected for genome wide expression analysis. Clinical parameters are available as Supporting Information Dataset S1 and Table S1.

The study patients had never been treated with statins before. They were instructed to adhere to their normal diet during the study. Patients with familial hypercholesterolemia and patients with serum total cholesterol>7.0 mmol/L in the initial screening were excluded. Other exclusion criteria were: use of concurrent lipid altering medication or antioxidant vitamins, renal or hepatic dysfunction, and use of medication known to affect metabolism of atorvastatin or simvastatin. The study protocol was accepted by the Ethics Committee of the Tampere University Hospital and written informed consents were obtained from all participants.

Design

The original study was a randomized, double blind and placebo-controlled trial with three treatment groups: placebo, atorvastatin 40 mg/day, and simvastatin 80 mg/day. Placebo was simvastatin-matched, and to ensure also blinding of atorvastatin, all study drugs were supplied in sealed, identical, numbered containers. The duration of the follow-up was eight weeks. Muscle biopsies were obtained at baseline and at the end of the treatment period. Biopsies were taken from the lateral portion of the quadriceps femoris muscle in local anesthesia at about the mid-point between the greater trochanter and the knee joint with a biopsy needle (Tru-Cut, Baxter, McGaw Park, Ill., USA). The muscle specimens were frozen within 1–2 seconds in liquid nitrogen and stored at −80°C until analyzed. The blood sampling was performed in the Department of Clinical Chemistry, Tampere University Hospital by an experienced laboratory technician. Venous blood was drawn from the antecubital vein in sitting position after a twelve-hour fast and after 15 minute rest just before blood sampling. The blood was drawn into tubes containing EDTA, and plasma was separated after cooling by centrifugation at 2000 rpm for 10 minutes. The samples were stored at −70° until analyzed. Investigators performing the gene expression and lipidomics analyses were blinded until the analyses were done.

Gene expression analyses

Microarray experiments were performed using Sentrix® Human-6 Expression BeadChips, analyzing over 46 000 known genes, gene candidates and splice variants (Illumina, San Diego, CA, USA) according to given instructions. The biopsy samples were homogenized using Ultra-Turrax (IKA Turrax T8/S8N-5G, IKA-Werke, Staufen, Germany). The total RNA was extracted using TRIzol (#15596-018, Invitrogen Corporation, Carlsbad, CA), DNase treatment and a second RNA purification by Qiagen kits (#74106, and, #79254, Qiagen GmbH, Hilden, Germany), all by given instructions.

A 200 ng aliquote of total RNA from each sample were amplified to cDNA using Ambion's Illumina RNA Amplification kit following the instructions (cat no I1755, Ambion, Inc., Austin, TX, USA). In vitro transcripiton (IVT) reaction of cDNA to cRNA was performed overnight (14h) including biotin-11-dUTP (PerkinElmer, cat no PC 3435-0402-Biotin-11-dUTP, >95%, NEL539001EA, PerkinElmer Life And Analytical Sciences, Inc., Boston, MA, USA) for labelling the cRNA product. Both before and after the amplifications the RNA/cRNA concentrations were checked with Nanodrop ND-1000 spectrophotometer (Nanodrop Technologies, Wilmington, DE, USA) and RNA/cRNA quality was controlled by BioRad's Experion Automated Electrophoresis System and RNA StdSens Analysis Kit (BioRad Laboratories, Inc., Hercules, CA, USA).

1500 ng of each sample cRNA was hybridized to Illumina's Sentrix® Human-6 Expression BeadChip arrays (Illumina, Inc., San Diego, CA, USA) at 55°C overnight (18 h) following the Illumina Whole-Genome Gene Expression Protocol for BeadStation (Doc. # 11176837 Rev. F, Illumina Inc.). Hybridized biotinylated cRNA was detected with 1 µg/ml Cyanine3-streptavidine (Amersham Biosciences #146065). BeadChips were scanned with Illumina BeadArray Reader.

Raw intensity data obtained from the Illumina™ platform were normalized with Inforsense KDE version 2.0.4 (Inforsense, London, UK) using quantile normalization method. The same software was also used for single-gene analyses including fold-change calculations and filtering the probes. The differences within the treatment group before and after the intervention were analyzed using the t-test statistic, with p-values calculated using 5000 permutations.

Pathway analysis of the expression data was performed using the Gene Set Enrichment Analysis (GSEA) implemented in javaGSEA application version 1.0 [13]. In order to avoid duplicates in the analysis, probes representing the same gene symbol in Illumina™ data were replaced with their average intensity before applying the GSEA. Gene sets for GSEA were taken from Database C2 of MSigDB version 1.0 of March 2005 [13]. Parameters used for the GSEA analysis are provided in Supporting Information Text S2. Gene expression data is available at Array Express web site (http://www.ebi.ac.uk/aerep/login; accession number E-TABM-116).

RT-PCR analysis

The microarray expression results recorded in the simvastatin group (n = 5, for one case there was not enough muscle RNA for PCR) were verified by Real-Time Quantitative TaqMan PCR. Previously purified cRNA was used as starting material for cDNA synthesis. A 1000 ng–18 µl aliquote of cRNA was mixed with 1 µl Promega Random Primer (C1181, Promega U.S., Madison, WI, USA) and incubated in +70°C for 10 min. The following reagents were added leading to 25 µl total reaction volume: 1 µl of 10 µM dNTP blend (F09892, Applied Biosystems, Foster City, CA, USA), 1 µl of Promega M-MLV Reverse Transcriptase 200 U/µl (M3682) and 4 µl of M-MLV RT 5× reaction buffer. Finally the incubations were performed in the following order: 10 min in RT, 50 min in 45°C, and, 10 min in 70°C.

10 µl volume was used for PCR reaction, consisting of 2 µl aliquote of 1∶10 diluted cDNA sample, and, Abgene ABsolute 2× QPCR ROX mix (AB-1139, Abgene, Epsom, UK). The primer concentrations were 300 nM, probe concentrations for Universal Probe Library (Exiqon, Vedbæk, Denmark) probes 100 nM and for ordinary long probes 200 nM. Finally the PCR reactions were performed in rtPCR system (ABI Prism 7700 Sequence Detection System, Applied Biosystems) having the following PCR procedure: 95°C for 15 min, and 40 cycles of 95°C for 15 s and 60°C for 1 min. The primer and probe sequences are available upon request.

Lipidomics analysis of plasma

An aliquot (10 µl) of an internal standard mixture containing 11 lipid classes, and 0.05 M sodium chloride (10 µl) was added to plasma samples (10 µl) and the lipids were extracted with chloroform/methanol (2∶1, 100 µl). After vortexing (2 min), standing (1 hour) and centrifugation (10000 RPM, 3 min) the lower layer was separated and a standard mixture containing 3 labeled standard lipids was added (10 µl) to the extracts. The sample order for LC/MS analysis was determined by randomization.

Lipid extracts were analysed on a Waters Q-Tof Premier mass spectrometer combined with an Acquity Ultra Performance LC™ (UPLC). The column, which was kept at 50°C, was an Acquity UPLC™ BEH C18 10×50 mm with 1.7 µm particles. The binary solvent system included A. water (1% 1 M NH4Ac, 0.1% HCOOH) and B. LC/MS grade (Rathburn) acetonitrile/isopropanol (5 2, 1% 1 M NH4Ac, 0.1% HCOOH). The gradient started from 65% A/35% B, reached 100% B in 6 min and remained there for the next 7 min. The total run time including a 5 min re-equilibration step was 18 min. The flow rate was 0.200 ml/min and the injected amount 0.75 µl. The temperature of the sample organizer was set at 10°C.

The lipid profiling was carried out on Waters Q-Tof Premier mass spectrometer using ESI+ mode. The data was collected at mass range of m/z 300–1200 with a scan duration of 0.2 sec. The source temperature was set at 120°C and nitrogen was used as desolvation gas (800 L/h) at 250°C. The voltages of the sampling cone and capillary were 39 V and 3.2 kV, respectively. Reserpine (50 µg/L) was used as the lock spray reference compound (5 µl/min; 10 sec scan frequency).

Data was processed using MZmine software version 0.60 [14]. Lipids were identified using internal spectral library. The normalization was performed using multiple internal standards as described in the Supporting Information Text S1. Only the identified lipid molecular species were included in further data analyses.

The Supporting Information Text S1, Figures S6–23 and Tables S8–11 also include general lipidomics platform characteristics such as internal and external standards used, calibration curves, dynamic ranges, recovery, variability, identification and quality control workflow, as well as illustrative spectra (MS and MS/MS) demonstrating how the specific species can be identified.

Lipid nomenclature

Lipids from the lipidomic analysis were named according to Lipid Maps (http://www.lipidmaps.org) [15]. For example, lysophosphatidylcholine with 16∶0 fatty acid chain was named as monoacyl-glycerophosphocholine GPCho(16∶0/0∶0). In case the fatty acid composition was not determined, total number of carbons and double bonds was marked. For example, a phosphatidylcholine species GPCho(16∶0/20∶4) is represented as GPCho(36∶4). However, GPCho(36∶4) could also represent other molecular species, for example GPCho(20∶4/16∶0) or GPCho(18∶2/18∶2). Such mass isomers may be separated chromatographically, as shown in Supporting Information Figure S20A for two lysophosphatidylcholine species GPCho(17∶0/0∶0) and GPCho(0∶0/17∶0).

Chemometric modeling and statistical analysis of lipidomics data

Partial least squares discriminant analysis (PLS/DA) [16], [17] was utilized as a supervised modeling method using SIMPLS algorithm to calculate the model [18]. As the total number of samples was insufficient for independent validation, no hold-out dataset was utilized for cross-validation. Instead, Venetian blinds cross-validation method [19] and Q 2 scores were used to optimize the model. Top loadings for latent variables associated with drug specific effects were reported. The VIP (variable importance in the projection) values [20] were calculated to identify the most important molecular species for the clustering of specific groups. Multivariate analyses were performed using Matlab version 7.2 (Mathworks, Inc.) and the PLS Toolbox version 3.5 Matlab package (Eigenvector Research, Inc.).

The regression of lipidomics data on muscle gene expression profiles was performed using the lasso method [21]. The lasso is a shrinkage regression method, similar to Ridge regression [22], which performs continuous variable selection causing some of the regression coefficients to be exactly zero. This reduces the variance of the regression estimates, which in the case of lipidomics data with large number of variables would otherwise be unacceptably high. Furthermore, the subset of lipids corresponding to non-zero coefficients can be considered as ‘the most important’ in explaining the muscle gene expression profiles. The lasso regression coefficients were calculated with the Least Angle Regression method [23] implemented in the R statistical language (package LARS). The corrected R 2 value and the Schwartz Criterion [24]were reported along with the measured and predicted gene expression values.

ResultsGene expression analysis reveals multiple upregulated pathways in the simvastatin group

In order to understand the pathways associated with statin response in muscle, we performed the whole genome microarray analysis of muscle biopsies. Microarray experiments were performed in 18 age-matched men (6 subjects from each group) who did not have any observed side effects such as muscle pain or CK elevations as a result of statin treatment. However, simvastatin treated men had substantial statin-induced unwanted and potentially toxic changes in muscle ubiquinone and mtDNA as reported earlier.

According to the used selection criteria for differentially expressed genes (1.5-fold change and p-value<0.05), expression of one gene was changed in the placebo group. Only modest changes were recorded in the atorvastatin group as expression of five genes was altered during the intervention. In the simvastatin group, however, expression of 111 genes changed (26 down-regulated and 85 up-regulated). Based on a hierarchical cluster analysis 20 genes were selected for further RT-PCR control. The following 5 genes were significantly upregulated: ALOX5AP (+3.6-fold, p = 0.041), CCL5 (+11.9-fold, p = 0.011), COL3A1 (+27.1-fold, p = 0.026), MYL5 (+8.0-fold, p = 0.021), MYBPH (+49.0-fold, p = 0.027).

As the recorded differences in single gene expressions were rather modest in general, we performed a Gene Set Enrichment Analysis [13] to identify globally affected metabolic pathways. The parameters of GSEA analysis are listed in Supporting Information Text S2. No pathways were affected significantly in the atorvastatin or placebo groups according to the criteria (False Discovery Rate q-value<0.25) recommended by Subramanian et al. [13]. However, in the simvastatin group 143 pathways were up-regulated (q<0.25) (Supporting Information Dataset S2). Due to the large number of affected pathways we limited our systematic analyses to the 23 most affected pathways (q<0.10) (Table 1).

Serum lipidomics reveals drug-specific changes

In order to investigate how the high dose statin treatment affects the plasma lipid profiles, we applied the UPLC/MS based lipidomics analysis, leading to a total of 132 identified lipid molecular species (data available as Supporting Information Dataset S3). Partial Least Squares Discriminant Analysis (PLS/DA) [17] revealed drug-specific changes in lipid profiles (Figure 1A). The PLS/DA model details are listed as Supporting Information Text S3 and Figure S1. The differences along the first latent variable (LV1), were associated with expected changes in triacylglycerols and cholesterol esters in agreement with the hypolipidemic effect expected from both drugs (Supporting Information Table S2). Specific differences between the simvastatin and atorvastatin lipid profiles were found in the third latent variable (LV3). Following VIP analysis, the most important lipid species were identified for each intervention group. The list of loadings in direction of atorvastatin-simvastatin differences (LV3) for most important lipids in simvastatin and atorvastatin groups is shown in Figure 1B. Notably, the main plasma lipid profile differences between the two statins can be considered as lipid-class specific, with specific upregulation of several phosphatidylethanolamines species and selective pools of long chain triacylglycerols. Similarly, downregulation of ether phosphocholines and cholesterol esters were observed in the simvastatin group compared to the atorvastatin group.

Combined lipidomics and gene expression

We wanted to identify if any of the lipidomic changes in plasma could be used as a marker of altered gene expression in muscle in the high dose simvastatin intervention group. Therefore we investigated if any of these gene expression changes were associated with the differences observed in the serum lipidome.

We selected a subset of genes based on GSEA analysis. Specifically genes from PLC, tubby, eicosanoid biosynthesis, and sodd pathways were chosen, based on their ranking as 2nd to 5th on FDR q-value. The top scored pathway “ST_T_Cell_Signal_Transduction” was not selected for further analysis since it is less pathway-specific than the other four and overlaps with the PLC pathway. The PLS/DA analysis on combined muscle gene expression (38 transcripts) and plasma lipid profile data revealed clear differences between the three treatment groups (Figure 2A). The PLS/DA model details are listed as Supporting Information Text S4. Simvastatin treatment was primarily associated with the gene expression changes in multiple genes involved in eicosanoid synthesis pathways as well as changes in multiple phosphatidylethanolamine and sphingomyelin molecular species (Figure 2B). Since the PLS analysis maximizes the product of variance matrix of measured variables (e.g. combined gene expression and lipid profile data) and correlation of measured data with properties of interest (e.g. treatment groups), our results indicate that there is a high degree of correlation between the upregulated genes (pathways) in skeletal muscle and specific lipids plasma changes in the simvastatin group.

The results from combined gene expression and lipid PLS/DA analysis raise the possibility that plasma lipid biomarkers may be found for the pathway changes observed in the muscle. In order to investigate this possibility, we performed regression analysis of plasma lipid profile data on a muscle selected marker gene expression profile. We chose arachidonate 5-lipoxygenase activating protein (ALOX5AP, Uniprot ID: P20292) as the marker for simvastatin dysregulated pathway changes in muscle. The ALOX5AP gene was selected based on high VIP scores in multiple PLS/DA analyses, PCR validated significant fold change in the simvastatin group, and its well known (pro-inflammatory) biological role [25]. As the main goal of this analysis was discovery of potential plasma molecular markers for statin induced muscle toxicity, we applied a shrinkage regression method lasso [21] aiming to find a subset of plasma lipids predictive of specific gene expression levels in skeletal muscle. Figure 3A shows the results of the lassomodel for NZ = 25 non-zero lipid variables. The variables and their coefficients are shown in Figure 3B. The lasso analysis was also performed for NZ = 5, 10, 15, and 20 non-zero variables (Supporting Information Tables S3–S7 and Figures S2–S5). Our analyses identified ALOX5AP gene expression in muscle had a high positive regression coefficients with plasma levels of phosphatidylethanolamine (42∶6) and negative for the cholesterol ester ChoE(18∶0), i.e. both type of lipids were selected as the non-zero variables in all regression analyses, with consistent results. Also in all analyses except NZ = 5, the sphingomyelins SM(d18∶1/24∶0) and SM(d18∶1/24∶1) showed negative regression coefficients (Figure 2B). Also the ether phosphocholines were selected with consistently negative coefficients, in agreement with PLS/DA analyses (Figures 1B and 2B).

Discussion

Our systems biology strategy using combined gene expression and lipidomics analyses revealed that simvastatin at high doses induces significant changes in the expression of multiple genes controlling metabolic and inflammatory pathways in a non-hepatic tissue. This observation markedly contrast with the minimum gene expression changes observed in skeletal muscle with atorvastin 40mg. Our studies also revealed novel plasma biomarker candidates for safety assessment of statin treatment before recorded changes in muscle metabolism become clinically evident.

Similar to an earlier report [26], expression of genes related to cholesterol metabolism or mevalonate pathway were only modestly affected by statins in the present study. Thus, our data do not directly support the view that statins would cause mitochondrial dysfunction by reducing ubiquinone, a mitochondrial coenzyme with a cholesterol synthetic pathway derived side-chain, due to inhibition of HMG-CoA reductase in the muscle. Similarly we were not able to provide evidence that statins would lead to inhibition of protein synthesis, signal transduction and metabolism due to decreased muscle mevalonate acid. Since our patients did not have any signs of clinical myopathy and muscle damage, we were not able to judge the significance of early proapoptotic markers during the course of the myopathy. However, in the GSEA analysis several pro-apoptosis pathways already appeared with significant FDR q-values at these early stages and, therefore, the present results support the role of pro-apoptosis pathways in statin myotoxicity. Furthermore, the hypothesis of an increased Ca2+ influx as a mediator of statin induced toxicity is supported by the significant up-regulation of phospholipase C pathway and by the dysregulation of genes encoding for calcium binding proteins (Supporting Information Dataset S2) in the present study. Another hallmark of high dose simvastatin effect in muscle is the activation of pro-inflammatory pathways such as eicosanoid synthesis. However, the present results cannot reveal the actual trigger leading to impaired mitochondrial function and induction of these proinflamatory pathways.

Atorvastatin and simvastatin treatments also resulted in specific plasma lipidome profiles. Thus, lipidomics analysis may have the potential of providing individualized specific lipid lowering agents suitable to individual patients after the significance of these novel lipid biomarkers is elucidated. The role of sphingolipids as independent predictors of coronary artery disease has been previously suggested [27]. However, our results demonstrate that plasma sphingomyelin changes in response to statin therapy varies for different sphingomyelin molecular species. This raises important questions about the biological significance of these molecular species.

We also identified ether phospholipids as another type of lipids differentially regulated by simvastatin. Interestingly plasmalogens, a most abundant sub-class of ether phospholipids, have been associated with protection against oxidative damage [28], [29]. Their observed decrease (and negative correlation with the ALOX5AP expression) following simvastatin treatment may thus be functionally linked to increased oxidative stress and inflammation in muscle. Phosphatidylethanolamines were specifically dysregulated by simvastatin. The significance of this observation remains to be elucidated but it is interesting that a link between the Leukotriene B4 (a lipid synthesised via ALOX5AP) and the release of arachidonate from phosphatidylethanolamine in human neutrophils has been established [30]. All together these data indicate that in parallel with specific gene expression changes in skeletal muscle, treatment with simvastatin was also associated with parallel changes in plasma of lipidic proinflamatory markers.

One limitation of the present study is the relatively small sample size due to obvious limitations in the number of muscle specimens obtained from patients. However, the conclusions of the results are strengthened by the combined genomic and lipidomic analyses. Although it may considered a potential weakness of the study, we decided not include at this stage any patients with acute proven myopathy. The rational for this is that analysis of gene expression profiles in specimens obtained from patients during acute muscle events (unpublished results) revealed hundreds of different genes affected in the context of muscle damage, making it rather difficult to analyze the results as directly related to the statin treatment or establish their potential use as early markers of myopathy. Therefore our strategy of investigating individuals with well-defined statin-induced mitochondrial defects during a randomized trial allowed us to identify markers with potential diagnostic and prognostic value. In conclusion, the combined analyses of gene expression and lipidomics profiles in asymptomatic statin treated individuals revealed that: a) simvastatin 80 mg induces specific gene expression and lipid changes compared to equally efficient atorvastatin treatments and b) that our combined transcriptomic lipidomic analysis provides bona fide sensitive biomarkers of statin induced metabolic changes in muscle potentially useful to identify patients at risk early enough to prevent actual muscle damage. These biomarkers are now available for further validation in patients with proven statin-induced myopathy.

Statin drugs represent the major improvement in the treatment of hypercholesterolemia that constitutes the main origin of atherosclerosis, leading to coronary heart disease. Besides tremendous beneficial effects of statins, various forms of muscular toxicity (myalgia, cramp, exercise intolerance, and fatigability) occur frequently. We hypothesized that the iatrogenic effects of statins could result from alterations in Ca2+ homeostasis. Acute applications of simvastatin on human skeletal muscle fibers triggered a Ca2+ wave of intra-cellular Ca2+ that mostly originates from sarcoplasmic reticulum (SR) Ca2+-release. In addition, simvastatin increased mitochondrial NADH content and induced mitochondrial membranedepolarization (EC50 = 1.96 μM) suggesting an altered mitochondrial function. Consequently on simvastatin application, a weak mitochondrial Ca2+ efflux (EC50 = 7.8μM) through permeability transient pore and Na+/Ca2+ exchanger was triggered, preceding the large SR-Ca2+ release. Increased SR Ca2+ content after acute application of statin is also suggested by the increased Ca2+ spark amplitude and by the effect of cyclopiazonic acid. We thus conclude that simvastatin induced alterations in mitochondrial function which lead to an increase in cytoplasmic Ca2+, SR-Ca2+ overload, and Ca2+ waves. Taken together, these statin-induced muscle dysregulations may contribute to myotoxicity.

In contrast to the current belief that cholesterol reduction with statins decreases atherosclerosis, we present a perspective that statins may be causative in coronary artery calcification and can function as mitochondrial toxins that impair muscle function in the heart and blood vessels through the depletion of coenzyme Q10 and ‘heme A’, and thereby ATP generation. Statins inhibit the synthesis of vitamin K2, the cofactor for matrix Gla-protein activation, which in turn protects arteries from calcification. Statins inhibit the biosynthesis of selenium containing proteins, one of which is glutathione peroxidase serving to suppress peroxidative stress. An impairment of selenoprotein biosynthesis may be a factor in congestive heart failure, reminiscent of the dilated cardiomyopathies seen with selenium deficiency. Thus, the epidemic of heart failure and atherosclerosis that plagues the modern world may paradoxically be aggravated by the pervasive use of statin drugs. We propose that current statin treatment guidelines be critically reevaluated.

Abstract

Our specific aims were to determine whether low serum 25 (OH) vitamin D (D2 + D3) (<32 ng/mL) was associated with myalgia in statin-treated patients and whether the myalgia could be reversed by vitamin D supplementation while continuing statins. After excluding subjects who took corticosteroids or supplemental vitamin D, serum 25 (OH) D was measured in 621 statin-treated patients, which consisted of 128 patients with myalgia at entry and 493 asymptomatic patients. The 128 myalgic patients had lower mean +/- standard deviation (SD) serum vitamin D than the 493 asymptomatic patients (28.6 +/- 13.2 vs 34.2 +/- 13.8 ng/mL, P < 0.0001), but they did not differ (p > 0.05) by age, body mass index (BMI), type 2 diabetes, or creatine kinase levels. By analysis of variance, which was adjusted for race, sex, and age, the least square mean (+/- standard error [SE]) serum vitamin D was lower in the 128 patients with myalgia than in the 493 asymptomatic patients (28.7 +/- 1.2 vs 34.3 +/- 0.6 ng/mL, P < 0.0001). Serum 25 (OH) D was low in 82 of 128 (64%) patients with myalgia versus 214 of 493 (43%) asymptomatic patients (chi(2) = 17.4, P < 0.0001). Of the 82 vitamin-D-deficient, myalgic patients, while continuing statins, 38 were given vitamin D (50,000 units/week for 12 weeks), with a resultant increase in serum vitamin D from 20.4 +/- 7.3 to 48.2 +/- 17.9 ng/mL (P < 0.0001) and resolution of myalgia in 35 (92%). We speculate that symptomatic myalgia in statin-treated patients with concurrent vitamin D deficiency may reflect a reversible interaction between vitamin D deficiency and statins on skeletal muscle.

Statins lower cholesterol and adverse cardiovascular outcomes, but this drug class increases diabetes risk. Statins are generally anti-inflammatory. However, statins can promote inflammasome-mediated adipose tissue inflammation and insulin resistance through an unidentified immune effector. Statins lower mevalonate pathway intermediates beyond cholesterol, but it is unknown whether lower cholesterol underpins statin-mediated insulin resistance. We sought to define the mevalonate pathway metabolites and immune effectors that propagate statin-induced adipose insulin resistance. We found that LDL cholesterol lowering was dispensable, but statin-induced lowering of isoprenoids required for protein prenylation triggered NLRP3/caspase-1 inflammasome activation and interleukin-1β (IL-1β)–dependent insulin resistance in adipose tissue. Multiple statins impaired insulin action at the level of Akt/protein kinase B signaling in mouse adipose tissue. Providing geranylgeranyl isoprenoids or inhibiting caspase-1 prevented statin-induced defects in insulin signaling. Atorvastatin (Lipitor) impaired insulin signaling in adipose tissue from wild-type and IL-18−/−mice, but not IL-1β−/− mice. Atorvastatin decreased cell-autonomous insulin-stimulated lipogenesis but did not alter lipolysis or glucose uptake in 3T3-L1 adipocytes. Our results show that statin lowering of prenylation isoprenoids activates caspase-1/IL-1β inflammasome responses that impair endocrine control of adipocyte lipogenesis. This may allow the targeting of cholesterol-independent statin side effects on adipose lipid handling without compromising the blood lipid/cholesterol-lowering effects of statins.

Introduction

Statins lower blood cholesterol and reduce the risk of adverse cardiovascular events, but these drugs can increase blood glucose and risk of diabetes (1). Statins inhibit HMG-CoA reductase, lowering cholesterol biosynthesis and promoting hepatic cholesterol uptake. Statins have cholesterol-independent actions that depend on HMG-CoA inhibition but are often termed pleiotropic effects. Statin-mediated lowering of mevalonate pathway intermediates can alter immunity, independent of cholesterol (2). Inhibition of mevalonate synthesis reduces the isoprenoid production required for protein prenylation, a posttranslational modification that occurs on hundreds of cellular proteins (3). Statin lowering of specific isoprenoids can limit farnesylation or geranylgeranylation (3).

Statin-mediated lowering of prenylation is generally associated with reduced inflammation, including lower levels of circulating interleukin-6 (IL-6) and tumor-necrosis factor-α (4). Statins paradoxically increase IL-1β and IL-18, and reduced geranylgeranylation is sufficient to increase IL-1β and IL-18 in monocytes (5). It is also known that statins increase caspase-1 activity in immune cells (6). Therefore, statins activate a caspase-1 inflammasome and IL-1β/IL-18 responses despite widespread anti-inflammatory actions of this drug class. We previously characterized how an inflammasome contributes to the balance of these immune effects and endocrine control of metabolism. We found that the nucleotide-binding oligomerization (NOD)-like receptor family, pyrin-domain containing 3 (NLRP3) contributes to statin-induced insulin resistance in adipose tissue (7). It was not known which metabolites in the mevalonate pathway promote statin-induced insulin resistance. Statin-mediated inhibition of the cholesterol biosynthesis pathway can influence immunity by altering intermediates such as 25-hydroxycholesterol (25-HC), which has been directly linked to NLRP3 inflammasome activation in macrophages (8). Statins can actually prevent inflammasome assembly and attenuate caspase-1–mediated IL-1β secretion by acutely depleting endoplasmic reticulum–resident cholesterol in macrophages (9). Statins can also attenuate NLRP3 activation and IL-1β release when lipopolysaccharide (LPS)-primed monocytes are activated with cholesterol crystals, but the same data show that statins alone increase active caspase-1 and IL-1β secretion (10). Here, we tested whether prenylation or cholesterol underpinned statin-induced adipose insulin resistance.

It is unknown which inflammasome effector propagates adipose insulin resistance as a result of statins. NLRP3/caspase-1 inflammasome regulation of IL-1β promotes insulin resistance in adipocytes and dysglycemia in rodents (11). Markers of NLRP3 inflammasome activation are higher in patients with type 2 diabetes, and IL-1 receptor antagonism can improve glycemia (12,13). IL-1β is a good candidate to test for statin-mediated insulin resistance, but caspase-1 targets beyond IL-1β or IL-18 can link NLRP3 inflammasome responses to metabolic defects in insulin-responsive tissues (14,15).

We show that statin-induced adipocyte insulin resistance occurs through IL-1β and the lowering of isoprenoids required for prenylation, independent of changes in cellular cholesterol or LDL-mediated pathways to insulin resistance. Statins cause cell-autonomous impairment in insulin-stimulated adipocyte lipogenesis rather than glucose uptake or inhibition of lipolysis.

Research Design and Methods

The McMaster University animal ethics review board approved all procedures. Male C57BL/6J mice were from The Jackson Laboratory (#000664). IL-1β−/− mice were from Yoichiro Iwakura (University of Tokyo, Tokyo, Japan) and bred in-house. IL-18−/− mice were from A.A.A. Explants from mouse gonadal adipose depots were exposed to statins, zoledronate, cholesterol derivatives, geranylgeranyl pyrophosphate (GGPP) or z-YVAD (18 h), LPS (final 4 h), or insulin (final 10 min). Lysates were immunoblotted (16), and IL-1β was quantified by ELISA (7). Adipocytes were separated from the stromal vascular fraction (SVF) (17). 3T3-L1 adipocytes and bone marrow–derived macrophages (BMDMs) were immunoblotted or analyzed by quantitative PCR (18).

3T3-L1 adipocyte lipolysis was determined as previously described (18). Lipogenesis was measured in lipids from adipose explants using [14C]U-glucose (2 μCi/mL) and 3T3-L1 adipocytes (1 μCi/mL) ± insulin (0.3 nmol/L) for the final 2 h (explants) or 1 h (3T3-L1). Glucose uptake was measured using [3H]-2-deoxyglucose (16).

Each experimental replicate represents a single adipose explant or 3T3-L1 culture well. Statistical significance was determined by ANOVA with Tukey post hoc test. Unpaired t tests were used to compare two conditions.

ResultsMevalonate Pathway Inhibition Lowers Insulin-Stimulated Lipogenesis in Adipose Tissue

We previously showed that NLRP3 was required for fluvastatin to impair insulin-stimulated Ser473 phosphorylation of Akt/phosphokinase B (PKB) in explanted mouse adipose tissue. Adipose tissue was primed with a dose of LPS (2 μg/mL, 4 h) that does not alter insulin signaling but allows investigation of inflammasome responses (7) (Fig. 1A). The bisphosphonate zoledronate (1 and 5 μmol/L), which inhibits the mevalonate pathway distal to HMG-CoA reductase, lowered insulin-stimulated Ser473 phosphorylation of Akt/PKB in LPS-primed explanted adipose tissue (Fig. 1A). Multiple statins impaired insulin signaling in adipose tissue, indicating a drug class effect on insulin sensitivity. We found that 1 μmol/L atorvastatin or 1 μmol/L pravastatin (Fig. 1B and C) but 0.1 μmol/L cerivastatin decreased insulin-stimulated Ser473 phosphorylation of Akt/PKB in LPS-primed adipose explants (Fig. 1D). Atorvastatin impaired insulin-stimulated lipogenesis in LPS-primed adipose explants (Fig. 1E). Compared with adipose explants from lean mice, explants from high-fat–fed obese mice had impaired insulin signaling where LPS, but not statin exposure, further decreased insulin action (Fig. 1F). These results show that inhibiting the mevalonate pathway at multiple steps or with multiple statins impaired insulin signaling wherein one functional outcome is lower insulin-stimulated adipose tissue lipogenesis as a result of statin exposure.

Statin Lowering of Prenylation Inhibits Insulin Signaling in Adipose Tissue

We next tested whether lowering cholesterol metabolites or isoprenoids underpinned statin-mediated insulin resistance. We found that supplementation with LDL cholesterol (0.01 and 1 mg/mL) or free cholesterol (1 and 20 μmol/L) did not cause a further reduction of (and did not restore) statin-mediated lowering of Ser473 phosphorylation of Akt/PKB in LPS-primed adipose explants (Fig. 2A and B). LDL cholesterol treatment alone (1 mg/mL) lowered insulin signaling independently of LPS priming or statin treatment (Fig. 2A). Lower 25-HC can lead to activation of caspase-1 and increase IL-1β in macrophages (8). However, we found that supplementation of LPS-primed adipose explants with 25-HC (1 and 20 μmol/L) did not cause a further reduction of (and did not restore) impaired insulin signaling as a result of atorvastatin (Fig. 2C). In contrast to all experiments using cholesterol derivatives, supplementation of LPS-primed adipose explants with the isoprenoid GGPP at 50 μmol/L (but not 5 μmol/L) restored atorvastatin-induced suppression of insulin-stimulated phosphorylation of Akt/PKB at Ser473and Thr308 (Fig. 2D and E). Importantly, GGPP restored insulin action as a result of statin exposure but not LDL cholesterol (1 mg/mL) (Fig. 2A). These data show that a statin-mediated reduction in a geranylgeranyl isoprenoid is required for impaired insulin signaling, which occurs independently of cholesterol or LDL-mediated effects on insulin action.

IL-1β Is Required for Statin-Induced Insulin Resistance

We next tested whether caspase-1 regulation of IL-1β or IL-18 was involved in impaired insulin action in adipose tissue (Fig. 3A). Inhibition of caspase-1 with 1 or 10 μmol/L z-YVAD restored atorvastatin-mediated lowering of Ser473 phosphorylation of Akt/PKB in LPS-primed adipose explants stimulated with insulin (Fig. 3B). Atorvastatin, pravastatin, or cerivastatin did not lower Ser473 phosphorylation of Akt/PKB in LPS-primed adipose explants derived from IL-1β−/−mice (Fig. 3C and D). Similar to wild-type mice, atorvastatin lowered Ser473 phosphorylation of Akt/PKB in LPS-primed adipose explants derived from IL-18−/− mice (Fig. 3E). Overall, these data show that statins require caspase-1 and IL-1β to impair insulin signaling in adipose tissue.

Statin Lowering of Prenylation Impairs Adipocyte-Autonomous Lipogenesis

We next sought to determine the contributions of adipocytes versus the immune cell-enriched SVF. When adipose tissue explants were LPS primed and treated with atorvastatin, we found that both adipocytes and SVF had higher IL-1β (Fig. 4A). IL-1β was more abundant in the SVF compared with adipocytes (Fig. 4A). Previous research in BMDM reported that statin-mediated lowering of isoprenoids can also activate the pyrin inflammasome to promote IL-1β processing (19). The requirement for pyrin in macrophages contradicts our previous report wherein statins engaged the NLRP3 inflammasome in adipose tissue (7). We assessed whether cell type was the underlying factor for this discrepancy. We found that transcript levels of NLRP3 were increased with LPS treatment in the adipocyte and SVF, although no increase in transcripts of pyrin could be detected (Fig. 4B). We found that transcript levels of pyrin were detectable and augmented by LPS priming in BMDMs and adipose tissue (containing the SVF), but pyrin transcripts were not detectable in 3T3-L1 adipocytes (Fig. 4C). NLRP3 transcripts were present in BMDM, adipose tissue, and 3T3-L1 adipocytes, and NLRP3 transcript levels were augmented by LPS priming in all three cell/tissue types (Fig. 4C). This prompted us to determine whether a cell-autonomous response underpinned NLRP3 statin-induced adipocyte insulin resistance.

holesterol concentration in 3T3-L1 adipocytes did not change during treatment with atorvastatin (Fig. 4D). Atorvastatin lowered insulin-stimulated Ser473 phosphorylation of Akt/PKB in 3T3-L1 adipocytes, and supplementation with the isoprenoid geranylgeraniol (GGOH) (25 μmol/L) restored this aspect of insulin signaling (Fig. 4E). Supplementation with farnesol did not restore atorvastatin-mediated lowering of insulin-stimulated Ser473phosphorylation of Akt/PKB in 3T3-L1 adipocytes (Fig. 4F). We next sought to determine the functional consequences of these adipocyte-autonomous statin-mediated responses. Atorvastatin did not alter lipolysis (with or without isoproterenol) in 3T3-L1 adipocytes (Fig. 4G). However, insulin-stimulated lipogenesis was decreased by atorvastatin, an effect that was not observed when 3T3-L1 adipocytes were supplemented with 25 μmol/L GGOH (Fig. 4H). Insulin-stimulated glucose uptake into 3T3-L1 adipocytes was not changed by atorvastatin or GGOH (Fig. 4I). Overall, these data show that atorvastatin lowers insulin-stimulated adipocyte-autonomous lipogenesis by lowering isoprenoids required for protein prenylation but not farnesylation.

Discussion

Statins lower cholesterol, risk of cardiovascular disease, and all-cause mortality, but this drug class can increase blood glucose (1). The mechanisms linking statins and increased risk of diabetes should be clarified because warning labels now include an increased risk of blood glucose and diabetes (Health Canada, RA-16949). It is not known whether statin-induced changes in glycemia are related to cholesterol lowering. We sought to determine whether statins impaired insulin action by lowering cholesterol or another mevalonate pathway metabolite. We found that LDL cholesterol and 25-HC were dispensable for statin-mediated insulin resistance in adipose tissue. Cell type is a key determinant of how statins engage specific inflammasomes. Others have shown that statins can attenuate ligand-induced NLRP3 or pyrin inflammasome responses in macrophages or monocytes (9,10,19). We show that isoprenoids required for protein prenylation were sufficient to prevent statin-mediated defects in insulin signaling in adipocytes. This is important because coadministration of statin and isoprenoids may mitigate adipose tissue side effects but not interfere with lipid/cholesterol-lowering benefits of statins. Determination of the prenylated proteins that alter inflammasome activation and insulin resistance is warranted. Targeting specific prenylation events may be superior to widespread reversal of isoprenoid lowering that could mitigate longevity and cardiac benefits of statins shown in Drosophila (20).

We show that statins impaired insulin-stimulated lipogenesis in an adipocyte-autonomous manner, which was prevented by supplementing isoprenoids required for prenylation. This is important because impaired insulin signaling does not always correlate with impaired insulin action. Given that statins did not impair glucose uptake in adipocytes, our results are consistent with statins impairing selective and heterogeneous effects of insulin resistance that manifest in lipid metabolism (21).

We show that statin-induced insulin resistance is a drug class effect. Cerivastatin was removed from the market because of side effects. Our results show that cerivastatin was the most potent statin in promoting adipose insulin resistance (and the most potent activator of IL-1β release from BMDMs [data not shown]). IL-18 was dispensable for statin-induced adipose insulin resistance. IL-1β was the key mediator of statin-induced adipocyte insulin resistance, which ultimately impaired lipogenesis. IL-1β alone lowers insulin-stimulated lipogenesis in rodent and human adipocytes (22). IL-1 receptor antagonism can improve glycemia, but the design of clinical trials using IL-1β inhibition, such as the Canakinumab Anti-Inflammatory Thrombosis Outcome Study (CANTOS) trial, does not necessarily always reveal reduced diabetes incidence (13,23). Statins are pervasive in these clinical trials and should be considered as a confounding variable. It is not clear why patients with familial hypercholesteremia on lifelong statins can have a low prevalence of type 2 diabetes as opposed to patients who are obese or with metabolic disease in whom stains can increase the risk of type 2 diabetes (24). Our results suggest that impaired adipocyte lipogenesis underpins the relationship between lipids and glucose during statin treatment because higher blood triglycerides are a known risk factor predicting statin-induced diabetes (25). Our results appear relevant to statin intolerance and diminishing returns of increased statin dose on triglyceride/lipid lowering. Statin engagement of an NLRP3/caspase-1/IL-1β response may limit the effectiveness of statins to lower blood triglycerides independent of cholesterol lowering. Targeting NLRP3/caspase-1/IL-1β may allow enhanced blood lipid lowering at a given statin dose and/or mitigate side effects at a statin dose that achieves lipid-lowering goals.

Abstract

Objective: In 150 hypercholesterolemic patients, unable to tolerate ≥1 statin because of myositis-myalgia, selected by low (<32 ng/ml) serum 25 (OH) vitamin D, we prospectively assessed whether vitamin D supplementation with resolution of vitamin D deficiency would result in statin tolerance, free of myositis-myalgia.

Research design and methods: We studied 74 men, 76 women, median age 60, 131 white, 17 black and 2 other. On no statins, 50,000 units of vitamin D was given twice a week for 3 weeks, and then continued once a week. After 3 weeks on vitamin D, statins were restarted. Patients were re-assessed on statins and vitamin D every 3 to 4 months, with serial measures of serum 25 (OH) vitamin D, creatine phosphokinase (CPK), LDL cholesterol (LDLC) and assessment of myositis-myalgia.

Main outcome measures: Percentage of patients myalgia-free on vitamin D plus reinstituted statins, serum 25 (OH) vitamin D, CPK, and LDLC on reinstituted statins and concurrent vitamin D supplementation.

Results: On vitamin D supplementation plus re-instituted statins for a median of 8.1 months, 131 of the 150 patients (87%) were free of myositis-myalgia and tolerated the statins well. Serum 25 (OH) vitamin D increased from median 21 to 40 ng/ml (p < 0.001), and normalized (≥32 ng/ml) in 117 (78%) of 150 previously vitamin D deficient, statin-intolerant patients. Median LDLC decreased from 146 mg/dl to 95 mg/dl, p < 0.001.

Conclusion: Symptomatic myositis-myalgia in hypercholesterolemic statin-treated patients with concurrent serum 25 (OH) vitamin D deficiency may reflect a reversible interaction between vitamin D deficiency and statins on skeletal muscle causing myalgia.

Coenzyme Q10 is an important factor in mitochondrial respiration. Primary and secondary deficiencies of coenzyme Q10 result in a number of neurologic and myopathic syndromes. Hydroxyl-methylglutaryl coenzyme A reductase inhibitors or statins interfere with the production of mevalonic acid, which is a precursor in the synthesis of coenzyme Q10. The statin medications routinely result in lower coenzyme Q10 levels in the serum. Some studies have also shown reduction of coenzyme Q10 in muscle tissue. Such coenzyme Q10 deficiency may be one mechanism for statin-induced myopathies. However, coenzyme Q10 supplements have not been shown to routinely improve muscle function. Additional research in this area is warranted and discussed in this review.

Coenzyme Q10 was discovered in 1957 by Dr. Fred Crane. During the past 50 years, coenzyme Q10 has been found to be a key component in mitochondrial bioenergy transfer. Its enzymatic processes facilitate electron transfer in the generation of adenosine triphosphate (ATP). It has also been shown to have important antioxidant properties. Physiologic concentrations of coenzyme Q10 do not fully saturate the mitochondrial receptors. Accordingly, even a small increase in the coenzyme Q10 concentration of mitochondrial membranes can lead to an increase in mitochondrial respiration.1 This observation may be the biochemical mechanism by which exogenous coenzyme Q10 administration has in some studies improved mitochondrial myopathies and cardiomyopathies. A diagram of the key role that coenzyme Q10 plays in mitochondrial membrane generation of ATP is shown in Figure 1. Mitochondrial complexes I through V are specialized protein complexes found in the inner mitochondrial membrane that facilitate the transfer of electrons in the mitochondrial utilization of oxygen. Note that this crucial cofactor, coenzyme Q10, lies at the intersection of electron transfers from both the citric acid cycle and in the reaction as nicotinamide adenine dinucleotide (NADH) is reduced to NAD+.

Coenzyme Q10 is a lipid compound with 10 isoprenoid units and is widely distributed in the human body. It is a lipophilic inner mitochondrial membrane cofactor that is used to shuttle electrons in the formation of ATP.2 The compound is synthesized in a number of reactions from mevalonic acid, whose production itself is inhibited by hydroxyl-methylglutaryl coenzyme A (HMA CoA) reductase inhibitors (Figure 2). Coenzyme Q10 has been shown to inhibit oxidation of proteins, DNA, and lipids in its reduced form, ubiquinol. Coenzyme Q10 in the serum is largely found bound to the lipoprotein transport of low-density lipoprotein (LDL) cholesterol and does not circulate in any appreciable concentration as an unbound form. Dietary supplementation of coenzyme Q10 increases levels of its reduced form within the circulating lipoproteins and inhibits LDL peroxidation. This inhibition of LDL peroxidation may play a key role in its antiatherogenic effects.1 Typical serum concentrations of coenzyme Q10 in values for healthy individuals are in the 0.8 to 1 µg/mL range. The half-life in human plasma is 30 to 35 hours.3

DEFICIENCY STATES OF COENZYME Q10

Multiple studies have shown that statins can decrease coenzyme Q10 levels. A portion of this decrease is related to the decrease in the levels of its lipoprotein transport carriers, which is induced by the therapeutic effect of the statins.4 Animal studies have revealed depletion of both tissue and blood levels of coenzyme Q10 after statin therapy in the dog model, hamster model, squirrel monkeys, and minipigs.4 In humans, exposure to atorvastatin, 80 mg for 14 to 30 days, caused a significant reduction in coenzyme Q10 levels of 34 subjects both at day 14 and at day 30. Baseline levels of coenzyme Q10 decreased from 1.26 to 0.67 µg/mL at 14 days and to 0.62 µg/mL at 30 days.5 Simvastatin, 20 mg daily, and pravastatin, 20 mg daily, have also been associated with a 40% reduction in coenzyme Q10 levels.6 Lamperti et al7 found muscle coenzyme Q10 levels that were at least 1 standard deviation below the mean in 9 of 12 patients who had statin-induced myopathy or elevated creatinine phosphokinase (CPK) levels. High-dose statin therapy resulted in decreased mitochondrial function in those patients who had low levels of ubiquinone in muscle.8

Although serum levels of coenzyme Q10 routinely decrease with statin therapy, not all studies confirm the potential mitochondrial dysfunction induced by statins. In humans, Laaksonen et al9 found no change in high-energy phosphate levels and coenzyme Q10 levels in muscle biopsies from patients before and after treatment with simvastatin, 20 mg/d for 6 months. The results did not differ from the muscle biopsy findings of matched controls taken at the same time.9

Aging appears to play an important role in causing low levels of coenzyme Q10.4 Older animals may be particularly affected by statin-induced coenzyme Q10 deficiency. Diebold et al10 found that mitochondrial conversion from adenosine diphosphate to ATP decreased by 45% in cardiac mitochondria in 2-year-old guinea pigs when compared with younger animals treated with lovastatin. In humans, aging may cause increased demands for coenzyme Q10. A study of the contractile force of myocardial trabecular tissue in humans noted a significantly lower coenzyme Q10 content in the tissue in patients older than 70 years. This lower coenzyme Q10 content of the tissue was associated with a significantly reduced contractile performance in vitro, which was reversed by pretreatment with coenzyme Q10.11 Differences in the contractile strength between young and senescent myocardial tissue from rats were similar.11 The study concluded that pretreatment with coenzyme Q10 improved the tolerance of the senescent myocardium to aerobic stress by improving ATP generation within the affected mitochondria as well as through its antioxidant role as a free radical scavenger.

Exercise may also induce a relative coenzyme Q10 deficiency because of increased demands on the mitochondria for ATP production. Oxidative stress as found in testing after vigorous exercise may cause depletion of muscle levels of coenzyme Q10. Such exercise results in increased uptake of coenzyme Q10 by the muscles.12

Finally, several primary deficiency states of coenzyme Q10 exist that result in encephalopathies, severe infantile multisystem disease, Leigh syndrome, myopathies, and cerebellar ataxia.2,5 Such coenzyme Q10 deficiency results from an autosomal recessive disorder with a clinically heterogeneous phenotype.

Go to:EFFECTS OF COENZYME Q10 SUPPLEMENTATION

Clearly, the biochemistry of coenzyme Q10 is an integral part of mitochondrial function. Deficiency states have been associated with multiple pathologic conditions. However, study results have been mixed on the clinical efficacy of coenzyme Q10 administration.

Studies Showing a Benefit of Coenzyme Q10 Administration

Mizuno et al13 found that coenzyme Q10 administration improved subjective fatigue sensation and physical performance during fatigue-inducing workload trials. In this double-blind, placebo-controlled trial, 17 healthy volunteers were randomized to either placebo or 100 mg or 300 mg of coenzyme Q10 for 8 days. Total coenzyme Q10 levels at baseline were 0.54 µg/mL. Supplementation with 100 mg/d of coenzyme Q10 resulted in a total serum level of 2.08 µg/mL. Supplementation with 300 mg/d of coenzyme Q10 resulted in a serum level of 3.11 µg/mL. After 30 minutes of exercise on a bicycle ergometer, the group randomized to 300 mg/d of coenzyme Q10 showed a statistically significant improvement in measures of fatigue using a visual analog scale.

A study of 18 male Japanese kendo athletes demonstrated that coenzyme Q10 supplementation decreased exercise-induced muscle injury as measured by CPK levels. These athletes were treated with 300 mg/d of coenzyme Q10 or placebo in a double-blind, randomized manner. Coenzyme Q10 levels at baseline were near 1 µg/mL and increased fourfold in the treated group, and CPK elevations in the treated group were statistically lower by approximately 50% than those in the nontreated group after 5 days.14

In a double-blind crossover trial of 25 Finnish top-level cross-country skiers, a statistically significant improvement was noticed in multiple measures of physical performance. After administration of 90 mg of coenzyme Q10 per day, plasma levels in the treated group rose from 0.8 to 2.8 µg/mL.15 The peak inspired oxygen consumption (V̇o2max) increased by 1.6 mL/kg per minute (P  =  0.02), the anaerobic threshold increased by 2.4 mL/kg per minute (P  =  0.003), and aerobic threshold increased by 2.6 mL/kg per minute (P  =  0.001). Additionally, 94% of the athletes, versus only 33% in the placebo group, thought that coenzyme Q10 had improved their performance in recovery time. No significant differences were noticed in lactic acid clearance.

Amadio et al16 studied 18 basketball players following administration of 100 mg/d of coenzyme Q10 for 40 days; they noted an 18% improvement in V̇o2max in the athletes. In 2008, Cooke et al12 found that coenzyme Q10 supplementation at 200 mg/d resulted in significantly increased coenzyme Q10 serum concentrations, which correlated with V̇o2max and treadmill time to exhaustion. There was also a significant correlation of serum coenzyme Q10 levels with muscle tissue levels of coenzyme Q10. This finding was present in both trained and untrained individuals. The study enrolled 22 trained athletes who had been exercising 8 hours per week in the course of 9 workouts per week for at least 2 years. The 19 untrained individuals who enrolled had not engaged in regular exercise for the past year.

Studies have also shown a benefit of coenzyme Q10 administration in the mitochondrial function of heart muscle. A randomized, multicenter study of 322 patients with congestive heart failure comparing patients receiving 2 mg/kg of coenzyme Q10 with those receiving placebo showed a significant decrease in hospitalization for heart failure. After 1 year, 73 patients in the treatment group versus 118 patients in the control group required hospitalization for heart failure.17 Statistically significant declines in pulmonary edema, cardiac asthma, and arrhythmias were also noted.

Studies Showing No Benefit of Coenzyme Q10 Administration

Other studies have not shown a clear benefit of coenzyme Q10 administration. In a study of 11 trained male triathletes, those receiving three daily doses of 100 mg of coenzyme Q10 did not significantly improve time to exhaustion.18 The trial was designed as a double-blind, crossover trial consisting of two 4-week periods of treatment separated by a 4-week washout between the 2 treatment periods. However, the study was confounded by the fact that other nutrients were also administered to the group receiving coenzyme Q10, including cytochrome C, inosine, and vitamin E. Additionally, the placebo that was administered, dicalcium phosphate, may have had an influence on the athletic performance of the control group. Finally, the coenzyme Q10 dose of only 100 mg three times daily may have been somewhat low. Coenzyme Q10 levels were not evaluated in this study; it is not known whether therapeutic levels were obtained from the dose of coenzyme Q10 that was administered.

A randomized, double-blind, placebo-controlled, crossover study of 11 young athletes and 8 older athletes receiving 120 mg of coenzyme Q10 per day failed to show any significant difference in V̇o2max following treatment with coenzyme Q10.19 The study consisted of two 6-week treatments periods separated by 4-week washout.

Finally, the mixed results of the studies are underscored by a 2003 review by Rosenfeldt et al,20 who found 6 studies showing improvement in exercise capacity with coenzyme Q10 supplementation and 5 studies showing no improvement. In the most recent extensive review, Marcoff and Thompson21 concluded that, given the conflicting nature of the studies, there is insufficient evidence to routinely recommend coenzyme Q10 therapy for statin-induced myopathy. However, they concluded that this therapy could be strongly considered, particularly considering the possible benefit and its low potential for side effects. They called for more research into this topic.

Go to:MEASUREMENTS OF MITOCHONDRIAL FUNCTION

The studies measuring activity of coenzyme Q10 on mitochondrial function have used a number of end points. Many studies have used V̇o2max as a measure of oxygen consumption. Some studies have used muscle biopsy and endomyocardial tissue to assess in vitro contractile ability with and without coenzyme Q10 availability. Phosphorus 31 and magnetic resonance spectroscopy have also been used to assess ATP activity. A study evaluating the effect of 100 mg of coenzyme Q10 on muscle energy metabolism, using magnetic resonance spectroscopy in middle-aged, postpolio patients showed a significant benefit.22Magnetic resonance spectroscopy has been used in a variety of settings to assess mitochondrial dysfunction in a number of human mitochondrial diseases.2

Muscle biopsy can allow for a number of sophisticated DNA, biochemical, and pathologic studies. Its use is more limited because of its more invasive nature.

Noninvasive measures of mitochondrial dysfunction can be performed by blood plasma evaluations of anaerobic metabolism. In mitochondrial dysfunction, ATP production is limited, thereby forcing the muscle fiber to apply anaerobic metabolism, resulting in increased lactate production. Exercise intolerance and fatigue thereby ensue.2 Venous lactate-pyruvate ratios have been shown to be clinically helpful in the evaluation of mitochondrial diseases. Chan et al23 reported that the venous lactate-pyruvate ratio improved after 6 months of supplementation with coenzyme Q10 in patients with mitochondrial myopathies.

Additionally, various exercise testing modalities provide an important method in determining mitochondrial dysfunction. Lactic acid, V̇o2max, aerobic threshold, and anaerobic threshold levels can be obtained by cardiopulmonary exercise testing and are helpful in diagnosing a variety of cardiopulmonary disorders.24 A number of protocols exist for assessing workload and mitochondrial function using treadmill testing and bicycle ergonometry. However, there is no unique mitochondrial dysfunction protocol for exercise testing. The relationship between oxygen utilization by the mitochondria and oxygen uptake by the lungs is shown in Figure 3 and provides the scientific basis for cardiopulmonary exercise testing.

THE ELDERLY POPULATION

Aging has a number of effects on both muscle activity and cardiac activity, including being associated with a steady decline in the absolute heart rate, which can translate into a decrease in cardiac output. Aging is also associated with a gradual decline in V̇o2max. The elderly also undergo a steady loss of fast-twitch muscle fibers with a relative increase in the proportion of slow-twitch fibers (type 1 fibers).25 Coenzyme Q10 supplementation in rats caused a significant increase in the coenzyme Q10 levels of slow-twitch fibers. Supplementation also reduced exercise-induced muscle injury by enhancing stabilization of the muscle cell membrane.26

Additionally, the elderly appear to be at greater risk for statin-induced myopathies, which can occur in up to 11% of these patients. Statins are commonly used medications in the elderly, given the high incidence of medical problems such as diabetes, cerebrovascular disease, and cardiovascular disease in this population.27,28

Go to:SUMMARY

Coenzyme Q10 is an important component of mitochondrial biochemistry, allowing for the production of ATP. HMA Co-A reductase inhibitors or statins inhibit one of the key steps in coenzyme Q10 synthesis. These drugs have been associated with a reduction in serum and muscle tissue coenzyme Q10 levels and may play a role in statin-induced myopathy. Given the low risk of toxicity and the potential benefit in treating statin-induced myopathy, a trial of 200 mg of coenzyme Q10 daily should be considered for these patients.

The elderly appear to be more susceptible to coenzyme Q10 deficiency. Athletes, who require the most efficient use of oxygen consumption by mitochondria for athletic performance, are also susceptible to mitochondrial dysfunction due to coenzyme Q10 deficiency. However, study results have been conflicting regarding the uniform effectiveness of coenzyme Q10 supplementation.

A population that would appear to gain the most benefit from coenzyme Q10 supplementation would be that population with all the mentioned characteristics. An elderly population of athletes receiving HMA Co-A reductase inhibitors would appear to be ideally suited to experiencing the greatest benefit from coenzyme Q10 supplementation, given the high risk of mitochondrial dysfunction from coenzyme Q10 deficiencies in this group.

Abstract

Statin-associated adverse effects, primarily muscle-related symptoms, occur in up to approximately one-third of patients in clinical practice. Recently, a Canadian Consensus Working Group outlined 6 key principles to assess and manage patients with goal-inhibiting statin intolerance, defined as a syndrome characterized by symptoms or biomarker abnormalities that prevent the long-term use of and adherence to indicated statin therapy, which includes a trial of at least 2 statins and precludes reversible causes of statin adverse effects. These principles ensure patients are appropriately receiving a statin and aware of both the benefits and risks of therapy. As well, they address factors that may increase the risk of statin-associated myopathy. A thorough assessment of patients’ clinical and laboratory history should be performed in any patient presenting with muscle symptoms on statin therapy, followed by a systematic dechallenge/rechallenge approach. In practice, most patients with statin intolerance due to muscle symptoms will be able to tolerate another statin. This is of particular importance because of the relative paucity of compelling evidence demonstrating a cardiovascular benefit with nonstatin therapies. Pharmacists are ideally situated to provide patient education, recommend changes to therapy and monitor patients with goal-inhibiting statin intolerance.

Introduction

Hydroxymethylglutaryl-coenzyme A reductase inhibitors (statins) are recommended in patients with established, or those at risk for, cardiovascular disease (CVD) to decrease the incidence of adverse cardiovascular events and mortality.1 Statin intolerance due to adverse effects, mostly muscle-related symptoms, is common in clinical practice. It is estimated that between 7% and 29% of patients taking statins experience some type of muscle-associated toxicity in real-world practice.2 This frequent clinical scenario led to the formation of a Canadian Consensus Working Group (CCWG) to evaluate the diagnosis, prevention and management of statin adverse effects and intolerance. In 2016, this group released the third iteration of a statement to inform and guide clinicians on how to approach patients with suspected statin-associated adverse effects.3

Go to:Terminology

The CCWG proposed terminology to provide consistency when describing muscle-related adverse effects3(Table 1). Myopathy is a general term that applies to any disease of the muscle, which is further subcategorized depending on whether the patient is symptomatic. Myalgia refers to patients with symptoms of muscle aches or weakness in the absence of an elevated creatine kinase (CK), whereas myositis is the presence of symptoms with a CK elevation. Rhabdomyolysis is muscle aches or weakness with a CK more than 10 times the upper limit of normal and is typically associated with myoglobinuria. The term hyperCKemia refers to patients with an elevated CK without muscle-related symptoms and is graded based on the degree of elevation.

In the most recent version of the CCWG statement, the term goal-inhibiting statin intolerance (GISI) was introduced as a pragmatic way of defining statin intolerance. GISI was defined as a syndrome characterized by symptoms or biomarker abnormalities that prevent the long-term use of and adherence to indicated statin therapy, which includes a trial of at least 2 statins (including atorvastatin and rosuvastatin, as appropriate) and precludes reversible causes of statin adverse effects (e.g., drug interactions, untreated hypothyroidism). The working group also suggested the term goal-inhibiting statin resistance (GISR), which refers to patients who are unable to achieve the expected lipid-lowering effect with maximally tolerated statin dose. In other words, GISR refers to an inadequate response or lack of efficacy, as opposed to adverse effects. The use of lipid targets as a therapeutic goal is controversial. However, guidelines from both the Canadian Cardiovascular Society (CCS)1 and the American College of Cardiology/American Heart Association4 recommend a specific lipid goal or expected percentage reduction in lipid parameters—even the simplified lipid guidelines published in Canadian Family Physician endorse a specific statin dose based on the patient’s level of risk.5

Management of GISI

In the most recent statement, the CCWG advocate 6 key principles in the management of patients with GISI3 (Figure 1). The following is a fictional case vignette of a typical clinical scenario to help illustrate these concepts.

Case vignette

A 62-year-old man presents to your pharmacy to refill his medications but admits he is reluctant to refill his statin because he is having frequent leg pains and read on the Internet that this is a common side effect of statins. He is currently taking atorvastatin 20 mg daily. His medical history includes type 2 diabetes mellitus, hypertension, dyslipidemia and chronic back pain. His other medications include metformin, an angiotensin-converting enzyme inhibitor, dihydropyridine calcium channel blocker and acetaminophen as needed. He is wondering if diet and exercise can replace his statin. He also read on the Internet that coenzyme Q10 and vitamin D help to reduce muscle pain caused by statins and wonders if he should start supplementation.

Step 1: Ensure the patient has a valid indication for statin therapy

The first step in approaching a patient with suspected statin-associated muscle symptoms is to confirm the patient has a valid indication for therapy. The most recent CCS guidelines for the management of dyslipidemia describe 5 “statin-indicated conditions” in which statins have been demonstrated to be efficacious.1 These include patients with established CVD (i.e., secondary prevention), certain patients with diabetes mellitus or chronic kidney disease and patients with an aortic abdominal aneurysm or genetic dyslipidemia (i.e., low-density lipoprotein cholesterol >5 mmol/L). As well, statin therapy is recommended for high-risk primary prevention patients and certain patients at intermediate-risk for CVD depending on their risk factors, as well as their values and preferences. The CCS guidelines advocate the use of the Framingham Risk Score to determine a patient’s cardiovascular risk in the setting of primary prevention.1

Case 

Your patient has a statin-indicated condition, as he has type 2 diabetes mellitus and is over 40 years of age.

Step 2: Identify factors that may limit or preclude use of statins

There are several factors that predispose patients to muscle-related adverse effects. These include advanced age (>80 years), female sex, Asian ethnicity, low body mass index, history of preexisting/unexplained muscle/joint pain, family history of myopathy, neuromuscular diseases (e.g., amyotrophic lateral sclerosis), severe renal/hepatic disease, untreated hypothyroidism and diabetes mellitus.3 In addition, there are several genetic polymorphisms that are associated with statin myopathy; however, these are not routinely investigated in clinical practice. There are several statin drug interactions that increase the risk of adverse effects.3,6 Atorvastatin, lovastatin and simvastatin are metabolized by the cytochrome P450 3A4 enzyme system and thus interact with strong inhibitors of this system (e.g., azole antifungals, large quantities of grapefruit juice, macrolide antibiotics, nondihydropyridine calcium channel blockers and protease inhibitors). As well, amiodarone may interact with any statin except pravastatin, and concomitant use of fibrates and niacin with a statin may also increase the risk of myopathy. High-dose statin therapy is more likely to be associated with statin muscle symptoms, although these symptoms can occur with any dose. Finally, red yeast rice supplements may increase statin toxicity, as they contain a compound called monacolin K, which is identical to lovastatin.

Case 

Your patient has a history of preexisting muscle pain (i.e., chronic back pain) and diabetes mellitus, which are nonmodifiable risk factors for statin-associated myopathy. However, he is not taking any interacting medications (including supplements) and does not drink grapefruit juice.

Step 3: Ensure the patient is fully informed regarding the benefits and risks of statin therapy

There is an unmet opportunity for pharmacists to aid patients in making an informed decision about both the benefits and risks of statin therapy, as it was stated by the CCWG that this is “often poorly executed by busy physicians or inaccurately interpreted by the patient.”3 Pharmacists should be wary that focusing only on possible adverse effects without a discussion of the benefits is potentially detrimental. As well, patients often identify unsubstantiated and/or misleading information on the Internet, where there is a seemingly disproportionate amount of negative information about statins as compared with other medications. To facilitate shared decision-making, it is imperative to understand the numbers regarding the potential benefits and risks of statin therapy. A meta-analysis of 10 randomized controlled trials of >70,000 patients without established CVD (i.e., primary prevention) showed that statin therapy, as compared with placebo, reduced all-cause mortality by 0.6% (number needed to treat [NNT] of 167), major coronary events by 1.3% (NNT of 77) and stroke by 0.4% (NNT of 250) over approximately 4 years.7 The authors of this meta-analysis did not investigate safety outcomes other than the risk of cancer, which was not significantly different between groups. In patients with stable coronary heart disease (i.e., secondary prevention), high-dose atorvastatin, as compared with low dose, reduced death and adverse cardiovascular events by 2.2% (NNT of 46) over approximately 5 years.8 However, treatment-related adverse events and drug discontinuation due to treatment-related adverse events were higher with high-dose atorvastatin (number needed to harm [NNH] of 44 and 53, respectively). Treatment-related myalgia was similar between groups, but patients on high-dose atorvastatin had a higher rate of persistently elevated alanine aminotransferase and/or aspartate aminotransferase (NNH of 100). Of note, the total rates of adverse events (i.e., not just those categorized as related to treatment) were not reported. In patients with a recent acute coronary syndrome, high-intensity statin therapy reduced death and adverse cardiovascular events by 3.9% (NNT of 26) when compared with moderate-dose therapy over approximately 2 years.9 There were no significant between-group differences for any of the prespecified safety endpoints including liver enzyme elevation, myopathy or cancer. Although neither trial reduced the risk of all-cause mortality, this may have been due to the use of an active treatment in both groups.

There is a risk of a so-called “nocebo effect” with statins, where patients with negative expectations about a therapy are more likely to experience an adverse effect. A recent retrospective analysis of a large statin randomized controlled trial demonstrated that patients reported a higher incidence of muscle-related adverse events with atorvastatin compared with placebo during the unblinded extension phase.10 In other words, patients reported more muscle-related adverse effects when they knew they were taking a statin.

Case 

You discuss with your patient the potential cardiovascular benefits of statin therapy compared with the potential risks. With respect to efficacy, you discuss with him the results of the Collaborative Atorvastatin Diabetes Study (CARDS), which compared atorvastatin 10 mg daily to placebo in patients aged 40 to 75 years with type 2 diabetes mellitus and another cardiovascular risk factor, such as hypertension.11 The results showed atorvastatin reduced the risk of adverse cardiovascular events (NNT of 32) and all-cause mortality (NNT of 67) compared with placebo, while about 5% of patients in each group reported muscle-related symptoms. He acknowledges these results seem important enough to continue treatment but not at the expense of having ongoing leg pain.

Step 4: Encourage dietary interventions and exercise to lower the patient’s cardiovascular risk and do not advocate supplements to reduce the risk of statin-associated myalgia

Lifestyle interventions, such as a heart healthy diet and exercise, are important components to reduce patients’ cardiovascular risk but are not as effective as and, in most cases, should not be viewed as a replacement for, statin therapy. Rather, these interventions should be used to augment therapy to minimize the statin dose or potentially avoid treatment in patients without established CVD. Recommendations should include a Mediterranean diet or increasing dietary intake of phytosterols or phytosterol-containing foods and 150 minutes of moderate-to-vigorous aerobic activity per week in bouts of 10 minutes or more.1

No vitamins or herbal supplements have definitively been shown to reduce the risk or improve the symptoms of statin-associated myopathy. The data for coenzyme Q10 supplementation in treating or preventing statin-associated muscle symptoms is inconsistent. A recent meta-analysis of 5 randomized controlled trials of patients with statin-associated myalgia did not demonstrate a significant improvement in pain scores with coenzyme Q10 vs placebo.12 As well, there are no data to support that coenzyme Q10 improves clinically meaningful outcomes, such as a reduction in incident myalgia or statin tolerability.13Myalgias have been reported in patients on statin therapy with low serum levels of vitamin D. However, these data are heterogeneous, and there are no randomized controlled trials of vitamin D supplementation to prevent or treat statin-associated myalgia.3 However, if a patient perceives benefit with coenzyme Q10 or vitamin D, it is not unreasonable to continue therapy. As previously mentioned, red yeast rice should not be used concurrently with a statin, as this supplement often has variable concentration and purity and thus may increase the risk of statin toxicity.

Case 

You recommend that your patient implement changes to his diet and exercise routine but explain these interventions are not adequate to replace his statin. You discuss with him the lack of evidence demonstrating a consistent benefit with coenzyme Q10 or vitamin D and that you would not recommend taking either supplement.

Step 5: Use a systematic challenge/dechallenge/rechallenge approach to patients with GISI

If statin-associated myopathy is suspected, it is essential to perform a thorough evaluation of the patient’s clinical and laboratory history and provide patient education as appropriate. Muscle symptoms are common and can be due to alternate causes such as exercise, soft-tissue disease, trauma or acute flu-like illness. Muscle complaints with statins are often described as a heaviness, stiffness, cramping or weakness during exertion, which are intermittent rather than continuous.14 These complaints are typically bilateral and affect multiple large muscle beds. If a patient has symptoms of isolated joint pain, it is unlikely to be secondary to the statin. Not all statins have the same incidence of myopathy—lipophilic statins (e.g., atorvastatin, fluvastatin, lovastatin, simvastatin) penetrate the muscle more readily and may have a greater potential for myopathy.15 The timing of muscle symptoms relative to when the statin was started or dose increased is an important component of the history. In a retrospective case series, patients were taking statins for an average of 6 months before developing muscle-related symptoms.16 The CCWG proposed a modified version of the statin-associated muscle symptoms (SAMS) score that can be used to assess patients’ symptoms3 (Table 2). Baseline bloodwork should be performed in all patients prior to initiating statin therapy—this should include lipid profile, glycosylated hemoglobin (to diagnose/rule out diabetes mellitus), CK (as some patients have an elevated CK based on their ethnicity, sex or physical activity), serum creatinine, alanine aminotransferase and thyroid-stimulating hormone (as kidney/liver disease or untreated hypothyroidism are risk factors for statin-associated myopathy).3 Routine monitoring of CK is not recommended; however, it is important to measure a baseline in case the patient develops muscle symptoms.

A variety of strategies can be used in the management of patients with statin-related myopathy. It is essential to balance the potential benefits and risks of continuing therapy on a case-by-case basis by incorporating the patient’s values and preferences. The CCWG created a systematic, algorithmic approach to managing statin-associated muscle symptoms or hyperCKemia3 (Figure 2). Systematically following this algorithm can be time-consuming and potentially frustrating for both patients and clinicians; however, it is important to stay resolute because of the paucity of therapeutic alternatives with evidence to support a reduction in adverse cardiovascular events. Based on published evidence, it is estimated that 70% to 90% of patients with statin intolerance will subsequently be able to tolerate another statin.3,17,18 In a retrospective case series of patients with statin-associated myopathy, 16 of 37 patients (43%) were able to tolerate another statin without recurrent symptoms.16 Some patients may actually tolerate a rechallenge of the same statin. A recent N-of-1 trial of 8 patients with statin-related myalgia were rechallenged with the same statin that was previously associated with myalgia vs placebo in a double-blind fashion.19 There was no significant difference in pain scores between the statin and placebo, and 5 patients (63%) were able to resume statin therapy after the trial. The appropriate time frame for follow-up is variable, depending on the patient and situation. Generally, it is recommended that patients be reassessed in 6 to 12 weeks if their CK is within normal limits or have mild (grade 1) hyperC- Kemia, although more frequent follow-up is likely warranted if the patient develops or has persistent symptoms.3 In cases of statin-associated myositis or moderate/severe hyperCKemia, the CCWG does not recommend a specific time frame for follow-up of CK and symptoms, but every 3 to 6 weeks is likely a reasonable time frame.3 In a retrospective case series of patients with statin-associated myopathy, resolution of muscle pain occurred an average of 2 months after discontinuing therapy.16

The CCWG outlined factors for which statin rechallenge may not be useful.3 Such cases include when the symptoms are plausible and resolve completely with cessation, are severe with objective weakness and/or hyperCKemia or if the patient refuses a rechallenge even at a low or intermittent dose. If the symptoms persist after discontinuing the statin for a reasonable period of time, this may be indicative of an underlying non–statin-related muscle condition, which should warrant further investigation but may not preclude future statin use.

Myalgia 

Not all patients with myalgia will require cessation or reduction of their statin therapy. If the symptoms are tolerable, one could consider continuing therapy (if the patient is willing) and have them monitor their symptoms. If the symptoms are intolerable, the statin dose could be reduced (including intermittent or nondaily dosing), switched to a different statin or temporarily discontinued. Intermittent nondaily dosing strategies include every other day or weekly dosing, which some patients may prefer, although adherence could be more challenging. A review of nondaily statin dosing strategies showed that at least 70% of patients with statin-associated myopathy were able to tolerate an intermittent dosing strategy.20 Addressing any reversible risk factors for myopathy is also paramount. If a patient has myalgia with a lipophilic statin (e.g., atorvastatin, simvastatin), it may be reasonable to switch to a less lipophilic statin (e.g., pravastatin, rosuvastatin).

Myositis 

The management of myositis follows the same principles as for myalgia regarding dechallenge/rechallenge and switching statins. In most cases, it is appropriate to hold the statin until the hyperCKemia resolves and the patient is asymptomatic.

Rhabdomyolysis 

In patients presenting with suspected rhabdomyolysis, statin therapy should be discontinued with prompt assessment of the patient’s renal function (i.e., serum creatinine, urine myoglobin). Rhabdomyolysis is typically managed with intravenous rehydration and, in severe cases, dialysis. Previous rhabdomyolysis is not necessarily a contraindication to statin therapy, and patients should be assessed on a case-by-case basis based on their individualized benefit vs risk.3 There appears to be a dose-dependent relationship with statin-associated rhabdomyolysis.21 Therefore, if a statin is rechallenged, it should be started at a low dose.

Case 

Your patient describes bilateral muscle cramping in his thighs. These symptoms started about 2 weeks ago. He has been taking atorvastatin for about 8 weeks, and there has been no recent change to his physical activity. His is referred for prompt bloodwork, which shows his CK is below the upper limit of normal. His muscle cramping is currently tolerable but quite bothersome. After discussing his options, he is willing to try a lower dose of atorvastatin at 10 mg daily. After 2 weeks, his muscle cramps are not improved. Therefore, you recommend he discontinue atorvastatin, and after a week his muscle cramps resolve. You then recommend he start rosuvastatin 10 mg daily, and after 6 weeks, he has had no recurrence of his muscle pain. Thus, his modified SAMS score is 7, which indicates that statin-associated myalgia was possible.

Step 6: If necessary, recommend nonstatin therapy to achieve the therapeutic goal

If a nonstatin drug is considered to achieve a lipid goal, the CCWG advocates that dual therapy is preferable to avoid unnecessary polypharmacy. Few therapeutic alternatives are available for patients with GISI or complete statin intolerance. While many nonstatin lipid-lowering agents have demonstrated a beneficial effect on lipid parameters, some have failed to show a reduction in adverse cardiovascular events when used in combination with a statin and thus should be avoided. Specifically, the combination of a fibrate or niacin with a statin has been demonstrated in large randomized controlled trials to be ineffective or, in the case of niacin, potentially harmful.22-25 On the other hand, there are recent data demonstrating that both ezetimibe and evolocumab (a novel proprotein convertase subtilisin/kexin type 9 [PCKS9] inhibitor), respectively, lower the risk of adverse cardiovascular events in combination with a statin in patients with established CVD.26,27 There are also short-term trials of ezetimibe vs a PCSK9 inhibitor in patients with complete statin intolerance. These trials were not adequately powered to evaluate cardiovascular events, and muscle-related symptoms were similar (approximately 20%-30% of patients) in each group.28,29

Case 

Six weeks after initiating rosuvastatin, your patient has a nonfasting lipid profile that demonstrates a non–high-density lipoprotein cholesterol (non-HDL-C) level of 2.8 mmol/L, which is above the target recommended by the CCS (<2.6 mmol/L). He does not meet the inclusion criteria for the studies that investigated ezetimibe or PCSK9 inhibitors with a statin and is reluctant to start an additional therapy. However, he is willing to try a higher dose of rosuvastatin. Therefore, you recommend an increase in his rosuvastatin dose to 20 mg daily. Six weeks later, he reports no recurrent myalgias, and his non–HDL-C is now below target.

Conclusion

Statin-associated muscle symptoms are common, occurring in up to roughly 30% of patients in clinical practice. Recently, a Canadian working group outlined 6 key principles to manage patients with statin intolerance. This systematic approach ensures patients are appropriately receiving a statin and are aware of both the benefits and risks of therapy and addresses factors that may increase their risk of statin-associated myopathy. A thorough assessment of patients’ clinical and laboratory history should be performed in any patient presenting with muscle symptoms on statin therapy, followed by a systematic dechallenge/rechallenge approach. In practice, most patients with statin intolerance due to muscle symptoms will be able to tolerate another statin. This is of particular importance because of the absence of compelling evidence demonstrating a cardiovascular benefit with nonstatin therapies. Pharmacists are ideally situated to provide patient education, recommend changes to therapy and monitor patients with goal-inhibiting statin intolerance.

Statins are widely used besides their myopathic side effects, ranging from mild myalgia to fatal rhabdomyolysis. Resveratrol is one of the most popular over the counter products used for similar purposes with statins. The aim of this study was to elucidate the myopathic effects of atorvastatin and coadministered resveratrol in male rat skeletal muscle via morphological analyses and immunohistochemistry studies. Control group received 1.5 ml of drinking water by oral gavage and 1 ml 15% ethanol (vehicle of resveratrol) i.p. for 14 days daily; atorvastatin group was treated with 40 mg/kg atorvastatin by oral gavage and 1 ml 15% ethanol i.p. for 14 days daily. Resveratrol + atorvastatin group was treated with 40 mg/kg atorvastatin by oral gavage and 20 mg/kg i.p resveratrol for 14 days daily. Atorvastatin treatment resulted with a moderate inducible nitric oxide synthase (iNOS) and endothelial nitric oxide synthase (eNOS) immunoreactivity in nucleus and strong immunoreactivity in fibers. Control group and resveratrol + atorvastatin group showed weak iNOS and eNOS immunoreactivity in nucleus and moderate immunoreactivity muscular fibers. Treatment with atorvastatin resulted in a significantly shortened fibrils, and resveratrol co-treatment reversed this effect. Resveratrol and atorvastatin co-treatment could be an alternative treatment to prevent the myositis adverse effects of atorvastatin on skeletal muscle.

INTRODUCTION

5-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase reaction is the rate limiting step of cholesterol biosynthesis and thus the primary mechanism of action of HMG-CoA reductase inhibitors (statins) is to lower cholesterol levels (Evans and Rees, 2002). Statins are widely used in the prevention of cardiovascular events. Although they are generally well tolerated, different grades of myopathy, ranging from mild myalgia to fatal

rhabdomyolysis has been reported (Abourjaily et al., 2003; Omar et al., 2002; Graham et al., 2004). The most serious risk associated with statins is myositis with rhabdomyolysis. The incidence of rhabdomyolysis has been estimated to be 0.44 to 0.54 cases per 10,000 person/years (Shek and Ferrill, 2001; Graham et al., 2004; Arora et al., 2006). The prevalence of milder muscle complaints like myalgia has been reported by statin

users range as high as 20% (Foley et al., 2004; O’Meara et al., 2004; Buettner et al., 2008). Some authors have also reported the incidence of myalgia approximately in 5 to 7% of all patients on statins (Arora et al., 2006). Although the risk of rhabdomyolysis with currently mar- keted statins is very low, symptomatic muscle weakness and pain are much more frequent. Several possible mechanisms such as depletion of secondary metabolic intermediates, induction of apoptosis and alterations of chloride channel conductance within myositis have been proposed for statin-associated myopathy (Pierno et al., 2009). But the reason why myopathy develops in some patients as a result of statin treatment is still not well understood and this side effect prevents patients and their physicians from complying with statin therapy guidelines.

Hypercholesterolemic patients may sometimes direct to various dietary components and natural compounds to regulate serum lipid concentrations because of these doubts on the safety of statins. Resveratrol is one of the most popular over the counter (OTC) products used for similar purposes. Resveratrol (trans-3,5,4’- trihydroxystilben) is a polyphenol (phytoalexin) naturally found mostly in red wine and different therapeutic plants. By in vitro experiments, it has been shown that the cardiovascular protective effects of resveratrol might be through a variety of mechanisms such as inhibition of smooth muscle cells proliferation, platelet aggregation, and the oxidation of low-density lipoprotein (LDL) cholesterol. Resveratrol also reduces the synthesis of lipids and eicosanoids, which promote inflammation and atherosclerosis (Soner et al., 2010). Such multiple protective effects of resveratrol increase its demand as an OTC product even for statin users.

The present study was designed to elucidate the effects of combined treatment of resveratrol on ator- vastatin-induced myopathy in male rat skeletal muscle via morphological analyses and immunohistochemistry studies.

MATERIAL AND METHODS

Animals and experimental protocol

Three groups of male Wistar-albino rats of 8 weeks, weighing 260 to 280 g were used in the experiments. Animals were housed identically in cages in an air conditioned room under a 12 h light dark cycle. Temperature and humidity were controlled within the limits 21 ± 2°C and 55 ± 15% relative humidity (RH). All animals became acclimatized for at least 7 days before the outset of the study. A standard diet and tap water were provided ad libitum. The experimental protocols were approved by the Animal Ethics Committee of Selcuk University, Meram Medical School. Control group received 1.5 ml of drinking water by oral gavage and 1 ml 15% ethanol (vehicle of resveratrol) i.p. for 14 days daily (n = 8); atorvastatin group was treated with 40 mg/kg atorvastatin

(Lipitor®, prepared daily and dissolved in drinking water) by oral gavage and 1 ml 15% ethanol i.p. for 14 days daily (n = 6). Resveratrol + atorvastatin group was treated with 40 mg/kg atorvastatin by oral gavage and 20 mg/kg i.p resveratrol for 14 days daily (n = 8). Rats were weighted every 5 days to rearrange dosing schedule and observed every day or as necessary. When the body weight loss exceed 15% of day 1, rats were excluded from the study group. On the 15th day, the following muscle tissues were sampled: trapezius, gastrocnemius, semitendinosus and biceps femoris.

Tissue processing and immunohistochemistry

Paraformaldehyde fixation of tissues continued for 24 h, at 4°C and processed for embedding in paraffin wax using routine protocols. 5 μm-thick coronal sections were cut using a microtome (Leica MR 2145, Heerbrugg, Switzerland); they were then dewaxed and rehydrated through a graded ethanol series using routine protocols. Sections were then washed with distilled water and phosphate buffered saline (PBS) for 10 min, then treated with 2% trypsin (Sigma Chemical Co., St. Louis, Missouri, USA) in 50 mM Tris buffer (pH 7.5), at 37°C, for 15 min. Sections were delineated with a marker pen (Dakopen, Glostrup, Denmark) and incubated in a solution of 3% H2O2 for 15 min to inhibit endogenous peroxidase activity. To reduce non-specific background staining, slides were incubated at room temperature for 30 min in 0.3% bovine serum albumin/1 × Tris-buffered saline. Then, sections were incubated with primary antibodies directed against inducible nitric oxide synthase (iNOS) (1:100 dilution; Abcam, Cambridge, UK), endothelial nitric oxide synthase (eNOS) (1:200 dilution; Abcam, Cambridge, UK) for 18 h at 4°C in a humid chamber. Sections were then incubated with biotinylated secondary antibody and then with streptavidin conjugated to horseradish peroxidase (Histostain plus peroxidase kit, Zymed Laboratories Inc., South San Francisco, CA, USA) for 30 min in accordance with the manufacturer’s instructions. Finally, sections were incubated with diaminobenzidine (DAB) for 5 min to reveal immunolabelling. All dilutions and thorough washes between stages were performed using PBS. Sections were counterstained with Mayer’s hematoxylin (Sigma Chemical Co., St. Louis, Missouri, USA).

After washing with tap water, sections were dehydrated through a graded ethanol series, cleared in xylene and mounted with entellan. Negative control samples were processed as described except that primary antibodies were omitted and replaced with PBS alone. Positive controls were represented by sections of a neuroblastoma specimen known to be positive for the markers of interest. The intensity of iNOS and eNOS immunohistochemical stainings was graded semiquantitatively according to the nuclear and cytoplasmic immunoreaction in sections as follows: (-) no immunostaining, (+) weak staining, (++) moderate staining, (+++) strong staining. Light microscope, equipped with a camera (Olympus BX-51 and Olympus C-5050 digital camera, Olympus Co., Tokyo, Japan) was used. The slides were examined by two investigators.

Morphometrical analysis

Longitudinal sections were cut and stained with Gomori’s trichrome. The first and second of the three consecutive serial sections were omitted and 12 sections from each subject were taken for quantification from the third. Preparations were screened systematically using a random start. Visualization of specimens at ×40 magnification was started from the top right corner of the preparation. Sections were examined under a light microscope

(Olympus BX-51, Olympus, Tokyo, Japan) equipped with a camera (Olympus C-5050 digital camera, Olympus). Image was transferred to a desktop computer system and length of muscle fibrils determined using image analysis software program (Image-Pro Express, Media Cybernetics, Bethesda, MD) for morphometric analysis. Evaluation of the specimen was performed by an experienced histologist blinded to the surgical groups.

Statistical analysis

The statistical significance of differences of groups was analyzed by one-way analysis of variance (ANOVA) or Student’s t-test. p-Values of < 0.05 were considered significant.

RESULTS

Observations

After the 10th day of atorvastatin treatment, rats showed piloerection, hunched posture, thin appearance with weight loss, pale appearance and decreased activity. Control group and resveratrol + atorvastatin group showed no significant alterations.

Immunoreactivity

Control group showed weak iNOS immunoreactivity in nucleus and moderate immunoreactivity in muscular fibers. 40 mg/kg atorvastatin for 14 days has resulted with a moderate immunoreactivity in nucleus and strong immunoreactivity in fibers. Co-administration of 20 mg/kg i.p resveratrol for 14 days with atorvastatin elicited a negative immunostaining in nucleus and weak immunoreactivity in fibers (Figure 1a, b and c). eNOS immunoreactivity results were similar in control group and in resveratrol + atorvastatin group, with weak immuno- reactivity in nucleus and weak in fibers. On the other hand, in atorvastatin group, moderate nuclear immuno- staining and weak muscle fiber staining were observed (Figure 1d, e and f).

Structural modification of rat skeletal muscles

Our results showed that the treatment of rats with atorvastatin have resulted with a significantly shortened muscle fibrils in trapezius, gastrocnemius, semitendi- nosus and biceps femoris muscles (20.48 ± 0.91, 18.07 ± 1.6, 19.53 ± 1.1, 20.52±1.0 μm, respectively) when com- pared with control group (34.22 ± 1.07, 32.01 ± 1.52, 33.0 ± 2.10, 31.65 ± 1.0 μm, respectively) (p < 0.05 for trapezius, gastrocnemius, semitendinosus and p < 0.01 for biceps femoris). Co-treatment with resveratrol preven- ted the shortening of muscle fibrils caused by statin treat- ment alone. Fibril lengths were 29.98 ± 2.4, 28.32±1.6, 27.84±1.9, 40.10±0.8 μm in resveratrol + atorvastatin treated groups for trapezius, gastrocnemius, semitendinosus and biceps femoris muscles, respectively (Figure 2a and b). A strong correlation between the atorvastatin and shortening of muscle fibrils has been found, and treatment with resveratrol has prevented the effect of atorvastatin.

DISCUSSION

Indications for statin therapy as recommended by The National Institute for Health and Clinical Excellence were defined as: ischemic heart disease (angina, myocardial infarction, chronic heart disease), cerebrovascular accidents (transient ischemic attack), hypercholestero- lemia, peripheral arterial disease and combination of risk factors (hypertension, type 2 diabetes mellitus, age > 70 years). They also recommend that adults with a history of cardiovascular disease (CVD) and adults with a 10-year risk of developing CVD equal to or greater than 20% should start statin therapy as primary prevention (NICE, 2008). These indications make statins one of the world’s most prescribed drugs.

Effects of statins on early cellular changes, inflammation markers, have been shown in rats. Our study has also evaluated muscle fiber lengths in early myositis in rats. Atorvastatin treatment has significantly shortened the fibrils of skeletal muscles. Resveratrol has completely reversed the shortening of fibrils, showing its protective effect on atorvastatin induced myositis.

Similar effect of statins has been shown in zebra fish and this effect has been attributed to an inhibition on biosynthetic pathway or an impaired production of mevalonate caused by statins. Such a destructive effect on myosin filament might result with impaired muscular functions which can be projected into clinical symptoms of myopathy (Huang et al., 2011). To our knowledge, this is the first study showing the shortening of myosin fibril in a rat toxicity model for skeletal muscles with deleterious muscle manifestations induced by atorvastatin. Since some muscle biopsies have documented statin-associated myopathy with normal creatine kinase (CK) as a reliable biomarker for muscle diseases has consequently been questioned (Gunst et al., 1998; Phillips et al., 2002).

CPK levels do not invariably correlate with clinical symptoms of myopathy because elevated CPK values can be associated with various diseases. Myosin fibril length could also be a marker for myositis in patients with normal CK levels. A change of the myosin fibril length associated with myopathy might allow a diagnosis of myopathy before the occurrence of clinical and laboratory symptoms such as elevated CPK levels. All isoforms of nitric oxide synthase are expressed in skeletal muscle of all mammals (Stamler and Meissner, 2001) and it has been shown that inflammatory cytokines upregulates iNOS in myositis (Williamset al., 1994; Park et al., 1996).

Our results showed that atorvastatin caused an increase in iNOS and eNOS and when resveratrol was co-administered, iNOS and eNOS upregulation has been recovered to control group. This possible down regulation of iNOS in resveratrol + atorvastatin group can be attri- buted to the effect of resveratrol which has been shown to inhibit iNOS induction in skeletal muscle (Centeno-Baez et al., 2011). Anti-inflammatory effects of resveratrol have been shown in several studies but for the first time, we have shown the effect of resveratrol on iNOS regulation when co-administered with atorvastatin.

Myotoxic effects of atorvastatin are known to be dose- dependent. As pointed out, impaired metabolism of sta- tins, pharmacokinetic interactions and/or genetic effects are all probable causes of the myotoxicity (Laaksonen, 2006). Co-administered agents effecting on similar metabolic pathways (eg fibrates) increases their myotoxic side effects besides their lipid-lowering action (Ballantyne et al., 2003). An alternative molecule that has the same protecting effect by a different mechanism is needed to decrease the adverse effects of atorvastatin without affecting its effects. With further studies, resveratrol could be an alternative to decrease the effects of statins on skeletal muscle.

Due to zinc's high distribution in muscle tissue and the observation that statins lower serum zinc levels, clincians should consider low zinc levels in muscle tissue being a mechanism of muscle pain, weakness, twitching and more that is commonly reported with statin use. *Zinc deficiency leads to detrimental consequences, especially in tissues with high demand such as skeletal muscle.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7284914/

Abstract

Patients previously not treated with a lipid-lowering agent (; mean age 49.15±3.28 years) were treated with either 10 mg/day of Simvastatin (), or Atorvastatin () for 4 months. Fourteen additional patients were recruited from the same clinic at the same hospital as a control group. The medication of these latter patients was unaltered for 4 months and the same parameters were measured as for the statin groups. Serum concentrations of zinc, copper, caeruloplasmin, selenium, glutathione peroxidase (GPx) and C-reactive protein (CRP) were measured together with their lipid profiles pre- and post-treatment. In addition to reducing serum total and low-density lipoprotein (LDL) cholesterol (), statin treatment was associated with a significant reduction in mean serum zinc (9%, ), copper (9%, ), caeruloplasmin (24%, ), and median CRP (45%, ). Similar changes were not observed in the control patients. No significant effects were observed for serum selenium, copper/caeruloplasmin ratio, or GPx () in either statin or control groups. 

Background: The 3-hydroxy-3-methylglutaryl-CoA reductase inhibitors (statins) reduce serum cholesterol level and cardiovascular morbidity and mortality. However, the effect of statins on glucose metabolism is unclear. Some studies have suggested that statins may cause hyperglycemia by increasing calcium concentration in the islet cells leading to decrease in insulin release or by decreasing GLUT 4-mediated peripheral glucose uptake.

Methods: We analyzed the data in 345,417 patients (mean age 61 +/- 15 years, 94% males, 6% diabetic, 20% statin users) from the Veterans Affairs VISN 16 database. We studied change in fasting plasma glucose (FPG) in this population over a mean time of 2 years between the first available measurement and the last measurement form the most recent recorded visit. Data were limited to patients who had 2 FPG measurements. Diagnosis of diabetes had to be present before the first FPG measurement.

Results: Among patients without diabetes, FPG increased with statin use from 98 mg/dL to 105 mg/dL, and among nonstatin users, FPG increased from 97 mg/dL to 101 mg/dL (increase in FPG with statin use P < 0.0001). Among patients with diabetes, FPG increased with statin use from 102 mg/dL to 141 mg/dL, and among nonstatin users, FPG increased from 100 mg/dL to 129 mg/dL (increase in FPG with statin use; P < 0.0001). After adjustment for age and use of aspirin, beta-blockers, and angiotensin-converting enzyme inhibitors, the change in FPG in nondiabetic statin users was 7 mg/dL (vs 5 mg/dL in nonstatin users, P < 0.0001) and for diabetic statin users it was 39 mg/dL (vs 32 in nonstatin users, P < 0.0001).

Conclusions: Statin use is associated with a rise of FPG in patients with and without diabetes. This relationship between statin use and rise in FPG is independent of age and use of aspirin, beta-blockers, and angiotensin-converting enzyme inhibitors.