There is an increased risk of developing T2DM in women with polycystic ovary syndrome (PCOS), and there is evidence that statins improve metabolic parameters in these patients. However, there are some data to show that statins increase the risk of incipient diabetes.
Materials and methods
We have previously shown that 12 weeks of atorvastatin improves insulin resistance when measured using HOMA-IR. This post hoc analysis was designed to look at the effect of atorvastatin on pancreatic β cell function using HOMA-β in the same study. In this randomised, double-blind placebo controlled study, 40 medication-naïve patients with PCOS were randomised to either atorvastatin 20 mg daily or placebo for 3 months. A 3-month extension study for both groups of patients was undertaken with metformin 1500 mg daily after completing initial 3 months of atorvastatin or placebo.
There was a significant reduction in HOMA-β (240 ± 3.2 vs 177 ± 2.3; P value <0.01) after 12 weeks of atorvastatin treatment, which was maintained by metformin in the subsequent 12 weeks. There were no changes in HOMA-β after the placebo or after subsequent metformin treatment.
There was no linear correlation between reduction in HOMA-β with improvement of free androgen index (FAI) (r2 = 0.02; P = 0.72), testosterone (r2 = 0.13; P = 0.49), SHBG (r2 = 0.22; P = 0.48), hsCRP (r2 = 0.19; P = 0.64), triglycerides (r2 = 0.09; P = 0.12), total cholesterol (r2 = 0.11; P = 0.32) or LDL-C (r2 = 0.19; P = 0.38).
Treatment with atorvastatin for 12 weeks in women with PCOS significantly reduced HOMA-β. This could be potentially due to fall in β-cell requirement with improvement of insulin resistance rather than a reduction of β-cell function.
Statin (HMG-CoA reductase) treatment is effective in the primary and secondary prevention of cardiovascular disease (CVD) events (1, 2) and is generally safe and well tolerated (2). In the West of Scotland Coronary Study (WOSCOPS), pravastatin treatment decreased the risk of type 2 diabetes (T2DM) by 30% (3). Emerging evidence, however, suggests that treatment with other statins slightly increase the risk of T2DM (4, 5). In pooled data sets, statin therapy was associated with a 9% increased risk of diabetes (6), and this effect was age and dose dependent (7). Population-based studies have also reported a 10–22% increased risk of diabetes with statins (8). However, there is also increasing evidence that reduction in insulin resistance especially with weight loss in patients with T2DM will result in normalisation of pancreatic β-cell requirement and even reversal of T2DM (9).
Statins have shown to improve certain features of polycystic ovary syndrome (PCOS) such as hyperandrogenemia (10, 11, 12, 13). PCOS is one of the most common endocrine disorders that affects 6–20% of reproductive-aged women (14, 15, 16) and is at an increased risk of T2DM (17, 18). Seventy-five to 90% of PCOS patients demonstrate insulin resistance (IR) above and beyond that predicted by body mass, race or age (19, 20) resulting in compensatory hyperinsulinaemia (21) and an increased risk for (T2DM) (22) and cardiovascular disease (23). However, the effect of statins in improving IR is variable (10, 11, 12, 13, 24). Insulin resistance in most of these studies used HOMA-IR, which is a derivative use to measure insulin resistance from insulin and glucose. We have previously shown that 12 weeks of atorvastatin improves insulin resistance when measured using HOMA-IR (10, 11). This post hoc analysis was designed to look at the effect of atorvastatin in insulin secretion using HOMA-β in the same study and its correlation with other hormonal and metabolic parameters.
Materials and methods
A randomised double-blind placebo-controlled study was undertaken using atorvastatin 20 mg daily (11). An extension arm followed this where immediately after stopping the study medication, 37 patients (19 patients from the atorvastatin group and 18 patients from the placebo group) who completed the study were given metformin 500 mg three times daily for 3 months (10). Clinical and biochemical assessments were performed at baseline, at the end of the 3-month period and at the end of 6 months after the extension phase.
The diagnosis of PCOS was based on all three diagnostic criteria of the Rotterdam consensus, namely clinical and biochemical evidence of hyperandrogenemia (Ferriman–Gallwey score >8; free androgen index >8 respectively), oligomenorrhoea or amenorrhea and polycystic ovaries on transvaginal ultrasound (25). Subjects had no concurrent illness, were not on any medication for the preceding 6 months and were not planning to conceive. All patients gave written informed consent. The study was approved by the Hull and East Riding Local Research Ethics committee.
Study blood tests and measurement were done after an overnight fast. Compliance was monitored by counting returned medication. Fasting venous blood was collected into serum gel and fluoride oxalate tubes. Samples were separated by centrifugation at 2000 g for 15 min at 4°C, and the aliquots were stored at –20°C. Analytical methods have been described previously (11). The insulin resistance was calculated using the homeostasis model assessment (HOMA) method (HOMA-IR = (insulin × glucose)/22.5). Pancreatic beta cell function (HOMA-β) was calculated using HOMA-β = (20 × insulin)/glucose-3.5. Data are reported as mean ± s.e.m.
The sample size calculation for the original study was based on the known effects of atorvastatin on CRP in patients with impaired fasting glucose with the assumption that a similar effect occurs in those with PCOS (11, 26). For 90% power, significance level of 5% and adjusting for a possible 20% dropout rate, 20 subjects were enrolled in each group.
Comparisons between both the groups from baseline were carried out using the paired t test for biochemical data and clinical observations. The Wilcoxon signed-rank test was applied to biochemical data that violated the assumptions of normality when tested using the Kolmogorov–Smirnov test. Between-group comparison of percentage changes was performed using independent samples t test. Intention-to-treat analysis was done. For all analyses, a two-tailed P ≤ 0.05 was considered to indicate statistical significance. Statistical analysis was performed using SPSS for Windows NT, version 14.0 (SPSS).
Thirty-seven patients completed the study. Two patients from the placebo group and one patient from atorvastatin group dropped out of the study due to non-compliance.
The mean age group of patients was 27.7 ± 1.4 years (atorvastatin 26.6 ± 1.2 vs placebo 28.8 ± 1.8). The BMI was comparable in both atorvastatin and placebo group (33.20 ± 1.4 vs 33.92 ± 1.4 kg/m2).
There was a significant reduction in serum insulin levels and HOMA-IR in patients taking atorvastatin whilst there was significant increase in both these parameters in patients randomised to placebo. In the extension phase, there were significant reduction in insulin and the HOMA-IR with metformin in the atorvastatin pre-treated group, whereas no significant changes in any of these parameters with metformin in the placebo pre-treatment group.
However, there was a significant reduction in HOMA-β (240 ± 3.2 vs 177 ± 2.3; P value <0.01) after 12 weeks of atorvastatin treatment compared to placebo, which was maintained by metformin in the subsequent 12 weeks (177 ± 2.3 vs 185 ± 3.3; P value 0.42). There were no changes in HOMA-β after the placebo (Table 1).
Comparison of anthropometric and hormonal parameters at baseline, 12 weeks of atorvastatin or placebo followed by 12 weeks of metformin.
|Atorvastatin group||Placebo group|
|Parameter||Baseline (V1)||12 weeks (V2) (Atorvastatin 20 mg daily)||24 weeks (V3) (Metformin 1.5 gm daily)||V1/V2 (P value)||% change V1–V2||V1/V3 (P value)||Baseline (V1)||12 weeks (V2) (Placebo)||24 weeks (V3) (Metformin 1.5 gm daily)||V1/V2 (P value)||% change V1–V2||V1/V3 (P value)|
|Weight (kg)||91.29 ± 3.4||91.20 ± 3.4||90.3 ± 3.1||0.42||0.1 ± 0.1||0.08||93.12 ± 4.8||93.05 ± 4.7||92.6 ± 3.2||0.67||0.03 ± 0.02||0.12|
|Testosterone (nmol/L)||4.1 ± 0.2||2.9 ± 0.1||2.7 ± 0.1||<0.01||−24.6 ± 2.6||<0.01||4.4 ± 0.2||4.3 ± 0.2||4.2 ± 0.8||0.73||−0.1 ± 1.5||0.18|
|SHBG (nmol/L)||31.1 ± 1.0||35.3 ± 1.2||36.6 ± 1.7||<0.01||13.7 ± 2.4||<0.01||31.9 ± 0.9||32.9 ± 0.9||32.2 ± 0.8||0.14||3.6 ± 2.0||0.8|
|FAI||13.4 ± 0.6||8.7 ± 0.4||7.6 ± 0.9||<0.01||−32.7 ± 2.6||<0.01||13.9 ± 0.6||13.3 ± 0.5||13.1 ± 0.9||0.07||−3.1 ± 2.0||0.14|
|Glucose (mmol/L)||4.8 ± 0.1||4.9 ± 0.1||4.6 ± 0.1||0.52||1.0 ± 2.0||0.47||4.7 ± 0.5||4.8 ± 0.1||4.6 ± 0.4||0.32||2.0 ± 1.0||0.33|
|Insulin (μIU/mL)||15.6 ± 1.8||12.4 ± 1.7||10.2 ± 1.8||<0.01||−21.4 ± 4.0||<0.01||14.4 ± 2.0||17.6 ± 2.4||16.4 ± 1.8||0.04||34.2 ± 19.7||0.32|
|HOMA-IR||3.3 ± 0.4||2.7 ± 0.4||2.0 ± 0.8||<0.01||−19.8 ± 5.2||0.04||3.0 ± 0.4||3.8 ± 0.5||3.4 ± 0.3||0.04||37.2 ± 20.4||0.36|
|HOMA-β||240. ± 3.2||177 ± 2.3||185 ± 3.3||<0.01||−26.3 ± 1.1||0.07||240 ± 4.2||270 ± 8.2||298 ± 3.6||0.10||12.5 ± 1.0||0.08|
|TC (mmol/L)||4.6 ± 0.2||3.4 ± 0.2||3.9 ± 0.5||<0.01||−26.4 ± 3.0||0.82||4.5 ± 0.2||4.6 ± 0.2||4.5 ± 0.3||0.13||3.1 ± 1.1||0.82|
|LDL-C (mmol/L)||2.9 ± 0.2||1.8 ± 0.2||2.8 ± 0.4||<0.01||−36.6 ± 5.0||0.04||2.7 ± 0.2||2.7 ± 0.1||2.9 ± 0.3||0.77||3.1 ± 0.9||0.70|
|HDL-C (mmol/L)||1.07 ± 0.1||1.08 ± 0.1||1.07 ± 0.2||0.17||−4.0 ± 3.9||0.20||1.1 ± 0.08||1.1 ± 0.09||1.1 ± 0.1||0.47||3.0 ± 3.2||0.70|
|TG (mmol/L)||1.34 ± 0.08||1.08 ± 0.1||1.01 ± 0.1||<0.01||−20.9 ± 3.8||0.05||1.39 ± 0.24||1.69 ± 0.27||1.34 ± 0.1||0.19||36.8 ± 21.0||0.82|
Atorvastatin group – Atorvastatin for 12 weeks followed by metformin for 12 weeks.
Placebo group – Placebo for 12 weeks followed by metformin for 12 weeks.
V1 – Baseline; V2 – 12 weeks from baseline on either atorvastatin or placebo; V3 – 24 weeks from baseline (12 weeks from visit 2 on metformin 1.5 g daily).
Data are presented as mean ± s.e.m. All serum results are obtained from fasting variables.
All variables were normally distributed.
To convert values for testosterone to nanograms per decilitre, divide by 0.03467.
To convert values for SHBG to micrograms per decilitre, divide by 34.7.
To convert values for glucose to milligrams per decilitre, divide by 0.056.
To convert values for insulin to picomoles per litre, multiply by 6.
To convert values for cholesterol to milligrams per decilitre, divide by 0.0259.
To convert values for triglycerides to milligrams per decilitre, divide by 0.0113.
P* – P value for percentage difference between both group using unpaired t test.
*P value <0.01.
FAI, Free Androgen Index; HDL-C, HDL cholesterol; LDL-C, LDL-cholesterol; TC, Total cholesterol; TG, Triglycerides.
There was no linear correlation between reduction in HOMA-β with improvement of FAI (r2 = 0.02; P = 0.72), testosterone (r2 = 0.13; P = 0.49), SHBG (r2 = 0.22; P = 0.48), hsCRP (r2 = 0.19; P = 0.64), triglycerides (r2 = 0.09; P = 0.12), total cholesterol (r2 = 0.11; P = 0.32) or LDL-C (r2 = 0.19; P = 0.38).
In women with PCOS, 12-week treatment with atorvastatin 20 mg resulted in a significant reduction in pancreatic beta cell requirement as assessed by HOMA-β. This was independent of the improvement seen in the inflammatory markers, IR, lipids and hyperandrogenemia.
There was 26% reduction in pancreatic beta cell requirement as assessed by HOMA-β with atorvastatin treatment in this study compared to 20% reduction in insulin resistance as measured by HOMA-IR. It has been shown that in patients with T2DM, severe calorie restriction reverse T2DM with reduction of IR and normalisation of pancreatic β-cell requirement (9). After 12 weeks of atorvastatin in this study, there was a significant improvement in insulin resistance as measured by HOMA-IR. This reduction in IR would have resulted in fall in the β-cell requirements. Another possibility is that the statin could have a direct deleterious effect on β-cells by multiple mechanisms. Statins have been shown to inhibit this glucose-induced calcium signalling-dependent insulin secretion (27). In addition, statins suppress the synthesis of ubiquinone (CoQ10), an essential factor in the mitochondrial electron-transfer system, resulting in the inhibition of insulin secretion due to reduced production of ATP (28). However, it is unlikely that the β-cell function fell more than expected by the reduction in IR since there was no significant hyperglycaemia after 12 weeks of atorvastatin therapy as measured by fasting glucose. This reduction in HOMA-β was maintained after subsequent metformin treatment. However, there were no significant changes in HOMA-β after placebo or subsequent metformin treatment.
This study used atorvastatin that is a lipid-soluble statin. There could be a differential effect of statins depending on lipid solubility rather than a class effect. Lipophilic statins such as simvastatin and atorvastatin are possibly absorbed by extra-hepatic cells; these statins can deregulate cholesterol metabolism, thus attenuating β-cell function (29). However, in the West of Scotland Coronary Study (WOSCOPS), treatment with pravastatin which is a water-soluble statin decreased the risk of diabetes by 30% (3). There are currently no studies reported looking into the effect of non-lipophilic statins in women with PCOS.
There was no significant correlation with reduction of lipids, inflammatory markers and hormonal parameters with reduction in HOMA-β. This shows that the effect of atorvastatin on HOMA-β is independent of its lipid-lowering effect.
The limitation of the study includes the fact that this is a post hoc analysis; however, all the biochemical analysis was done immediately after the study so that it should not affect the validity of analysis. For measurement of pancreatic beta cell function, hyperglycaemic insulin clamp studies or intravenous glucose tolerance test is ideal; however, in this study, we calculated HOMA-β using fasting glucose and insulin measurements. HOMA-β correlates well with other methods of assessment of pancreatic beta cell function over a wide range of glucose tolerance status (r = 0.69 vs hyperglycemic clamp; r = 0.62–0.90 for normal glucose tolerance; r = 0.73–0.88 for impaired glucose tolerance; r = 0.69–0.90 for diabetes) (30, 31). However, more elaborate studies are needed to ascertain the effect of atorvastatin on beta cell function including insulin clamp studies.
In conclusion, treatment with atorvastatin for 12 weeks in women with PCOS significantly reduced HOMA-β. This could be potentially due to fall in β cell requirement with improvement of IR rather than suppression of β cell function, since there was no significant hyperglycaemia or hyperinsulinemia after atorvastatin therapy.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
This research did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector.
Author contribution statement
T S, A M C, E S K and S L A conceived the study, performed the analysis and drafted the manuscript.
Cholesterol Treatment Trialists CKearneyPMBlackwellLCollinsRKeechASimesJPetoRArmitageJBaigentC. Efficacy of cholesterol-lowering therapy in 18,686 people with diabetes in 14 randomised trials of statins: a meta-analysis. Lancet 2008 371 117–125. (doi:10.1016/S0140-6736(08)60104-X)
Cholesterol Treatment TrialistsCBaigentCBlackwellLEmbersonJHollandLEReithCBhalaNPetoRBarnesEHKeechA. Efficacy and safety of more intensive lowering of LDL cholesterol: a meta-analysis of data from 170,000 participants in 26 randomised trials. Lancet 2010 376 1670–1681. (doi:10.1016/S0140-6736(10)61350-5)
FreemanDJNorrieJSattarNNeelyRDCobbeSMFordIIslesCLorimerARMacfarlanePWMcKillopJH. Pravastatin and the development of diabetes mellitus: evidence for a protective treatment effect in the West of Scotland Coronary Prevention Study. Circulation 2001 103 357–362. (doi:10.1161/01.CIR.103.3.357)
WatersDDHoJEDeMiccoDABreaznaAArsenaultBJWunCCKasteleinJJColhounHBarterP. Predictors of new-onset diabetes in patients treated with atorvastatin: results from 3 large randomized clinical trials. Journal of the American College of Cardiology 2011 57 1535–1545. (doi:10.1016/j.jacc.2010.10.047)
CederbergHStancakovaAYaluriNModiSKuusistoJLaaksoM. Increased risk of diabetes with statin treatment is associated with impaired insulin sensitivity and insulin secretion: a 6 year follow-up study of the METSIM cohort. Diabetologia 2015 58 1109–1117. (doi:10.1007/s00125-015-3528-5)
LimELHollingsworthKGAribisalaBSChenMJMathersJCTaylorR. Reversal of type 2 diabetes: normalisation of beta cell function in association with decreased pancreas and liver triacylglycerol. Diabetologia 2011 54 2506–2514. (doi:10.1007/s00125-011-2204-7)
SathyapalanTKilpatrickESCoadyAMAtkinSL. The effect of atorvastatin in patients with polycystic ovary syndrome: a randomized double-blind placebo-controlled study. Journal of Clinical Endocrinology and Metabolism 2009 94 103–108. (doi:10.1210/jc.2008-1750)
ManiHLevyMJDaviesMJMorrisDHGrayLJBankartJBlackledgeHKhuntiKHowlettTA. Diabetes and cardiovascular events in women with polycystic ovary syndrome: a 20-year retrospective cohort study. Clinical Endocrinology 2013 78 926–934. (doi:10.1111/cen.12068)
AzzizRCarminaEDewaillyDDiamanti-KandarakisEEscobar-MorrealeHFFutterweitWJanssenOELegroRSNormanRJTaylorAE. The androgen excess and PCOS society criteria for the polycystic ovary syndrome: the complete task force report. Fertility and Sterility 2009 91 456–488. (doi:10.1016/j.fertnstert.2008.06.035)
ShawLJBairey MerzCNAzzizRStanczykFZSopkoGBraunsteinGDKelseySFKipKECooper-DehoffRMJohnsonBD. Postmenopausal women with a history of irregular menses and elevated androgen measurements at high risk for worsening cardiovascular event-free survival: results from the National Institutes of Health – National Heart, Lung, and Blood Institute sponsored Women’s Ischemia Syndrome Evaluation. Journal of Clinical Endocrinology and Metabolism 2008 93 1276–1284. (doi:10.1210/jc.2007-0425)
PuurunenJPiltonenTPuukkaKRuokonenASavolainenMJBloiguRMorin-PapunenLTapanainenJS. Statin therapy worsens insulin sensitivity in women with polycystic ovary syndrome (PCOS): a prospective, randomized, double-blind, placebo-controlled study. Journal of Clinical Endocrinology and Metabolism 2013 98 4798–4807. (doi:10.1210/jc.2013-2674)
CostaACasamitjanaRCasalsEAlvarezLMoralesJMasramonXHernandezGGomisRCongetI. Effects of atorvastatin on glucose homeostasis, postprandial triglyceride response and C-reactive protein in subjects with impaired fasting glucose. Diabetic Medicine 2003 20 743–745. (doi:10.1046/j.1464-5491.2003.00993.x)
YadaTNakataMShiraishiTKakeiM. Inhibition by simvastatin, but not pravastatin, of glucose-induced cytosolic Ca(2+) signalling and insulin secretion due to blockade of L-type Ca(2+) channels in rat islet β-cells. British Journal of Pharmacology 1999 126 1205–1213. (doi:10.1038/sj.bjp.0702397)
MabuchiHHigashikataTKawashiriMKatsudaSMizunoMNoharaAInazuAKoizumiJKobayashiJ. Reduction of serum ubiquinol-10 and ubiquinone-10 levels by atorvastatin in hypercholesterolemic patients. Journal of Atherosclerosis and Thrombosis 2005 12 111–119. (doi:10.5551/jat.12.111)