Effect of Pharmacological Interventions on Lipid Profiles and C-reactive Protein in Polycystic Ovary Syndrome

A Systematic Review and Meta-Analysis

Mohammed A. Abdalla; Najeeb Shah; Harshal Deshmukh; Amirhossein Sahebkar; Linda Östlundh; Rami H. Al-Rifai; Stephen L. Atkin; Thozhukat Sathyapalan


Clin Endocrinol. 2022;96(4):443-459. 

In This Article


This systematic review provides an overview of the current evidence on the effect of pharmacological interventions on the lipids profile in women with PCOS. In the current review, we found that when metformin and atorvastatin were administered at various doses, compared with placebo, there were significant reductions in the mean CRP, TC, TGs and LDL-C. Saxagliptin, pioglitazone and rosiglitazone also showed significant reductions in the mean TC, TGs and LDL-C when compared with metformin or with placebo.

Lipid Profile Outcomes and CRP

Metformin significantly reduced the mean TC, TGs, LDL, but no effect on HDL was seen. In a systematic review and meta-analysis of 12 RCTs, metformin showed a significant effect for LDL-C reduction, but no effect was seen for the other parameters of the lipid profiles.[53] However, an RCT that compared metformin with placebo reported a significant increase in the mean HDL-C and a decrease in the mean TC.[545,55] The lipid lowering mechanism of action of metformin is that it activates the AMP-activated protein kinase (AMPK), which regulates the sterol regulatory element-binding protein-1 (SREBP-1) and inhibits the hepatic lipogenesis.[56] Statins reduce cholesterol production by competitively inhibiting the 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase, the rate limiting enzyme in cholesterol biosynthesis.[57]

Metformin at various therapeutic doses showed no effect on CRP when compared with other agents. The subgroup analysis also did not indicate any significant effect of Metformin at various doses and durations on CRP when compared with placebo. This is the converse to a meta-analysis of 20 RCTs[58] that assessed the effect of Metformin on CRP that reported a significant reduction in CRP. However, in the above study there was a significantly high level of heterogeneity among the studies; therefore, care, must be taken when interpreting the results of the study. Dawson et al., in an open clinical trial of exenatide (5 mcg BID administered for 4 weeks then titrated to 10 mcg for 12 weeks) reported a significant reduction in CRP from baseline (8.5 ± 1.4 to 5.6 ± 0.8 mmol/L p = 0.001).[59] Conversely, in this study we did not observe any effect for exenatide on CRP when compared with metformin. No effect on CRP was seen in this study when rosiglitazone was compared with metformin that differs to a study of rosiglitazone 4 mg QD administered for 12 months that showed a significant reduction in CRP.[60]

The review was conducted based on a systematic search for the related databases and grey sources. It also included RCTs and crossover trials only with the exclusion of both observational and non-randomized studies.

To date, this is the most inclusive systematic review and meta-analysis of the effect of pharmacological interventions on lipid profiles in women with PCOS.

One of the limitations of this systematic review is that a language filter was applied and only RCTs reported in English language were included. This could have significantly affected the inclusion of several studies published in foreign languages. Retrieving such studies requires translation to English that could be challenging and may also influence the methodology of this review. Moreover, we only included fully published studies and there may be unpublished trials that could not be retrieved. The majority of the RCTs reported in this review had small sample size and lacked statistical rigour used to identify sample size. Additionally, most of the RCTs had a short duration thus, the long-term effect of the various pharmacological interventions on the lipid profiles in women with PCOS is not clear.

This systematic review recognizes the poor quality of the included RCTs, which is also shown in the summary of evidence of the GRADE score in the supplementary material. Because of the design of some clinical trials (open-label), there was a substantially high level of performance bias. In some studies, the reporting and the selection bias were inadequately evaluated that led to the adjudication of an unclear RoB in 69% of the included RCTs. In addition, only 49% of the RCTs reported information of the method used to blind the participants and the outcome assessor and 45% were judged to have an unclear risk of attrition bias. For the lipid profile outcomes, the grade of evidence was rated as very low, low, or moderate due to the unclear or high risk of performance bias. There was lack of blinding for the participants and the outcome assessors, lack of allocation, unclear risk of attrition bias, unclear risk of selective reporting and considerable heterogeneity.

This study highlights that there is a lack of robust RCTs evaluating various pharmacological agents used in the treatment of PCOS. Moreover, currently available RCTs assessing the effectiveness of these pharmacological interventions are of low or very low quality. Therefore, the present results do not allow a definite conclusion and recommendation for clinical practice. Furthermore, these RCTs are of small sample size that may not have had the power to exclude false negative outcome. Thus, this review acknowledges the need for RCTs with rigorous design to facilitate better-informed clinical decisions to draw recommendations and help develop guidelines.