The impact of the recommended first-line treatment with metformin on C-reactive protein (CRP) levels in patients with polycystic ovary syndrome (PCOS) is still controversial. We conducted a meta-analysis of studies reporting the impact of metformin on serum CRP levels in women with PCOS. The weighted mean differences (WMDs) and the corresponding 95% confidence intervals (CIs) were used to assesse the effects. GRADE approach was used to assesse the quality of the evidence. A total of 20 studies that included 433 women with PCOS were analyzed. CRP levels significantly decreased after metformin treatment (WMD = -1.23mg/L, 95%CI: -1.65 to -0.81, I2 = 93% and P < 0.001 for heterogeneity). The decreased levels of CRP were observed both in lean (BMI<25 kg/m2) and obese (BMI>25 kg/m2) patients. Interestingly, the degree of decreased CRP levels was not depended on metformin dosage, but more significantly in patients treated beyond 6 months (WMD≥6months = -1.47mg/L vs. WMD<6months = -0.94 mg/L). Decreased CRP levels are not associated with the status of IR and androgen in patients with PCOS. However, the quality of evidence was very low because of the limitations and inconsistency of the included studies. Therefore, metformin shows the potential effects on CRP levels in women with PCOS. However, considering the very low quality of evidence for the analysis, the effect of metformin on CRP levels are still very uncertain, and large-scale randomized-controlled study is needed to ascertain the long-term effects of metformin in PCOS.
Keywords: CRP; meta-analysis; metformin; polycystic ovary syndrome.