We undertook a comprehensive search of all published original articles evaluating the diagnostic value of prostate-specific antigen (PSA) for the detection of polycystic ovary syndrome (PCOS). Finally, 25 studies with a total of 1865 participants (1104 cases and 761 controls) met the inclusion criteria. Mean serum PSA levels were significantly higher in PCOS women compared to healthy women [Standardized mean difference (SMD)=1.73 ng/ml, 95% CI=1.22 to 2.25, p<0.001]. Only 5 studies reported accuracy indicators and were included in the bivariate random-effects model; with a total of 509 samples that included 332 (65.22%) confirmed PCOS women. The sensitivity of serum PSA for the diagnosis of PCOS varied between 66.3 and 90.6%, and, the specificity ranged from 62.5–86.7%. Meta-analysis revealed a pooled sensitivity of 72.3% (95% CI 67.1–77.00%) and specificity of 74.6% (95% CI 67.5–80.8%) with an area under the summary ROC curve equal to 0.82 (95% CI 0.75–0.83). The summary estimates were 2.92 (95% CI 1.98–4.31) for positive LR and 0.37 (95% CI 0.26–0.51) for negative LR. The pooled estimate of diagnostic odds ratio (DOR) of serum PSA was 9.01 (95% CI 4.35–18.64). Considering the methodological weaknesses existed in studies, lack of sufficient data regarding the accuracy indicators, and the complexity of the exact pathophysiology of PSA secretion in women, the precise position of PSA, as a diagnostic test for the detection of PCOS, in clinical practice is undetermined, and further studies are still needed.
Key words
prostate-specific antigen - insulin resistance - polycystic ovary syndrome
