Female age and the incremental effect of clomiphene on first ovulation in polycystic ovary syndrome: a secondary analysis of the PCOSAct randomized trial

Clomiphene remains widely used for ovulation induction in women with polycystic ovary syndrome, but whether female age modifies its incremental ovulatory benefit remains uncertain. We examined whether the clomiphene placebo contrast in first documented ovulation varied with age. This secondary analysis used the clomiphene placebo comparison from PCOSAct, a multicenter factorial randomized trial of 1,000 Chinese women with polycystic ovary syndrome. The primary analysis included 987 women and 769 first ovulation events during 16 weeks. Weekly first ovulation hazards were modeled using discrete time complementary log-log regression, with age entered continuously and a treatment by age interaction. Models adjusted for body mass index, acupuncture allocation, and study site. Marginal standardization estimated cumulative ovulation and time gained across ages 23–33 years. Each additional year of age was associated with a 9.2% smaller relative clomiphene placebo contrast (interaction ratio, 0.908; bootstrap 95% CI, 0.866–0.950; P < 0.001). There was no evidence that the interaction varied over time or departed from linearity. At ages 23, 28, and 33 years, standardized week 16 ovulation probabilities were 91.9%, 90.2%, and 88.4% under clomiphene, compared with 52.9%, 65.7%, and 78.2% under placebo. Corresponding absolute differences were 39.0, 24.5, and 10.2% points. Clomiphene advanced first ovulation by an estimated 41.8, 29.1, and 15.3 days, respectively. The narrowing contrast reflected increasing standardized ovulation under placebo, while ovulation under clomiphene remained high. Sensitivity analyses were directionally consistent. Female age modified the incremental short term effect of clomiphene on first documented ovulation. Across ages 23–33 years, where age specific estimates were most stable, clomiphene remained beneficial, but its additional benefit decreased with age. These findings may inform counseling about expected ovulatory response but do not imply that reproductive aging is beneficial or establish effects on pregnancy or live birth. This study is a secondary analysis of the PCOSAct trial, a randomized, multicenter, 2 × 2 factorial clinical trial conducted in mainland China. The parent PCOSAct trial was registered at ClinicalTrials.gov (identifier: NCT01573858; date of registration: April 10, 2012).