Fernandez · Human reproduction (Oxford, England) 2021 · cross-sectional cohort study · n=974

Diagnosis delayed: health profile differences between women with undiagnosed polycystic ovary syndrome and those with a clinical diagnosis by age 35 years.

Cited 48 times in the scientific literature.

Level 4 - case-series / case-control

Cross-sectional analysis of a cohort

PubMed 33963388 · doi:10.1093/humrep/deab101 · record verified 2026-08-26

What was done

Cross-sectional analysis of a community-based cohort of 974 women born in Adelaide, South Australia, assessed at around 35 years of age. Researchers compared 56 women with a prior clinical diagnosis of PCOS, 64 women with undiagnosed PCOS identified during study participation using Rotterdam criteria, and the remaining women in the cohort without PCOS. Logistic regression was used to identify independent predictors of prior clinical diagnosis across sociodemographic, reproductive, metabolic, and psychological characteristics.

What was found

Over half of women meeting Rotterdam criteria for PCOS were previously undiagnosed (64 of 120, 53.3%). Women with a prior diagnosis were four times more likely to report fertility difficulties than undiagnosed women (OR 4.05, 95% CI 1.74-9.45), despite similar rates of attempting or achieving pregnancy. Women with a prior diagnosis reported higher rates of menstrual irregularities (95% vs 81%) and excess body hair (63% vs 45%) than undiagnosed women. Metabolic problems were elevated in both PCOS groups compared to controls. Clinical depression prevalence was 50% higher in both PCOS groups than in women without PCOS (P = 0.021), but did not predict prior clinical diagnosis.

Why it matters

More than half of women with PCOS remained undiagnosed by age 35, with clinical diagnosis predominantly triggered by fertility problems rather than metabolic dysfunction or clinical depression, both of which were equally elevated in undiagnosed women.

Limits

The small numbers of diagnosed (n = 56) and undiagnosed (n = 64) women limited statistical power to detect modest differences. The cross-sectional design cannot establish causality, relies partly on self-reported histories, and some undiagnosed PCOS cases may have remained misclassified in the non-PCOS control group.

Cited by