Objective: Reduced growth before birth is known to associate with a smaller ovarian volume in adolescents and women without androgen excess. We studied whether prenatal growth relates also to ovarian size and Polycystic Ovary (PCO) morphology in non-obese adolescents and young women with ovarian androgen excess.

Design: A cross-sectional analysis of standardized case notes over a 2 yr period.

Patients: Non-obese adolescents and young women (age 17 yr; N=86) seen for ovarian androgen excess, as confirmed by 17-hydroxy-progesterone hyperresponse to a gonadotropin-releasing hormone agonist.

Measurements: Endocrine-metabolic assessment in fasting state, together with a vaginal ultrasound scan to verify the presence or absence of PCO. Birthweight and gestational age were derived from medical records.

Results: PCO prevalence by ultrasound was 38%. Absence of PCO was associated with a shift (P<0.0005) of the birthweight distribution toward lower values. Patients with a birthweight <3.0 Kg were 6-fold more likely to have no PCO than to have PCO. Birthweight was across a wide range [1.5–4.0 Kg] associated with ovarian volume in hyperandrogenic patients with non-cystic ovaries (r=0.60; p<0.00001) and was, in a multiple regression analysis, the prime variable linked to ovarian volume (=0.57, p<0.00001), explaining 32% of its variance.

Conclusion: The ovarian size and the development of a PCO morphology in non-obese adolescents and young women with ovarian androgen excess relates to prenatal growth. These findings indicate that there are two subgroups of non-obese patients with ovarian androgen excess: one with a normal birthweight distribution and with PCO, and one with lower birthweights and without PCO.