BACKGROUND

PCOS is the most common endocrine disorder in reproductive age women. The origins of PCOS are unknown but experimental and limited human evidence suggests that greater prenatal exposure to androgens may predispose to PCOS. Experimental evidence suggests that maternal stressors may affect reproductive function in the offspring via changes in prenatal androgen exposure. In this present study, we aim to investigate whether maternal stressful life events during pregnancy are associated with polycystic ovary morphology (PCOM) or polycystic ovary syndrome (PCOS) in adolescent offspring.

METHOD

In a large population-based pregnancy cohort study (The Raine Study) continuously followed from prenatal life through to adolescence we examined the association between maternal stressful life events during pregnancy in both early and late gestation, and subsequent circulating concentrations of ovarian and adrenal androgens, PCOM and PCOS in the normal menstrual cycle of offspring age 14-16 years. Maternal stressful life events were prospectively recorded during pregnancy at 18 and 34 weeks using a 10-point questionnaire. Female offspring (n = 223) completed a questionnaire about their menstrual cycles, underwent a clinical examination for hirsutism (Ferriman-Gallwey score) and transabdominal pelvic ultrasound examination to determine ovarian morphology according to standardized criteria for classification of PCOM. Plasma samples were obtained at day 2-6 of the normal menstrual cycle for measurement of androgens. PCOM was defined according to the international consensus definition, 2003 and the evidence-based guideline for the assessment and management of PCOS, 2018. PCOS was diagnosed according to Rotterdam criteria and National Institute of Health (NIH) criteria. Multivariate linear and logistic regression analyses were used to examine the associations between maternal stressful life event exposure and ovarian morphology (PCOM), circulating ovarian and adrenal androgens (clinical and biochemical hyperandrogenism (hirsutism)) and presence of PCOS.

RESULTS

Of 223 recruited adolescent girls, 78 (35.9%) and 68 (31.3%) had PCOM by the 2003 and 2018 criteria respectively, while 66 (29.6%) and 37 (16.6%) had PCOS, using Rotterdam and NIH criteria, respectively. Most girls (141/223, 63.2%) were exposed to at least one stressful life event in early gestation and around half (121/223, 54.3%) were exposed to at least one stressful life event in late gestation. Maternal stressful life events in early gestation were associated with a statistically significant lower prevalence of PCOM when applying the 2003 criteria [adjusted odds ratio [aOR] and 95% confidence intervals (CI): 0.74 (95% CI: 0.55; 0.99)], and a similar association was detected when applying the 2018 PCOM criteria (aOR, 0.69, 95% CI: 0.50; 0.95)]. Maternal stressful life events in early gestation were also associated with lower circulating concentrations of testosterone (β = -0.05, 95% CI: -0.09; -0.004) and androstenedione (β = -0.05, 95% CI: -0.10; -0.002) in the offspring. No similar effects for PCOM or circulating androgens were detected in late gestation. No statistically significant associations between maternal stressful life events in early or late gestation with PCOS (neither Rotterdam nor NIH criteria) in adolescence were detected. The prospective collection of maternal stressful life events during both early and late gestation and direct measurement of PCOM, PCOS and circulating androgens in adolescence and key co-variates implies minimal possibility of recall, information bias and selection bias.

CONCLUSION

Maternal exposure to stressful life events in early gestation is associated with significantly reduced circulating ovarian and adrenal androgen concentrations in adolescence (testosterone and androstenedione), and an indication of fewer cases of polycystic ovary morphology (PCOM) defined by the 2003 international consensus definition and by the 2018 international evidence-based guideline, but has no effect on polycystic ovary syndrome (PCOS), diagnosed using either Rotterdam or NIH criteria.