UCSF

Objective(s)

To identify clinical predictors of future eating disorder symptoms in women with polycystic ovary syndrome (PCOS).

Design

Prospective cohort study.

Setting

University center.

Patient(s)

One hundred sixty-four women with PCOS by the Rotterdam criteria.

Intervention(s)

Participants were characterized at a baseline visit between 2006 and 2017. A questionnaire including the validated Eating Disorder Examination-Questionnaire (EDE-Q) was self-administered at follow-up.

Main outcome measure(s)

EDE-Q global score (0-6, higher scores indicate more severe symptoms).

Result(s)

One hundred sixty-four women completed the follow-up survey an average of 5.3 years after the baseline visit. Compared with a normative population, women with PCOS had higher EDE-Q global scores (2.3 vs. 1.5) and scored higher on all subscales. Within the PCOS cohort, the following baseline clinical characteristics were independently predictive of scoring in the highest EDE-Q global score tertile: body mass index, waist circumference, hyperandrogenemia, high sensitivity C-reactive protein, and depression scores. Obesity at baseline conferred a 6.9-fold increase in the odds of elevated EDE-Q score (adjusted odds ratio = 6.89; 95% confidence interval, 2.70, 17.62), while a positive depression screen conferred 3.6-fold increased odds (adjusted odds ratio = 3.58; 95% confidence interval, 1.74-7.35). Compared with white women, nonwhite women were at risk of higher EDE-Q scores.

Conclusion(s)

Women with PCOS are at risk of disordered eating attitudes and behaviors, which may interfere with attempts at lifestyle interventions. Clinicians should screen women with PCOS for eating disorder psychopathology, especially those with obesity or depression. An exclusive focus on weight loss may have unintended consequences.