Association between maternal body mass index in early pregnancy and anorexia nervosa in daughters

BACKGROUND: The etiology of anorexia nervosa is poorly understood. Although genetic factors play a major role, maternal factors, and obstetric complications are possible environmental causes. We investigated the association between maternal overweight and obesity in early pregnancy and risk of anorexia nervosa in daughters. METHODS: A retrospective cohort study including all live singleton females born in Sweden from 1992 through 2002. Anorexia nervosa diagnosis was identified by using the nation-wide Swedish Patient and Cause of Death Registers. Multivariable Cox hazards regression was used to estimate adjusted hazard ratios (aHRs) and 95% CIs, adjusting for confounders. Stratified Cox regressions were applied to data on siblings to adjust for unmeasured familial confounding. RESULTS: Among 486,688 live singleton females, 2,414 (0.50%) were diagnosed with anorexia nervosa through 2012. The aHR of anorexia nervosa decreased linearly by maternal BMI (p-value for trend < .001). When compared with daughters of normal weight mothers (body mass index [BMI] 18.5-24.9), aHRs (95%CI) of anorexia nervosa were 0.74 (0.65-0.84) in daughters of overweight mothers (BMI 25-29.9) and 0.61 (0.47-0.78) in daughters of mothers with obesity Grade I (BMI 30-34.9). In sibling control analysis, no associations were observed between maternal BMI and aHRs of anorexia nervosa in offspring. CONCLUSIONS: The rate of anorexia nervosa decreased with maternal overweight and obesity in a dose-response manner. However, the sibling control analysis suggests that these findings are not consistent with causals effects of maternal BMI on anorexia nervosa in offspring.