MENTAL HEALTH DISORDERS SUBSEQUENT TO GESTATIONAL DIABETES MELLITUS DIFFER BY RACE/ETHNICITY
Contract grant sponsors: Robert Wood Johnson Foundation Harold Amos Medical Faculty Development Program; Massachusetts General Hospital Executive Committee on Research/Multicultural Affairs Office Physician-Scientist Development Award; Contract grant number: NIH R03DK096152 (to R.B.-L.)
Contract grant number: NIH K24DK094872 (to R.T.).
Abstract
Introduction
The relationship between gestational diabetes mellitus (GDM) and postpregnancy mental health disorders has been inconsistently reported. Additionally, race/ethnicity data are limited. We sought to elucidate the intersection of these relationships.
Methods
We analyzed 18,109 women aged 18–40 with self-reported race/ethnicity. Women with (n = 659) and without (n = 14,461) GDM were followed for a median of 4.4 (interquartile range 1.4–6.8) and 4.0 (1.5–6.4) years, respectively, for incident mental health disorders. Multivariable repeated measures analyses were conducted to examine associations between GDM and postpregnancy mental health disorders, race/ethnicity, and the interaction of these factors.
Results
Women with compared to women without GDM were older (mean ± standard deviation, 32 ± 5 vs. 30 ± 5 years; P < .001) and had higher body mass index (29.0 ± 7.2 vs. 25.3 ± 5.2 kg/m2; P < .001). GDM was associated with increased risk for depression and anxiety after adjusting for age and pregnancy complications; however, loss of significance in the fully adjusted model for depression (odds ratio [95% CI]: 1.29 [0.98, 1.70]; P = .064) and anxiety (1.14 [0.83, 1.57], P = .421) suggested that clinical and socioeconomic factors influence this relationship. Hispanic compared to white women had a greater risk for depression (1.40 [1.15, 1.72]; P = .001), even after multivariable adjustment. The interaction between GDM and race was evident in complication-adjusted but not fully adjusted models.
Conclusions
The incidence of mental health disorders subsequent to GDM was attenuated after adjustment for clinical and socioeconomic factors. Moreover, race/ethnicity influenced this relationship. Further investigation is warranted to clarify potential underlying mechanisms.
