Background: Women with obesity or gestational diabetes (GDM) and their children are at risk for type 2 diabetes. Given that obesity and fat distribution are correlated with insulin resistance, we examined whether there are differences in total adiposity and fat distribution among women who were lean (NW) or obese (Ob) with no GDM history, verses women with prior GDM, and their children. We hypothesized that dyads in the GDM group would have greater adiposity and more centralized fat distribution compared to dyads in the NW and Ob groups.

Methods: Mother-child dyads were enrolled 4-10 years after a pregnancy complicated by GDM (N=73), or without GDM and NW (N=76) or Ob (N=76). Body weight, height, and anthropometric measures of adiposity were collected. ANCOVA were used to assess group differences in BMI, BMIz (in children), sum of skinfolds (SS), waist to hip ratio (WHR), and central (suprailiac + subscapular) to peripheral (thigh + tricep) skinfold ratio (CvP). Models were adjusted for age and race for mothers, along with tanner stage and sex for children.

Results: Women with prior GDM had greater BMI and SS compared to women in the NW group (P<0.0001), but did not differ from Ob. Despite similar total adiposity for GDM and Ob groups, the GDM group had greater WHR and CvP as compared to Ob (P<0.05). Among children, BMIz and WHR were greater in the GDM versus NW offspring (P<0.05), with children in the Ob group having a mean BMIz and WHR that was in between NW and GDM.

Conclusion: Results support the hypothesis that women with prior GDM exhibit more central adiposity despite having similar BMI as women with Ob without prior GDM. Among children, BMIz and central adiposity increased progressively when comparing offspring of NW, Ob, and GDM mothers. The results indicate that both Ob and GDM in mothers during pregnancy may affect total and central adiposity in children. While additional research is needed, attention to maternal health before and during pregnancy is vital to mitigate against a cardiometabolic disease phenotype in children.


S. Martin: None.


American Heart Association (336101011); University of Alabama at Birmingham (DK079626, DK056336)

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at