Li · Pediatrics 2016 · prospective cohort study · n=2734

The Association of Maternal Obesity and Diabetes With Autism and Other Developmental Disabilities.

Cited 345 times in the scientific literature.

Level 3 - non-randomized controlled study

Non-randomized controlled prospective birth cohort study

PubMed 26826214 · doi:10.1542/peds.2015-2206 · record verified 2026-08-30

What was done

The authors analyzed 2,734 children from the Boston Birth Cohort who had at least one postnatal visit at Boston Medical Center between 1998 and 2014. Autism spectrum disorder (ASD) and other developmental disabilities (DDs) were determined from physician diagnoses in electronic medical records. Cox proportional hazards regression models adjusted for potential confounders compared ASD and DD risks across six groups defined by maternal prepregnancy obesity and diabetes status (pregestational diabetes, gestational diabetes, or none).

What was found

Among 2,734 children, 102 were diagnosed with ASD. Individually, maternal prepregnancy obesity and pregestational diabetes (PGDM) were each associated with ASD risk. In combined analysis, significantly increased ASD risks were observed only among mothers with both obesity and PGDM (hazard ratio 3.91, 95% confidence interval 1.76-8.68) and those with both obesity and gestational diabetes (hazard ratio 3.04, 95% confidence interval 1.21-7.63). Intellectual disabilities (IDs), but not other DDs, showed a similar elevated risk associated with combined obesity and PGDM, mostly accounted for by co-occurring ASD and ID.

Why it matters

This study demonstrates that maternal prepregnancy obesity and diabetes exert a synergistic effect on the risk of offspring ASD and intellectual disability. It highlights maternal metabolic health before and during pregnancy as a potential target for reducing neurodevelopmental disorder risk.

Limits

The study is limited to a single medical center cohort. Subgroup analyses had modest sample sizes (102 total ASD cases distributed across six exposure groups), resulting in wide confidence intervals. Specific residual confounding factors cannot be fully evaluated from the abstract alone.

Cited by