Maternal diabetes and hypertensive disorders in association with autism spectrum disorder.

This was a case-control association study, so it cannot show that maternal diabetes or hypertension caused autism or developmental delay. The analysis excluded 1,205 participants, and that excluded subset was less likely to be non-White and less likely to have attained a higher education. The authors could not fully account for sample selectivity because limited data were available for non-participants, especially general population controls. Residual confounding by unknown or unmeasured variables cannot be ruled out. Exposure classification may have included misclassification from poor maternal recall or incomplete medical records. Medical record availability was lower for the developmental delay group (62%) than for autism (69%) or general population (67%). Sample sizes were insufficient to evaluate the causal pathway through preterm birth, and the study could not fully assess heterogeneity in severity or timing of maternal hypertension and diabetes. Children with developmental delay who had some autism characteristics were excluded from this analysis.

The authors reported no conflicts of interest. The article is tagged as supported by NIH and U.S. government research funds.

Do not read the association between maternal hypertensive disorders and autism or developmental delay as proof that hypertension caused the child's diagnosis. This was a case-control study, and the authors could not rule out residual confounding or exposure misclassification.