BACKGROUND. Previous studies have suggested that language is affected in infants of diabetic mothers, yet there have been no systematic investigations to address this question.
OBJECTIVE. Our goal was to compare infants of diabetic mothers and controls on language outcomes from ages 18 months to 7 years.
METHODS. This was a case-control longitudinal design with 2 birth cohorts: 1835 singletons from the Quebec Longitudinal Study of Child Development (born October 1997 to July 1998) and 998 twins from the Quebec Newborn Twin Study (born November 1995 to July 1998). Cases were 221 infants of diabetic mothers (105 singletons and 116 twins), and controls were 2612 children (1730 singletons and 882 twins) for whom at least 1 language measure from ages 18 months to 7 years was available. Exclusion criteria were gestation of <32 weeks. The outcome measures were McArthur Communicative Development Inventory expressive and receptive vocabulary and grammar at 18 months and 30 months, the Peabody Picture Vocabulary Test receptive vocabulary at 48 months and expressive and receptive vocabulary at 60 months, and Early Development Instrument teacher-assessed communication at 72 months and 84 months (kindergarten and first grade).
RESULTS. Analyses of variance (controlling for gender, socioeconomic status, and perinatal factors) revealed effects of gestational diabetes on expressive language at 18, 30, and 72/84 months. Infants of diabetic mothers scored 0.27 to 0.41 SD lower than controls and were 2.2 times more at risk of a language impairment. Genes and maternal education both moderated the effect of gestational diabetes on expressive language during this period.
CONCLUSION. Gestational diabetes hinders expressive language in offspring into middle childhood. Genes are strongly associated with the risk of delays in infants of diabetic mothers, and offspring of educated mothers are less affected.