Autism risk in small- and large-for-gestational-age infants

We sought to determine whether small-for-gestational age (SGA) and large-for-gestational age (LGA) birthweights increase autism risk.

This was a retrospective cohort analysis comparing children with autism (n = 20,206) within a birth cohort (n = 5,979,605). Stratification by sex and birthweight percentile (SGA, <5th or 5-10th percentile; appropriate size for gestational age [GA], >10th to <90th percentile; LGA, either 90-95th or >95th percentile) preceded Cochran-Mantel-Haenszel analysis for GA effect, and multivariate analysis.

Autism risk was increased in preterm SGA (<5th percentile) infants 23-31 weeks (adjusted odds ratio [aOR], 1.60; 95% confidence interval [CI], 1.09-2.35) and 32-33 weeks (aOR, 1.83; 95% CI, 1.16-2.87), and term LGA (>95th percentile) infants 39-41 weeks (aOR, 1.16; 95% CI, 1.08-1.26), but was decreased in preterm LGA infants 23-31 weeks (aOR, 0.45; 95% CI, 0.21-0.95).

SGA was associated with autism in preterm infants, while LGA demonstrated dichotomous risk by GA, with increased risk at term, and decreased risk in the premature infants. These findings likely reflect disparate pathophysiologies, and should influence prenatal counseling, pediatric autism screening, and further autism research.