Objectives: To assess whether measures of temperament predict ASD status.
Methods: Using the ALSPAC cohort of 14,062 livebirths born to pregnant women enrolled during 1991-2, mothers prospectively reported aspects of child temperament at 1 month using 14 questions, and at 6 and 24 months using over 80 questions from the Carey Temperament scales for infants and toddlers. Children who had been given a diagnosis of ASD by 11 years of age were compared with their non-ASD peers.
Results: No differences were observed between ASD children and controls at 1 and 6 months. But at 24 months, ASD children scored higher on adaptability and persistence (P<0.0001) and on mood (P<0.01) but scored lower on distractibility and threshold (P<0.0001). Children were 40% to 80% more likely (or less likely) to have an ASD diagnosis for a one SD increase in score.
Conclusions: Differences between diagnostic groups on temperament measures up to age 24 months were small, and the likelihood of the measures being of use in predicting children at high risk of ASD is minimal.