Limited data exist about the potential developmental delays in appropriately grown twins; furthermore, twins may be at higher risk of developmental delay than singletons. Small-for-gestational age is a risk factor for developmental delay and is based on singleton birthweight references, which may misclassify small-for-gestational age in a subset of appropriately grown twins. This study aimed to evaluate the risk of developmental delay in twins classified as small-for-gestational age according to the twin and singleton birthweight references (<10th percentile). In a birth cohort (2008-2010) of twins (n=1790) and singletons (n=3829) where parents completed Ages & Stages Questionnaires for child development between 4 and 36 months, we used a US population-based birthweight reference to categorize singletons and twins as small-for-gestational age. Uncertain small-for-gestational age twins were defined as small-for-gestational age by a singleton reference (<10th percentile) and not by a twin reference, and twin-reference small-for-gestational age twins were defined as small-for-gestational age by a twin reference. Adjusted generalized linear mixed-effects models were used to estimate the odds of failure on any Ages & Stages Questionnaires domain and on each of the 5 domains (fine motor, gross motor, communication, personal-social, and problem-solving domains); random intercepts accounted for repeated measures and twin clustering. Compared with non-small-for-gestational age twins (>10th percentile), uncertain small-for-gestational age twins did not have higher odds of Ages & Stages Questionnaires failure (adjusted odds ratio, 1.28; 95% confidence interval, 0.91-1.80). Compared with non-small-for-gestational age singletons, both twin-reference and uncertain small-for-gestational age twins had higher odds of Ages & Stages Questionnaires failure, with the highest risk conferred to twin-reference small-for-gestational age twins (twin-reference adjusted odds ratio, 3.14 [95% confidence interval, 1.94-5.10]; uncertain adjusted odds ratio, 2.35 [95% confidence interval, 1.69-3.26]; P<.01 for trend). Results remained consistent when limiting analyses to term births (≥37 weeks' gestation). Although a singleton reference may overestimate small-for-gestational age in twins, the findings indicated that a singleton birthweight reference may be appropriate for twins because it identifies more twins at risk of developmental delay than a twin reference.
Read full abstract