Estimated effects of in utero cocaine exposure on language development through early adolescence

The potential longitudinal effects of prenatal cocaine exposure (PCE) on language functioning were estimated from early childhood through early adolescence in a large, well-retained urban sample of 451 full-term children (242 cocaine-exposed, 209 non-cocaine-exposed) participating in the Miami Prenatal Cocaine Study (MPCS). The sample was enrolled prospectively at birth, with documentation of prenatal drug exposure status through maternal interview, and toxicology assays of maternal and infant urine, and infant meconium. Age-appropriate versions of the Clinical Evaluation of Language Fundamentals (CELF) were used to measure total, expressive, and receptive language at ages 3, 5, and 12 years. Longitudinal latent growth curve (LLGC) modeling of the data revealed an association between PCE (measured dichotomously as yes/no) and lower functioning in expressive and total language scores, after considering other sources of variation including child's age at testing, sex, prenatal exposure to alcohol, marijuana, and tobacco, and additional medical and social-demographic covariates. Analyses of level of PCE showed a gradient, i.e. dose-dependent, relationship between PCE level and expressive, receptive, and total language scores in the models controlling for age, child's sex, and other prenatal drug exposures. With additional covariate control these findings were most stable for the total language score. The evidence supports an inference about an enduring stable cocaine-specific effect on children's language abilities, with no effect on language growth over time in the longitudinal trajectory of language development.