History of in utero cocaine exposure in language-delayed children.

To determine whether children with language delays are more likely to have been exposed to cocaine in utero than children with normal language development, a case-control study was undertaken. Based on routine office screening in our primary-care clinic over a 1-year period, we identified 29 consecutive children, aged 24 to 48 months, as language-delayed. They were compared with an approximate 2:1 match of children without language delay who had been seen in the clinic on the same days and who were of similar age. There was more reported cocaine use during pregnancy (six of 29, 21%) among the language-delayed children than among the controls (five of 71, 7%). This difference is statistically significant (P < 0.05, chi 2 = 3.92; odds ratio = 3.4 +/- 2.2). Discriminant analysis revealed that both cocaine and nicotine exposure were associated with delayed language development--with an unexpected negative, i.e., an antagonistic, protective, interactive effect (F[3,96] = 4.66, R2 = 12.7%, P < .005); neither gender nor caretaker contributed to language development in this sample. These results suggest that children with language delay detected in a clinical setting are more likely to have been exposed in utero to cocaine than children with normal language development. Prenatal cocaine exposure should be a risk factor in monitoring development in children.