Cocaine use during pregnancy and intrauterine growth retardation: new insights based on maternal hair tests.

Prenatal cocaine use is more accurately measured by maternal hair assay than by urine toxicology screening or self-report. To investigate the consequences of improved measurement, the authors ascertained cocaine use during pregnancy by maternal hair test, urine test, and self-report in a sample of 691 patients recruited from one New York City hospital in 1990-1992. Associations with intrauterine growth retardation, head circumference, and length of gestation were investigated. A positive hair test at delivery was not more strongly associated with birth weight (-38.1 g; 95% CI: -164, 88.3) or head circumference (-1.73 mm; 95% CI: -5.91, 2.44) than a positive urine test at delivery (-182 g (95% CI: -295, -69.8) and -6.11 mm (95% CI: -9.99, -2.24), respectively). Cocaine concentration in hair (which was higher if urine tests were positive) had a dose-response relationship with birth weight: a 27-g decrease (95% CI: -51.9, -1.04) with each log-unit increase in concentration. Birth weights were similar among infants of never users and infants of users who stopped using cocaine before delivery. Heavier use of cocaine, but not lighter use, was associated with intrauterine growth retardation, and exposure in late pregnancy was necessary to the association. Although maternal hair tests were instrumental in clarifying these relations, their clinical use is probably not warranted.