“Crack Kids” Revisited

IN THIS ISSUE OF THE JOURNAL, SINGER AND COLLEAGUES 1 report the findings from their prospective cohort study assessing the relationship between prenatal cocaine exposure and cognitive and developmental outcomes in 218 cocaine-exposed infants and 197 unexposed infants. After controlling for prenatal exposure to other drugs, gestational age and size at birth, and a number of caregiver characteristics, the authors found that infants who had in utero cocaine exposure scored on average 6 points lower than the comparison group on the Mental Scale of the Bayley Scales of Infant Development at 24 months of age. Rates of clinically important developmental delay on the scale were doubled in the cocaine-exposed group compared with the unexposed group (13.7% vs 7.1%, respectively). The study by Singer et al is the only 1 of 10 peerreviewed, adequately controlled, large-scale, prospective longitudinal studies to show an unequivocal negative association between toddlers’ developmental test scores and prenatal exposure to cocaine. Discrepancies among similar controlled longitudinal masked studies are of great scientific interest if the factors contributing to such discrepancies can be elucidated. Do children from the geographical area (ie, Cleveland) in the study by Singer et al experience more ill effects of prenatal cocaine exposure than those studied elsewhere? Since cocaine is an illegal substance, the extent of either its potency or its possible contamination by other potential toxins cannot be determined. Thus, it is impossible to know whether the drug used by women in Cleveland differs significantly from that in other studies from other cities where no such ill effects were found. Moreover, as the authors point out, their clinical selection criteria may have biased their sample toward a group of mothers who were heavy cocaine users. Other cohorts described in the literature are apparently more heterogeneous in the amount of cocaine use. Also, the cocaine-exposed sample in the study by Singer et al includes a substantial number of premature infants who are often excluded from other study samples. Prematurity, with or without prenatal substance exposure, is often associated with persistent developmental impairments. Although the authors controlled for gestational age, prenatal cocaine exposure may increase the risk of delayed development at 2 years of age only in children most heavily exposed or primarily in children who have a second serious biological risk factor such as prematurity. Although considering residual confounding in observational studies is important, Singer and colleagues do control for a number of important covariates, but others are not measured, including potentially protective factors such as participation in home visiting and other early intervention programs. Singer and colleagues have made an important contribution to an evolving research field, raising many new issues for future investigation and replication. However, on the basis of the long and ignoble history of “crack babies,” a major concern is how these findings will be interpreted and applied by other researchers, physicians, and the public. The social and biological correlates of prenatal cocainepolydrug exposure vary from sample to sample, so that one must avoid applying sweeping generalizations from any single sample to all cocaine-exposed children nationwide. Understanding factors that potentiate or moderate the potential impact of a toxic exposure is critical to clinical care and public policy because these factors may be amenable to programmatic interventions. The popular perception of the “crack kid” preceded and exceeded the findings of Singer et al and undoubtedly shaped physicians’ attitudes and public policies. Why do such perceptions persist? The answer provides a cautionary tale of the interface between science, clinical care, attitudes about drugs, the media, and the law. Addiction, illegality, prenatal toxicity, and poor outcomes are linked in the public and professional mind. In reality, scientific evidence for prenatal toxicity and teratogenicity is equivocal for some drugs and stronger for others. Inaccurate public expectations of correspondence between illegality and toxicity lead to distortions in interpreting and applying scientific findings. For example, although the study by Singer et al also finds some adverse effects related to prenatal tobacco and cocaine exposure, these results are