Illicit Drug Use: Cocaine


Epidemiology of Cocaine Use in Pregnancy

Although the epidemic of cocaine and crack use that began in the 1980s has waned, the problem of prenatal exposure to cocaine persists. According to the National Institute on Drug Abuse (NIDA,1996), approximately 2.3 percent of women of childbearing age have used cocaine in the past year and many of these women continue to use when pregnant. This figure may be higher in certain population subgroups and lower in others. In 1994, using blood drawn from a cohort of neonates, about 0.1 percent of all births were reported to have been exposed to cocaine, with a higher incidence among older women, those delivering without prenatal care, and inner-city populations. Most women reporting cocaine use also used tobacco, alcohol, and cocaine, and some combined the use of cocaine or crack with heroin.

Because of concerns raised during the “crack baby” period, extensive examination occurred of the teratogenic potential of this drug in both animal models and clinical studies. Although in 1993 one could conclude that inadequate data existed to support conclusions about the effects of this drug, during the latter half of the 1990s many studies were published that provide considerable understanding of this area, at least during infancy and the preschool period ().

Growth Effects

Cocaine exposure has been associated with lower gestational age and reduced growth parameters at birth in a number of studies. Because cocaine users have many other characteristics that may be associated with such outcomes, interpretation of these effects can be difficult. However, Kliegman and colleagues () found that cocaine exposure was associated with preterm birth as well as lower birth weight even when associated factors were controlled for statistically. Richardson et al. () controlled for the effects of prenatal care by comparing the effects of cocaine use for both those who had prenatal care and those who did not. They found that cocaine had a significant impact on both gestational age and birth weight in each group even when the effects of alcohol, marijuana, and tobacco were controlled. Even when growth effects are observed, interpreting the relationship may not be straightforward. While examining the relationship between gestational age and cocaine exposure in neonates, Brown and colleagues () found that lower birth weight was characteristic only of full-term cocaine-exposed infants, suggesting that such effects occurred in the third trimester. In contrast, Richardson and colleagues () found that growth effects in their sample were attributable to exposure during the first and second trimesters. Finally, even when statistically significant effects are found during the neonatal period, cocaine-exposed children do not have “clinically significant” growth failure and often appear to have a postnatal “catch-up” in growth. For instance, while comparing preterm and full-term cocaine-exposed infants to socioeconomic status (SES)-matched contrast groups, Coles and colleagues () found that growth differences could no longer be observed by eight weeks of age and there were no differences in growth rate over 24 months for weight, length, or ponderal index.

Motor Development

Early studies of cocaine effects identified reflexive behavior and motor development as areas of concern. Schneider and Chasnoff () compared 30 full-term four-month-old infants exposed to cocaine (and other drugs) to 50 unexposed infants using the Movement Assessment of Infants (MAI). Exposed infants were found to have higher risk scores on motor tone, primitive reflexes, and volitional movements. Swanson and colleagues () found poorer mean scores on the volitional movements subscale of the MAI as well as the total risk score among four-month-old exposed infants compared to controls. Fetters and Tronick () followed 28 cocaine-exposed and 22 control infants to 15 months and found a negative drug effect on motor performance at this age. The authors note, however, that both cocaine-exposed and contrast groups of children performed more poorly than would be expected from the age norms.

Arendt and his colleagues () used the Psychomotor Index of the Bayley Scales of Infant Development (BSID) and the MAI as well as other measures of sensorimotor development in a sample of inner-city children exposed to cocaine. They found small but significant effects of cocaine and other drug exposure on a variety of motor indicators both early in infancy and at 12 months. At 24 months, children from this sample were reassessed using the Peabody Developmental Motor Scales with the cocaine-exposed group performing significantly lower on both fine and gross motor development indices. The effects appeared to be more significant in the fine motor rather than the gross motor area.

Later in infancy motor differences are not described by most investigators. This discrepancy may be the result of differences in the measurement tools used. Those studies reporting effects often used the MAI, while those that did not used the BSID. As more longitudinal data is published, it will be possible to evaluate the implications of observed differences in motor function for later development.

Behavioral Effects

Initially, severe consequences were anticipated in this area of development, although the evidence to support such effects was not strong. Studies of newborns provided conflicting information about the immediate impact of maternal cocaine use during gestation. In a meta-analysis, Held and colleagues () critically reviewed Brazelton Neuro-behavioral Assessment Scale (BNBAS) studies of infants. It was concluded that while effects could be found reliably on motor performance, abnormal reflexes, orientation, and autonomic regulation, the effect size was small and tended to diminish over the first month of life. As well-controlled studies of later development are reported, evidence of direct teratogenic effects on cognition have been limited, although children born to drug-using mothers in low SES populations continue to be at risk for nonoptimal development in many domains. For instance, in a follow-up study that examined outcomes at four to six years, Chasnoff and colleagues () reported that differences in developmental functioning in their clinical samples can be accounted for by environmental factors, principally, caregiver behavior. Singer et al. () reported that prenatal cocaine and alcohol exposure as well as maternal postpartum psychological distress directly impacted the BSID Mental Development Index (MDI) while Psychomotor Index (PI) scores were affected only by cocaine. Kilbride and colleagues () reported that at 36 months no effects on cognition, psychomotor skills, or language were observed in exposed children who had received case management services, compared to those who did not receive services and a nonexposed contrast group. Kilbride et al. also found that those exposed children who remained with their mothers and did not receive services had lower verbal scores on intelligence tests and measures of language development. In contrast, Richardson () found that in three-year-olds, cocaine exposure was associated with lower scores on some of the subtests of the Stanford-Binet (fourth edition), including composite IQ scores and short-term memory scores, although all children scored within the typical range of development. A previous study of a different cohort of children by the same author did not show effects on cognition, demonstrating the extent to which these outcomes are dependent on sampling and other methodological considerations.

Language Development

A number of studies have identified deficiencies in the early language development of children born to cocaine-using women. Bland-Stewart et al. () compared semantic content category in a small sample of low-SES infants exposed to cocaine with a contrast group matched for social class and ethnicity and found some restriction in the development of semantic representations (meaning) in the children of cocaine users. No effect was observed in the structural features of language, that is, mean length of utterance (MLU) and utterance type, or for general language and cognitive functioning. In contrast, Hurt and colleagues () found no differences in language functioning at two-and-a-half years when cocaine-exposed and contrast children from the same SES group were compared using the Preschool Language Scale (PLS), a standardized measure of early language development. In reviewing the literature in this area, Mentis () suggested that there is not yet sufficient evidence for definitive statements about the language development of this group of children. While language development may be disrupted, the factors affecting such development are numerous and their interaction is complex. She also suggested that deficits may be specific to certain areas of language function and are only evident under stressful conditions.

Play Behavior

Play behavior is often assessed as an indicator of children’s functional status that does not require standardized testing. Play behavior has been examined in a number of studies of cocaine- and polydrug-exposed children that followed an initial study by Rodning et al. () that reported alterations in the usual play patterns. Subsequent studies have been inconsistent in reported outcomes. Metosky and Vondra () reported differences in play analogous to Rodning and colleagues (), while several other investigators have not found evidence of differences in the play of toddlers that can be attributed to the direct effects of cocaine when associated factors are controlled. These outcomes suggest that such behavioral observations may be accounted for by environmental factors or group differences.

Arousal Regulation and Attention

The most persuasive evidence for a behavioral effect of cocaine concerns the impact on physiological arousal (e.g., heart rate, respiration) in early infancy and, by extension, on temperament and social/emotional development. Mayes () provides an animal model of this phenomenon that suggests that dopamine regulation has been impacted. Several investigators have identified increased irritability in young infants and alterations in psychophysiology, including heart rate and respiration. These effects appear to persist beyond the neonatal period. At eight weeks, Bard and colleagues () identified cocaine-related alterations in baseline heart rate and respiration as well as differences in response to moderate stress that appeared to be drug-related. Karmel, Gardner, and their colleagues identified cocaine-related differences in attention and arousal modulation in newborns that persisted through four months of age. At four months, Bendersky and Lewis () found that exposed infants were less able to modulate arousal. Other systems, such as sleep, appear to be impacted during the toddler period. That cognition may also be affected in some manner is suggested by reported effects on early attention. Mayes et al. () reported that cocaine exposure affected three-month-old children’s ability to complete a procedure measuring attention. Coles and colleagues () found differences in attentional response associated with prenatal cocaine, but not other drug exposure at eight weeks. However, as these authors note, the caregiving instability associated with maternal drug use independently accounted for more variance in attentional response than did the direct effect of cocaine. These findings raise concerns about the vulnerability of exposed children. In addition to physiological dysregulation associated with prenatal exposure to cocaine and other drugs, exposed children are clearly also at environmental risk to an increased incidence of developmental psychopathology.


Prenatal cocaine exposure is a marker for a number of risk factors that appear to have negative consequences for the infant and developing child. No specific “cocaine syndrome” has yet emerged and many of the problems previously anticipated have not manifested. However, the weight of the evidence suggests that cocaine exposure may produce an increased vulnerability to certain environmental stressors. The interactions of these factors may have long-term negative consequences for children.