Skip over navigation

Journal Issue: Drug-Exposed Infants Volume 1 Number 1 Spring 1991

Complex Developmental Issues of Prenatal Drug Exposure
Diana Kronstadt

Effects of Prenatal Drug Exposure on Child Development

Given all these cautions, the following summarizes what we know about the developmental effects of cocaine, marijuana, opiates, tobacco, and alcohol on child development. Most of the research reviewed considers children's development through age 4 years, with a few studies looking at children through age 7.

Cocaine and Polydrug Exposure

Two primary groups of researchers have been studying the long-term effects of cocaine upon children: Judy Howard and her colleagues at UCLA, and Ira Chasnoff and his associates at the Center for Perinatal Addiction at Chicago's Northwestern Hospital. In both cases, the exposed babies are not really "crack" or "cocaine" babies but rather "polydrug" babies. The UCLA group, for example, has studied 18 toddlers who were most often exposed to PCP, along with cocaine and heroin, marijuana, and/or alcohol.8 Infants in the Chicago study were born exposed to cocaine as well as to marijuana and/or alcohol.9 It is impossible to say which of the effects described below (and often misattributed to cocaine alone in the popular press) are due to the cocaine, which to the other substances used, and which to the combination.

Infants (Birth to 1 Year)

Although the medical difficulties of drug-exposed infants are described in detail elsewhere in this journal (see the article by Zuckerman), some are also mentioned here because of their particular implications for child development. Low birth weight, growth retardation in utero, and small-sized heads at birth are all potential consequences of cocaine use during pregnancy (see the article by Zuckerman in this issue).10 These conditions may contribute to the risk for developmental problems, such as cerebral palsy, seizure disorder, and mental retardation. Not all infants with these conditions go on to develop the more serious problems; the risk for developing these is strongly influenced by the child's caregiving environment.3

Some researchers have postulated central nervous system damage as the underlying cause of later developmental problems of polydrug-exposed children. The irritability, tremulousness, and irregular sleep patterns of some newborn cocaine-exposed infants are taken as markers of such damage. The only study to document specific central nervous system damage11 indicated that over one third of a sample of 28 cocaine-exposed infants had bleeding in the cerebral area of the brain, a finding relatively common in very low birth weight or preterm infants born to mothers who have not taken such drugs during pregnancy.12 The dosage levels or timing of cocaine ingestion which cause such problems and the reasons why some, but not all, exposed infants appear to suffer them is not known. Nor is it known if such damage has functional significance for later developmental problems, or if the impairments can be ameliorated through intervention programs such as those used for other vulnerable infants.

With respect to behavior, cocaine-exposed newborns are not a homogeneous group, but the following paragraphs describe a cluster of characteristics that has been noted in many newborn infants identified as cocaine-exposed.13 Again, we do not know why some, but not all, infants display these characteristics, or why some infants show the symptoms with greater intensity and persistence than do others.

Immediately after birth, some cocaine-exposed infants are often in great distress. Jittery and suffering tremors, the infants are irritable and sensitive to the mildest environmental stimulation. Their muscles are unusually stiff, and they may show a prolonged persistence of early reflexes. Often, they cry a great deal. They do not fall asleep readily, but once asleep are easily awakened. The distress of these newborns is obvious, but they are unable to calm themselves. Sometimes cocaine-exposed infants display the opposite characteristics: they sleep much of the time and appear to shut down as if to avoid environmental stimulation.

Although the problems associated with motor development, such as increased muscle tone and persistence of reflexes, usually diminish during the first year, irritability, sleep and feeding problems, and difficulty with calming may continue into the second year for some infants.

Most newborns cycle through periods of sleeping, wakefulness, and crying. As they develop, infants spend more time in quiet alert states and less time sleeping or crying. In the quiet and alert state, infants make contact with their caregivers. They are increasingly attracted and responsive to the face, touch, and voice of their mothers. 14 In contrast, many cocaine-exposed infants are initially not capable of achieving the calm state necessary to participate in this mutual interaction with their mothers or primary caregivers.14 They may withdraw from the caregiver's face and touch.

Toddlers (Age 12-30 months)

Researchers at UCLA's Department of Pediatrics8 and the Center for Perinatal Addiction at Northwestern Hospital in Chicago15 have each been monitoring the development of a group of drug-exposed children since birth. As described above, Judy Howard and her colleagues at UCLA have followed a sample of 18 full-term and preterm infants born to highly impoverished women who used cocaine and other drugs during pregnancy. The babies were compared with non-drug-exposed, preterm infants from similar environments. Both groups of families received intervention services, including pediatric care, case management, parent education, and home visits.

Comparisons between the two groups indicate that developmental scores for the drug-exposed group in a structured setting were lower than for the non-drug-exposed comparison group. Scores for both groups were within the low-average range. In an unstructured, free play setting, drug-exposed toddlers' play was less age-appropriate and more constricted and impulsive than that of non-drug-exposed toddlers. This play pattern, consisting of throwing and batting objects, is comparable to that observed in children with neurological impairments.16 In addition, drug-exposed toddlers appear to be less securely attached to their caregivers than were the comparison group. Although the sample of drug-exposed toddlers included those living with biological mothers, extended family members, and in foster care homes, it is noteworthy that of the eight residing with their biological mother, seven were insecurely attached.

Researchers at Chicago's Center for Perinatal Addiction at Northwestern Hospital have been following a group of over 200 infants who were born prenatally exposed to cocaine as well as to marijuana and/or alcohol. Like those of the UCLA researchers, their results indicated that drug-exposed toddlers were within the normal range in structured developmental assessments of cognitive and motor abilities. However, about 30% to 40% had language or behavioral problems of varying severity. These included delayed language development, lack of tolerance for frustration, distractibility, and difficulty organizing their behavior.

Preschool and Beyond

At the Center for Perinatal Addiction, preliminary results regarding outcomes for children at age 3 are just being analyzed.9 A sample of about 20 of these 3-year-olds continued to test within the normal range on structured developmental tests. In fact, at least 60% of the drug-exposed 3-year-olds were found to be normal with respect to language and behavioral organization. It is estimated that 30% to 40% of the drug-exposed 3-year-olds had specific problems of widely varying severity that included difficulties with expressive language articulation, focusing attention, and organizing their own behavior.

As described above, pregnant women in the Chicago program receive prenatal care and nutrition services as well as drug treatment and social services referral. Because the researchers have been unable to find a suitable non-drug-exposed control group, these results do not tell us what developmental problems the substance-exposed youngsters would have shown if their families had not received any intervention. Would their performance have been better, worse, or the same? We cannot tell. Nor can we tell if the children's developmental delays are the result of in utero substance exposure or are due to their living situations.

No other research results are available regarding the development of preschool and older children whose primary prenatal drug exposure includes cocaine. However, some of these preliminary findings are confirmed by the clinical impressions of professionals in special education who treat children presumed to have been prenatally exposed.

For example, staff in a special Los Angeles Unified School District program for 3- and 4-year-old children prenatally exposed to drugs report the children show high sensitivity to their environments, irritability, agitation, hyperactivity, speech and language delays, poor task organization and processing, emotional problems related to difficulty with attachment and separation, passivity, apathy, aggression, and poor social skills.

Teachers in the Los Angeles program view these behaviors as the consequences of multiple risk factors, including the prenatal exposure, early difficulties with attachment and children's subsequent lack of trust, and children's unstable home and community environments. The program staff caution that there is no typical profile of a drug-exposed child, and that children show a continuum of impairment ranging from mild symptomatology in one area to severe problems in all areas of development.17

Lack of a non-drug-exposed comparison group and the presence of chaotic home environments make it very difficult to attribute these characteristics solely to the prenatal drug exposure. Furthermore, some of the characteristics listed clearly indicate the difficulty in defining a profile for these children. For example, "apathy" and "agitation" are unlikely to be present in the same child at the same time, nor are both likely to be the primary characteristics of any one child.


Heroin and methadone are opiates. Methadone, given to heroin-addicted, pregnant women in registered treatment programs, contributes to better pregnancy outcomes, perhaps due to the medical management and lifestyle changes made by the pregnant woman in treatment as compared to the pregnant heroin addict on the street.18

More research has been conducted concerning the developmental outcomes of children exposed in utero to heroin and methadone than for most other drugs. Much of that research was initiated or conducted in the 1970s, when heroin was seen as the most serious drug threat then facing the nation. In some cases, children have been followed through school age.19

Newborn infants of opiate-using mothers may go through withdrawal, called "neonatal abstinence syndrome," which consists of central nervous system and digestive system symptoms that may include irritability, poor feeding, poor weight gain, ineffective sucking, yawning, sneezing and tremulousness, and sometimes seizures. They are often of low birth weight and have small head circumference,20 conditions associated with increased risk for later developmental problems. Most withdrawal symptoms disappear by age 2 months, but the irritability may persist during the first year or longer, contributing to caretaking difficulties similar to those encountered by parents of cocaine-affected infants.21 (For further description of early neonatal symptoms, see the article by Zuckerman in this issue.)

As mentioned above, researchers have followed opiate-exposed children through infancy, into the preschool years, and beyond. 22-35 The extensive literature has been reviewed by several authors.36-38 In general, findings, but not their interpretations, are fairly consistent.

Methadone-Exposed Children

If we consider overall scores on standard developmental tests (such as the Bayley Scales of Infant Development), most studies find small or no differences in the development of children who were exposed to methadone prenatally when compared either with national norms32-39 or with groups of similar, but non-drug-exposed children.31,34,35,40 In most cases, methadone-exposed children through 2 years of age score within normal ranges. However, some studies do find differences between these and other children when subscales of the developmental tests rather than overall scores are considered, or when other non-standard tests are used. Some studies, for example, find that methadone infants do not perform as well as comparison group infants on the mental or motor development subscales of the Bayley.23,24,27,30,33 These differences are not consistently found either across studies or at different points in time within the same study. In one study, for example, methadone-exposed children who were tested every 2-3 months, from birth to age 18 months, performed more poorly than non-exposed children only on a motor development subscale at 12 and 18 months.22

Most of these studies involve children from low socioeconomic backgrounds and have found that the developmental scores of all of the children (substance-exposed and non-exposed) sink below national norms as the children grow older.34,39 This general decline appears to begin by about 18 months to 2 years of age and is similar to a decline in scores observed in other studies involving children from low socioeconomic backgrounds.33

Studies that follow children through age 5 find similar results: either no differences between methadone-exposed and other children29 or no differences on general scores but deficits on subscales or on non-standardized tests.28

Heroin-Exposed Children

Geraldine Wilson and her colleagues at Baylor College of Medicine have studied the development of heroin-exposed children. Their results have not always been consistent. In a study of 1-year-olds, for example, methadone-exposed children were worse off than either heroin-exposed or non-drug-exposed children, but all groups were within normal ranges on a standard developmental test.26 In a study of 3- to 6-year-olds, heroin-exposed children displayed physical, intellectual, perceptual, and behavioral problems when compared with similar non-drug-exposed children.25 A third study found no differences between groups of heroin-exposed, methadone-exposed, and non-exposed children at ages 3-5 years. Follow-ups with a subgroup of the heroin-exposed children at school age indicated that 65% of the children had repeated one or more grades or needed special educational services.19

Perhaps even more striking were differences among the three groups in parenting and home environments. By their first birthday, 48% of infants of untreated heroin users were living with their biological parents, as compared with 80% of the methadone-maintenance children, and 100% of the comparison group.26 By the time the children reached preschool ages, only 9% of children of untreated heroin users were still cared for by their biologic mother, but almost 50% of the methadone- maintained mothers still retained custody.19 Given this fact, it is impossible to determine the extent to which placement in one or more foster or relatives' homes accounts for some or all of the differences in development that these children displayed.

Most researchers agree that a poor environment will magnify any weaknesses caused by the opiate exposure.24 Where researchers disagree is in whether or not the drug exposure so damages children that a ceiling to their development is created—a ceiling that children cannot clear, no matter how enriched their subsequent environment. At this time, most researchers appear to agree that postnatal experiences are probably more important than prenatal. For example, one group states, "From the present data, it appears that the infants' environment and subsequent lack of stimulation had a more direct influence on 2-year development than maternal drug use during pregnancy." 41


Although some studies of marijuana use during pregnancy report lower birth weight in newborns, these findings are not consistent.42 Most of what we know about the longer-term effects of marijuana exposure on child development derives from the work of Peter Fried and his colleagues at Carleton University in Ontario, Canada.43-47 Their Ottawa Prenatal Prospective Study has followed substance-exposed (marijuana, alcohol, and/or cigarettes) and non-substance-exposed children through the first 4 years of their lives.

Approximately 700 women enrolled in the study between 1979 and 1985. They were interviewed about drug use prior to and during each trimester of pregnancy. Children of a subgroup of the women, including those who used marijuana or smoked during pregnancy or who drank alcohol more than the average for the group were compared with children of a group who did not smoke tobacco or use marijuana and who abstained or drank little alcohol. Tests at 12, 24, and 36 months of age indicated no negative effect of marijuana exposure on mental, motor, or language development.47,48 Indeed, at 36 months of age, children of moderate marijuana users inexplicably displayed superior motor performance.47 At 48 months, however, heavy maternal marijuana use during pregnancy was associated with poor performance on memory and verbal tests.47 Because findings similar to those at 36 and 48 months were not observed at any other ages, these results can only be regarded as tentative. The children are due to be tested again at 60 and 72 months of age.


Frequent or heavy use of alcohol during pregnancy can result in a constellation of characteristics known as fetal alcohol syndrome (FAS). Children with FAS have growth deficiencies, specific facial malformations, and mental retardation.49 Some children who are alcohol-exposed may not have full FAS but only one or two signs that may then be called "fetal alcohol effects" (FAE).50 However, we do not know how much alcohol leads to problems, when during pregnancy alcohol intake is riskiest, or why some exposed children develop FAS, while others do not.

Researchers have investigated the development of FAS or FAE children separately from that of children whose mothers may have had less frequent or less heavy alcohol use during pregnancy. Studies that have followed FAS and FAE children into adolescence and adulthood find that FAS adults have lower IQ scores than FAE adults.51 The range of scores in both groups is quite broad, however, with some adults scoring above average. Therefore, it is probably neither possible nor wise to predict any one individual child's development solely on the basis of an initial diagnosis of FAS or FAE.51

Studies that have considered more moderate maternal use of alcohol are subject to varying interpretations. In a series of papers based on the Seattle Longitudinal Study on Alcohol and Pregnancy, a study which has followed a group of about 500 children over 7 years, Ann Streissguth and her colleagues at the University of Washington have traced several developmental problems to prenatal exposure to alcohol. Studies of 8-month-olds, for example, indicated that infants of mothers who drank more heavily (although the whole sample was primarily composed of social drinkers rather than alcoholics) had lower mental and motor development scores than infants of less heavy drinkers. Nevertheless, the differences were slight, and overall scores of the sample were above average.52,53

At age 4, children of heavier drinkers displayed poorer performance on some measures of motor development54 and lower scores on a test to assess their ability to maintain attention to a task.55 Other studies link prenatal alcohol exposure to lower IQ scores at age 4.56

The researchers have followed the children through age 7½ years. They conclude that binge drinking during pregnancy, and especially during the first trimester, is associated with deficits in memory, in arithmetic problem solving, and in attention.57-59 The best predictor of children's performance on tests of IQ, achievement, and classroom behavior was the father's educational level, but prenatal alcohol exposure was still a significant determinant of children's development.57 The authors believe that the negative effects of prenatal alcohol exposure are heightened in families with poorly educated parents or with larger numbers of older children.57

Interpreting these findings is difficult. Not all researchers have found similar decrements in performance in alcohol-exposed children, or at least not on all tests.47 Streissguth and her colleagues conclude that no level of alcohol exposure during pregnancy is safe,54 although in other papers they trace more serious decrements to binge drinking early in the first trimester.58 Even when differences are found, they are usually small, with most of the children scoring within normal ranges on standardized tests.54 It is not clear what functional significance these subtle differences may have.


Cigarette smoking during pregnancy has consistently been associated with lower birth weight in newborns; the greater the number of cigarettes smoked, the lower the birth weight. In addition, many studies have documented the relationship between smoking during pregnancy and decreased length and head circumference in newborns.42 As described above with respect to the effects of cocaine and polydrug exposure, these conditions may be associated with increased risk for later learning and developmental problems.

Quite a few studies have been conducted to trace the effects of maternal smoking during pregnancy on subsequent child development. Some have examined children as old as 16 years of age. In their 1989 review of over 30 papers on the subject, Rush and Callahan conclude that there is a "regular and consistent pattern of lower IQ and ability, and less advanced verbal, reading, and mathematical skills associated with maternal smoking during pregnancy."60 In addition, smoking appears related to behavioral and temperamental difficulties in the children, with children of smokers appearing to have more behavioral problems, hyperactivity and inattention, and poorer social adjustment.

Although the patterns of association are strong, Rush and Callahan emphasize that smokers may differ from nonsmokers in terms of behavior, personality, and social status. The observed differences in child development may therefore be due to those parental or environmental differences rather than to the cigarettes. However, they cite a few studies in which some of these environmental differences were taken into account, with the finding that the decrements in children's performance are still observed, indicating that cigarettes may indeed have a persistent effect.

Summary of Drug Effects on Child Development

In sum, there is little evidence that prenatal substance exposure, whether to cocaine, marijuana, opiates, tobacco, or alcohol, is linked with large deficits on standardized developmental tests. Indeed, much of the recent concern regarding cocaine- and polydrug-exposed children has been related to their performance on nonstandardized tests.8 If we rely on findings from the effects of other drugs on child development as a guide to what we can expect from cocaine- and polydrug-exposed children, then we should expect small differences between exposed and similar non-exposed children, with both groups probably scoring within low-normal ranges. However, to the extent that the cocaine-exposed children are from low socioeconomic backgrounds, then we should also expect to see the performance of substance-exposed children decline in comparison with national norms as they grow older, because that is what we would expect to see from other children from similarly disadvantaged backgrounds. The importance of postnatal environments for these children cannot be emphasized enough. The following sections describe some of the ways in which parents and structured, early intervention programs can help or hinder the development of substance-exposed children.