Children of Addicted Women

Abstract

The purpose of this article was to review follow up studies of children with prenatal drug exposure from preschool through adolescence. Specifically, the authors focus on the effects of prenatal exposure to cocaine, methamphetamine, and opiates on behavior and development. The largest number of studies have examined cocaine-exposed children. The authors identified 42 studies that suggest that there are unique effects of prenatal cocaine exposure on 4- to 13-year-old children, particularly in the areas of behavior problems, attention, language, and cognition. In addition, studies make reasonable attempts to control for possible confounding factors. Systematic research on the long-term effects of prenatal methamphetamine exposure is just beginning but seems to be showing similar effects to that of cocaine. The literature on the on the long-term effects of children with prenatal opiate exposure is more substantial than the methamphetamine literature but it is still relatively sparse and surprising in that there is little recent work. Thus, there are no studies on the current concerns with opiates used for prescription mediation. There is a growing literature using neuroimaging techniques to study the effects of prenatal drug exposure that holds promise for understanding brain/behavior relationships. In addition to pharmacological and teratogenic effects, drugs can also be viewed from a prenatal stressor model. The author discuss this “fetal origins” approach that involves fetal programming and the neuroendocrine system and the potential implications for adolescent brain and behavioral development.

INTRODUCTION

In the previous article, Bandstra et al. reviewed the effects of prenatal drug exposure through infancy and early childhood. Our review will consider prenatal drug exposure effects from preschool through adolescence, specifically the outcomes of exposure to cocaine, methamphetamine, and opiates related to behavior and development. We first provide a review of the published findings of exposure to each drug. We then describe how the effects of these drugs can be viewed from a prenatal stressor model involving fetal programming and the neuroendocrine system and the implications for adolescent brain development.

PRENATAL COCAINE EXPOSURE

Unlike the rather sparse prenatal methamphetamine and opiate literatures, our knowledge about the long-term effects of prenatal cocaine exposure is based on a reasonably large and growing corpus of studies. We were able to identify 42 published, empirical studies of the behavior and development of children with prenatal cocaine exposure between the ages of 4 and 13. These studies were published between 1996 and 2008, most (n = 34) between 2004 and 2008, many of which were longitudinal follow-up reports of the cohorts in the previous Bandstra review. We grouped the studies into categories based on the following outcomes: behavior problems, psychopathology, substance use, IQ, language, cognition (including attention), motor function, school and achievement, quality of relationships, and psychophysiology. In addition, there are 10 neuroimaging studies in this population. We will summarize the findings by category with effects for cocaine considered positive when adjusted for covariates (if applied) and then review covariate and related methodological issues.

Behavior Problems

Behavior problems formed the largest category and included 18 studies,^1-18 10 of which showed negative effects of prenatal cocaine exposure.^{2-4,6-8,10,13,16-18} Of the 18 studies,8 used the parent report Child Behavior Checklist,^{1,2,5,11,12,16-18} 4 used the teacher version (Teacher Report Form),^9,13-15 4 used other tests (1 child report [Youth Risk Behavior Survey]⁴ and 3 teacher report [PROBES-14]^6-8), and 2 used behavioral observations (frustration; aggression elicitation).^3,10 For the 12 studies that used the comparable measures (Child Behavior Checklist or Teacher Report Form), only four showed adverse cocaine effects on major subscales (e.g., externalizing) or minor subscales,^2,13,16 with two studies reporting interaction of cocaine exposure, gender, and alcohol exposure (girls exposed to cocaine but not alcohol were more aggressive, whereas boys exposed to both cocaine and alcohol were more delinquent).^13,16

Psychopathology

Only two studies have examined psychopathology in these children,^12,19 probably because the age of most of the cohorts limited these examinations. One study showed an association between cocaine exposure and ADHD and oppositional defiance disorder¹² and the other studied showed no effects on suicide ideation.¹⁹

Substance Use

Only 1 study included substance use as an outcome in children. In this study of preadolescents at 10.5 years of age, cocaine-exposed children were more likely to have smoked cigarettes.⁴

IQ

IQ was reported in 14 studies^{5,11,14,15,20-29} and, perhaps surprisingly, negative effects of prenatal cocaine exposure were only observed in five of those.^{5,21,22,24,28} Of the five studies, gender effects were observed in two studies with lower IQ in boys.^5,22 Among the seven studies reporting analysis of summary IQs and subtests,^{5,14,21-23,27,28} six reported significant cocaine effects on Perceptual Reasoning,²⁸ Verbal Comprehension,²¹ Verbal Reasoning (boys only),⁵ Short Term Memory,⁵ Picture Completion,²³ Object Assembly,²³ Abstract/Visual Reasoning (boys only),²² Visual-Spatial Skills,²⁷ General Knowledge,²⁷ and Arithmetic Skills.²⁷ In only two cases, summary and subtests scores were reduced by prenatal cocaine exposure but the full IQ score was not.^23,27

Language

There were eight studies of language in cocaine exposed children.^{5,11,24,26,30-34} Only one study did not find negative exposure effects,¹¹ and one study reported adverse effects for boys only.⁵ In a study that examined the trajectory of overall language function from 3 to 7 years, language deficits associated with cocaine exposure were stable across all age points with no increase or decrease over time.³¹

Cognition

Cognition is a somewhat heterogeneous category. The largest single category among cocaine exposure studies is sustained attention, which was examined in seven studies,^{14,15,26,35-38} with all but one study¹⁴ reporting adverse exposure effects. Six studies used a continuous performance or vigilance task^{14,15,26,35,36,38} that yields scores for omission (poor sustained attention) and commission errors (impulsivity); of these 6 studies,2 reported exposure effects on omission errors,^35,36 2 reported both omission and commission errors,^15,26 and 1 reported the effects on both outcomes.³⁸ Other attention tests (contrary tapping/delay task efficiency) also reported poor attention in exposed children.³⁷

Nine studies examined the relationship of cocaine exposure to other cognitive processes, including visual motor integration,^20,26 task persistence observation during a mastery paradigm,³⁶ central processing (PROBES) for all subjects.⁶ or for boys only,⁷ visual-spatial short term memory (Gordon Maze),^39,40 maintaining set with interference (Stroop Task),⁴¹ visual attention span and sequencing (Knox Cube),²⁶ and set switching (Trail Making).¹⁵ Of the nine studies, all but two studies reported harmful exposure effects, both on visual motor integration.^15,20 The diversity of findings is similar to the subscales on IQ tests in that no specific ability emerges across the relevant studies that relate to cocaine exposure. However, in this case, the study-specific selection of paradigms makes replication less likely.

Motor

Of three studies assessing motor abilities, one showed more motor problems in cocaine exposed boys⁷ and two showed no cocaine effects on fine motor behavior.^20,26 Unlike the infancy studies described by Bandstra et al. in the previous article in this supplement, relatively few studies examined motor function in older exposed children.

School Performance and Academic Achievement

Of the seven studies of school performance and academic achievement,^{14,17,18,24,25,28,42} three showed negative effects of prenatal cocaine exposure.^24,25,42 Two effects were on achievement tests…;^24,25 the other showed an increase in referrals for special education and academic support services.⁴² In addition, the one study that anticipated school success by examining school readiness reported negative prenatal cocaine exposure effects.²⁶

Relationships

No effects of prenatal cocaine exposure were found in the study of the mother-child relationship using observation of play.¹¹ No other relationships, such as relationships to peers or teachers, were examined in any study.

Psychophysiology

Only two physiological studies have been reported, and both showed cocaine effects. One showed changes on event-related potentials during the Stroop executive function test of working memory.⁴¹ In the other study, skin conductance, which measures electrodermal activity and reflects autonomic arousal, was altered during stress in children with prenatal cocaine exposure.¹⁸

Covariates

As has been well documented, early studies of the effects of prenatal cocaine exposure were difficult to interpret because of methodological problems, including small sample size and poor control for confounding factors (e.g., prenatal exposure to other substances and measures of the postnatal caregiving environment) that could also explain what has been ascribed to cocaine exposure. The 42 studies reported above were based on reasonably large sample sizes, with 20 studies based on samples of over 300 participants with the maximum over 1,000 participants. Of course, some studies are based on the same cohorts. There are 14 different cohorts for a total of 4,419 children. Large population studies also allow statistical control for many prenatal and postnatal confounding variables. Studies vary in the criteria used to determine what potentially confounding variables are, such as variables on which the exposed group and the comparison group show statistical differences or variables that have a predetermined relationship (e.g., P < .10) with both the predictor variable (prenatal cocaine exposure) and the outcome variable (e.g., IQ). Thus, in the description below, the fact that a variable was not controlled could simply mean that it did not need to be controlled. For example, a study of full-term infants would not control for gestational age. Lack of control could also mean that a factor was not measured.

Prenatal Confounding Variables

Use of legal or illegal substances other than cocaine during pregnancy (tobacco, alcohol, marijuana, or opiates) was controlled in 33 of the 42 studies. Measures of fetal growth, including birthweight, prematurity, head circumference, birth length, and neonatal medical problems, were controlled in 20 studies, and 12 studies controlled for prenatal care.

Postnatal Confounding Variables

Poverty or socioeconomic status was controlled in 27 of the 42 studies. A variety of measures have been used to control for specific aspects of the home environment. The Home Observation Measure of the Environment (Home scale) was used as a covariate in 15.⁴³ Placement in foster care was used in 23 studies. Related measures of the household included the number of children in the home, the number of regular caregivers, chaos, social support, and early case management. Measures of the caretaker included age, education, marital status, parity, employment, IQ, and self-esteem. Parental psychological problems or symptoms were controlled in 15 studies, and maternal IQ was controlled in 16 studies. Parent’s postnatal use of drugs was controlled in 22 studies. Measures of the child included daycare, head start, enrichment programs, services, and blood lead level. Child measures that could be outcomes were also included as covariates, including IQ, physical growth, and depressive symptoms. Two studies controlled for child abuse, one for domestic violence, and six for community violence. Race was used as a covariate in 10 studies and gender was used in 28 studies.

Dose-Response Relationships

One question that is often asked is whether the effects of prenatal cocaine exposure are related to the level of cocaine use. There are no studies in this group in which results of toxicology analysis (e.g., urine or meconium) were reported in relation to prenatal cocaine exposure. However, studies have used maternal interviews to derive measures of cocaine use patterns, such as the frequency of use per week. Of the 42 studies,10 found effects of heavy cocaine use related to worse outcome: six on behavior problems, two on language, and two on IQ.

Summary

The review of these 42 studies suggests that there are unique effects of prenatal cocaine exposure on 4- to 13-year-old children across a range of domains, including behavior problems, attention, language, and cognition. The evidence is weaker that IQ, school performance, and achievement are affected. In the domains of psychopathology, physiology, and motor performance, only a few studies have been reported. It is important to note that, in general, the effects that have been observed have been conducted on reasonably large samples with adjustment for confounding variables, suggesting that these are “true” cocaine effects. There is also evidence for dose response relationships, especially with behavior problems.

It is also interesting to note dimensions that have not been well studied. Substance use by the children and more studies of psychopathology will undoubtedly be undertaken as the age of children in longitudinal cohorts increases. It is somewhat surprising that there is only one study of parent-child relationships because this was an important focus in the infancy research, and that there are only two psychophysiological studies. There are also few studies using direct observation of behavior. Maternal report, paper and pencil forms, or computerized tests were used in 39 of the 42 studies.

In addition to determining the unique effects of cocaine, it is also important to determine how cocaine interacts with other factors in predicting child outcome. There are no studies on the interaction between genetic markers and prenatal cocaine exposure. The field of genetics provides an interesting metaphor in that we now know that most diseases do not obey the laws of classical Mendelian genetics and this is likely even more the case for behavioral phenotypes. For most outcomes, the genetic “main effect” is relatively small, allowing the study of gene-environment interactions to take center stage. This approach is starting to appear in the prenatal cocaine exposure literature. For example, studies have examined mediators or moderators of prenatal cocaine exposure effects, including gender, alcohol, head circumference, and foster care. The impact of puberty on this complex matrix has yet to be studied. More sophisticated statistical modeling is starting to be used to study more complex relationships between prenatal cocaine exposure and child outcome, including structural equation modeling, generalized estimated equations, and hierarchical linear modeling.

Neuroimaging

Quantitative neuroimaging studies represent another way in which the field of prenatal cocaine exposure has matured. We chose to review these 10 studies separately because this is a new area for this field and many are preliminary with relatively small sample sizes and were not able to control for confounding variables as in the 42 studies reviewed above. Quantitative neuroimaging studies in children with prenatal cocaineinclude anatomical magnetic resonance imaging (MRI), diffusion tensor imaging, functional MRI, and magnetic resonance spectroscopy (MRS).

In anatomicor structural MRI, the volumes of cortical gray matter, white matter, subcortical structures, corpus callosum, and cerebellum have been studied. In a study of 41 11-year-old children, Behnke et al. reported a decrease in the volume of the right anterior cerebellum in cocaine-exposed children.⁴⁴ Singer et al. reported a decrease in gray matter volume in the right parietal and left occipital lobes, as well as a decrease in corpus callosum volume in 35 cocaine-exposed children aged 7 to 8.⁴⁴ They also found a decrease in the corpus callosum, which contains the largest fiber tract in the brain. A mean decrease in the caudate volumes in 49 cocaine-exposed children at 13 to 15 years of age has also been reported.⁴⁵ We⁴⁶ reported a volumetric decrease in cortical gray matter, thalamus, and putamen in a sample of 22 cocaine-exposed and comparison children between 9 and 11 years of age. We also found an inverse “dose response” relationship between prenatal exposure level and these volumes. The caudate and putamen findings in these last two studies^45,46 are noteworthy because these are major dopaminergic areas. Finally, although lower mean cortical gray matter and total parenchymal volumes were reported by Rivkinin in a study of 35 children aged 10 to 14, cocaine effects did not remain⁴⁷ when adjusted for prenatal exposure to other drugs.

In a study using diffusion tensor imaging, cocaine-exposed children (age: 10.6 years, N = 53) had increased average diffusion in left frontal callosal and right frontal projection fibers.⁴⁸ They found no differences in fractional anisotropic values. A brain MRS study of children exposed to cocaine in utero showed increased Creatine in the frontal white matter, with normal N-acetyl aspartate in the absence of any visible volumetric brain abnormalities on MRI.⁴⁹ It was suggested that the increased Creatine could be due to glial proliferation or an abnormal energy metabolism. Using perfusion fMRI, Rao et al. reported a 10% decrease in global cerebral blood flow in 49 adolescents with prenatal cocaine exposure.⁵⁰ They found an increase in cerebral blood flow in the anterior and superior brain regions and suggested that this may be due to development of compensatory mechanisms for reduced global cerebral blood flow during neural ontogeny. In an fMRI study of 34 adolescents, both fMRI activation patterns during task and performance on a separate executive function test were similar between cocaine exposed and comparison groups.⁵¹ We⁵² measured fMRI activation during a response inhibition task in 24 children aged 8 and 9 years. Groups did not differ in task performance by design. Cocaine-exposed children showed greater activation in the right inferior frontal cortex and the caudate during response inhibition, suggesting that prenatal cocaine may affect the development of brain systems involved in the regulation of attention and response inhibition.

Recent advances in MRI quality have greatly contributed to our understanding of prenatal cocaine exposure effects on the developing human brain. Longitudinal studies with careful control of other factors (exposure to other drugs and postnatal environmental factors) will add to the excitement of these findings. It will be important to determine whether these imaging findings are related to the child’s behavior and performance on psychological tests outside of the scanner. It is also important to be mindful of the fact that imaging findings do not denote causal mechanisms, but rather associations between a condition and brain involvement, not that the condition was caused by said brain findings. For example, we might speculate that the increases in fMRI,⁵² Creatine,⁵³ diffusion,⁴⁸ and cerebral blood flow⁵⁰ could represent compensatory mechanisms related to other (as yet unknown) areas of brain involvement. Nor do we know the developmental course of these findings, their presence or absence at birth, the effects of prenatal cocaine exposure on normative processes of brain development and maturation, or the role of the postnatal environment in shaping brain architecture, structure, and function.

PRENATAL METHAMPHETAMINE EXPOSURE

Systematic research on the long-term effects of prenatal methamphetamine exposure is limited.⁵⁴ As mentioned in the previous Bandstra review in this supplement, the only published findings from the NIH longitudinal Infant Development Environment and Lifestyle (IDEAL) study are from the infancy period. IDEAL is a multisite (Iowa, Oklahoma, California, and Hawaii) longitudinal study of children with prenatal exposure to methamphetamines. Preliminary findings from approximately half of that sample at age 3 show (Figure 1) no differences between methamphetamine-exposed and comparison children on the Bayley Scales (MDI shown here) at ages 1, 2, or 3 years of age.

FIGURE 1.

Mean Bayley MDI scores in methamphetamine exposed and comparison children at ages 1, 2, 3.

On the Child Behavior Checklist at age 3, the number of children with total problem behavior scores above the clinical cutoff was similar for the methamphetamine exposed (28 of 133; 21%) and comparison group (25 of 138; 18%). This finding should be considered in light of our study of cocaine exposed children. Differences in IQ between cocaine exposed and controls did not emerge until 4.5 years of age and increased from ages 4.5 to 7.⁵⁵ Adverse cocaine effects on the Child Behavior Checklist were not observed until age 5 but remained stable from ages 5 to 7.² It is possible that stimulant drugs, such as cocaine and methamphetamine, affect areas of the brain whose functions become more easily tested as children are forced to deal with settings that place more demands on executive function abilities and require higher cortical functioning, such as school. Continued follow-up is critical to determine whether the methamphetamine effects continue to parallel the cocaine effects.

The only published work on prenatal methamphetamine exposure in school-age children draws from a series of studies from Sweden examining 66 amphetamine-exposed children who have been studied to the age of 14.⁵⁶ At age 4,⁵⁷ there were no differences in physical growth or health. Mean IQ was lower than a separately selected control group from the population. The amphetamine group had more disturbed or problem children if the mother was still addicted. Also at age 4,⁵⁸ child adjustment correlated with maternal alcohol and drug use, maternal stress, and paternal criminal convictions. At age 8,⁵⁹ the extent of prenatal exposure could be related to psychometric outcome, aggression, peer problems, adjustment, and general assessment, as well as alcohol exposure and pregnancy attitudes. Maternal psychiatric treatment, alcohol abuse, and the number of custodians correlated with aggressive behavior and general assessment. The children had problems with advancement in school due to delays in math and language, and at age 14 they had difficulties with physical fitness activities.⁵⁶ The limitations of this work include lack of a control group, small sample size, and prenatal polydrug use (30% of the mothers used heroin,81% used alcohol with 1/3 meeting criteria for alcohol abuse, and 80% smoked more than 10 cigarettes a day). The study is also based on self-report and the route of administration of amphetamine was injection, which is less common today in the United States.

Neuroimaging

Two MRI studies of children exposed prenatally to methamphetamine have been reported.^53,60 Smith et al.⁵³ used proton magnetic resonance spectroscopy (¹H MRS) to evaluate neurochemical alterations in 12 children between 7 and 8 years old who were exposed prenatally to methamphematine and other substances (alcohol and tobacco) and 14 controls. Increases were found in total creatine in the basal ganglia of the methamphetamine exposed group, suggesting possible alterations in cellular energy metabolism; however, no differences were found in N-acetyl-containing compounds that would indicate neuronal damage or loss. In the second study,⁶⁰ volumetric MRI differences in brain regions were found between 13 methamphetamine-exposed (and other drug) children and a group of 15 non-drug exposed controls at 7 and 8 years of age. Methamphetamine was associated with reduced volumes of the globus pallidus, and hippocampus. However, the reductions in the methamphetamine-exposed group were not related to performance on attention and verbal memory tests.

PRENATAL EXPOSURE TO OPIATES

The literature on the long-term effects of children with prenatal opiate exposure is more substantial than the methamphetamine literature but it is still relatively sparse⁶¹ and surprising in that little work has been done since the thorough reviews published ten years ago.^61,62 The opiates studied include heroin and methadone. Currently, there is concern about the abuse of opiate-based prescription medication (e.g., Vicodin or hydrocodone and OxyContin or oxycodone) by pregnant women, as well as the recently approved drug, buprenorphine. No studies have yet been published on the effects of these drugs on the post-infancy child. Methodological similarities and differences exist between the literature studying prenatal cocaine (and to some extent methamphetamine) and opiate exposure. Both literatures face the problem of how to control for virtually the same set of confounding factors when trying to isolate the effects of the target drug. Most pregnant women who use cocaine or opiates are polydrug users and most of their children are raised in high-risk social environments. One major difference between the prenatal cocaine/methamphetamine and opiate studies is that the opiates studies tend to have small sample sizes (20 subjects per group is typical), which precludes statistical adjustment for confounding variables, whereas cocaine studies deal with larger population samples, allowing more valid adjustment for confounding variables.

Some studies were cross-sectional rather than longitudinal. In one such study,⁶³ 3 to 6 year old children exposed to heroin performed more poorly than the comparison groups on the general cognitive index and on the perceptual, quantitative, and memory subscales of the McCarthy Scales. There were three comparison groups—drug-nave subjects in drug-using households, a group with multiple high risk factors, and a socioeconomic comparison group—but there were only 15 to 22 children in each group, including the heroin exposed group. No effects of heroin exposure on measures of speech and language, activity level, mental maturity, or developmental status were found. In a study of 126 offspring of heroin addicts between 8 and 17 years old,⁶⁴ school absence, school failure, and behavior problems in school were more common than in comparison group from the same neighborhoods. Cubas and Field⁶⁵ compared 20 children aged 6 to 13 with prenatal exposure to methadone and a control group of 20 non-exposed children. No differences were found on cognitive tests, although methadone exposure was associated with lower IQ scores. Methadone-exposed children exhibited greater anxiety, aggression, and rejection than did those in the control group, and their mothers reported more behavior problems.

looseness-1In three longitudinal studies, no differences were found on the McCarthy scales related to prenatal opiate exposure. These include: a study comparing methadone-exposed (N = 26), heroin-exposed (N = 25), and a drug-free comparison group (N = 41) in 3 to 6 year old children;⁶⁶ a study of 27 methadone-exposed and 17 non-drug exposed children at age 4 years;⁶⁷ and a study of methadone-exposed (N = 33) and non-drug exposed comparison children (N = 30) at age 5 years.⁶⁸ In the latter study, during the testing session, the drug-exposed children were more active, energetic, immature, and showed more task irrelevant activity during the testing session.⁶⁸ In a longitudinal follow-up of the previously reported sample,⁶³ 32 prenatally exposed 6- to 11-year-old children were compared with 12 controls.⁶⁹ Behavior problems in school and psychiatric referrals were related to maternal opiate use, although the number of cases was small (7 psychiatric referrals in the exposed group and 1 in the comparison group). In a 10-year follow-up of 36 methadone-exposed children and a similar group of comparison children, Hans⁶² found that drug-exposed children were somewhat more likely to receive ADHD and disruptive behavior diagnoses and made more errors on an attentional (continuous performance) task.

As mentioned above, it is known that many of these children grow up in high-risk environments.⁶² Similar learning and behavioral problems have been reported in school-age children either exposed to drugs in utero⁷⁰ or living with drug-dependent parents who were not exposured in utero.^71,72 Some studies have attempted to isolate the effects of in utero opiate exposure by using different types of comparison groups. An early study compared 6- to 15-year-old, opiate-exposed children to a narcotic-environment group (no in utero exposure but caregivers used opiates) using measures of neurological and behavioral functioning.⁷³ Opiate-exposed children had impairments in perceptual, motor, and attentional functioning, whereas the narcotic environmental group did not. In more recent work, lower scores were found on the McCarthy scales at age 4.5 in a Norwegian study of 64 exposed and 52 comparison children under conditions of minimal postnatal social risk.⁷⁴ Social risk was “controlled” by studying children of heroin users (and other drugs) placed in the care of foster or adoptive parents who were specially selected to provide care for infants at risk. In an Israeli study attempting to control the postnatal environment, the development of children aged 5 to 12 years born to mothers with heroin dependency who were raised at home or adopted was studied in comparison with children with environmental deprivation alone (i.e., low parental socioeconomic status and evidence of neglect), children born to fathers with heroin dependency,⁷⁵ and control individuals of average socioeconomic status.⁷⁶ Children born to parents with heroin dependency who were raised at home showed impaired verbal, performance, and reading and arithmetic skills. Children born to mothers with heroin dependency but who were adopted at a young age had normal intellectual and learning abilities, but some had reduced function on performance IQ. Children born to parents with heroin dependency had a high rate of ADHD, including those who were adopted; the highest rate of ADHD was seen in children born to mothers with heroin dependency who were raised at home.

One physiological study attempted to control for opiate-related environmental factors, comparing a group of children born to mothers who began drug use after the child was born to a group of children with intrauterine opiate exposure for environmental influences associated with parental drug use.⁷⁷ Heart rate variability (vagal tone) was measured during an attention task. When distracters were added to the attention task, opiate-exposed boys failed to suppress vagal tone compared to both control groups. However, both the opiate-exposed boys and the environmental controls made fewer correct responses than non-drug-exposed controls on this task. Thus, physiological responses to increased attention demands may be affected by prenatal opiate exposure as well as environmental influences.

In a neuroimaging study of this population,⁷⁸ morphometric cerebral characteristics were studied in 9- to 11-year-old children with prenatal exposure to opiates and other drugs (n = 14) compared to unexposed controls (n = 14). Compared to controls, the drug-exposed children had smaller intracranial and brain volumes, including smaller cerebral cortex, amygdala, accumbens area, putamen, pallidum, brainstem, cerebellar cortex, cerebellar white matter, and inferior lateral ventricles, as well as a thinner cortex of the right anterior cingulate and lateral orbitofrontal cortex. There was some association between volumes of the right anterior cingulate, the right lateral orbitofrontal cortex, and the accumbens area and cognitive ability and behavior problems.

UNDERSTANDING THE EFFECTS OF PRENATAL DRUG EXPOSURE

One determinant of the long-term effects of prenatal drug exposure is the pathophysiology of the drug itself. Typically, we think of drugs as teratogens that cause malformations or defects depending on the drugs’ mechanisms of action. For example, the effects of cocaine on blocking the reuptake of neurotransmitters, especially dopamine, have been well documented. However, in addition to teratogenic effects, there may also be non-teratogenic effects of drugs. We⁷⁹ have described a model in which cocaine, as well as other substances, acts as an intrauterine stressor that disrupts the neuroendocrine environment and the genetic programming of fetal-placental development.

Increasing evidence from preclinical, prospective clinical and epidemiological studies suggest that, at least in the case of disease, early development does echo throughout life.^80-82 The literature on the influence of prenatal stress on the offspring suggests that many biological factors acting during prenatal life are associated not only with the development of common adult cardiovascular and metabolic disorders but also with neurobehavioral abnormalities^83-93 and behavioral disorders.^94-97 These findings have given rise to the concept of “fetal origins of adult disease” or the “developmental origins of health and disease.” Although original studies related low birthweight to adult disease, it is generally accepted that low birth weight per se is not at the heart of these disorders, but that there are common factors that influence intrauterine growth, as well as adult physiological systems.⁸² The “fetal origins” observations are due, in part, to environmental factors acting early in life that “program” developing systems, altering structure and function and probably behavioral expression. Prenatal drug exposure may be one of those factors that alter the set-points or “hard-wire” physiological systems. Stress hormones such as catecholamines and glucocorticoids can alter regulation of the neuroendocrine environment by acting on the hypothalamic-pituitary-adrenal (HPA) axis, which results in an altered set point for physiologic, metabolic, and behavioral outcomes.⁹⁸ Because they are an important feature of the stress response, glucocorticoids have become prominent candidates as mediators of the effects of “programming.”

Fetal Programming and Brain

The brain is particularly sensitive to prenatal programming including effects on the HPA axis. Pregnant dams exposed to stress show increased adrenocorticotrophic hormone (ACTH) and corticosterone (CORT).^{82,99, 100} Prenatal stress also increases CORT and ACTH levels in adult offspring.^{80-82, 96,98} The effect of prenatal stress on adult hippocampal corticosteroid receptor density^95,101-104 may have implications for emotional reactivity. Prenatally, stressed rats have a high degree of “emotionality,”¹⁰⁵ indicated by decreased locomotion and increased defecation.^105-108 They also show less play,¹⁰⁹ more defensive freezing,¹¹⁰ less movement in an activity wheel,¹¹¹ and increased vocalizations.¹⁰⁰ Prenatal stress affects cognitive abilities including operant discrimination,¹¹² reversal task in a water maze,^113,114 and memory.¹¹⁵ In other animal studies, maternal stress during pregnancy results in offspring that are more irritable, anxious, and difficult to control.^82,116-119

Brain neurotransmitter systems and glucocorticoids interact to modulate both behavior and HPA activity.¹²⁰ Disturbances in HPA regulation and brain monoamine levels have been associated with affective and anxiety disorders in humans.^121-124 Also, in human studies, poor health outcomes, such as low birthweight, preterm birth, and intrauterine growth retardation, have been related to prenatal stress.^125,126 Moderate to severe stressful life events during midgestation was related to birth weight and small head circumference, suggesting effects on the brain.¹²⁷ Maternal first-trimester exposure to the stress of war has also been associated with an increased risk of the offspring developing schizophrenia in adult life.¹²⁸ Similar to the “emotionality” reported in animal studies, human infants exposed to stress in utero show high reactivity, activity, and irritability.^129-131 Psychological and behavioral abnormalities, including learning and behavior problems, have also been reported in children exposed to prenatal stress.^132-135 The effects of prolonged exposure to chronic stress, called allostatic load,¹³⁶ and, concomitantly, prolonged activation of the neuroendocrine stress axes has also been related to physical disease and behavioral disorders.¹³⁷

Drugs and Gene Expression

Our model⁷⁹ argues for the effects of an adverse intrauterine environment, created by maternal use of cocaine and other substances, on catecholamines and glucocorticoids. As a stressor, a drug such as cocaine programs the HPA axis as well as behavior, due in part to plasticity of brain monoamine systems. Specifically, a drug as a stressor alters the expression of key candidate genes and gene networks important to placental function in late gestation. We focus on the norepinephrine transporter (NET) and a steroid metabolic enzyme, 11β-HSD-2 and placental gene networks.

Placental NET and 11ß-HSD-2 are pivotal placental genes that program the intrauterine neuroendocrine environment during development by protecting the fetus from excess catecholamines and glucocorticoids, which have harmful effects on the fetus.¹³⁸ Lower placental 11ß-HSD-2 activity is associated with smaller fetuses in rats¹³⁹ and humans.^140-143 Mutations of 11ß-HSD-2 are also associated with low birth weight in human infants¹⁴⁴ and increased fetal cortisol levels are associated with IUGR¹⁴⁵ 11ß-HSD-2 modulates the programming effects of prenatal glucocorticoid exposure.^146,147 The HPA axis is highly sensitive to the effects of glucocorticoids on perinatal programming.^82,148-150 High levels of maternal glucocorticoids have been shown to disrupt intrauterine growth, postnatal HPA axis function, and neurobehavioral outcome in offspring. Placental expression of this key enzyme is potently downregulated at the RNA, protein, and functional level by norepinephrine, which is in turn regulated by NET. Downregualtion of NET is associated with adverse intrauterine events, including maternal/placental disorders such as preeclampsia and exposure to drugs such as cocaine and nicotine.^151,152 Reduced placental NET expression from drug exposure may lead to increased circulating catecholamines, downregulation of 11β-HSD-2 and chronic fetal hyper-cortisolism, leading to altered neuroendocrine (HPA axis) activity and dysregulated neurobehavioral functioning. Because catecholamines are released during stress, this may be a mechanism linking prenatal stress with altered fetal development mediated by effects on 11β-HSD-2.

The altered expression of these two key candidate genes is likely associated with changes in networks of genes involved in critical placental functions that maintain physiological homeostasis in utero and otherwise promote intrauterine growth, development, and preparation for postnatal life. Epigenetic mechanisms (i.e., environmentally induced changes in gene expression without altering DNA sequences) are thought to be involved. Our preliminary findings⁷⁹ showed that these changes in placental gene expression are associated with altered methylation of placental genomic DNA, particularly in promoter regions. We found that dysregulation of 11β HSD2 related to cocaine or nicotine use during pregnancy was associated with hyper-methylation, suggesting DNA silencing. Significant interest exists in the concept that modifications of DNA mechanistically affect “adaptation to the environment.” Complex processes such as the development of memory are now being attributed to chromatin remodeling, epigenetic modifications of DNA, and lasting alterations in gene expression.¹⁵³ Drugs of abuse induce “adaptations” in brain regions. This may occur through alterations in gene expression, altered chromatin remodeling, and alterations in the number and projections of neurons in specific regions. Recently, cocaine was shown to induce neuroadaptations through altered gene expression.¹⁵⁴ It is becoming widely accepted that chromatin remodeling is an important regulatory mechanism underlying gene environment interactions, learning, and drug-induced neural and behavioral plasticity. Similar alterations in DNA methylation and histone acetylation are seen following intrauterine growth restriction.¹⁵⁵ It is thus likely that altered expression of these key candidate genes and associated changes in networks of genes involved in placental function follows intrauterine exposure to stress, cocaine, and other substances. Some of these changes may lead to permanent “epigenetic” alterations in placental gene expression through DNA methylation and chromatin remodeling. Because of the unique nature of intrauterine development, these observations suggest mechanisms whereby an altered placental environment can have effects on neurodevelopmental outcome.

Behavioral Dysregulation

In our model,⁷⁹ these disruptions in the placental environment alter HPA set points, resulting in neuroendocrine and behavioral dysregulation in the newborn that can lead to the type of childhood behavioral, emotional, and neurocognitive deficits related to prenatal drug exposure described earlier. These deficits may, in turn, have implications for adolescent psychopathology, including early onset of substance use in adolescence. Behavioral dysregulation can begin in utero and is proposed to be a dynamic developmental process as alterations in the quality of the environment (prenatal and postnatal) modify behavioral expression. Behavioral dysregulation is evidenced during infancy as neurobehavioral and neuroendocrine disorganization. During childhood, indicators of behavioral dysregulation reflect a deficient capacity to control behavior and regulate emotion. These phenotypes are important because they appear to be prognostic indicators for substance use.¹⁵⁶ The earlier a drug is used, the greater the likelihood of its abuse during adolescence and adulthood.^157,158 These phenotypes reflect major domains of psychological function, cognition, affect, and behavior and include impulsivity, reactive aggression, sensation seeking, excessive risk taking,^159,160 irritability, negative affect, difficult temperament,^161-163 conduct disorder, attention deficit hyperactivity disorder, oppositional defiant disorder, anxiety, depression,¹⁶⁴and impaired executive function.^165-167 This collection of disturbances in emotion regulation and behavior control is included in the construct of neurobehavioral disinhibition and is thought to reflect disturbances in the prefrontal cortex.¹⁶⁸ Some of these disturbances have already been reported in the prenatal cocaine exposure studies reported above and may become more visible as these youths enter adolescence. Disruption of neuroendocrine homeostasis in utero by prenatal drug exposure can be observed at birth and may lead to lasting behavioral dysregulation that increases vulnerability to substance use, resulting in early onset substance use in adolescents.

Adolescent Brain Development

The understanding that addiction is a developmental disease means that it is critical to consider the influence of specific developmental periods.¹⁶⁹ In addition to fetal development, we know that brain development continues well beyond childhood and adolescence.^170,171 The “immaturity” or abnormal development of the adolescent brain may be related to risk taking behavior, including substance use.

Extensive maturational changes that occur in brain development in adolescence may confer vulnerability to substance use through disruption of the HPA system.^172-174 Maturation of high-order association cortices, including changes in prefrontal cortex, occurs later than low-order sensory cortices during late adolescence.⁷⁵ Neurophysiologic,^175,176 and fMRI¹⁷⁷ studies have shown disruption of the frontal cortex in youths at high risk for substance use onset. These youths show deficits related to substance use onset, including poor affect modulation, behavior control, and executive function processes thought to reflect a neuromaturational disturbance subserved by neural systems in the anterior cerebral cortex.^160,178-180 Stress reactivity is related to pubertal development and brain regions, such as the hippocampus, pre-frontal cortex, and amygdale, and anatomic areas that are highly sensitive to stress hormones and regulate emotionality and continue to mature during the peripubertal period.¹⁸¹ Glucocorticoid receptors are present in both subcortical (paraventricular nucleus and other hypothalamic nuclei, the hippocampus and parahippocampal gyrus) and cortical structures, with a preferential distribution in the prefrontal cortex.^120,182-184 The distribution of glucocorticoid receptors in the primate brain are more closely related to the human brain in terms of neocortex development.^185,186 Non-human primate studies suggest that the behavioral effects of prenatal cocaine exposure may not be manifested until later in development.^187,188 Chronic exposure to stress or allostatic load leads to atrophy and impaired neuronal function in the hippocampus and medial prefrontal cortex,^189-191 regions that are responsible for executive function and adaptation to stress. Cocaine and other substance related to “fetal programming” could have long-term effects on HPA axis activity, resulting in a physiological endophenotype of cortisol reactivity that could be a harbinger of substance use. Although there are studies of cortisol reactivity related to prenatal drug exposure in infancy, as described in the previous article by Bandstra et al., we did not find any studies relating prenatal cocaine, methamphetamine, or opiate exposure to cortisol in older children.

The advent of the “Genome Era” has dissolved disciplinary boundaries and transformed the research landscape. The sequencing of the human genome has lead to advances in our understanding of the cellular, molecular, and biochemical mechanisms that regulate development, learning, behavior, and the molecular pathways that could be altered due to prenatal substance exposure. These are exciting times as advances in neurobiology, including genetics, epigenetics, fetal programming, and neural plasticity, will provide a better understanding of the pathophysiology and long-term implications of prenatal drug exposure that should have implications for public policy, preventive intervention and treatment.