Parenting and family factors

The 1996 National Household Survey on Drug Abuse found that of the 75 million US children under the age of 18 who were represented in the survey, 4% lived with at least one parent who was dependent on illicit drugs, 6% lived with at least one parent in need of illicit drug treatment, 14% lived with one or two parents who reported illicit drug use in the past year, and 11% lived with one or two parents who reported past month drug use. For comparison, 8% lived with at least one parent who was dependent on alcohol (Huang, Cerbone, & Gfroerer, 2004). According to the Office of Applied Studies (2002), National Survey of Drug Use and Health (NSDUH) data show that in 2001, more than 6 million children in the United States lived with at least one alcohol or drug-abusing parent including approximately 10% of children under the age of 6, 8% of children between 6 and 11 years of age, and 9% of children between 12 and 17 years of age.

For 2002–2003, analysis of NSDUH data shows that 16.1% of mothers living with a child 12–17 years old reported significant drug or alcohol use; this means that approximately 3,864,000 US children lived with a substance-abusing mother but does not include those who lived with a substance-abusing father or other adult (Office of Applied Studies, 2005b). Although it is common for drug-abusing women to report that they are concerned that their drug abuse may harm their children, there is some evidence that having children is associated with a lower likelihood of seeking drug abuse treatment (McMahon, Winkel, Suchman, & Luthar, 2002).

The research literature on prenatal drug exposure focuses heavily on the behavior and characteristics of the abusing mothers and their influence on their children. This is in part because PNDE is a direct result of maternal behavior and in part because the fathers of the children often have a smaller role in the children's development. However, at the very least the fathers contribute to their children's genetic inheritance and to the extent that they are involved in raising their children, they are an influence (e.g., Fals-Stewart, Kelley, Fincham, Golden, & Logsdon, 2004). Unfortunately, substance-abusing women are often involved with substance abusing men; there is assortative mating for both antisocial personality and substance abuse (Sakai et al., 2004).

Many substance-abusing parents try to protect their children from harm associated with their drug abuse (Richter & Bammer, 2000), and many provide adequate care (e.g., Kearney, Murphy, & Rosenbaum, 1994). However, although some children of drug-abusing parents are well adjusted and become successfully functioning adolescents and adults, a significant number develop problems. Parental drug abuse has the potential to negatively affect multiple aspects of development, including cognitive and social domains, as well as drug use outcomes (e.g., Chronis et al., 2003; Compton, Cottler, Ridenour, Ben-Abdallah, & Spitznagel, 2002; Conners et al., 2004; Fals-Stewart et al., 2004; Langbehn, Cadoret, Caspers, Troughton, & Yucuis, 2003; Nuñes et al., 2000). A range of problems is associated with having drug-abusing parents.

Parenting attitudes and behaviors

Parenting quality is commonly assessed by parenting attitudes and parenting behaviors. Research comparing drug-abusing and nondrug-abusing parents has not found clear differences in parenting attitudes between these groups. Early findings suggested that parenting attitudes between drug-abusing and nondrug-abusing parents tend to be similar (e.g., Bauman & Dougherty, 1983). More recent research reports that parenting attitudes of drug-abusing parents depend on their level of environmental risk (i.e., exposure to domestic and community violence, depression, incarceration, and other environmental stressors), which suggests that parenting attitudes may vary depending on particular contextual factors (Kettinger, Nair, & Schuler, 2000).

In 1992, approximately 15% of US children under the age of 18, more than 9.5 million children, were living in households with one or more alcohol-abusing or -dependent adults (Grant, 2000). In 2002, almost 5 million alcohol-abusing or -dependent adults had at least one child younger than age 18 living in their home, (Office of Applied Studies, 2004). There is a sizeable research literature on the impact of parental alcoholism on the parenting and environmental consequences for their children, and a wide range of detrimental effects has been demonstrated (e.g., Olson, O'Connor, & Fitzgerald, 2001; Weinberg, 1997). Many of these same factors are associated with being the child of a drug abuser.

Research has consistently shown a lower quality of parenting by drug-abusing parents (e.g., Hans, 2000, 2002; Hans, Bernstein, & Henson, 1999; Mayes, 1995; Tarter, Blackson, Martin, Loeber, & Moss, 1993). Results suggest that drug-abusing parents tend to be less positive (Fitzgerald, Kaltenbach, & Finnegan, 1990), less responsive (Burns, Chethik, Burns, & Clark, 1991, 1997; Hans et al., 1999), use less positive reinforcement (Bauman & Dougherty, 1983), use more overly harsh and punitive discipline (Bauman & Dougherty, 1983; Hans et al., 1999; Miller, Downs, & Gondoli, 1989; Miller, Smyth, & Mudar, 1999), be less supportive (Johnson et al., 2002), and provide less supervision and monitoring than nondrug-abusing parents (Chilcoat, Breslau, & Anthony, 1996; Miller et al., 1989; Tarter et al., 1993). Furthermore, these parents spend less time with their children (Miller et al., 1989; Tarter et al., 1993). Child neglect is more common among parents who have a substance use disorder (Dunn, Mezzich, Janiszewski, Kirisci, & Tarter, 2001; Ondersma, 2002). Cash and Wilke (2003) found that severity of drug use predicted incidence of neglectful behaviors among drug-abusing women.

Evidence suggests that parents' substance abuse negatively impacts the development of attachment relationships and parent–child bonds. Rodning, Beckwith, and Howard (1991) found that the majority of their sample of infants who were prenatally exposed to PCP and cocaine were insecurely attached, and that all infants whose mothers who continued to abuse drugs postnatally demonstrated insecure attachments. Alternatively, half of the infants whose mothers were abstinent during the first 15 months of life formed secure attachment relationships with their mothers. In the multisite, longitudinal Maternal Lifestyle Study (MLS), Seifer et al. (2004) found that at 18 and 36 months, only children who had been prenatally exposed to both cocaine and opiates had a lower level of attachment security, although the difference compared to a similar nonexposed population was slight. Child temperament factors may play a role in this. Furthermore, continued postnatal alcohol use was associated with higher rates of insecurity and disorganization at 18 but not 36 months of age. Espinosa, Beckwith, Howard, Tyler, and Swanson (2001) found that cocaine-using mothers with clinical levels of depression had toddlers who were more likely to be disorganized/disoriented in their attachments. Hans et al. (1999) found that illicit drug users had more difficulty fostering emotional connections to their children, and viewed their relationships as detached and disengaged. Similarly, Catalano, Haggerty, Fleming, Brewer, and Gainey (2002) found that children of substance abusers are not as close to their parents compared to children from the general population. They also found that parent–child bonds were associated with more drug use by children whose parents continued drug use but with less drug use for children whose parents discontinued use.

Although parent drug abuse seems to be detrimental to parent–child interactions, the relationship is not a simple unidirectional one; it is a transactional system that evolves between the parents and the children over time (Cicchetti & Rogosch, 2002). Child characteristics and behaviors affect responses elicited from parents, and therefore influence the pattern of parent–child interactions, which consequently affects their own development. Some child and family characteristics may act as buffers, or protective factors. For example, Hoffman and Cerbone (2002) found that family cohesion weakened the association between parent drug abuse and adolescent drug abuse. In contrast, other child characteristics may accentuate certain risks. For example, difficult temperaments, which are frequently observed in children of drug abusers, have been linked with poor parenting practices (Mun, Fitzgerald, Von Eye, Puttler, & Zucker, 2001).

Prenatal research suggests it is likely that the atypical transactional patterns between drug-abusing parents and their children begin at birth. Several studies have found that both prenatally cocaine and opiate exposed infants have feeding problems (e.g., Hans, 1992; Minnes, Singer, Arendt, & Satayathum, 2005). However, LaGasse et al. (2003) concluded from MLS data that the critical issue seems to be the feeding interaction rather than just feeding behavior of the infants, and they hypothesize that this may reflect problems in the developing mother–infant relationship. Other findings also support this hypothesis. For example, Molitor, Mayes, and Ward, (2003) found a lower degree of emotional engagement and responsiveness in the mother–child interaction of 18-month-old prenatally exposed children of cocaine using mothers and determined that this was related to the behavior and responses of both the children and the mothers. Tronick and colleagues found small but significant negative engagement effects in the interactions of MLS prenatally cocaine exposed infants and their mothers (Tronick et al., 2005). Johnson et al. (2002) reported a deficit in the mutual regulation in mother–child interactions between prenatally cocaine exposed 3 year olds and their mothers who continued to abuse drugs.

Child maltreatment

Although parent drug abuse and child maltreatment are separate aspects of family dysfunction, they are clearly related (Felitti et al., 1998; Fleming, 1998; Murphy et al., 1991; Stein, Leslie, & Nyamathi, 2002). Studies suggest that there are 3- to 12-fold increases in rates of child maltreatment reports among families with prenatally exposed children (Jaudes, Ekwo, & Voorhis, 1995; Kelley, 1998; Wasserman & Leventhal, 1993). Reviews of juvenile and protective services court records indicated that between 43 and 79% of youth who reported maltreatment also reported having at least one parent with a substance abuse problem (Famularo, Kinscherff, & Fenton, 1992; Murphy et al., 1991). In a study of 639 children removed from their homes for maltreatment, Besinger, Garland, Litrownik, and Landsverk (1999) reported evidence of parent substance abuse among 79% of the cases. Dunn et al. (2001) found that neglectful parenting is more likely among mothers who experienced neglect themselves and is exacerbated if they have a substance use disorder. These reports are consistent with a number of other studies that have found a high rate of concomitant reports of childhood maltreatment and parent substance abuse (e.g., Leventhal et al., 1997; Melchert, 2000).

Comorbid psychopathology

Psychopathology, which is highly common among substance abusing parents (Hans et al., 1999; Luthar & Cushing, 1999; Nair, Schuler, Black, Kettinger, & Harrington, 2003), is another factor associated with child maltreatment and other problems in families affected by drug abuse. Luthar and Cushing (1999) found that in their sample of opioid and cocaine addicted parents 90% of mothers and 80% of fathers had experienced a comorbid psychiatric disorder. Analysis of combined 2002–2003 data from NSDUH found that 3.2% of mothers with both significant alcohol or illicit drug use and serious mental illness lived with children aged 12–17, approximately 800,000 children. Having a mother with a serious mental illness increased the odds of the adolescents' drug or alcohol use by 44%, and considering the influence of both maternal mental illness and heavy alcohol or illicit drug use, the odds of substance use among youths increased 93% when the mother was a substance abuser (Office of Applied Studies, 2005b).

Evidence suggests that the effect that comorbid psychopathology has on drug-abusing parents may be dependent on the drug being used and on their particular comorbid diagnosis. For example, maternal depression has been linked to lack of maternal attentiveness in cocaine using women (Ball, Mayes, DeToso, & Schottenfeld, 1997). Personality disorders are at particularly high rates among substance-abusing women, and are associated with insensitive and unresponsive parenting in drug-using women (Hans et al., 1999) and to disorganized parent–child attachment (Espinosa et al., 2001). In a sample of opioid-dependent women, Hans and colleagues found that unresponsive and negative parenting was most attributable to mothers' comorbid psychopathology (Hans et al., 1999).

Comorbid parent psychopathology in families affected by drug abuse has been linked with a number of deleterious child outcomes. When the comorbid parent illness is antisocial personality disorder, families are more likely to experience forced parent–child separations (Hans, 2002; Hans et al., 1999). Compared to children with fathers who are diagnosed only as substance abusers and children in control families, children of substance abusing fathers with antisocial personality disorder demonstrated more externalizing and internalizing disorders (Moss, Lynch, Hardie, & Baron, 2002). Langbehn et al. (2003) found that for adoptees with a biological parent with both substance abuse and antisocial personality disorder, the risk of drug use and dependence, as well as aggression, was increased. In a sample of 78 children of cocaine and opioid addicted mothers, Luthar and colleagues (Luthar & Cushing, 1999; Luthar, Cushing, Merikangas, & Rounsaville, 1998) found that poorer functioning children had mothers with greater psychiatric disturbances, and almost two-thirds of the children had been diagnosed with a major psychiatric disorder by preadolescence. Accornero and colleagues (Accornero, Morrow, Bandstra, Johnson, & Anthony, 2002; Accornero, Morrow, Xue, Anthony, & Bandstra, 2005) found that cocaine use during pregnancy predicted maternal cocaine use 5 years later, but did not predict the child's internalizing or externalizing behavior as measured by parent report using the Child Behavior Checklist (CBCL; Achenbach, 1991). Although prenatal exposure to cocaine was not related to child behavior problems, recent maternal drug use and psychological distress were related to child behavior problems (Accornero et al., 2002). Interestingly, the subsequent reports of trained observers at 7 years indicated that there was a behavioral impairment consequence of prenatal drug exposure (Accornero et al., 2005).

Psychopathology by itself is known to detrimentally affect parenting, and a co-occurring drug use disorder can only worsen a mother's problems and dysfunction. Regardless of which disorder has a more primary role in influencing maternal behavior, having a mother who has both a drug abuse disorder and a comorbid psychopathology increases the risks for negative outcomes for the children. Not only are the mother's comorbid disorders likely to interact and exacerbate each other, but each may have independent negative effects on the child (e.g., Hans et al., 1999). Fineman, Beckwith, Howard, and Espinosa (1997) suggested that “it is the psychological symptoms of drug abuse that appear to have a greater relationship to parenting, rather than the drug itself, or the frequency and severity of drug use (p. 315). Even if this were a valid characterization of the maternal psychiatric symptomatology, it does not take into consideration the many harmful pragmatic and environmental consequences that are typically associated with maternal drug abuse.

Drug involved lifestyle

Abusing drugs influences many aspects of the abuser's life, and a drug-involved lifestyle can have a great impact on children's developmental trajectories. For example, children in families affected by drug abuse have been found to exhibit greater internalizing and externalizing symptoms than children of families without drug abuse as well as compared to children in families affected by alcohol abuse (Fals-Stewart et al., 2004). This might be due to characteristics of the lifestyle specifically associated with the illegal behaviors involved in illicit drug use.

Acute versus chronic effects

In some cases, PNDE consequences include acute outcomes observed primarily in the neonatal period. These generally transient effects are usually assumed to be related to postnatal withdrawal from the abusable drug, and their resolution may be improved with intervention. Postnatal outcomes of opiate exposure include a range of CNS, autonomic (ANS), and gastrointestinal symptoms often referred to as neonatal withdrawal (or abstinence) syndrome (Bada, Bauer, et al., 2002; Chiriboga 2003). The symptoms typically diminish over the first month, although it is not clear to what extent any subtle behavioral symptoms persist. For example, poorer feeding behavior is evident at 1 month (LaGasse et al., 2003), as are alterations in auditory brain response (Lester et al., 2003). Prenatal cocaine exposure is associated with an increased risk for a constellation of CNS and ANS outcomes including a transient neonatal dysphoric condition with marked state dysregulation (Bada, Bauer, et al., 2002; Bauer et al., 2005). The symptoms, which often manifest as irritability, impulsivity, and aggression, tend to diminish over time, and are usually resolved within a few days to a month, although the ultimate extent of resolution is unclear. For example, differences on a range of neurobehavior tests measuring level of arousal, quality of movement and self-regulation, excitability, hypertonia, and reflexes have been found at 1 month (Lester et al., 2002). High levels of prenatal marijuana exposure have been associated with some acute neonatal symptoms, but the symptoms appear to be much less severe than those associated with cocaine and opiate exposure (Mayes & Fahy, 2001). Although the acute neonatal symptoms of prenatal cocaine and opiate exposure typically diminish over time, even if there is recovery to normal levels it is not clear whether the child's developmental trajectory has been detrimentally altered. The PNDE findings reviewed in the following sections focus on outcomes that are not thought to be the direct, acute effects of drug withdrawal.

Prenatal marijuana exposure

Developmental follow-up has been longest for prenatal marijuana exposure (PNME) and as a result, the long-term consequences of PNME are more definitively described by empirical research (Fried & Smith, 2001; Goldschmidt, Richardson, Cornelius, & Day, 2004; Richardson, Ryan, Willford, Day, & Goldschmidt, 2002). Although exposure-related differences have been found for a number of outcomes, impairment involving prefrontal cortical function has been most clearly indicated, although the effects are generally subtle.

An emergent theme of the Ottawa Prenatal Prospective Study is that PNME has a negative influence on offspring beyond the toddler stage in terms of aspects of executive function but not on overall measures of intelligence (Fried, 2002a). Based on their review of the literature, Fried and Smith (2001) have proposed that prenatal marijuana exposure is associated with long-term deficits in aspects of executive function, particularly attentional behavior and visual analysis/hypothesis testing. Providing further support for this, Fried, Watkinson, and Gray (2003) report that, consistent with findings for their cohort at earlier ages, at 13 to 16 years of age, PNME was associated with poorer performance on tasks requiring attention, visual memory, analysis, and integration, and had a detrimental impact on the application of intellectual and visuoperceptual skills to analytical and problem solving tasks (see also Fried, 2002b).

Lester et al. (2002) examined neurodevelopmental outcomes of PNME at 1 month using the NICU Network Neurobehavioral Scale (NNNS), which assesses neurologic, behavioral, and stress/abstinence neurobehavioral function, and yields multiple scale summary scores, including attention, arousal, regulation, excitability, and lethargy (Lester & Tronick, 2004). Some marijuana exposure effects related to poorer regulation and higher excitability were found. Two long-term cohort studies of prenatal marijuana exposure have published data for multiple ages on a variety of related cognitive and behavioral outcomes, including response inhibition and impulsivity (Smith, Fried, et al., 2004; Richardson et al., 2002). A variety of measures were used across ages and cohorts. In one of the cohorts, prenatal marijuana exposure was associated with poorer self-regulatory abilities in children 6 years old (Fried, O'Connell, & Watkinson, 1992) and 9 to 12 years old (Fried, Watkinson, & Gray, 1998), with higher parental ratings of impulsive/hyperactive activity in 6-year-olds (Fried, Watkinson, & Gray, 1992), and with deficits in sustained attention for 13- to 16-year-old adolescents (Fried & Smith, 2001). In the other cohort, prenatal marijuana exposure was related to increased inattentiveness and impulsivity at 6 years of age (Leech, Richardson, Goldschmidt, & Day, 1999), as well as hyperactivity, impulsivity, and inattention at 10 years (Goldschmidt, Day, & Richardson, 2000; Richardson et al., 2002). Level of exposure associations were seen in some of these findings. It has been reported that among 18- to 22-year-olds for whom functional magnetic resonance imaging data were collected during a task requiring response inhibition, increased levels of PNME were associated with specific neural activity patterns, suggesting a long-term impact of PNME on the neural system involved in response inhibition when the exposed offspring are in late adolescence and early adulthood (Smith, Fried, et al., 2004).

Other aspects of development may also be affected by PNME. For example, Faden and Graubard (2000) found increased fearfulness, poorer motor skills, and shorter length of play at age 3. Some detrimental effect on language skills at 4 years has been found for PNME (Lewis, Singer, et al., 2004); however, significant deficits in reading or language outcomes in 9- to 12-year-olds were not found (Fried et al., 1997). Although PNME does not seem to impact composite measures of global intelligence (Fried & Smith, 2001), it does seem to impact academic achievement (Goldschmidt et al., 2004) and specific aspects of cognition at different ages. For example, negative effects on learning and memory at age 10 have been reported (Richardson et al., 2002).

PNME has also been associated with behavioral and psychiatric problems. Gray, Day, Leech, and Richardson (2005) found that children who were prenatally exposed to marijuana had increased symptoms of depression at 10 years of age. Goldschmidt et al. (2000) reported that PNME was associated with increased delinquency and externalizing problems at age 10. Porath and Fried (2005) found that for 16- to 21-year-old youth, those who were prenatally exposed to marijuana were at increased risk for initiation of cigarette smoking, daily cigarette smoking, and marijuana use, relative to youth whose mothers did not report using marijuana during pregnancy. Evidence also indicated a dose–response relationship between prenatal exposure to marijuana and use of cigarettes and marijuana.

Prenatal cocaine exposure

Findings are most extensive for cocaine, which is indicative of the emphasis in prenatal drug exposure research over the past 15 years. Published data for prenatal cocaine exposure provide findings for assessment ages ranging from the neonatal period to preadolescence (e.g., Eyler et al., 2001; Savage, Brodsky, Malmud, Giannetta, & Hurt, 2005). Research involving exposure to methamphetamine in utero began relatively recently, and some early reports are available (Arria et al., 2006; Scott, Fleming, Bennett, & Graves, 2005; Smith, Lagasse, et al., 2004). Research reports indicate that there is a modest but consistent effect of prenatal cocaine exposure (PNCE) on neurobehavioral function in neonates (Morrow et al., 2001). Some of these effects appear to persist to older ages.

Although numerous studies have reported that PNCE infants have poorer motor performance at least through the first 2 years, the MLS found that the deficits in motor development identified at 1 month generally recovered to normal levels of function by 18 months (Miller-Loncar et al., 2005). Alternatively, Arendt et al. (2004) reported finding lower visual and motor skills at age 7 associated with PNCE. Similarly, inconclusive findings are reported for other assessments of developmental level. Behnke, Eyler, Garvan, Wobie, and Weir (2002) reported lower development at 6 months on Brazelton qualifier scores for PNCE infants and Singer et al. (1997) found lower scores on standard assessments of mental and motor development at 1 to 2 years old comparing PNCE children to nonexposed children with similar demographics (see also Singer, Arendt, Minnes, Farkas, & Salvator, 2002). In contrast, although Mayes, Cicchetti, Acharyya, and Zhang (2003) found that prenatally cocaine exposed children demonstrated delayed mental development through 3 years of age, the trajectories were similar to those from noncocaine exposed comparison groups. There is some evidence that males who were prenatally exposed to cocaine may be more vulnerable to deficits in motor development and emotional regulation (Lewis, Misra, et al., 2004), which may account for some of the inconsistency in findings.

Reports related to self-regulatory problems associated with PNCE have been more consistent. Schuetze and Eiden (2006) examined the association between prenatal cocaine exposure and physiological regulation at 4 to 8 weeks of age. A dose-dependent effect of cocaine exposure was found for resting heart rate (HR) and respiratory sinus arrhythmia (RSA). Resting HR was used to reflect autonomic function and neurological integrity, and RSA was seen as a physiological marker that provides an index of capacity to self-regulate. Findings indicated a dose-dependent effect of cocaine exposure on RSA, with heavy exposure levels associated with lower RSA. Tronick et al. (2005) reported that PNCE is associated with higher rates of autonomic stress indicators.

Several developmental studies of prenatal cocaine exposure have explored the possibility that exposed children may experience compromised reactivity, arousal modulation, and attention regulation because of exposure influences on monoaminergic neurotransmitter systems (e.g., Bennett, Bendersky, & Lewis, 2002). Examining neurodevelopmental outcomes at 1 month using the NNNS, the MLS found effects of cocaine exposure and level of cocaine exposure, particularly with regard to poorer regulation and higher excitability (Lester et al., 2002). The authors interpreted the results as demonstrating subtle effects reflecting a neurobehavioral vulnerability that has the potential to be exacerbated by the caregiving environment.

Mayes (2002) proposes that prenatal cocaine exposure may influence aspects of arousal regulation and prefrontal cortical function, which in turn, may relate to later attentional problems, anxiety, and conduct disorders. She discusses candidate mechanisms for prenatal cocaine effects including influences on monoaminergic neurotransmitter systems development, changes in neural growth factors, and effects on other neurotransmitter systems. Presenting evidence from animal and human studies relevant to connections between prenatal cocaine exposure and arousal/attention systems, Mayes supports the hypothesis that there is a relationship between cocaine exposure and impairments in arousal regulation and aspects of prefrontal cortical function. She also discusses multiple prenatal and postnatal influences that may occur in addition to prenatal drug exposure. Other researchers who have reported early infancy arousal modulation problems in studies of PNCE have delineated important interrelationships between exposure and such factors as infant term–preterm status, childrearing environment characteristics, and stressful situations (e.g., Bard, Coles, Platzman, & Lynch, 2000; Bendersky & Lewis, 1998; Molitor et al., 2003).

Studies of infant temperament provide relevant information on PNCE and self-regulation from a different perspective. Several studies have found associations between PNCE and difficult infant temperament involving problems of reactivity and affect regulation (Alessandri, Sullivan, Bendersky, & Lewis, 1995; Eiden, 2001; Lester, Lagasse, & Brunner, 1997). There is evidence that these difficulties persist beyond infancy. For example, Bandstra, Morrow, Anthony, Accornero, et al. (2001) reported that PNCE is associated with deficits in emotional regulation, inhibitory control, and sustained attention measured at age 7. Other studies have also reported finding associations of PNCE with inattention, behavioral regulation, and impulse control (Bendersky, Gambini, Lastella, Bennett, & Lewis, 2003; Heffelfinger, Craft, White, & Shyken, 2002; Leech et al., 1999; Savage et al., 2005).

Compromised infant self-regulation may be a contributor to poorer attachment. Several studies suggest that PNCE is associated with a subtle but potentially significant impact on the child's social and emotional development with differences in maternal–child interactions deriving from the interaction of the behavior and lower engagement of both mother and child (Beeghly, Frank, Rose-Jacobs, Cabral, & Tronick, 2003; Minnes et al., 2005; Tronick et al., 2005). Although the specific deficits may resolve over time (e.g., Seifer et al., 2004) it is not clear whether the trajectory of the child's socioaffective development has been altered.

PNCE does not seem to be associated with lower scores on measures of general intelligence or development (Frank et al., 2002, 2005; Messinger et al., 2004), but there is evidence of some subtle cognitive effects including specific cognitive impairments and a lower likelihood of above average IQ (Cone-Wesson, 2005; Schroder, Snyder, Sielski, & Mayes, 2004; Singer et al., 2004). Cocaine-exposed boys may be particularly affected demonstrating a gender effect of poorer short-term memory and lower verbal reasoning scores (Bennett et al., 2002). A number of studies assessing prenatally cocaine exposed children at different ages have reported subtle deficits in language development (Bandstra et al., 2002, 2004; Delaney-Black, Covington, Templin, Kershaw, et al., 2000; Morrow et al., 2004; Singer, Arendt, et al., 2001), but there is evidence that the impairments may be ameliorated by influences found in a strong home environment (Cone-Wesson, 2005; Lewis, Singer, et al., 2004; Singer et al., 2004).

There have not been longitudinal studies of sufficient duration to assess whether PNCE is associated with an increased likelihood of drug abuse. However, there has been research on childhood externalizing disorders and delinquency among prenatally cocaine-exposed children that are risk factors for substance abuse. Reported findings of PNCE-related behavior problems are less common when parents are the source of information about their children's problems than when teachers or other observers are the source. For example, Bennett et al. (2002) did not find an association between prenatal cocaine exposure and mother-reported externalizing problems in 4-year-old children, using the CBCL. However, in the same cohort, cocaine exposure, male gender, and living in a high-risk environment were associated with aggressive behavior at age 5 years, as reflected in a composite aggression variable that included teacher and parent ratings, child self-report of reactions to hypothetical interpersonal conflict situations, and a behavioral observation (Bendersky, Bennett, & Lewis, 2006).

Linares et al. (2006) found that 6-year-old children who were exposed to cocaine in utero reported more symptoms of oppositional defiant disorder and attention-deficit/hyperactivity disorder. Prenatal cocaine exposure was not related to caregiver-reported CBCL scores. However, CBCL scores did differ for cocaine-exposed children depending upon whether the ratings were done by adoptive or foster caregivers in comparison to maternal or relative caregivers, with the former caregiver group reporting more behavioral problems. Recently, findings from this cohort, based on child self-reported mental health symptoms at age 9 years, showed that prenatal cocaine exposure is associated with conduct disorder symptoms for girls only (Singer et al., 2005). In another cohort study, the Miami Prenatal Cocaine Study, Accornero et al. (2002) found that prenatal cocaine exposure was not related to maternal report CBCL scores for child behavior problems when the children were 5 years of age. However, based on trained-examiner behavioral observations for children from this same cohort at age 7 years, prenatal cocaine exposure was associated with less optimal behavioral regulation (e.g., lower attention, restlessness, impulsivity, temper outbursts) in a structured testing situation (Accornero et al., 2005).

Delaney-Black et al. (2004) reported that level of prenatal cocaine exposure and gender were associated with behavioral outcomes at 6 years of age, as reflected on a teacher-report scale involving 14 problem behavior areas. Boys, but not girls, with any prenatal cocaine-exposure scored higher on hyperactivity. Boys with persistent prenatal exposure (i.e., use by the mother up until delivery) showed more problems in central processing, motor skills, handling abstract concepts, and passivity to the environment. Prenatally exposed boys were twice as likely to have clinically significant scores for externalizing behavior and delinquent behavior (Delaney-Black, Covington, Templin, Ager, et al., 2000). Additional analyses at age 7 years indicated that PNCE was associated with more externalizing behavior for both boys and girls and that relationships between prenatal cocaine exposure and behaviors using a teacher report (Nordstrom Bailey et al., 2005) and caregiver report (Sood et al., 2005) were affected not only by gender but also by whether the child was also prenatally exposed to alcohol.

Lester et al. (2005) utilized structural equation modeling procedures to consider pathways to behavior problems at 7 years of age. Measurements included prenatal drug exposure, neurobehavioral disorganization at 1 month of age, difficult temperament at 4 months, and behavior problems at ages 3 and 7 years as reflected on the CBCL. Birth weight and social class were included among the other variables considered in the analyses. Results indicated a direct relationship between prenatal cocaine exposure and behavior problems at age 7 (consistent with a behavioral teratogenic model), and they also indicated indirect pathways between prenatal exposure and behavior problems (i.e., pathways from prenatal exposure to neurobehavioral disorganization at 1 month to difficult temperament at 4 months to behavior problems at 3 years and then at 7).

Prenatal opiate exposure

There is relatively little information about the long-term consequences of prenatal opiate exposure (PNOE). Reporting on MLS findings of assessments at 1 month using the NNNS, Lester et al. (2002) found some effects of PNOE on neurodevelopmental outcomes related to self-regulation and excitability. Seifer et al. (2004) reported that PNOE is associated with mildly poorer mother–child attachment at 18 months. Hans and Jeremy (2001) found that PNOE is associated with moderately poorer mental and motor development Bayley scores through 2 years. Based on their review of the literature, Mayes and Fahy (2001) concluded that PNOE is associated with greater impulsivity, poorer attention, and an increased incidence of behavior and conduct problems, although they acknowledge that the studies do not provide information to determine the extent to which the outcomes are the product of persistent drug exposure effects as opposed to familial and environmental factors.

PNDE outcomes

Children who are exposed to opiates, cocaine, or marijuana are often subject to prenatal consequences of other related harmful factors. For example, the use of illegal drugs while pregnant can lead to a range of maternal, prenatal, and perinatal health complications (Bishai & Koren, 1999; White, 2001) and is associated with an increased risk for complicating medical conditions such as HIV, hepatitis, sexually transmitted diseases, traumas related to violence, and other long-term health consequences (Bauer et al., 2005; Chiriboga, 2003; White, 2001). It is also important to note that drug abusers are typically multidrug users (Lester et al., 2004). Women who use illegal drugs while pregnant may use more than one type of illicit drug and they are also likely to use tobacco, alcohol, or both in addition (Bauer et al., 2005; Ebrahim, Decoufle, & Palakathodi, 2000; Lester, Boukydis, & Twomey, 2000). As previously discussed, both tobacco and alcohol have been shown to have serious harmful consequences of their own and pregnant women who use illicit drugs in addition compound the risks for the children (e.g., Eiden, 2001). The outcomes of multisubstance use may not necessarily be a simple sum of the consequences of the individual substance effects, there may be an interaction. Many drug abusers use more than one illegal drug and prenatal exposure to multiple illicit drugs is associated with even greater risks for negative outcomes (e.g., LaGasse et al., 2003).

Other variables may influence the outcomes of PNDE. Beyond the issue of which drugs the pregnant mother uses are the issues of the quantity and the points in the pregnancy when the drugs were used, and all of the drug use factors may interact with other health variables. Although many studies have found dose-related PNDE, some have not. PNDE may have different effects during different stages of fetal development, and this may account for some of the variability in dose–outcome relationships. The illegality of the abused drugs may also be a factor. The purity and composition of illegally purchased drugs can vary influencing outcomes in unpredictable ways. As mentioned above, the illegal lifestyle aspects of illicit drug abuse may affect the children of drug abusers. However, separate from the lifestyle factors, the biochemical effects of the drugs themselves seem to be significant contributors to the outcomes for exposed children. This is indicated by animal research, the human research as reviewed, and by studies of the outcomes of PNDE associated with the use of legal, medically supervised psychotropic drugs (e.g., Bercovici, 2005), including methadone and buprenorphine (Dashe et al., 2002; Jones et al., 2005).

Review of the PNDE research literature indicates that in addition to the increased risks associated with having parents who are substance abusers, there are persistent adverse consequences associated with gestational drug exposure that impact the child's development and long-term function. These chronic consequences of PNDE are generally small differences that reflect some neurobehavioral vulnerability. In general, it does not seem that PNDE by itself is associated with severe, pervasive deficits attributable to drug teratogenesis. When found, PNDE effects are usually specific rather than global and subtle rather than gross. For the most part, findings from prenatal exposure studies are not in terms of clinical or diagnostic categories, but more typically, results are in terms of behavioral measures. The severity of these problems varies as does their resolution over time, and in many cases a strong positive home environment can significantly ameliorate the negative exposure outcomes. Unfortunately, PNDE is often only one factor adding additional risk to a disadvantaged child's development, which is in jeopardy due to parental substance abuse and other detrimental factors that are likely to accumulate and interact (Lewis, Misra, et al., 2004). Even when prenatally exposed children do appear to recover from early functional impairments (e.g., Moe & Slinning, 2001) they may be at increased risk for later cognitive and behavioral problems when they are older because of exacerbating familial and environmental factors. Rather than fostering resilience and helping drug-exposed children to overcome the adverse consequences of PNDE, their families and environments are more likely to exacerbate the effects of the impairments (Lester et al., 2002; Olson et al., 2001).

Several researchers have suggested that some effects of PNDE may be latent or delayed, eventually manifesting because affected aspects of brain functioning only become evident at older ages. It may also be the case that deficits will become evident only when the PNDE child or adolescent is experiencing challenging or stressful conditions (Mayes, 2002) or in situations where there are more demands for performance and control (Bennett et al., 2002). It is also possible that PNDE effects may manifest more in certain contexts, for example, situations involving “hot cognition,” the process of thinking, and decision making under conditions of high arousal and/or strong emotion (e.g., Dahl, 2004).

PNDE findings may be important in attempting to understand trajectories from prenatal exposure to later developmental outcomes, but they may or may not have clinical significance at a particular assessment point. For example, researchers using data from the Maternal Health Practices and Child Development Project originally reported that prenatal tobacco exposure significantly predicted tobacco use at 10 years of age even after controlling for maternal tobacco use at 10 years, other prenatal substance exposure, maternal and child psychological characteristics, and peer use at age 10 (Cornelius, Leech, Goldschmidt, Lebow, & Day, 2000). However, when assessing the cohort at age 14, the researchers found that although prenatal tobacco exposure still significantly predicted tobacco use at 14 years of age, when proximal intervening variables such as current maternal and peer tobacco use were controlled for, prenatal exposure was no longer a significant predictor (Cornelius, Leech, Goldschmidt, & Day, 2005). There are a number of possible explanations for this seeming inconsistency but one possibility is that prenatal tobacco exposure predicted tobacco use at an age when it was more of a deviant behavior (age 10) but was less of a predictor at an age when the behavior was more common (age 14).

Prenatal exposure to marijuana is associated with negative outcomes including an increased risk for substance abuse, and prenatal cocaine exposure is associated with negative outcomes including some precursors of substance abuse. Insufficient research on gestational exposure to opiates is available to determine the long-term consequences, but some dysregulation precursors have been reported. Given that there is evidence that PNDE increases the exposed children's vulnerability to drug abuse beyond the elevated risk resulting from postnatal exposure to their parents' substance abuse, it is important to consider whether research on drug abuse etiology suggests mechanisms through which the increased vulnerability might occur.

Paths to drug abuse

Research indicates that there are several major comorbidities that are strongly associated with both alcohol and drug abuse disorders. The most common and the most clearly substantiated of these are antisociality, psychopathology, and cognitive/behavioral dysregulation. From an etiological perspective, they seem to be more than just risk factors. These commonly co-occurring conditions typically precede the development of substance use disorders and are considered by some to be causes of the subsequent substance abuse. Others have viewed these comorbid conditions as having a less determinative relationship. They see the comorbidities as being at least markers of risk for a later substance use disorder, probably contributors, and possibly earlier emerging consequences of underlying (or at least unmeasured) causes that also contribute to the later emerging substance abuse. Lacking more conclusive understanding of the nature of these comorbidities, it is heuristically useful at this point to view them as commonly occurring and not necessarily mutually exclusive paths to drug abuse.

Underlying mechanisms

A great deal of research has been devoted to finding the underlying “cause” of drug abuse, and the three paths described above have each been proposed by some as a cause of drug abuse disorders. Although it is generally acknowledged that vulnerability to drug abuse results from person–environment interactions, some feel that in terms of the individual factors, the fundamental cause of drug abuse must at least in part be underlying biological processes. Support for this comes from research that has demonstrated a genetic component to drug abuse etiology, from animal behavior studies of the reinforcing properties of drugs, and from neuroscience research that has identified neuropsychiatric mechanisms for drug effects and models for escalating and compulsive drug use. Research has identified a number of basic antecedent functional and/or biological factors that are hypothesized to be underlying mechanisms of drug abuse. They are inherent individual characteristics that are presumed to be more characteristically preexisting and predetermined than the paths described above and more intrinsic to the basic underlying function of the individual. The three most substantiated candidates will be described briefly.

Animal research

Animal studies have found that gestational exposure to tobacco, alcohol, and many drugs of abuse has significant harmful effects on CNS development with lasting functional consequences for the offspring (Ferriero & Dempsey, 1999; Mayes, 1999). Of particular relevance here, research on prenatally ATOD-exposed animals indicates that as adults they are more responsive to that same drug as well as to other psychoactive drugs. This suggests that the reward system and reward-driven behaviors may be altered by intrauterine exposure in ways that predispose animals to increased self-administration of drugs, an outcome that has been found with prenatal exposure to cocaine, opiates, and other drugs (Lidow, 2003; Malanga & Kosofsky, 2003). Although research has not determined the exact mechanisms that account for the effect, animal research provides evidence that prenatal exposure, independent of familial behavior and environmental factors, can result in increased behavior related to drug reinforcement. There may be a similar consequence of prenatal exposure in humans with both substance abuse and treatment implications (Stanwood & Levitt, 2004).

Animal studies indicate that intrauterine drug exposure contributes to increased reward-driven behaviors and increased self-administration of drugs. This neurobehavioral evidence goes beyond other findings that substance abuse while pregnant has neurodevelopmenatal consequences for the offspring, and it provides increased experimental evidence in support of the hypothesis that PNDE may increase vulnerability to drug abuse independent of genetic variability, parenting influences, and other environmental effects. Although this research is neither a full analog to drug abuse vulnerability in humans nor conclusive in its implications, it does provide noteworthy corroboration.