The primary goal of this study was to examine the level of sleep problems in a sample of cocaine-exposed 7-month-old infants and to determine if maternal psychopathology mediated any existing association. We found that prenatal exposure to heavier amounts of cocaine was significantly related to more severe sleep difficulties. These results indicate that the disorganized-state regulation noted in cocaine-exposed neonates (DiPietro et al., 1995
; Hume, O'Donnell, Stanger, Killam, & Gingras, 1989
) is still present by 7 months of age. Since increased sleep problems are disruptive to family life (Richman, Douglas, Hung, Lansdown, & Levere, 1985
) and are associated with both maternal fatigue and child abuse (Chavin & Tinson, 1980
), these findings suggest that the heavily cocaine-exposed mother–infant dyads are at an increased risk for long-term emotional and behavioral problems. Future studies should examine sleep problems in cocaine-exposed children into the preschool years to see if the sleep difficulties persist and if they are indeed associated with nonoptimal developmental outcomes.
We also found that levels of maternal anxiety mediated the association between cocaine exposure and sleep difficulties. Specifically, exposed infants who had mothers with higher levels of anxiety at 7 months of infant age showed higher levels of sleep problems. Given the correlational nature of these findings, it is not clear if this is a causal pathway. One explanation for these findings is that although dysregulated behavioral states during the neonatal period seem to reflect disrupted autonomic nervous system development, the postnatal environment of the cocaine-exposed infant has a significant impact on regulatory problems during later infancy. Specifically, mothers who use heavier amounts of cocaine have higher levels of anxiety which, in turn, impact the sleep behaviors of their infants beyond the first month of life. Thus, interventions which include services to reduce maternal anxiety may have a positive impact on infant regulation. Another explanation for these findings is that infants who were prenatally exposed to heavier amounts of cocaine display disrupted regulatory processes beginning at or immediately after birth, which then increases maternal anxiety. In the latter scenario, the increased sleep difficulties seen in heavily exposed 7-month-old infants demonstrate continuity in disrupted behavioral-state organization from the neonatal period. This explanation is supported by our findings that sleep behaviors at 7 months of age were associated with increased physiological arousal and dysregulation during periods of rest at both 1 and 7 months.
Importantly, there was no association between maternal cocaine use during pregnancy and maternal cognitions about their infants' sleep behaviors. Although there was a slight association between cocaine use during pregnancy and scores on the Doubt subscale of the MCISQ, this difference disappeared after controlling for maternal alcohol consumption. Thus, mothers who used cocaine during pregnancy do not appear to think differently about the sleep behaviors of their infants than do mothers who did not use cocaine during pregnancy. Thus, although parental cognitive factors have been associated with the development of various behavior problems in children (see Morrell, 1999b
), these findings suggest that maternal cognitions about sleep are unrelated to the sleep difficulties found among heavily exposed infants.
Within the cocaine group, there were significant differences in perinatal risk between infants who remained in parental care at 7 months of age and those who did not. Specifically, the infants in nonparental care were exposed to heavier amounts of cocaine, had decreased fetal growth, fewer prenatal visits, and less optimal scores on the OCS; however, despite their higher perinatal risk, infants who were in nonparental care at 7 months had less severe sleep problems than did infants who remained in parental care. Since the nonmaternal caregivers had significantly fewer symptoms of psychopathology than the cocaine-using women who retained custody of their children, one possible explanation for this finding is that the quality of the caregiving environment has a significant impact on infant sleep. Another possible explanation for these findings is that the mothers who retained custody of their children were more likely to perceive sleep problems among their infants than were other caregivers. In fact, there has been considerable discussion regarding the possible influence of maternal psychiatric symptoms such as depression on maternal reports of children's behavior (e.g., Briggs-Gowan, Carter, & Schwab-Stone, 1996; Chilcoat & Breslau, 1997
; Fergusson, Lynskey, & Horwood, 1993
). Sophisticated analytic models have demonstrated support for both the hypothesis that children of mothers with psychiatric problems have higher behavior problems and the hypothesis that mothers with a history of psychiatric disorder overreport negative behavior in their children (Chilcoat & Breslau, 1997
). Thus, it is possible that higher sleep problems noted among the cocaine-exposed infants in the care of their biological mothers were due to the higher rates of maternal psychiatric symptoms in this group, method bias, or both; however, our findings of an association between sleep problems at 7 months of age and physiological regulation during sleep at 4 to 8 weeks of age suggest that at least in early infancy, exposed infants do have autonomic nervous system functioning that differs from comparison infants. It is therefore likely that exposed infants exhibit differential physiological regulation early in life, which in turn contributes to maternal perceptions of more difficult behaviors later in infancy. Future studies are needed to examine whether the differential physiological regulation seen at 4 to 8 weeks of age predicts continued physiological dysregulation later in infancy and whether physiological regulation at 7 months of age is associated with maternal reports of infant sleep.
It is not surprising, then, that maternal psychopathology was so strongly related to maternal cognitions about their infants' sleep. Higher levels of caregiver psychopathology were associated with higher levels of maternal anger related to the demands of an infant, doubt about the adequacy of their parenting, beliefs that feeding is important for soothing an infant at night, and concerns about SIDS. Since maternal cognitions about infant sleep are associated with infant sleep problems (Morrell, 1999b
), treatment for psychiatric symptoms may ameliorate the development of sleep problems.
There are other limitations of this study. First, the number of infants exposed to heavier amounts of cocaine as well as the number of cocaine-exposed infants in nonparental care were relatively small, increasing the chance that a few extreme data points may have influenced the findings. Second, it is unclear if the finding of sleep problems among cocaine-exposed infants, particularly those who remain in parental care, at 7 months of age is a transient problem that may lessen as a result of normal developmental processes. Longitudinal studies have shown that sleeping problems tend to decline with age (Kataria, Swanson, & Trevarthan, 1987
). Thus, future studies should use a longitudinal approach to examining sleep behaviors in drug-exposed infants to determine if the increased sleep problems among cocaine-exposed infants at 7 months of age is a transient issue or whether they predict a more persistent problem. This is particularly important given the finding by some studies that persistent sleep problems are associated with psychological disturbances in early childhood (Herzog, 1980
; Lozoff et al., 1985
; Mahler et al., 1975
), including both internalizing and externalizing problems (Atkinson, Vetere, & Grayson, 1995
; Seifer et al., 1996
). Third, our measures of sleep behaviors are based on maternal report, which is susceptible to bias, particularly among women with increased psychopathological symptomatology; however, our objective findings of dysregulated physiological behavior during sleep and rest among infants reported to have more sleep problems suggest that these findings reflect the actual presence of sleep disturbances among cocaine-exposed infants rather than just altered maternal perceptions. Thus, future studies should include objective measures of sleep when exploring the development of these behaviors among substance-exposed infants. Finally, although care was taken in the present study to identify substance use in this sample, the accurate assessment of substance use is difficult. Pregnant women are often hesitant to divulge information regarding the use of substances during pregnancy, particularly illicit substances such as cocaine. To address this issue, multiple indices of cocaine use were used, including self-report using the reliable Timeline Followback Interview as well as analysis of hair and urine samples. Each of these measures has its own limitations; however, when used in combination, the likelihood of accurately identifying cocaine use is increased. Note that measures of the other substance use during pregnancy were based entirely on self-report.
Despite these limitations, the present findings are important because they provide additional support for the influence of prenatal exposure to cocaine on infant regulation and indicate that one pathway from cocaine exposure to altered regulation in older infants may be through the caregiving environment. In particular, these findings highlight the importance of targeting psychiatric symptoms in addition to substance use in clinical interventions. In addition, this study adds to an increasing literature suggesting that cocaine-exposed infants who remain in parental care experience developmental outcomes that differ from those of infants in nonparental care. Finally, similar to the findings of other studies (Jacobson et al., 1996
; Lester et al., 2003
), our findings suggest that altered developmental outcomes become apparent at heavier levels of exposure and highlight the importance of considering a dose-dependent response when examining the development of cocaine-exposed children.