We investigated brain activation patterns in school-aged children with prenatal cocaine exposure and non-exposed controls during a response inhibition task. The hemodynamic response to stimuli during response inhibition differed between exposed and non-exposed children. Cocaine-exposed children showed greater activation in frontal and striatal regions, areas that prior research has shown to be associated with controlled attentional processes in typically developing children [18
]. In contrast, non-exposed children showed greater activations in regions of the occipital cortex and the fusiform gyrus. The regions that were found to differentiate the cocaine-exposed from non-exposed children were generally similar to regions identified in prior research to be related to response inhibition in children and adolescents [17
]. However, research has reported developmental changes across adolescence and into adulthood that reflect increased activation of frontal and striatal regions associated with response inhibition [28
]. Developmental changes may reflect differential or alternative recruitment of brain regions to support cognitive functions with maturation [28
]. While the current study was not longitudinal, these data do raise the hypothesis that prenatal cocaine exposure may affect long-term brain maturational processes underlying response inhibition.
These results are preliminary evidence that prenatal cocaine exposure is associated with functional differences in neural systems underlying response inhibition. Exposed children appeared to recruit fronto-striatal networks during response inhibition to a greater extent than non-exposed children. Thus, one interpretation of these results is that prenatal exposure may affect brain regions that are involved in the cognitive control and regulation of attention. In this way, these results are consistent with the theory that prenatal cocaine exposure affects developing monoaminergic systems and executive functions [1
]. This does not necessarily mean that our findings are reflective of functional deficits in fronto-striatal networks. As noted earlier, recent research has found that prenatal cocaine exposure may be related to elevated frontal creatine [6
], elevated diffusion in frontal regions measured by diffusion tensor imaging [7
], increased frontal cerebral blood flow [9
], and increased left prefrontal activity measured by fMRI during a working-memory task [5
]. Thus, one hypothesis going forward is that functional differences seen in this population may reflect the engagement of compensatory mechanisms during executive tasks. Other alternative explanations of these data are also possible. For example, the differences in the hemodynamic response in right frontal and striatal regions in the exposed group, as compared to occipital and temporal activations in the non-exposed group, could indicate more automatic processing of visual information in the non-exposed group, and thus may reflect a differential recruitment of brain regions rather than regional dysfunction, per se. Alternatively, groups may have differed in the degree of salience attributed to the task stimuli or to the degree to which they monitored their performance for errors.
These differences were not due to performance differences between the groups. Neuroimaging studies may be able to identify subtle but meaningful differences in cognitive processing even in the absence of performance differences on cognitive tasks. In fact, differences in brain function in the absence of performance deficits on a particular task or behavioral probe may have special value to the study of developmental psychopathology and the development of at-risk children [28
These findings are also consistent with known long-term effects of cocaine in adults. Long-term effects of cocaine use on executive functions and measures of brain functioning in frontal regions subserving these functions, even after periods of abstinence, have been documented [12
]. Long-term cocaine use in adults is associated with reduced prefrontal cortical volume [13
], decreased brain maturation in the frontal and temporal lobes [32
], reduced white matter integrity in the frontal lobes [33
], and decreased gray matter concentrations in orbitofrontal and anterior cingulate regions [34
]. Although there is the possibility that our groups differed in patterns of functional activation in the anterior cingulate, these results were found in exploratory post hoc analyses, and thus cannot be considered definitive. Nonetheless, these results do raise the hypothesis that prenatal cocaine exposure may in some ways affect similar brain regions affected by chronic cocaine use in adults. Alternatively, these results raise the possibility that prenatal cocaine exposure may be associated with a differential recruitment of brain regions underlying cognitive inhibition.
Our results are also consistent with animal models, which predict some specificity of prenatal cocaine exposure effects on brain regions high in dopamine receptor concentrations [14
]. However, the effects of prenatal cocaine are not limited to dopamine. For example, repeated cocaine administration results in reduced GABA-mediated inhibition of medial prefrontal neurons in adult rats [36
]. Recent findings from a mouse model have demonstrated delayed tangential migration of GABA-ergic interneurons to cortical areas, including the medial prefrontal cortex, following prenatal cocaine exposure [37
]. Thus, the effects of prenatal cocaine exposure on neurotransmitters are not limited to dopamine, and such effects (e.g. on the functional maturation of GABA circuitry) may contribute to some of our fMRI findings.
Past research has indicated that there may be reduced recruitment of brain regions associated with executive functions (e.g. working memory) as task performance becomes more automatic [38
]. While the present study did not directly address this issue, it is possible that the groups differed in the degree to which they recruited more controlled attentional mechanisms to complete the nogo task. Thus, one hypothesis may be that prenatal cocaine impacts the integrity of brain regions that support the development of executive functions and that such effects may be reflected in subtle though functionally significant differences in cognitive, social, or adaptive functioning in this population.
There are limitations to this study. The sample size is small and most children were exposed to drugs other than cocaine. The design of the MLS allowed for exposure to alcohol, marijuana, and tobacco in the unexposed groups because it is well documented that most women who use cocaine also use these other substances. Cocaine-exposed children in the other imaging studies were also exposed to other substances [5
]. The fact that we did find group differences with other substance exposure in both groups does point to the potential role of prenatal cocaine on the development of executive function-related brain systems. However, we could not fully explore the relative effects of cocaine and prenatal exposure to other substances. Also, the presence of other drug exposures in the comparison group allowed for a preliminary test of brain differences that may be more specifically attributable to cocaine exposure. However, the lack of a typical control group does not allow us to rule out the possibility that the children with other drug exposures were atypical in brain response as well.
Another potential limitation is that although we controlled for socioeconomic status, we cannot rule out other postnatal environmental factors that could have influenced our findings. We also cannot rule out the degree to which other child characteristics may have affected our findings. While the children in this study were representative of the overall MLS sample, it is possible that selection biases (e.g. toward children who could complete the scan session; away from those children who declined to participate) may limit the generalizability of findings to the population of children with prenatal cocaine exposure. Related to this issue is whether the presence or absence of attention deficit hyperactivity disorder (ADHD) affected the findings. There were 5 children in this sample who later received diagnoses of ADHD at age 11 years, 3 in the cocaine exposed group and 2 in the non-cocaine-exposed group (diagnosis determined by the Diagnostic Interview Schedule for Children) [40
]. This sample composition did not allow for a full exploration of the effects of ADHD or other diagnostic classifications on the fMRI results, and this is recognized as a limitation to this study. However, the regional differences in brain activation between groups remain when these children with ADHD were excluded from the analyses on a post hoc basis. Finally, we did not exclude children based on history of prematurity for this preliminary study. Recent research has reported differences in brain structure and function in premature children, particularly in the temporal lobe [41
]. While the behavioral task employed in this study is different than those used in recent research with premature infants, it is acknowledged that this type of heterogeneity in our sample is a limitation. Thus, future work in this area should be designed to address these issues, either with statistical controls (requiring larger sample sizes) or more focused sampling and selection criteria.
The limitations identified above underscore the fact that these results need to be viewed as preliminary in nature. Nonetheless, if replicated, our findings would suggest that prenatal cocaine exposure is associated with long-term effects on brain development and could help explain the cognitive deficits that have been reported in this population, for example, in the area of executive function. It would also be important for future research to determine whether these deficits in brain function are related to other developmental outcomes such as school performance or psychopathology, including early onset of substance use in these children.