This study is the first to use DTI to investigate frontal white matter development in children with PCE. We found that children with PCE had significantly higher values of Dav (all directions) in 2 frontal white matter regions, the left frontal callosal and right frontal projection fibers, compared with controls. There was also a trend for higher Dav values in the right frontal callosal fibers in the group with PCE. Because Dav values in children are known to decrease with age and development, these differences could suggest less integrity and/or slower maturation of these areas in cocaine-exposed children.
No significant between-group differences were found for FA (more directionality) in the 4 frontal white matter ROIs examined. That group differences were found for Dav
but not FA could be because of a number of factors, including the age of our sample, the areas studied, and the fact that, theoretically, Dav
and FA are independent measurements40
and may reflect different physiologic processes.33
Our sample consisted of children between 10 and 12 years old, and volumetric studies indicate that myelination of frontal white matter tracts continues well into adolescence and beyond.41,59–61
In addition, FA is sensitive to the direction of greatest diffusion and is, therefore, limited in characterizing tracts with crossing fibers or fibers that are incoherently oriented. For example, relatively low FA values are typically found in the centrum semiovale and U-shaped fibers, although they are highly myelinated.62
The white matter tracts that we chose may contain fibers that lack sufficient directional coherence for FA to provide a sensitive measure of their relative development. This is particularly likely for the frontal projection fibers, which include afferent and efferent fibers between cortical and subcortical areas, including the dorsolateral prefrontal cortex, caudate, dorsomedial nucleus of the thalamus, and reticular formation. Also, as shown in , the frontal projection fiber tract that we chose for analyses is somewhat curved. Finally, it may be that Dav
is more sensitive than FA to the changes associated with PCE. A study of patients with multiple sclerosis found that Dav
was more sensitive than FA in detecting disease-related white matter changes.63
Multiple regression revealed that the variability of Dav
in the left frontal callosal fibers in our sample was because of prenatal exposure to alcohol and marijuana in addition to an interaction between PCE and marijuana. The negative effects of alcohol, marijuana, and tobacco on brain development and executive functioning are well known.64
The interaction between PCE and marijuana as seen in suggests that prenatal exposure to both cocaine and marijuana is worse than exposure to cocaine alone as measured by Dav
. Although it seems from the figure that children with prenatal marijuana exposure but not cocaine exposure have better (lower) Dav
, there were only 2 children in this group, casting doubt on the reliability of this finding.
This study also examined performance on 2 executive functioning measures in children with PCE. We found that the cocaine-exposed group was significantly slower on average than nonexposed children in completing a timed task that involves shifting between sequencing numbers and sequencing letters (TMT part B). Cocaine-exposed children also performed more poorly, although not significantly, on a timed task that requires inhibition of reading a color word in favor of naming the competing color ink in which the word is printed (Stroop color-word task). Poorer executive functioning in the cocaine-exposed children is consistent with other reports in the literature of visual attention and motor inhibition difficulties in this population.2–7
A multiple regression analysis showed that performance on the TMT part B was significantly predicted by PCE and an interaction with PCE and prenatal tobacco exposure. Paradoxically, however, children with PCE and higher levels of prenatal tobacco exposure were faster rather than slower at completing the task than children with PCE and lower levels of tobacco exposure.
Finally, this study demonstrates a brain-behavior relationship between frontal white matter anisotropy and executive functioning performance. In the sample as a whole, better performance on both executive functioning measures was associated with greater anisotropic diffusion (FA) in the left frontal callosal fibers. Faster performance on the set-shifting task (TMT part B) was also associated with greater anisotropic diffusion (FA) in the right frontal projection fibers.
The association between Stroop color-word scores and FA in left frontal callosal fibers may be related to the verbal nature of the test and the fact that children may require bilateral hemispheric coordination to complete the task. Adleman et al53
showed that developmentally specific Stroop-related activation of the left prefrontal cortex begins in adolescence and increases through early adulthood. The immature left hemisphere specialization in children could also explain why no association was found between Stroop performance and FA in the left frontal projection fibers that emanate from the prefrontal cortex. The association between the TMT part B performance and left frontal callosal FA may also be related to the language demands of the task (interpreting letters and numbers), whereas the association between TMT part B performance and right frontal projection FA may be because of the visual-spatial nature of the task (drawing lines to sequence symbols scattered randomly on a page). Functional MRI studies that evaluate activation patterns associated with the various task components of the TMT could further elucidate these brain-behavior relationships.
The significant correlations between one of the executive functioning measures and FA but not Dav
is not unexpected. The few studies that have correlated cognitive performance with DTI measures have only used anisotropy measures.42
Use of anisotropy measures is logical as a marker of myelination and axonal thickness, particularly in adults for whom brain maturation is considered complete. Notably, the single DTI study of cognitive performance in children ages 7 to 13 years found significant correlations with FA in a temporoparietal area but not in a frontal area. We chose to examine possible correlations with Dav
in part because there are so few data on the association between DTI and cognition in children and none in children with PCE. In addition, we thought that Dav
may provide as useful an index of frontal lobe white matter development as FA in our sample of 10.5-year-olds, whose frontal lobes are expected to be less myelinated relative posterior cortical areas.
The study findings are presented as the first pieces of a puzzle designed to elucidate the effects of PCE on the developing brain. Many questions remain unanswered. For example, it is curious to us that the between-group differences were found for Dav in the left frontal callosal and right projection fibers but the significant correlations between one of the executive functioning measures was with FA for these same 2 structures. Also, the state of the science does not allow us to speculate as to laterality of the group differences in Dav (ie, higher Dav in callosal fibers on the left and the projection fibers on the right in the cocaine-exposed group). In addition, the paradoxical finding that PCE in combination with higher levels of prenatal tobacco exposure was associated with better executive functioning warrants further investigation and replication.
In future studies, we would like to evaluate the relative contribution of postnatal environmental factors on brain development and cognitive functioning in cocaine-exposed children. Environmental factors have been found to influence the development of executive functioning in typically developing children65
and the development of dopaminergic innervation of the prefrontal cortex in laboratory animals.66,67
The deleterious effects of trauma, maltreatment, and abuse on brain development have been shown in terms of total cerebral volume and changes in evoked related potentials and electroencephalogram studies.68
For children with PCE, a number of studies have indicated that postnatal environmental factors, such as maternal psychological functioning, the caregiving environment, and early intervention, may have effects that are equal to, if not more important than, prenatal drug exposure on child functioning. 9,10,69–71
The possible effects on brain development of stressors associated with poverty and out-of-home placement early in life for the children with PCE in our cohort72
merit close consideration as well.