We appreciate the comments on our article by Drs. Savage and Rauch proposing that our null results might be explained by the fact that cognitive abilities dependent on prefrontal function do not develop until later in childhood and, thus, were not apparent at the time of our evaluation of children with OCD. They suggest an emerging deficit in frontal lobe function that mirrors the clinical phenomenology of the illness, noting that the obsessions may not fully develop until frontal lobe processes reach full maturity, sometime after age 12. This is a worthy hypothesis. In a study by Levin and colleagues (1991), healthy older children (ages 13–15 years) showed the expected developmental gains on a measure of verbal fluency and on the Tower of London task (a.k.a. Tower of Hanoi). However, on the Wisconsin Card Sorting Test and the Go–No-Go task, both of which were also included in our study, children performed at the adult level by age 12. Noting that our group was strikingly similar in age to that of the group of Levin et al. (12.3 years [SD=2.9] versus 11.1 years [SD=1.1], respectively), we do not feel that developmental level itself can explain our null results, especially with respect to the Wisconsin Card Sorting Test.

OCD symptoms, even early in the illness, might influence problem-solving efficiency rather than problem-solving accuracy measured within a time limit. A controlled study of adults with OCD +(1) demonstrated no between-group differences in number of responses, rate of perseverative responses, and rate of perseverative errors on the Wisconsin Card Sorting Test. However, OCD patients required significantly more time to respond to each trial and to complete the entire test. Careful consideration suggests that within our group of frontal lobe tests, we can make distinctions between those that were timed but brief and highly structured (e.g., Stroop Color/Word Test, Trail Making Test B, Controlled Oral Word Association test) and those that had no obvious time constraints but required deductive reasoning (e.g., Tower of Hanoi, Wisconsin Card Sorting Test). The latter tests emphasize rule acquisition and the problem-solving process, whereas the former require clearly defined responses (e.g., ink color, alternation between numbers and letters) while inhibiting other clearly defined but inappropriate responses. One explanation for our null results may be that we failed to measure problem-solving efficiency during the completion of unstructured problems that require, among other things, the discovery of rules.

Neuroimaging studies have begun to elaborate the structural brain changes associated with OCD in young children. As Drs. Savage and Rauch suggest, it remains for carefully designed longitudinal studies to determine the neuropsychological correlates and functional significance of these brain changes early in the illness and over the course of subsequent development.