The current study was designed to assess specific aspects of neurocognitive functioning of children in the PANDAS subgroup. PANDAS subjects differed from matched controls on only one of the measures — the response accuracy in a test of attention and suppression. These results are consistent with previous studies of OCD, which have shown few or no deficits in attention (e.g., Boone, Ananth, Philpott, Kaur, & Djenderedjian, 1991
; e.g., Flament et al., 1990
; Hollander et al., 1993
). Our results suggest that there may be an effect of age of symptom onset on attentional functioning, which has been studied directly by at least one study by Roth, Milovan, Baribeau, and O'Connor (2005)
, who studied children with onsets either above or below age 12. However, the results of this study are not directly applicable to our study group because the age of onset for most PANDAS subjects was below 12 years, and none were older than 15 at onset (see ). The results of the present investigation are also in keeping with previous reports for the WCST, where OCD subjects consistently perform as well as matched controls (for review, see Kuelz, Hohagen, & Voderholzer, 2004
). Last, studies of adults with OCD on the TOH have produced mixed results, with some (Schmidtke, Schorb, Winkelmann, & Hohagen, 1998
) showing no deficits, and others (Cavedini, Cisima, Riboldi, D'Annucci, & Bellodi, 2001
; Mataix-Cols et al., 1999
) showing deficits compared to controls. Deficits may be a result of dysfunction within the orbitofrontal-striatal-thalamocortical circuitry, as has been demonstrated in functional imaging studies of OCD (Saxena et al., 2001
), or it may represent a separate deficit in speed and special working memory (Kuelz, Hohagen, & Voderholzer, 2004
Our results are reflective of the neuropsychological literature of TS in the executive functioning and attentional functioning domain. In a review of the literature, Como (2001)
found the only reliable deficit in executive functioning (including both tests of executive functioning and attention) was increased time on the CPT. When the PANDAS group was divided by primary psychiatric diagnosis, those with tics and OCD or TS performed significantly worse than their controls, while those with OCD (without tics) did not show a significant difference in performance from their matched controls. However, caution should be taken when interpreting these results because of the small sample size of each diagnostic group. Nevertheless, these results may be explained in part by Baron-Cohen et al. (Baron-Cohen, Cross, Crowson, & Robertson, 1994
), who have theorized that children with TS exhibit a dysfunction of the “Intention Editor,” explaining their inhibitory deficits, manifested in involuntary movements and utterances and obsessive thoughts. Further, results of intracortical transcranial magnetic stimulation in children with TS have shown reduced intracortical inhibition and a shortened cortical silent period, providing a neurological basis for decreased inhibitory control (Moll et al., 2001
). However, Ozonoff, Strayer, McMahon, and Filloux (1998)
have shown that ADHD comorbidity may account for attentional dysfunction in TS. Studies of executive functioning in children with TS have shown no significant deficits (Ozonoff & Jensen, 1999
; Verté, Geurts, Roeyers, Oosterlaan, & Sergeant, 2005
; Yeates & Bornstein, 1996
Within the response execution task, ADHD and age were significant predictors of RT; subjects with PANDAS and ADHD and younger subjects exhibited a significantly longer RT than their matched HV. Both of these results are supported by previous findings using this task by Casey, Durston, and Fossella (2001)
. In their study of 108 typically developing children, age and RT were negatively correlated. In addition, in a group of 26 nonmedicated children with ADHD between the ages of 6 and 16 years, the authors found decreased performance on the response execution task as compared to matched controls. With regards to the response selection task, the differences between low- and high-IQ subjects may be explained, in part, by the small sample sizes. Previous studies of cognitive functioning in OCD and TS have not sustained effects of IQ on attention or executive functioning (Como, 2001
; Greisberg & McKay, 2003
Results of our analyses on the effects of medication agree with previous studies that have found little to no neuropsychological differences between medicated and unmedicated pediatric subjects with OCD and/or TS. For TS, one study (Bornstein & Yang, 1991
) found no differences in performance on a wide battery of neuropsychological tests (including the WCST) between medicated and unmedicated children, even though most of the medicated children were taking a neuroleptic (clonidine, pimozide, or haloperidol) at the time of testing. A similar study (Channon, Pratt, & Robertson, 2003
) found that medicated children scored lower on only one test of multitasking. Likewise, a study (Mataix-Cols, Alonso, Pifarré, Menchón, & Vallejo, 2002
) that compared SRI-medicated SRI-free children diagnosed with OCD found no significant differences between groups in a continuous performance task and the WCST; the groups only differed significantly in the number of reversions in the TOH. The literature and results of our current investigation generally argue against a deleterious effect of medication use on neuropsychological performance; however, to our knowledge, there have been no long-term outcome studies evaluating this question.
Strengths of our study included a well-defined set of subjects in the PANDAS subgroup, which were carefully matched to control subjects. Although these tests may not be specific enough to assess important facets of cognitive functioning in subjects with PANDAS, they have proven useful in assessing deficits in attention in subjects with ADHD (Casey, Durston, & Fossella, 2001
). Our results from the executive functioning task reveal effects of comorbid ADHD on cognitive functioning in subjects with OCD and/or TS, which warrant further study.
Limitations and Future Directions
This study is limited by the small sample sizes for some measures, and the secondary analyses are further limited by the division of the cohort into smaller samples. The limited sample size can be explained by two factors: (a) some patients may not have been available to complete all tests because of time limitations imposed by the schedule of their primary treatment protocol, and (b) the battery of administered tests differed across the primary protocols into which the subjects were enrolled; the tests presented here represent those for which we had the largest sample sizes. In addition, data for some patients were no longer available at the time of analysis, so that sample sizes for individual tasks vary. Of particular note, the sample size for the response execution task was considerably smaller (n = 22) than those of the other tasks. The power of this analysis and several others, including the regression analysis, to detect small differences is thereby reduced, requiring additional caution in interpreting the results.
Lack of functional neuroimaging data also limited the specificity of our results. Otto (1992)
points out that it is difficult to draw conclusions about specific dysfunction within the frontostriatal circuitry from cognitive test results without imaging data because caudate dysfunction can mimic frontal lobe dysfunction, other compensatory mechanisms in the brain may mask basal ganglia or prefrontal dysfunction, and cognitive dysfunctions may be secondary to the disorder itself.
Future investigations would be strengthened by the addition of neuroimaging data, such as event-related brain potential (ERP). Studies of ERP in subjects with OCD have consistently shown hyperarousal of the cortex during attentional tasks (de Groot, Torello, Boutros, & Allen, 1997
; Gohle et al., 2008
; Herrmann, Jacob, Unterecker, & Fallgatter, 2003
; Towey et al., 1990
). In subjects with TS, ERP has begun to help unravel the neurological substrates that separate subjects with and without comorbid ADHD (Drake et al., 1992
; Oades, Dittmann-Balcar, Schepker, Eggers, & Zerbin, 1996
; Zhu et al., 2006
). Other methods of functional imaging, such as functional Magnetic Resonance Imaging (fMRI), may also be considered.