Background: Patients with obsessive compulsive disorder (OCD) often show deficits in inhibitory control, which
may underlie poor control over obsessions and compulsions. Several functional magnetic resonance imaging
(fMRI) experiments utilizing a variety of tasks have investigated the neural correlates of inhibitory control in
OCD. Evidence from existing meta-analyses suggests aberrant activation of regions in fronto-striatal circuits
during inhibitory control. However, new fMRI articles have since been published, and a more rigorous meth-
odology for neuroimaging meta-analyses is now available.
Objectives: First, to reevaluate the evidence for abnormal brain activation during performance of inhibitory
control tasks in OCD while adhering to current best practices for meta-analyses, and second, to extend previous
findings by separately assessing different subprocesses of inhibitory control.
Method: We systematically searched Web of Knowledge, ScienceDirect, Scopus, PubMed and the functional
BrainMap database for fMRI articles that compared activation during performance of inhibitory control tasks in
patients with OCD and healthy control (HC) subjects. Thirty-five experiments from 21 articles met our criteria for
inclusion. We first performed activation-likelihood-estimation meta-analyses to elucidate brain areas in which
case-control activation differences converged across articles and tasks. We then aimed to extend previous work
by separately evaluating experiments requiring inhibition of a prepotent response without execution of an
alternative response (i.e., response inhibition) and experiments requiring inhibition of a prepotent response and
execution of an alternative response (i.e., cognitive inhibition).
Results: The 35 experiments included a total of 394 patients and 410 controls. We did not find evidence of
abnormal brain activation in OCD during inhibitory control when pooling data from all experiments. Analysis
restricted to cognitive inhibition experiments showed abnormal activation of the dorsal anterior cingulate cortex
(dACC; P = .04, cluster-level familywise error-corrected, cluster volume of 824 mm3). We did not have sufficient
data to evaluate response inhibition experiments separately.
Conclusion: Findings of abnormal brain activation in OCD from different inhibitory control tasks do not appear to
converge on the same brain regions, but the dACC may be implicated in abnormal cognitive inhibition. Our
findings highlight a need for experiments that specifically target subprocesses of inhibitory control to achieve a
more differentiated understanding of the neural correlates for impaired inhibitory control in OCD.