Shin, N.Y.^a, Lee, T.Y.^b, Kim, E.^b, Kwon, J.S.^a,b,c,*

^aInterdisciplinary Cognitive Science Program, Seoul National University, Seoul, South Korea
^bDepartment of Psychiatry, Seoul National University, College of Medicine, 28 Yeongon-dong, Chongno-gu, Seoul, 110-744, South Korea
^cDepartment of Brain and Cognitive Sciences, Seoul National University, College of Medicine, Seoul, South Korea

^*Corresponding author

Background
Substantial empirical evidence has indicated impairment in the cognitive functioning of patients with obsessive-compulsive disorder (OCD) despite inconsistencies. Although several confounding factors have been investigated to explain the conflicting results, the findings remain mixed. This study aimed to investigate cognitive dysfunction in patients with OCD using a meta-analytic approach.

Method
The PubMed database was searched between 1980 and October 2012, and reference lists of review papers were examined. A total of 221 studies were identified, of which 88 studies met inclusion criteria. Neuropsychological performance and demographic and clinical variables were extracted from each study.

Results
Patients with OCD were significantly impaired in tasks that measured visuospatial memory, executive function, verbal memory and verbal fluency, whereas auditory attention was preserved in these individuals. The largest effect size was found in the ability to recall complex visual stimuli. Overall effect estimates were in the small to medium ranges for executive function, verbal memory and verbal fluency. The effects of potentially confounding factors including educational level, symptom severity, medication status and co-morbid disorders were not significant.

Conclusions
Patients with OCD appear to have wide-ranging cognitive deficits, although their impairment is not so large in general. The different test forms and methods of testing may have influenced the performance of patients with OCD, indicating the need to select carefully the test forms and methods of testing used in future research. The effects of various confounding variables on cognitive functioning need to be investigated further and to be controlled before a definite conclusion can be made.