Cognitive deficits are a matter of emotional context: Inflexible strategy use mediates context-specific learning impairments in OCD

Abstract

The present study examines the interplay between cognitive deficits and emotional context in obsessive-compulsive disorder (OCD) and social phobia (SP). Specifically, this study examines whether the inflexible use of efficient learning strategies in an emotional context underlies impairments in probabilistic classification learning (PCL) in OCD, and whether PCL impairments are specific to OCD. Twenty-three participants with OCD, 30 participants with SP and 30 healthy controls completed a neutral and an OCD-specific PCL task. OCD participants failed to adopt efficient learning strategies and showed fewer beneficial strategy switches than controls only in an OCD-specific context, but not in a neutral context. Additionally, OCD participants did not show any explicit memory impairments. Number of beneficial strategy switches in the OCD-specific task mediated the difference in PCL performance between OCD and control participants. Individuals with SP were impaired in both PCL tasks. In contrast to neuropsychological models postulating general cognitive impairments in OCD, the present findings suggest that it is the interaction between cognition and emotion that is impaired in OCD. Specifically, activated disorder-specific fears may impair the flexible adoption of efficient learning strategies and compromise otherwise unimpaired PCL. Impairments in PCL are not specific to OCD.

Keywords:

• Probabilistic classification learning
• Emotional context
• Learning strategies
• Obsessive-compulsive disorder
• Social phobia

The authors express their appreciation to Matijn Meeter for his valuable assistance with the strategy analyses. This study was supported by a grant from the German Research Council (Deutsche Forschungsgemeinschaft) to C. Exner and W. Rief.

The authors express their appreciation to Matijn Meeter for his valuable assistance with the strategy analyses. This study was supported by a grant from the German Research Council (Deutsche Forschungsgemeinschaft) to C. Exner and W. Rief.

Notes

¹ The original sample comprised of 30 participants diagnosed with OCD (see also Exner et al., 2014). However, seven of these participants met criteria for comorbid social phobia and where thus excluded from the present analyses.

² All participants in partial remission had met full criteria for OCD within the past year. At the time of the study, they reported experiencing obsessions and/or compulsions for less than one hour per day, thus no longer meeting full criteria.

³ See footnote 1 above.

⁴ To test whether other possible anxiety dimensions explained the results, we included SPS scores as additional mediator into the model. This did not change results and SPS scores were no significant mediator (indirect effect = 0.22, SE = 2.18, 99% bias corrected and accelerated CI −7.5 to 6.8).