Performance monitoring in nicotine dependence: Considering integration of recent reinforcement history.

INTRODUCTION: Impaired monitoring of errors and conflict (performance monitoring; PM) is well documented in substance dependence (SD) including nicotine dependence and may contribute to continued drug use. Contemporary models of PM and complementary behavioural evidence suggest that PM works by integrating recent reinforcement history rather than evaluating individual behaviours. Despite this, studies of PM in SD have typically used indices derived from reaction to task error or conflict on individual trials. Consequently impaired integration of reinforcement history during action selection tasks requiring behavioural control in SD populations has been underexplored.

METHODS: A reinforcement learning task assessed the ability of abstinent, satiated, former and never smokers (N=60) to integrate recent reinforcement history alongside a more typical behavioural index of PM reflecting the degree of reaction time slowing following an error (post-punishment slowing; PPS).

RESULTS: On both indices there was a consistent pattern in PM data: Former smokers had the greatest and satiated smokers the poorest PM. Specifically satiated smokers had poorer reinforcement integration than former (p=0.005) and never smokers (p=0.041) and had less post-punishment slowing than former (p<0.001), never (p=0.003) and abstinent smokers (p=0.026).

CONCLUSIONS: These are the first data examining the effects of smoking status on PM that use an integration of reinforcement history metric. The concordance of the reinforcement integration and PPS data suggest that this could be a promising method to interrogate PM in future studies. PM is influenced by smoking status. As PM is associated with adapting behaviour, poor PM in satiated smokers may contribute towards continued smoking despite negative consequences. Former smokers show elevated PM suggesting this may be a good relapse prevention target for individuals struggling to remain abstinent however prospective and intervention studies are needed. A better understanding of PM deficits in terms of reinforcement integration failure may stimulate development of novel treatment approaches.