Nicotine deprivation and trait impulsivity affect smokers' performance on cognitive tasks of inhibition and attention

Abstract

Increased nicotine deprivation and impulsivity have been associated with relapse but the degree to which they together influence cognitive processing has not been explored. We examined the effects of increasing levels of nicotine deprivation on cognitive processing, and assessed the relationship of trait impulsivity with these effects in daily smokers (n=30). Using a within-subject design with three deprivation conditions (nondeprived, 5-hr, 17-hr), volunteers completed the Conners' Continuous Performance Task-II and the Cued Go/No-Go Task. Trait impulsivity was assessed at intake with the Barratt Impulsiveness Scale (Patton et al., 1995). Mixed-model regression analyses revealed deprivation slowed reaction time, increased errors, increased variability in responding, and increased failures of inhibitory control. Performance at 17 hours of deprivation was most likely to be affected. Significant deprivation and impulsivity interactions indicated impulsiveness was negatively correlated with deprivation-associated performance decrements. Less impulsive smokers were more affected by deprivation, demonstrating greater impairment. Research is needed to understand mechanisms by which impulsivity confers greater risk for relapse. Our results suggest deprivation may not increase relapse risk among impulsive smokers by increasing impairment of cognitive processing.

Figure 1

Mean task performance (+SE) outcome measures of the Cued Co/No-go Task (CGNG) across nicotine deprivation conditions. The left panel presents mean reaction times to the go target following the go cue and the right panel presents mean reaction times to the go target following the no-go cue. Asterisks denote ps < .05, from the mixed model findings (listed in Table 1), for within-subject comparisons.

Figure 2

Mean task performance (+SE) outcome measures of the Conners’ Continuous Performance Task (CPT-II) across nicotine deprivation conditions. Panel A presents mean omissions, panel B presents mean reaction time, and panel C presents mean reaction time standard error. Asterisks denote ps < .05, from the mixed model findings (listed in Table 1), for within-subject comparisons.