Neuropsychological Bases for Individual Differences in Impulsivity: Examining the Role of Working Memory and Executive Attention

Introduction: Impulsivity and lack of self-control are well-recognized precursors of unhealthy risk-taking during adolescence (Moffitt et al., 2011), including use of substances and risky sexual involvement. Despite the rising interest in working memory (WM) training as a prevention target for reducing adolescent impulsivity (Diamond & Lee, 2011), the associations between WM and impulsivity reported in studies of adolescents tend to be small in magnitude, and the mechanisms linking WM to different dimensions of impulsivity remain unclear. This study is the first attempt at unpacking the association between WM and impulsivity dimensions of Acting-Without-Thinking (AWT) and Delay Discounting (DD) by examining effects of WM and its specific components such as Executive Attention (EA) on reductions in impulsivity across adolescence. Further, in evaluating the utility of WM as a potential intervention target, we examine its effects on impulsivity controlling for more pervasive risk factors such as socioeconomic status (SES) as well as maturational changes in WM and impulsivity.

 

Methods: Using structural equation modeling, we analyzed three annual waves of data from 387 community adolescents (Baseline age = 10-12 years; 52% female; low-mid SES), and examined the role of WM and executive attention (EA) – a key component of WM – as predictors of change in AWT and DD during the three year follow-up period. AWT was assessed using a self-report questionnaire (Eysenck et al., 1984), while DD was assessed using a hypothetical monetary choice procedure (Green et al., 1994). EA was assessed using the Stop Signal Task and WM was assessed using four separate complex WM tasks, including Corsi Block Tapping, Object-two-back, Digit Span Backwards and a spatial WM task. SES, assessed using the Hollingshead Two-Factor Index, as well as age were included as controls.

Results: Analyses revealed that reduction in AWT (impulsive action) was more closely related to EA, while reduction in DD (impulsive choice) was better predicted by the broader construct of WM that measured other, more demanding, processing abilities besides EA. The longitudinal effects of WM and EA remained significant even after accounting for the pervasive influences of age and socioeconomic status that were linked to the development of both WM and impulsivity.

Conclusions: The results point to the potential of cognitive training interventions targeting WM and EA as a way to enhance developmental improvements in self-control during adolescence. Future research should focus on evaluating the effects of WM interventions on real-world risk behaviors in adolescents.