Working Memory, Impulsivity and Sexual Risk-Taking in Adolescence

Introduction: Weakness in working memory (WM) and associated lack of self-control has been associated with a wide range of adolescent risk-taking behaviors (Moffitt et al., 2011), including early sexual debut (Khurana et al., 2012). This weakness undermines the ability to exert inhibitory control and consider long-term consequences of behavior –characteristics that underlie two forms of impulsivity, Acting without Thinking (AWT) and Inability to Delay Gratification (IDG). Sensation Seeking (SS), an impulsivity dimension not characterized by WM weakness, is not expected to be as strongly related to adverse forms of risk-taking. To test this hypothesis, we examined the longitudinal effect of WM and three impulsivity dimensions (AWT, IDG, SS) on involvement in risky sexual behaviors (RSB) associated with increased risk for STI’s/HIV and unplanned pregnancies. 

Method: A cohort of 327 urban adolescents (53% female; baseline age=14.4+/-0.87 yrs) was interviewed at baseline and one year later. WM was assessed using –Corsi Block-Tapping, Letter two-back, Digit Span Backwards, and a spatial WM task. SS was assessed using the Brief sensation seeking scale, AWT using the Junior Eysenck Impulsivity Scale, and IDG using a hypothetical monetary choice procedure (Green et al., 1994). Involvement in RSB was assessed at both time points using a cumulative index ranging from 0(virgin), 1(sexually experienced, used a condom during last vaginal intercourse) to 2(sexually experienced, last intercourse was unprotected).

Results: Weakness in WM (at baseline) predicted increase in RSB over the one year follow-up, the effect being entirely mediated by AWT and IDG, B(SE)=.04(.02), p<.01. Controlling for the effect of AWT and IDG, SS did not uniquely predict RSB. Adolescents’ perceived parental monitoring also had a direct protective effect on RSB, B(SE)=-.13(.05), p<.01, controlling for age, sex, race-ethnicity and socio-economic status [Model Fit: Chi Sq.=65.5, p<.01; RMSEA=.05 (90%CI=.03-.06); CFI=.93]. Re-analysis using the sexually-experienced sub-sample (n=91) to specifically predict change in non-condom use behaviors revealed the same pattern of results.

Conclusion: Weakness in WM can serve as an early indicator of early sexual debut as well as of unprotected sexual activity. Impulsivity dimensions characterized by weak executive control (i.e. AWT and IDG) are more strongly related to RSB and other adolescent risk behaviors (Magid et al., 2007) than SS, a finding that has important implications for preventive interventions. Promising evidence from WM training programs (Diamond & Lee, 2011) suggests that similar approaches may be useful in improving adolescents’ ability to exert greater self-control and make thoughtful decisions in sexual and other health behavior domains.