Category: Addictive Behaviors
Sexual risk taking is associated with many negative outcomes such as HIV infection, unplanned pregnancy, and sexual assault. Research indicates that alcohol use is a significant predictor of risky sexual behavior (George et al., 2014; Purdie et al., 2011), and reducing alcohol use might be important in reducing risky sex. Sexual risk taking may occur when deliberative cognitive control processes are weaker than automatic approach tendencies. Approach biases toward alcohol cues are associated with increased drinking behavior and may also impact an individual’s decision-making process regarding condom use. Joystick cognitive bias modification (CBM) interventions using a modified version of the approach avoidance task (AAT) are emerging as effective treatment tools by modifying implicit processes for many disorders and problematic behaviors. Namely, if an individual is trained to attend to or attend away from specific information, symptoms may be reduced and behavior may be changed. Previous experimental research has been successful in modifying various cognitive biases away from alcohol stimuli and subsequently reducing drinking behavior (Wiers et al., 2010; 2011).
Participants in the current study were college students between the ages of 18-25, who drink moderately, and have engaged in sexual intercourse without a condom with a casual, non-committed partner during the three months prior to data collection. Participants were randomly assigned to either an active training group or a sham training control group. Participants in the active training group were explicitly trained to make avoidance movements (pushing a joystick) away from alcohol stimuli and approach movements (pulling a joystick) toward pictures of condoms.
The current study tested associations between the implicit approach-avoidance biases and drinking and risky sexual behavior at baseline. Avoidance bias towards condoms was significantly associated with greater engagement of risky sexual behavior at baseline (r = .33, p = .024). However, alcohol use was not significantly associated with baseline approach bias. In addition, we examined the effects of the intervention on alcohol and condom approach biases. To test whether the training affected implicit approach-avoidance tendencies for alcohol and condoms, two 2 (Training: experimental/sham) x 3 (Time: pretest/posttest/3-month follow-up) mixed ANOVA analyses were run. Results indicated significant treatment condition effects for both alcohol (F = 4.37, p < .001) and condoms (F = 3.92, p = .049). Currently, we have enrolled 58 participants in the study and anticipate reaching an N of 100 by August 2017. Upon completion of data collection, we will examine the effects of the intervention on alcohol and condom approach biases, the effects of the intervention on alcohol and condom use at the 3- month follow-up, if cognitive biases post-intervention mediate changes in corresponding behavior at 3-month follow-up, and if changes in alcohol use mediate changes in risky sex using an autoregressive path model with the maximum likelihood robust estimator using full-information maximum likelihood estimation. Mediation will be tested by estimating bias-corrected bootstrapped confidence intervals.