تعصبات شناختی و مصرف الکل در نوجوانی و بزرگسالی جوان: نقش تعدیل جنس، کنترل توجه و کنترل مهاری

Abstract The present study investigated the cross-sectional associations between cognitive biases (i.e., attentional bias and approach bias) and alcohol use and investigated the moderating role of gender, attentional control and inhibitory control. The sample consisted of 94 adolescents and young adults (52.1% boys) between 15.3 and 20.8 years old (Mage = 18.0, SD = 1.1) who reported drinking alcohol in the past three months. A stronger approach bias was related to higher alcohol use, albeit only among boys. Furthermore, the association between attentional bias and alcohol use was moderated by attentional control; the lowest alcohol use was found in adolescents with low attentional bias and high attentional control, suggesting protective effects of both variables. The present study replicates and extends the results of studies on cognitive biases and addiction in adolescence and young adulthood.

1. Introduction In adolescence and young adulthood, it is normative to experiment with alcohol (Shedler & Block, 1990). However, some adolescents develop alcohol use problems; posing risks for their future development (Burrow-Sanchez, 2006). To improve prevention and intervention, it is important to gain insight into the processes that contribute to alcohol use problems. According to the incentive sensitization theory (Robinson & Berridge, 1993), substance-related cues (e.g., beer glasses, a pub, etc.) in the environment acquire the ability to grab the user’s attention, because they become associated with the effects of the substance through the process of classical conditioning (Schoenmakers, Wiers, & Field, 2008). Consistent with this theory, several studies have indicated an attentional bias for substance-related stimuli in substance users with the strength of the bias proportional to the frequency and quantity of use (e.g., Field & Cox, 2008). Cognitive biases for disorder-relevant cues are, however, not limited to selective attention. In the domain of problematic substance use there is also evidence for the existence of biases in the evaluation of substance-related cues (e.g., Field, Eastwood, Bradley, & Mogg, 2006). Stimulus valence can be assessed either explicitly (e.g., by means of valence ratings) or implicitly (e.g., by means of a reaction time task, in which valence is derived from latencies to approach and avoid stimuli). There is growing evidence that individuals with substance use problems show greater preference for substance-related cues both on explicit and implicit indices of stimulus valence (e.g., Bradley, Field, Mogg, & De Houwer, 2004), which is consistent with incentive models of addiction (e.g., Robinson & Berridge, 2001), according to which drug addicted patients perceive drug-related cues as attractive. Most research on these attentional and approach biases has been conducted in adult samples of addicted patients or heavy drinkers. Only a few studies have examined cognitive biases in preadult samples (e.g., Peeters et al., in press and Van Hemel-Ruiter et al., 2011). Moreover, with the exception of the study of van Hemel-Ruiter et al. (2011) all other studies have been conducted in samples of adolescents recruited in schools with high prevalence of alcohol and drug use (e.g., Special Education schools). These samples are presumed to be at-risk samples for alcohol use problems (e.g., Peeters et al., in press). However, it is important to investigate whether automatic cognitive processes are also related to alcohol use in adolescents and young adults who are not assumed to be at-risk. Indeed, if cognitive biases are associated with drinking behavior, then youth with high cognitive biases can be identified as at risk youth and thus become candidates for prevention programs. The first aim of the present study was to examine whether attentional and/or approach bias are related to alcohol use in a community sample of adolescents and young adults. Numerous studies have demonstrated gender differences in drinking patterns. In adolescence, boys report higher levels of alcohol use compared to girls (Schulte, Ramo, & Brown, 2009). However, to our knowledge, none of the studies on the associations of attentional and/or approach bias with alcohol use in adolescence and young adults have investigated gender differences in these associations. Given the sex differences in drinking behavior during adolescence and the lack of studies investigating gender differences in the relation between attentional and/or approach bias and alcohol use in this age group, the second aim of the present study was to test whether the associations between cognitive biases and alcohol use are moderated by gender. Recently, dual-process models have been formulated (e.g., Stacy & Wiers, 2010) which view addictive behaviors as the joint outcome of automatic appetitive processes (such as cognitive biases) and controlled regulatory processes. From this view, problematic alcohol use may result from a fundamental imbalance between automatic appetitive processes and weakened ability and motivation to regulate these appetitive impulses (Stacy & Wiers, 2010). Previous studies both in young adults (e.g., Houben & Wiers, 2009) and adolescents (e.g., Thush et al., 2008) have shown that automatic cognitive processes are related to higher alcohol use especially when individuals have low regulatory capacities. Across studies regulatory capacities have been operationalized in different ways, with studies in adolescents mainly assessing executive control using working memory or response inhibition tasks (e.g., Thush et al., 2008). However, it would also be useful to investigate whether trait aspects of self-regulation, assessed by a questionnaire, influence cognitive biases. Indeed, self-reports of self-regulation are more easily collected, so, to the extent that they yield effects, they may offer a more feasible or efficient means of identifying ‘at risk’ youths. In addition, it would be useful to investigate whether different aspects of self-regulation are related to different cognitive processes. In this respect, the temperament trait effortful control as conceptualized by Rothbart (1989) is noteworthy. Effortful control is a self-regulatory capacity that emerges in childhood, allowing the child to gain active control over behavior and emotional responses. It includes inhibitory control, activation control and attentional control. In accordance with dual-process models of problematic substance use, the third aim of the present study was to investigate whether the associations of attentional bias and approach bias with alcohol use are moderated by effortful control. Given that attentional control reflects the ability to voluntarily focus or shift attention when needed, it seems reasonable to assume that the predictive value of attentional bias for alcohol use would especially be pronounced in adolescents with weak attentional control. In the same line, given that inhibitory control reflects the ability to inhibit behavior if necessary, it is reasonable to suppose that the predictive validity of approach bias would especially be pronounced in adolescents with low levels of inhibitory control.

. Results 3.1. Descriptives and correlations Table 1 presents means, standard deviations and gender differences for all measures. Gender differences emerged on the AUDIT, indicating that boys reported higher alcohol use compared to girls. In addition, girls had higher RTs on the various conditions of both cognitive tasks, indicating slower responding. Table 1. Internal consistencies, means and standard deviations on questionnaires and on cognitive tasks. α Total group M (SD) Boys M (SD) Girls M (SD) F AUDIT .77 3.82 (3.17) 4.45 (2.74) 3.13 (3.49) 4.17⁎ Attentional control .57 3.64 (1.04) 3.75 (1.10) 3.53 (0.97) 1.02 Inhibitory control .56 3.91 (0.93) 3.99 (0.88) 3.82 (0.99) 0.79 Dot probe congruent 445.66 (53.51) 432.59 (57.20) 459.88 (44.91) 6.54⁎ Dot probe incongruent 441.89 (51.19) 427.61 (51.21) 457.45 (46.93) 8.62⁎⁎ Attentional bias −3.77 (19.96) −4.98 (20.61) −2.44 (19.37) 0.38 SRC congruent 791.19 (128.74) 749.20 (129.28) 836.92 (112.68) 12.20⁎⁎⁎ SRC incongruent 800.17 (132.93) 760.51 (113.49) 843.35 (140.14) 9.99⁎⁎ Approach bias 8.97 (83.07) 11.31 (76.77) 6.43 (90.23) 0.08 ⁎ p ⩽ .05. ⁎⁎ p ⩽ .01. ⁎⁎⁎ p ⩽ .001. Table options Table 2 shows correlations between questionnaires and attentional bias and approach bias scores. Attentional control was negatively correlated with AUDIT (r = −.23, p = .03), indicating that lower levels of attentional control were associated with higher levels of alcohol use. In addition, there was a positive correlation between approach bias and AUDIT (r = .24, p = .02), indicating that adolescents with a stronger approach bias towards alcohol related stimuli reported higher levels of alcohol use. Table 2. Correlations between questionnaires, attentional bias and approach bias. 1 2 3 4 5 AUDIT – −.23⁎ −.20 .09 .24⁎ Attentional control – .47⁎⁎ −.10 −.15 Inhibitory control – −.07 −.08 Attentional bias – .10 Approach bias – ⁎ p ⩽ .05. ⁎⁎ p ⩽ .001. Table options 3.2. Prediction of AUDIT scores by attentional bias and approach bias and the moderating role of gender Table 3 shows the results of the hierarchical regression analyses for the prediction of alcohol use by attentional bias and approach bias, and their interaction with gender. Gender was entered in Step 1, attentional bias and approach bias scores in Step 2, and the bias × gender interactions in Step 3. Results indicated that AUDIT scores were predicted by male gender and a stronger approach bias. In addition, the association between approach bias and AUDIT scores was moderated by gender; stronger approach bias predicted higher levels of alcohol use only in boys (Fig. 1). The simple slope for approach bias for girls was β = 0.02, t(87) = .44, p = .66, whereas for boys it was β = 0.15, t(87) = 3.57, p < .001. Table 3. Hierarchical linear regression analysis predicting alcohol use by cognitive biases and their interaction with gender. R2 R2 change β Step 1 .09⁎⁎ .09⁎⁎ Step 2 .15⁎⁎ .06⁎ Step 3 .23⁎⁎⁎ .08⁎ Gender −.30⁎⁎ Attentional bias .09 Approach bias .27⁎⁎ Attentional bias × gender −.17 Approach bias × gender −.21⁎ ⁎ p ⩽ .05. ⁎⁎ p ⩽ .01. ⁎⁎⁎ p ⩽ .001. Table options Approach bias×Gender interaction predicting alcohol use. Fig. 1. Approach bias × Gender interaction predicting alcohol use. Figure options 3.3. Moderating role of attentional control and inhibitory control We investigated whether the associations between attentional bias and alcohol use and between approach bias and alcohol use were moderated by attentional control and inhibitory control, respectively. A hierarchical linear regression analysis was performed, with gender in Step 1, attentional bias, approach bias, attentional control and inhibitory control in Step 2, the hypothesized interactions in Step 31 (i.e., attentional bias × attentional control and approach bias × inhibitory control) and the non-predicted interactions in Step 4 to investigate whether the effects of attentional control and inhibitory control are specific to attentional bias and approach bias respectively (i.e., attentional bias × inhibitory control and approach bias × attentional control) (Table 4). Given that the ΔR2 of the final step was not significant, we only report the result of Step 3. Results showed that AUDIT scores were predicted by male gender and strong approach bias, as before. Furthermore a significant attentional bias × attentional control interaction emerged, indicating that at low levels of attentional control there was no difference between adolescents with high or low levels of attentional bias in the prediction of alcohol use. In contrast, alcohol use was lower only for adolescents with low attentional bias and high attentional control ( Fig. 2). The simple slope for attentional bias at low attentional control was β = −0.03, t(85) = 0.61, p = .54, whereas for high attentional control it was β = 0.10, t(85) = 2.51, p = .01. Table 4. Hierarchical linear regression analysis predicting alcohol use by cognitive biases and their interaction with attentional control and inhibitory control. R2 R2 change β Step 1 .09⁎⁎ .09⁎⁎ Step 2 .21⁎⁎⁎ .12⁎ Step 3 .28⁎⁎⁎ .07⁎ Step 4 .29⁎⁎⁎ .01 Gender −.33⁎⁎⁎ Attentional bias .12 Approach bias .23⁎ Attentional control −.19 Inhibitory control −.10 Attentional bias × attentional control .22⁎ Approach bias × inhibitory control .16 ⁎ p ⩽ .05. ⁎⁎ p ⩽ .01. ⁎⁎⁎ p ⩽ .001. Table options Attentional bias×attentional control interaction predicting alcohol use. Fig. 2. Attentional bias × attentional control interaction predicting alcohol use.