Mediating mechanisms of the relation between anxiety and cognitive control in Spanish-speaking young adults

Abstract

<strong>Background and aims:</strong> For several decades, the scientific literature has focused on the direct effect of anxiety on cognitive control, paying less attention to the intermediate mechanisms underlying this relation. Although there is an ever-increasing trend of research findings that show how anxiety negatively influences different aspects of cognitive control, such as reversal learning and task-switching cost, there is still much to learn about the cognitive processes that have a mediating role in this relation. Thus, we investigated whether cognitive avoidance, cognitive flexibility, dispositional mindfulness, and decentering exert a mediating role in the relation between trait anxiety and cognitive control. We translated into Spanish and validated five self-report scales originally written in English that measure trait anxiety, cognitive flexibility, cognitive avoidance, dispositional mindfulness, and decentering. We also designed three cognitive computerized tasks that measure cognitive interference towards emotional stimuli, task-switching cost, and reversal learning using E-Prime 3.0. <strong>Method:</strong> We recruited 149 Spanish-speaking young adults (ages 18-28, <em>M</em>=21.3, <em>SD</em>=2.49; 110 females) and administered an online survey consisting of the State-Trait Anxiety Inventory - Trait (STAI-T, Cronbach&rsquo;s <em>&alpha;</em>=0.92), the Cognitive Avoidance Questionnaire (CAQ, Cronbach&rsquo;s <em>&alpha;</em>=0.90), the Cognitive Flexibility Inventory (CFI, Cronbach&rsquo;s <em>&alpha;</em>=0.82), the Mindful Attention Awareness Scale (MAAS, Cronbach&rsquo;s <em>&alpha;</em>=0.85), and the Experiences Questionnaire (EQ, Cronbach&rsquo;s <em>&alpha;</em>=0.68). We also administered the three computerized tasks remotely with E-Prime Go 1.0. We obtained electronic consent from each participant prior to administering the survey and cognitive computerized tasks. <strong>Results:</strong> The global fit indices of the measurement model with four correlated errors were generally favorable, <em>CFI</em>=.955, <em>RMSEA</em>=.069, 90% <em>CI</em> [.048, .088], <em>SRMR</em>=.058. Regarding the structural component, all unstandardized and standardized direct effects of trait anxiety towards mediating variables (e.g., cognitive avoidance, cognitive flexibility, decentering, and dispositional mindfulness) were significant at .05 level. However, the unstandardized and standardized direct effects of trait anxiety and the mediator variables towards the outcome variables (e.g., cognitive interference towards emotional stimuli, task-switching cost, and reversal learning) were not statistically significant. Furthermore, a negative nonsignificant association was found between task-switching cost and reversal learning, <em>r</em>=-.059, <em>p</em>=.420. <strong>Conclusion:</strong> The patterns of results observed suggest that there are some processes associated with emotional regulation that are interrelated, but that did not demonstrate a significant relation with purely cognitive aspects. This implies that Spanish-speaking young adults did not have a generalized cognitive problem because the cognitive computerized tasks did not show any significant relation with the personality characteristic variables. In contrast, the observed linkages among variables show that the participants&rsquo; problems tend to be affective in nature. We suggest that this investigation could be replicated using alternative versions of the task-switching alternating and reversal learning tasks with negative emotional stimuli.