Event-Related Potentials as Predictive Indexes of the Anxiety-Related Attention Bias in Adolescents

• In Event: 3-035 - Rising From the Ashes? Paradigm Shifts in the Study of Attention Biases in Adolescent Anxiety

Integrative Statement

Anxiety disorders are the most common mental illness among adolescents (Merkangas, 2010), a period of critical neural development. A key cognitive mechanism posited to underlie anxiety is the anxiety-related attention bias (AB), characterized by exaggerated attention towards threat. AB can be quantified via the dot-probe task, which measures behavioral responses to probes following threatening versus neutral stimuli presented simultaneously. While frequently considered the gold standard measure of AB, the dot-probe task yields a limited index of AB in terms of average reaction times that fail to capture more nuanced and rapid attention processes, and show mixed links with anxiety symptoms (Roy, Dennis, & Masia-Warner, 2015). This study used a biobehavioral approach to quantify AB among anxious adolescents, with the goal of identifying metrics of AB showing predictive power beyond the traditional static AB measure. To this end, we focused on two target event-related potentials (ERPs): the anterior N2 and the posterior N2, which reflect allocation of attention control resources to visual stimuli and targets, respectively (e.g. Folstein & Petten, 2008). In adults, previous studies have provided evidence for the anterior N2 (e.g. Dennis-Tiwary, et al., 2016) and a component of the posterior N2 (i.e. N2pc; Kappenman, et al., 2014), as sensitive measures of differential attention control to threat versus neutral during the dot-probe task. However, less is known about these ERPs in relation to AB in anxious adolescents. Further, few studies have simultaneously examined attention processing of faces and probes, thus capturing the full time-course of processing and responding throughout the task. We hypothesized that greater N2 amplitudes to threatening faces and subsequent probes, indicating exaggerated attention to threat, will uniquely predict greater parent-reported anxiety symptoms, beyond the behavioral AB metric. Participants included 18 (11 female) 12-to-14-year-olds (M = 14.02, SD = 0.69) with clinically-significant anxiety. EEG was recorded while adolescents completed a dot-probe task, during which angry and neutral faces served as cues which were simultaneously presented side-by-side. On each trial, participants were instructed to respond to the direction of a probe replacing either the angry or neutral face. Behavioral AB scores were computed by subtracting reaction times to threat from neutral probes, with positive scores indicating a bias toward threat. The anterior N2 was generated from 300-360 milliseconds at FCz following face onset, and the posterior N2 was generated from 230-320 milliseconds at PO7/PO8 following probe onset. Linear regressions were conducted as follows: 1st step: behavioral AB scores; 2nd step: anterior or posterior N2; DV: parent-reported anxiety symptoms. Results showed that greater anterior N2 to threat versus neutral faces predicted greater total anxiety [β = -51; t(17) = -2.23, p = .041; Figure 1], and greater posterior N2 to threat probes predicted greater social anxiety [β = -.70; t(17) = -2.78, p = .015; Figure 2]. In contrast, behavioral AB scores did not significantly predict adolescent anxiety (p’s > .05). These findings highlight the importance of using biologically-based approaches for characterizing anxiety in youth, potentially speeding progress toward developmentally-sensitive and treatment-relevant biological signatures.

Authors

• Samantha Denefrio, The Graduate Center, City University of New York
  Presenting Author

• Tracy Dennis-Tiwary, Hunter College, The City University of New York
  Non-Presenting Author