Category: Adult Anxiety
Processes related to approach and avoidance have mostly been conceptualized as distinct. However, threat and reward are often juxtaposed in real-life, arguing for an integrated view of approach-avoidance (Stein & Paulus, 2009). We showed that conditioned anxiety influences reward learning on a threat-relevant reward-task, with individuals in the threat-relevant condition showing increased reward learning (Walker & Zoellner, 2017). We also found that anhedonia, which has been linked to deficits in reward functioning (Pizzagalli et al., 2014), moderated the effect of conditioned anxiety on reward learning, such that higher anhedonia predicted higher learning but only in the threat-relevant condition. The purpose of the current study was to examine the specificity of this moderating effect by investigating whether broader constructs of anxiety or depression could also account for this relationship. Lower reward responsiveness, higher behavior inhibition, higher avoidance, and lower estradiol level have each also been linked to increased avoidance or decreased approach and, thus, were examined as potential moderators as well (Jovanovic, & Norrholm, 2015; Carver & White, 1994).
Ninety-nine undergraduate females underwent fear conditioning, using the screaming lady paradigm (Lau et al., 2009) and were randomized to threat-relevant or threat-irrelevant conditions of an adapted probabilistic reward task involving differential reinforcement (Pizzagalli et al., 2005). Key dependent variable was response bias, operationalized as systematic preference for the response paired with the more frequent reward. Moderation effects of anhedonia (SHAPS; Snaith et al., 1995), trait and state anxiety (STAI; Spielberger et al., 1983), depression (QIDS-SR; Rush et al., 2003), reward responsiveness (BIS/BAS; Carver & White, 1994), behavior inhibition (BIS/BAS; Carver & White, 1994), avoidance (MIA; Chambless et al., 1985), and menstrual cycle phase were examined.
Anhedonia moderated the effect of conditioned fear on response bias, such that higher anhedonia predicted higher response bias in the threat-relevant group (β = .26, t(95) = 2.52, p = .01) but not in the threat-irrelevant group. Depression, trait and state anxiety, reward responsiveness, behavior inhibition, avoidance, and early follicular phase did not significantly moderate the effect of conditioned fear on reward learning. Analyses using salivary estradiol prior to the task will also be examined for potential moderating effects.
University of Washington