Category: Adult Anxiety
Individuals with social anxiety disorder (SAD) are known to have significant anxiety surrounding feared social events (Coles & Heimberg, 2011; Morrison et al., 2016; Price & Anderson, 2011). Prior research has shown that, compared to healthy controls, those with SAD show higher levels of anxiety in anticipation of and immediately after a speech (Morrison et al., 2016). However, predictors of anxiety trajectory during a prolonged period after the occurrence of a stressful event and whether anxiety returns to baseline have yet to be explored.
In this study, we aimed to examine the effect of trait negative affect (NA) on differences in anticipatory anxiety using latent trajectory models (LTMs) of state anxiety (SA) prior to a speech and attenuating anxiety after a speech. Participants were 121 undergraduate students randomized into either the induced anticipatory anxiety (IAA) condition or a no induced anxiety (NIA) group, who then completed computerized tasks and a speech task. The IAA group was verbally reminded that they would complete a speech task several times during the computer task prior to the speech. The NIA group did not receive any reminders about the speech. We hypothesized that NA would predict a higher intercept (baseline anxiety), steeper positive slope 1 (S1; change in anxiety from baseline to speech), and less steep negative slope 2 (S2; change in anxiety post-speech) in the IAA condition compared to NIA.
Trait NA was measured with the Positive and Negative Affect Scale (PANAS; Watson et al., 1998). SA was measured at multiple timepoints using the Brief State Anxiety Measure (BSAM; Berg et al., 1998). Due to two data collection efforts being combined, 85 of the participants only received 4 BSAM measures and 36 of the participants received 7 BSAM measures.
Model fit was assessed in Mplus and missing data was handled with the MLR estimator. We examined a piecewise LTM that demonstrated adequate fit with Group x Negative Affect interaction as a predictor. There was a main effect of Negative Affect on the intercept (p < .01), indicating higher levels of baseline anxiety were predicted by greater NA. Contrary to our hypotheses, NA did not predict S1 (p>.05) and no significant variance was found for S2, implying little variation in post-speech rates of anxiety attenuation among subjects.
Our results suggest clinical implications regarding interventions for managing anxiety in context of feared events. If post-event rate of decline in anxiety is not variable, perhaps it is more important to instead target the accumulation of anxiety in anticipation of the event.