82 Chapter 4 Mean HR was calculated for the baseline measurement (2 minutes recorded 15 minutes before each session) and for the 2-minute samples during each exposure, resulting in a maximum of four samples per speech per participant (0-2 minutes, 2-4 minutes, 4-6 minutes, 6-8 minutes). Because not every participant presented a speech lasting more than 6 minutes, we had only 3 HR samples for some participants. Statistical analyses HR To examine whether enhanced exposure affected psychophysiological reactivity we posthoc analyzed heart rate responses with the same mixed models as fear levels in session 1 and session 2), including baseline HR as a fixed factor. Details results Adverse events (AEs) Compared to the placebo (P) group, the testosterone (T) group reported fewer AEs, including mild headache (T: n = 1, P: n = 4), mild nausea (T: n = 0, P: n = 2), and mild stomachache (T: n = 0, P: n = 1). Around 1.5 weeks after the post-treatment assessment, one participant in the T-group, who was on hormonal birth control, reported a breakthrough bleeding lasting several days. On request of her GP, our attending psychiatrist de-blinded her randomization (with the condition remaining blinded for the researchers) three days before the one-month follow-up. No serious AEs were reported. Details analyses of the effects of testosterone-administration on exposure (session 1) Fear The model showed that the linear and quadratic terms were significant predictors (each added significantly to the model fit). SUD scores reduced over time, confirming that the enhanced exposure resulted in the expected within-session reduction in fear: Estimate (linear) = -81.96 (16.76), F(1,50) = 23.89, p <.001, Estimate (quadratic) = −85.12(13.99), F(1,50) = 36.95, p < .001. Of main interest here, was the Group x Time effect. The interaction was not significant: Estimate(linear) = -.86(16.76), F(1,51) = .003, p = .959, Estimate(quadratic) = 7.01(13.99), F(1,51) = .251, p = .619, which showed that the fear patterns observed for the enhanced session did not differ per group. Accordingly, both groups followed the same quadratic pattern (Figure chapter 4 – appendix 3, panel A).
RkJQdWJsaXNoZXIy MTk4NDMw