Examining experiential avoidance as a mediator of the relation between anxiety sensitivity and depressive symptoms.

Initial evidence suggests that experiential avoidance (EA) mediates the relation between anxiety sensitivity (AS) and depression. We examined the AS-EA-depression pathway, examining both concurrent, and prospective (cross-lag), mediation models. Utilizing data from a study that examined the effects of exercise on AS ( = 60), we modeled depressive symptoms, EA, and AS over four time points. Time-varying predictors were disaggregated into between-subjects (each person's mean level of the predictor) and within-subjects change (each person's deviations, at each time point, from their mean level on the predictor) components. Tests of the concurrent relations were partially consistent with predictions, with mean EA levels, but not within-subjects changes in EA, partially mediating the relation between AS and depression symptom severity. However, the prospective, cross-lag mediation model, in which AS predicted future EA controlling for previous EA, and EA predicted future depression, controlling for previous depression, yielded no significant effects. These results suggest that observed between-subjects mediation findings, found here and in previous studies, may not replicate using more stringent, quasi-causal, cross-lag mediation analyses. These results highlight the importance of estimating causal pathways in mediation analyses. Clinical implications and directions for future research are discussed.