A key feature of the dynamic model is its emphasis on the complex interplay between tonic and phasic processes. As indicated in Figure 2, distal risks may influence relapse either directly or indirectly (via phasic processes). The model also predicts feedback loops among hypothesized constructs. For instance, the return to substance use can have reciprocal effects on the same cognitive or affective factors (motivation, mood, self-efficacy) that contributed to the lapse. Lapses may also evoke physiological (e.g., alleviation of withdrawal) and/or cognitive (e.g., the AVE) responses that in turn determine whether use escalates or desists. The dynamic model further emphasizes the importance of nonlinear relationships and timing/sequencing of events. For instance, in a high-risk context, a slight and momentary drop in self-efficacy could have a disproportionate impact on other relapse antecedents (negative affect, expectancies) [8]. Furthermore, the strength of proximal influences on relapse may vary based on distal risk factors, with these relationships becoming increasingly nonlinear as distal risk increases [31]. For example, one could imagine a situation whereby a client who is relatively committed to abstinence from alcohol encounters
