Anticipation of affect in dysthymia: behavioral and neurophysiological indicators

Abstract

Anticipation for future affective events and prediction uncertainty were examined in healthy controls and individuals with dysthymia (DYS) using behavioral responses and the contingent negative variation (CNV) and post-imperative negative variation (PINV) event-related potential (ERP) components. Warning stimuli forecasted the valence of subsequently presented adjectives ("+", positive; "=", neutral; "-", negative), and participants indicated whether each adjective would describe them over the next two weeks. Controls expected fewer negative, and individuals with DYS expected fewer positive, adjectives to apply to them. CNV amplitudes were enhanced in controls prior to positive versus other adjectives. Response times and PINV amplitudes were greater following neutral compared to other adjectives, and PINV was larger overall in dysthymics compared to controls. In sum, healthy controls and individuals with DYS exhibit different behavioral and neurophysiological biases in anticipation for future affective events. These results are discussed in the context of cognitive theories of depression.

Figure 1

Average (and standard deviation) self-reference ratings for positive (pos), neutral (neu), and negative (neg) adjectives, separated by group. Significant differences are indicated for within-group valence comparisons (V; p < .01) and between-group comparisons (G; p < .001); self-reference ratings for marked valence conditions (V) were significantly less than self-reference ratings for the two alternative valence conditions in both groups.

Figure 2

Grand average ERP waveforms following positive, neutral, and negative warning symbols, separated by group and lead. Data from 800-3000 ms were characterized as CNV.

Figure 3

Average (and standard deviation) CNV amplitude for positive (pos), neutral (neu), and negative (neg) adjectives, separated by group. Significant differences are indicated for within-group valence comparisons (V; p < .05) and between-group comparisons (G; p < .05); CNV prior to positive adjectives was significantly larger than the two alternative valence conditions in control participants.

Figure 4

Grand average ERP waveforms following positive, neutral, and negative adjectives, separated by group and lead. Data from 4000-5000 ms were characterized as PINV.