However, among never-quitters, the effect of stimulus valence on startle responses dissipated as the sessions progressed. The effect of stimulus valence on the startle response, while significant at baseline and at 1–2 days postcessation, became nonsignificant at sessions that occurred at 3–14 days postcessation, precessation baseline: F(3, 333) = 15.48, p < .0001; 1–2 days postcessation: F(3, 333) = 16.21, p < .0001; 3–5 days postcessation: F(3, 333) = 1.92, nonsignificant [ns]; 10–14 days postcessation: F(3, 333) = 1.41, ns. Furthermore, never-quitters exhibited a significantly larger reduction in startle potentiation, relative to ITI, to unpleasant stimuli (simple slope = —3.94) than to cigarette, simple slope = —1.81; t(1612) = 3.16, p < .001, pleasant, simple slope = —1.38; t(1612) = 3.96, p < .001, and neutral, simple slope = —1.63; t(1612) = 3.52, p < .001, stimuli across the preand postcessation sessions. Across groups, never-quitters’ reduction in startle potentiation to unpleasant stimuli across session was significantly larger than that of controls, simple slope = —1.38; t(1612) = 2.25, p < .05, abstainers, simple slope = —1.36; t(1612) = 2.15, p < .05, and relapsers, simple slope = —1.35; t(1612) = 2.00, p < .05.
