Table 5. Reward predictors of leisure activity engagement (leisure vs functional activity engagement)^a.

	Whole sample (n=1383)			Cb (n=365)			MDc (n=257)			HDd (n=761)
Predictors (t−1)	β	95% CI	P value	β	95% CI	P value	β	95% CI	P value	β	95% CI	P value
Intercept	−.633	−0.943 to −0.323	<.001	−.514	−1.161 to 0.133	.12	−.973	−1.688 to −0.258	.008	−.620	−1.032 to −0.208	.003
Anticipatory pleasure	.194	0.069 to 0.319	.003	.426	0.13 to 0.722	.005	.232	−0.105 to 0.569	.18	.129	−0.028 to 0.286	.10
Expectation	−.014	−0.022 to −0.006	<.001	−.032	−0.05 to −0.014	.001	−.005	−0.021 to 0.011	.55	−.013	−0.021 to −0.005	.003
Motivation (interest)	−.056	−0.176 to 0.064	.35	−.142	−0.422 to 0.138	.32	.006	−0.29 to 0.302	.97	−.079	−0.228 to 0.07	.30
Motivation (prefer)	.105	0.023 to 0.187	.011	.165	0.002 to 0.328	.048	.090	−0.116 to 0.296	.39	.087	−0.023 to 0.197	.12

^aTime-lagged predictors of engagement in leisure activities in the whole sample, controls, moderate depression, and high depression groups. Time-lagged logistic regressions, n = number of assessments. Leisure versus functional activity outcomes were coded as 1 and 0, respectively. Controlled for assessment time and mood (negative affect and positive affect). When age and assessment period were added as extra covariates, results remained the same.

^bC: controls.

^cMD: moderate.

^dHD: high.