Unravelling the relationship between exercise‐induced affective responses and daily physical activity in people with chronic diseases: Spillover or confounding effect?

OBJECTIVES: Exercise-induced affective responses (ARs) may encourage people with chronic diseases (PCD) to engage in future physical activity (PA). This study examined whether AR timing during exercise predicts daily moderate-to-vigorous PA (MVPA) in PCD and explored two mediating pathways: the reflective affect processing pathway (via remembered pleasure, forecasted pleasure and affective attitudes) and the self-efficacy pathway (via self-efficacy towards exercise and PA). DESIGN: Prospective correlational design. METHODS: = 66.6 years, 79% women) with chronic diseases reported ARs and perceived exertion four times during an exercise session. Post-exercise measures included remembered and forecasted pleasure, BMI, affective attitudes and self-efficacy. Daily MVPA was measured using accelerometers over eight days. Total PA (light-intensity PA and MVPA) was examined as a secondary outcome. RESULTS: The last AR predicted daily MVPA (β = .24, p = .041), but this association became non-significant after adjusting for confounders (BMI, perceived exertion). In contrast, the last AR predicted total PA independently of confounders (β = .30, p = .008). Exploratory mediation analyses partially supported the reflective affect processing pathway: ARs predicted affective attitudes via remembered and forecasted pleasure, though affective attitudes did not predict either PA outcome. The self-efficacy pathway was supported for MVPA only, albeit with a modest indirect effect (β = .05). CONCLUSIONS: The last AR did not predict daily MVPA after adjusting for confounders, questioning a spillover effect from exercise sessions to MVPA. Secondary analyses suggest the last AR may predict total PA regardless of confounders, hinting at an intensity-specific spillover. Findings should be interpreted cautiously given the limited sample size and the exploratory nature of secondary analyses.