The effect of regular aerobic exercise on positive-activated affect: A meta-analysis

The purpose of this meta-analysis was to examine the effect of regular aerobic exercise on self-reported positive-activated affect (PAA). Samples from 105 studies (1980–2008) were included yielding 370 effect sizes (ESs) and 9840 participants. Studies were coded for the following moderators: baseline affect, exercise frequency, intensity, time, program duration, exercise dose, study quality, and study source. The analysis employed multiple measures of affect and corrected for statistical artifacts using the meta-analytical methods of Hunter and Schmidt (1990, 2004). The overall mean corrected ES ( d ¯ corr ) and standard deviation (SD corr) were .57 and .48, respectively. Two clear moderator effects were found: the inverse association between baseline PAA and ES and the positive association between study quality and ES. The effect also varied with exercise frequency (positive relation) and exercise intensity (negative relation). Exercise dose was only a weak moderator, but the results indicate the following aerobic exercise program as optimal for improving PAA: low intensity (∼30% VO 2R), 30–35 min, 3–5 days/wk for 10–12 weeks. Similar effects were found for published and unpublished studies (source). Control conditions produced little change in PAA ( d ¯ corr = .03 , SD corr = .11 ) . Regular aerobic exercise results in moderate increases in self-reported PAA, but the effects vary by baseline affect and study quality. Exercise-related variables produced weaker moderating effects. PAA was unchanged for control conditions. A more comprehensive understanding of exercise-related affect will emerge when researchers examine the interaction of acute and chronic responses.