Twelve studies (647 participants) were included in the review. Studies were considered low to moderate quality; none described randomisation procedures or blinding methods.
Ejection fraction: Starting exercise sooner after myocardial infarction and use of longer exercise programmes resulted in increased benefit.
End systolic and diastolic volumes: There were greater reductions in end systolic and diastolic volumes when exercise training started early and lasted for longer.
The two variables of time from myocardial infarction and length of programme accounted for a significant proportion of between-study variation in outcomes (Q=25.5, p<0.01, R²=0.76 in 12 trials for ejection fraction, Q=23.9, p<0.05, R²=0.79 in nine trials for end systolic volume and Q=27.4, p<0.01, R²=0.83 in 10 trials for end diastolic volume). The trials were comparable when these variables were taken into account.
Sensitivity analyses did not markedly alter the direction of the findings.