Objective: The purpose of this review was to assess whether body weight con
founds the relationships between physical activity and its health benefits.
Methods: The review includes 80 reports from population-based studies (Evi
dence Category C) of physical activity or fitness and cardiovascular diseas
e (CVD) or coronary heart disease (CHD). Results: Eleven of 64 reports on a
ctivity found no relationship between physical activity and disease. Of the
remaining 53 reports, 11 did not address the possible confounding effects
of body weight, nine cited reasons that weight differences should not expla
in their observed associations, and 33 statistically adjusted for weight (a
s required). Only three of these changed their associations from significan
t to nonsignificant when adjusted. Ten of 16 reports on cardiorespiratory f
itness and CHD or CVD used statistical adjustment, and none of these change
d their findings to nonsignificant. Population studies show that vigorously
active individuals also have higher high-density lipoprotein (HDL)-cholest
erol concentration, a major risk factor for CHD and CVD, than sedentary ind
ividuals when statistically adjusted for weight. In contrast, intervention
studies, which relate dynamic changes in weight and HDL, suggest that adjus
tment for weight loss largely eliminates the increase in HDL-cholesterol in
sedentary men who begin exercising vigorously. Adjusting the cross-section
al HDL-cholesterol differences for the dynamic effects of weight loss elimi
nates most of the HDL-cholesterol difference between active and sedentary m
en. Conclusion: Population studies show that the lower incidence of CHD and
CVD and higher HDL of fit, active individuals are not because of lean, hea
lthy individuals choosing to be active (i.e., self-selection bias). Neverth
eless, metabolic processed associated weight loss may be primarily responsi
ble for the HDL differences between active and sedentary men, and possibly
also their differences in CHD and CVD.