Studies of the association between physical activity and breast cancer
have yielded inconsistent findings. These findings may be related to
a true null association or an inability to measure physical activity w
ith enough precision to measure a protective relation. The authors rev
iewed and critiqued physical activity measurement methods used in publ
ished studies of the association between physical activity and breast
cancer. The authors examined the quality of physical activity measures
in 20 published studies. A summary score was created to rank the qual
ity of the activity score. Studies with higher scores had a more preci
se measure of physical activity Physical activity measurement methods
were different in each study. Activity was measured by job classificat
ion, occupational tasks, participation in competitive athletics, and r
ecreational and leisure-time pursuits. The recall period for physical
activity ranged from a lifetime to the past year. Comparison of qualit
y scores showed no associations between the precision of activity meas
ures and the study results. Future studies of physical activity and br
east cancer should utilize standardized methods to measure physical ac
tivity. Researchers should be encouraged to choose a measure based on
hypotheses regarding physical activity and breast cancer mechanisms. S
tudies also should extend to subgroups of women with differences in ot
her breast cancer risk factors, such as body mass, menopausal status,
and hormone replacement status. Cancer 1998;83:611-20, (C) 1998 Americ
an Cancer Society.