Regular physical activity has been hypothesized to prevent breast cancer, largely by reducing circulating levels of sex hormones. The mechanisms by which physical activity reduces exposure to hormones vary by period of life. Young girls participating in strenuous athletic training, such as running and ballet dancing, have delayed menarche, which is known to reduce risk of breast cancer, and even moderate-intensity physical activity may delay menstruation. This effect of activity at young ages may be reflected in lower body weight and body fat, both of which are determinants of delayed menstruation. A later menarche is associated with a later onset of regular ovulatory cycles as well as lower serum estrogen concentrations during adolescence. Once menstruation has been established, anovulatory and irregular menstrual cycles may be more frequent among moderately and strenuously active women than among inactive women, although disagreement exists regarding the degree to which the intensity of physical activity influences menstrual abnormalities. Furthermore, a substantial degree of ovarian dysfunction may occur, even among physically active women who appear to have normal menstrual cycles. Among older women, levels of past and current physical activity influence fat stores, and which after menopause are the locus of conversion of androstenedione to estrogen.

Despite the evidence that higher levels of physical activity are associated with lower levels of circulating ovarian hormones, epidemiologic studies relating physical activity to risk of breast cancer show inconsistent results. Methodologic differences in physical activity assessment are likely to have contributed to these inconsistencies. Studies have differed in the ages at which physical activity was assessed; methods for measuring intensity, frequency, and duration of physical activity; definition and categorization of physical activity levels; and age of breast cancer diagnosis. Furthermore, ranges of physical activity typically studied are very limited in comparison with the levels of hard labor typically practiced by women in traditional agrarian societies.

The strongest reduction in breast cancer risk associated with increased physical activity was reported in a population-based case control study of women younger than age 40. The relative risk was 0.42 (95% confidence interval, 0.27 to 0.64) for women with a lifetime average of 3.8 hours or more of physical activity per week compared those with an average of 0 hours per week.

This has been the only study explicitly devoted to the relationship between physical activity and breast cancer, and it used a detailed physical activity assessment instrument to quantify the average number of hours per week of recreational physical activity over the reproductive life span, beginning at menarche. Activities such as housework, gardening, and easy walking not for the explicit purpose of physical exercise were not counted in the measure of physical activity. These researchers concluded from their various analyses that lifelong physical activity is the critical factor of interest with regard to breast cancer risk.

In contrast to the detailed measurement instrument used in the study described in the previous paragraph, a relatively simple measure of physical activity was used in a prospectivecohort study of Norwegian women aged 20 to 54 years at baseline. Over a period of 3 to 5 years, women were administered two surveys about their patterns of physical activity during leisure hours. The relative risk was 0.63 (95% confidence interval, 0.42 to 0.95) for consistently active women compared with consistently sedentary women, which is the second-strongest inverse association reported in the literature. This study is also the only prospective cohort study of those reported and to find a substantial inverse association between physical activity and breast cancer risk.

Most studies fall between these two with regard to the detail of physical activity measurement and categorization. For instance, in two population-based case control studies conducted among younger women, physical activity both early in life and in the period immediately before the interview was assessed. However, neither of these studies found an inverse association between physical activity (in either period) and breast cancer risk, despite defining physical activity categories in various ways.

Because of the potential public health significance of an association between a modifiable lifestyle risk factor such as physical activity and breast cancer, studies need to address important methodologic issues surrounding physical activity measurement. These issues include the resolution of whether a critical time period exists during which increased physical activity exerts its strongest effect on breast cancer risk, or whether, as hypothesized by Bernstein et al., lifetime physical activity is the critical exposure of interest. A second important issue relates to the quantification of physical activity and how information on frequency, intensity, duration, and time span of activity should be combined into a single measure. A third issue pertains to the validity of women’s reports of past physical activity. In case control studies, random error in recall of past activity levels that is not dependent on disease status would be expected, on average, to dilute any inverse association that might truly exist. If errors are differential by disease status, however, findings may be biased in either direction away from their true point estimates.

In studies of occupational physical activity, the results can potentially be confounded by reproductive characteristics, because women in physically active jobs may be more likely to be of lower socioeconomic status and thus may be more likely to have a lower-risk reproductive profile. Several studies of occupational activity were unable to control for such potential confounders. Finally, although a hormonal mechanism has been postulated, few data exist relating physical activity over sustained periods of time to lower endogenous ovarian hormone levels. Available studies have been very short term, have been based on small numbers of women, and often have been limited to comparisons between young women who engage in high levels of activity and inactive young women.

Although the relationship between physical activity and risk of breast cancer remains unsettled, indirect evidence relating higher physical activity levels to decreased risk of postmenopausal breast cancer is strong because of the important role of activity in controlling weight gain, an important cause of postmenopausal breast cancer. This, in addition to many other benefits of staying lean and fit, provides sufficient justification for including regular physical activity in daily life.

Walter C. Willett, Beverly Rockhill, Susan E. Hankinson, David J. Hunter and Graham A. Colditz

W. C. Willett: Harvard Medical School, Boston, Massachusetts; Department of Nutrition, Harvard School of Public Health, Boston, Massachusetts
B. Rockhill: Harvard Medical School, Brigham and Women’s Hospital, Boston, Massachusetts
S. E. Hankinson: Departments of Medicine and Epidemiology, Harvard Medical School and Harvard School of Public Health, Boston Massachusetts
D. J. Hunter: Departments of Epidemiology and Nutrition, Harvard School of Public Health, Boston, Massachussetts
G. A. Colditz: Department of Medicine, Harvard Medical School, Boston, Massachussetts

References