Observational epidemiologic research has provided conclusive evidence that increased levels of at least moderately intense physical activity are associated with a decreased risk of breast cancer in both premenopausal and postmenopausal women. Women who remain physically active throughout their lifetimes benefit from a 30% to 40% decreased risk of breast cancer (5). There is also some evidence for a dose-response relation between increasing levels of activity and decreased breast cancer risk (5). Despite the strong evidence for a risk reduction, this effect has not been observed consistently in all epidemiologic studies conducted. Several methodologic reasons could explain the inconsistencies, including the methods used to measure physical activity, the type of activity assessed, the time period in life used to represent the etiologically relevant period of exposure, the type of adjustment for other important confounders, and the consideration of effect modification (5). One main reason for the inconsistencies across studies could be the fact that the relation between physical activity and breast cancer risk may differ by menopausal status. The purpose of this review was to examine whether menopausal status may be an important effect modifier in the association between physical activity and breast cancer. The specific questions to be addressed are: (1) What are the findings on physical activity and breast cancer stratified by menopausal status? and (2) What are possible reasons for effect modification by menopausal status?
EVIDENCE ON PHYSICAL ACTIVITY AND BREAST CANCER
All Studies Combined
To date, 23 cohort studies and 32 case-control studies have been conducted that have examined occupational, recreational, or household activity and breast cancer risk. Physical activity was assessed using self-administered or interview-administered questionnaires or job titles. Most of the studies included recent activity, but some studies did assess some aspect of lifetime activity, for example, usual occupational title or activity performed at different ages in the past. One study measured all types of physical activity performed throughout a lifetime, from childhood to the time of the interview, with a complete assessment of dose (i.e., frequency, intensity, and duration) of activity (4). No studies included any objective measures of physical activity or physical fitness. Of these 55 separate studies, 17 cohort and 23 case-control studies (40 studies) had evidence of decreased risk with increased levels of activity when examining the entire study population combined.
Most of these studies assessed menopausal status though the use of self-report; however, some used younger age groups (age younger than 45, 50, or 55 years, depending on the study) to represent the premenopausal population. The use of age groups as a proxy for menopausal status is a recognized limitation when examining the influence of menopausal status given the range in age of menopause that exists within populations. For the purposes of this review, stratification by age group was the only option available to begin assessing the influence of a possible differential effect of physical activity in premenopausal and postmenopausal women. Clearly, very careful determination of age at menopause is needed for valid assessments of the influence of menopausal status on breast cancer risk.
Twenty studies examined the impact of physical activity separately in both premenopausal and postmenopausal women within their study populations. A total of 26 studies were restricted to only premenopausal women, and 27 studies examined only postmenopausal women. The studies that included women of both menopausal subgroups are likely the most informative, because the assessment of physical activity and other study design and data collection methods are the same within each subgroup.
Studies of Premenopausal and Postmenopausal Women
Of the 20 studies that included both menopausal groups, three studies found no effect of physical activity in either menopausal group, four showed a significant decrease in risk in both groups, and five others found similar decreases that missed significance in one or both groups (Fig. 1). For eight studies, the results differed across the two groups, with decreases found only in the postmenopausal subgroup. A slightly larger number of studies, 11 of the 20, had a greater risk reduction in the postmenopausal group than in the premenopausal group. Furthermore, all of the case-control studies published since 2001 have found larger risk decreases in the postmenopausal women.
Premenopausal Breast Cancer Studies
There were 26 studies that examined the association of physical activity and breast cancer risk in premenopausal women (Fig. 2). Half of these studies (12 of 26) found no significant association between physical activity levels and breast cancer risk. Of the remaining 14 studies, seven found statistically significant risk reductions with increasing levels of physical activity and six found nonsignificant risk decreases. Only one study found an increased risk with increasing activity, and this increase was not statistically significant.
Postmenopausal Breast Cancer Studies
Results among postmenopausal women more strongly supported the association between increased activity and reduced risk of breast cancer (Fig. 3). Of 27 studies conducted in postmenopausal women, 16 demonstrated a statistically significant decreased risk with increased activity, a further six had nonstatistically significant risk reductions, and the five remaining studies observed no effect of activity on risk. No studies found increased risks of postmenopausal breast cancer with increased activity levels. Hence, most studies, 22 of 27 investigations, did find risk reductions among the most active postmenopausal women as compared with the least active within their studies.
Summary on Evidence for Effect Modification by Menopausal Status
From this review of the 40 epidemiologic studies of physical activity and breast cancer that permitted some assessment of a possible effect modification by menopausal status, it seems that a stronger and more consistent effect is observed for postmenopausal women. This conclusion is somewhat hampered by the limitations of these studies notably, the fact that menopausal status was not uniformly defined and applied to stratify the study populations in these studies. Hence, more confirmatory research will be needed that uses a clear definition of menopausal status to subdivide the participants. This possible interaction of menopausal status with physical activity on breast cancer risk also needs clarification from some of the underlying biologic mechanisms.
BIOLOGIC RATIONALE FOR AN EFFECT MODIFICATION BY MENOPAUSAL STATUS
There may be reason to expect that physical activity would have a different effect on breast cancer risk in premenopausal and in postmenopausal women. The natural history of breast cancer may differ in these two groups (6). Premenopausal breast cancer is more likely to be invasive disease that more often leads to death than breast cancer diagnoses after menopause (11). The median survival time for women with premenopausal breast cancer is lower than for women diagnosed after menopause (6). The type of effective treatment for breast cancer also differs by age and menopausal status and other key factors, such as tumor size and lymph node status (6). There are special concerns for premenopausal breast cancer that need to be considered, including the higher risk of local recurrence with breast-conserving surgery, fertility impairment, premature menopause, and more psychosocial distress and problems (11). Finally, the cause of the disease also differs by menopausal status at the time of diagnosis.
Table 1 lists some of the main established risk factors for breast cancer and the type of association with premenopausal and postmenopausal breast cancer; Figure 4 demonstrates the differential impact of these risk factors by menopausal status (12). The incidence of breast cancer increases with age, doubling every 10 yrs until menopause, when the rate of increases slows dramatically (10). Because the rate of increase in incidence decreases at menopause, substantial research has focused on understanding the hormonal and other biologic mechanisms that alter with menopause that could explain this change in slope of age-specific incidence rates. A possible unifying model for the cause of breast cancer suggests that breast cancer risk is increased as endogenous estrogen and progesterone levels are increased (8). Reproductive factors that extend the time period of estrogen exposure increase risk, for example, early menarche and late menopause, whereas factors that decrease estrogen exposure during the childbearing, premenopausal years, including pregnancy and breastfeeding, decrease risk (8). Some recent research has suggested that there may be a differential effect of these established risk factors for premenopausal and postmenopausal breast cancer. Early age at menarche and later age at first full-term pregnancy were shown to exert a stronger risk in premenopausal women, whereas the number of full-term pregnancies had a stronger protective effect on postmenopausal breast cancer (2). At the same age, premenopausal women are at a higher risk of breast cancer than postmenopausal women (8).
Endogenous hormones also influence breast cancer risk by binding to specific receptors in the breast. Some breast cancer tissues have estrogen receptor (ER) and progesterone receptor proteins (6). The incidence of ER-positive breast cancers increases with age and is significantly higher in women after menopause (6). Young women with ER-positive tumors have poorer prognoses than ER-negative women, an opposite pattern than that observed for older women (11). Unlike ER, progesterone receptor is more likely to be higher or positive in young or premenopausal women, likely because of greater estrogen stimulation (6).
Exogenous hormones also are implicated in the cause of breast cancer, with no large effect observed for oral contraceptives because these mimic the normal hormonal stimulation of the breasts (1). Hormone replacement therapy, however, now is more clearly associated with increased breast cancer risk, because the estrogen exposures in women taking hormonal therapy for menopause are much higher than those of untreated postmenopausal women (1). Furthermore, the effect of hormone replacement therapy is greater in lean women than in overweight women (8).
Genetic factors also have a greater influence on breast cancer risk in premenopausal women than in postmenopausal breast cancer. Having a family history of breast cancer, Jewish heritage, and being a carrier of BRCA-1, BRCA-2, or P53 have been identified as established, nonmodifiable risk factors that have a stronger effect for premenopausal than for postmenopausal breast cancer (8).
Epidemiologic evidence exists that some modifiable risk factors also differ by menopausal status. The influence of obesity and adiposity on breast cancer risk varies by menopausal status: obesity has a protective effect against breast cancer in premenopausal women, but is associated with increased risk in postmenopausal women (3). The biologic rationale for this difference in effect of obesity on breast cancer risk is based on the source of endogenous estrogens before and after menopause. Before menopause, the main source of estrogen production is in the ovaries and endogenous estrogen levels are tightly controlled by sex hormone binding-globulin (SHBG). In obese premenopausal women, ovarian estrogen production is inhibited by leptin that increases with increasing fat stores, thereby decreasing breast cancer risk. After menopause, adipose tissue becomes the main site of estrogen production. In obese postmenopausal women, these estrogen levels increase because SHBG levels decrease. Obesity also increases insulin levels, which in turn decrease SHBG levels. Prospective cohort studies have shown that high circulating levels of estrogens and androgens are associated with increased breast cancer risk in postmenopausal women (14).
Both insulin and insulin-like growth factors have strong mitogenic effects on breast cancer cells. Hyperinsulinemia and the insulin-resistance syndrome may be mechanisms whereby obesity increases breast cancer risk, but the associations are unclear because insulin is a risk factor for breast cancer independent of obesity, adiposity, and menopausal status (13). Increased plasma insulin-like growth factor 1 levels seem to increase breast cancer risk in premenopausal women also without any relationship to body weight. No association between insulin-like growth factor 1 has been found consistently for postmenopausal women (9).
Physical activity may influence premenopausal and postmenopausal breast cancer risks differentially through similar hormonal mechanisms as seem to exist with obesity. Physical activity reduces insulin resistance and consequently the levels of circulating insulin. Physical activity also can reduce insulin-like growth factor levels that can increase production of SHBG, resulting in diminished availability of free sex hormones. Hence, greater levels of physical activity can result in lower levels of endogenous sex hormones, insulin, and insulin-like growth factor that, in turn, could result in lower breast cancer risk. This risk reduction could have a stronger effect after menopause because physical activity not only would be reducing endogenous hormone exposure, but also would be decreasing obesity, which has a separate causal association with breast cancer risk (3). Kaaks (7) proposed a unifying model for breast cancer that combines overnutrition, obesity, low physical activity, chronic changes in the endocrine secretion of steroid hormones, and reduced SHBG production. In this model, a biologic rationale is provided for how postmenopausal obese and inactive women have an increased breast cancer risk as compared with comparable premenopausal women. Additional epidemiologic and biologic data are needed to clarify the exact nature of these mechanisms and how the differential effect on breast cancer risk occurs in premenopausal and postmenopausal women.
Numerous questions remain on the influence of physical activity on breast cancer incidence and survival after breast cancer. It is unknown if physical activity influences stage of disease at diagnosis, if there is an interaction with type of treatment, and if menopausal status does modify the risk between physical activity and breast cancer. Studies are currently on-going that are examining the influence of physical activity on breast cancer survival and are specifically examining the influence of numerous personal, medical, and treatment variables on the association between lifetime physical activity and breast cancer risk. More research clearly is warranted on the role of specific biologic mechanisms on the association between physical activity and breast cancer in premenopausal and postmenopausal women. Research on animal models of exercise and breast cancer is providing some insights into the possible mechanisms that may be operative that can be investigated further in humans. This additional animal and human research ultimately will provide the necessary scientific basis for clear and precise exercise guidelines for breast cancer risk reduction.
The author thanks Marla Orenstein for assistance in writing this paper and Xuechao Chen for the preparation of the figures. Dr. Friedenreich is supported by a Canadian Institutes of Health Research New Investigator Award and an Alberta Heritage Foundation for Medical Research Health Scholar Award. This article, with a more complete list of references than that provided here, can be obtained from the author.
1. Bernstein, L. Epidemiology of endocrine-related risk factors for breast cancer
. J. Mammary Gland Biol. Neoplasia
2. Clavel-Chapelon F., and M. Gerber. Reproductive factors and breast cancer
risk. Do they differ according to age at diagnosis? Breast Cancer Res. Treat.
3. Friedenreich, C. M. Review of anthropometric factors and breast cancer
risk. Eur. J. Cancer Prev.
4. Friedenreich, C. M., H. E. Bryant, and K. S. Courneya. Case-control study of lifetime physical activity
and breast cancer
risk. Am. J. Epidemiol.
5. Friedenreich, C. M., and M. R. Orenstein. Physical activity
and cancer prevention: Etiologic evidence and biological mechanisms. J. Nutr.
6. Harris, J. R., M. E. Lippman, M. Morrow, and K. E. Osborne. Diseases of the Breast
, 2nd ed. New York: Lippincott Williams & Wilkins, 2000.
7. Kaaks, R. Nutrition, hormones, and breast cancer
: Is insulin the missing link? Cancer Causes Control
8. Key, T. J., P. K. Verkasalo, and E. Banks. Epidemiology of breast cancer
. Lancet Oncol.
9. Krajcik, R. A., N. D. Borofsky, S. Massardo, and N. Orentreich. Insulin-like growth factor I (IGF-I), IGF-binding proteins, and breast cancer
. Cancer Epidemiol. Biomarkers Prev.
10. McPherson, K., C. M. Steel, and J. M. Dixon. ABC of breast diseases. Breast cancer
—Epidemiology, risk factors, and genetics. BMJ.
11. Shannon, C., and I. E. Smith. Breast cancer
in adolescents and young women. Eur. J. Cancer
12. Singletary, S. E. Rating the risk factors for breast cancer
. Ann. Surg.
13. Stephenson, G. D., and D. P. Rose. Breast cancer
and obesity: An update. Nutr. Cancer
14. Toniolo, P. G. Endogenous estrogens and breast cancer
risk: The case for prospective cohort studies. Environ. Health Perspect.
105(Suppl 3):587–592, 1997.