Greater Survival After Breast Cancer in Physically Active Women With High Vegetable-Fruit Intake Regardless of Obesity

John P. Pierce            Journal of Clinical Oncology, Vol 25, No 17 (June 10), 2007: pp. 2345-2351

Purpose: Single-variable analyses have associated physical activity, diet, and obesity with survival after breast cancer. This report investigates interactions among these variables.

Patients and Methods: A prospective study was performed of 1,490 women diagnosed and treated for early-stage breast cancer between 1991 and 2000. Enrollment was an average of 2 years postdiagnosis. Only seven women were lost to follow-up through December 2005.

Results: In univariate analysis, reduced mortality was weakly associated with higher vegetable-fruit consumption, increased physical activity, and a body mass index that was neither low weight nor obese. In a multivariate Cox model, only the combination of consuming five or more daily servings of vegetables-fruits, and accumulating 540+ metabolic equivalent tasks-min/wk (equivalent to walking 30 minutes 6 d/wk), was associated with a significant survival advantage (hazard ratio, 0.56).

The approximate 50% reduction in risk associated with these healthy lifestyle behaviors was observed in both obese and nonobese women, although fewer obese women were physically active with a healthy dietary pattern (16% v 30%). Among those who adhered to this healthy lifestyle, there was no apparent effect of obesity on survival. The effect was stronger in women who had hormone receptor–positive cancers.

Conclusion: A minority of breast cancer survivors follow a healthy lifestyle that includes both recommended intakes of vegetables-fruits and moderate levels of physical activity. The strong protective effect observed suggests a need for additional investigation of the effect of the combined influence of diet and physical activity on breast cancer survival.

In this study, we examined the interactions between selected lifestyle factors and mortality in breast cancer survivors who might change lifestyle to improve prognosis. These women volunteered for a diet trial within 4 years of diagnosis and were randomly assigned to the comparison group of the ongoing Women's Healthy Eating and Living Study. Our goal was to establish a parsimonious set of lifestyle predictors for use in the final analysis of the Women's Healthy Eating and Living Study, which is investigating the effect of a plant-based diet on additional breast cancer events. First, we looked for univariate effects on survival of lifestyle variables measured at baseline and, using only variables with significant associations, we developed a combined lifestyle variable. We tested the combined variable in a multivariate model controlling for obesity and cancer and treatment characteristics.

In these breast cancer survivors who were interested in lifestyle change and were predominantly nonsmokers, the 30% of women who were physically active and consumed at least 5 servings of vegetables and fruits each day had an estimated 10-year mortality rate of 7%, approximately half that of any other combination of PA and dietary pattern. Of particular importance, this halving of risk was seen in women who were not obese as well as in those who were obese. Indeed, it appeared that both being physically active and having a healthy dietary pattern attenuated the increase in risk observed among the obese. It was noted, however, that obese women were approximately one half as likely as nonobese women to be both physically active and to have a healthy dietary pattern.

This study is not the first to suggest that a combination of healthy lifestyle behaviors is associated with improved survival, although it is the first, to our knowledge, in a population of breast cancer survivors. The European Healthy Aging study of 70- to 90-year-old men and women noted that being a nonsmoker, adhering to a Mediterranean diet, being physically active, and consuming alcohol in moderation were associated with a 60% reduction in cancer deaths as well as 60% reduction in all-cause mortality. Similarly, the Nurses' Health Study noted that 82% of coronary events could be attributed to lack of adherence to a similar combination of behaviors. Longitudinal follow-up of a population-based random survey of men and women in Hawaii demonstrated that a combination variable constructed from smoking status, BMI, dietary pattern, and alcohol use was strongly associated with mortality, including mortality from cancer. Accordingly, most recent guidelines for chronic disease prevention, including cancer, cardiovascular disease, and obesity, emphasize the importance of adhering to multiple lifestyle behaviors. Given that the number of cancer survivors is increasing, and breast cancer survivors make up the largest group of cancer survivors,it is important to consider the potential impact on prognosis of lifestyle changes such as diet and PA on women who have survived initial treatment.

PA recommendations usually include duration (eg, 30 min/d), intensity (moderate, 3 to 6 METs; vigorous, 6+ METs) and frequency (eg, vigorous, 3 times/wk; moderate, 5+ times/wk). Accordingly, the absolute minimum recommendation for moderate activity would require 450 MET-min/wk. In this study, we chose a level 20% higher than this absolute minimum for classifying a participant as physically active; this allowed us to use an equivalent cut point to that used in the Nurses' Health Study. However, unlike the Nurses' Health Study, in the univariate analysis only, we found a significant linear association in which more activity was associated with increased benefit. It is important to note that the brief self-reported questionnaire has been shown to overestimate activity when compared with an accelerometer measure in our study population.

The survival advantage for the high VF/high PA lifestyle was present in women who had ER-positive tumors, but not in those who were ER-negative, suggesting that the mechanism for action for this effect might be reproductive gonadal hormones, as has been suggested previously for each of the component lifestyle factors. It is likely that this observed effect is not simply an effect of adjuvant tamoxifen, which was prescribed to 45% of the study sample. The effect observed may vary in current populations because study participants were all diagnosed before the introduction of adjuvant aromatase inhibitor therapy.

In summary, breast cancer survivors who consume a healthy diet and are physically active may increase their years of survival after diagnosis. Specifically, we have shown that those who reported eating a minimum of 5 VF servings daily and performing weekly PA equivalent to 30 minutes of walking at a moderate pace for 6 days a week had a higher 10-year survival rate than those who did not adhere to these lifestyle practices. The improved survival rate was observed in women who were obese as well as those who were not obese. Adhering to these two health behaviors reduced the probability of death in the follow-up period by 50%, whereas this effect was not seen if breast cancer survivors were adherent to only PA or VF dietary pattern. These findings suggest the need for additional study of the combined protective effect of diet and PA on breast cancer survival.