Population attributable risk for breast cancer: diet, nutrition, and physical exercise

J Natl Cancer Inst. 1998 Mar 4;90(5):389-94. doi: 10.1093/jnci/90.5.389.


Background: The percentages of breast cancer cases in a given population attributable to specific risk factors (i.e., the population-attributable risks) can be calculated; determination of such risks associated with potentially modifiable risks factors, such as diet (e.g., levels of consumption of fruits, vegetables, vitamins, etc.), alcohol consumption, exercise, and body weight, are necessary to focus prevention strategies.

Methods: With the use of data from a case-control study conducted in Italy from June 1991 through April 1994 on 2569 breast cancer case subjects and 2588 control subjects, we calculated multivariate odds ratios and population-attributable risks for breast cancer in relation to dietary beta-carotene and vitamin E intake, alcohol consumption, physical activity, and, for postmenopausal women, body mass index.

Results: Among all subjects, the following attributable risks for the indicated risk factors were observed: 10.7% (95% confidence interval [CI] = 4.4%-17.0%) for high alcohol intake (i.e., >20 g/day), 15.0% (95% CI = 7.4%-22.9%) for low beta-carotene intake (i.e., <3366 microg/day), 8.6% (95% CI = -0.4%-17.5%) for low vitamin E intake (i.e., <8.5 mg/day), and 11.6% (95% CI = -0.1%-23.3%) for low levels of physical activity. The risks associated with alcohol and beta-carotene intake were larger among premenopausal women, and the risk associated with physical activity was larger among postmenopausal women. Being overweight accounted for 10.2 % (95% CI = 0.2%-20.2%) of breast cancer cases in postmenopausal women. Beta-carotene plus alcohol accounted for 28.1% (95% CI = 16.8-39.4) of the cases. Beta-carotene and physical activity accounted for 32% (95% CI = 14.3-49.8), and these three factors together accounted for 33% (95% CI = 19.9-46.1) of the breast cancer cases in the overall dataset.

Conclusion: Exposure to a few selected and potentially modifiable risk indicators explained about one third of the cases of breast cancer in this Italian population, indicating the theoretical scope for prevention of the disease.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Aged
  • Alcohol Drinking*
  • Antioxidants / administration & dosage
  • Body Mass Index*
  • Breast Neoplasms / epidemiology*
  • Breast Neoplasms / etiology*
  • Case-Control Studies
  • Exercise*
  • Feeding Behavior*
  • Female
  • Humans
  • Italy / epidemiology
  • Middle Aged
  • Multicenter Studies as Topic
  • Odds Ratio
  • Postmenopause
  • Risk
  • Risk Factors
  • Vitamin E / administration & dosage
  • beta Carotene / administration & dosage


  • Antioxidants
  • beta Carotene
  • Vitamin E