Current issues of ACP Journal Club are published in Annals of Internal Medicine


Etiology

Neither dietary fat or fibre were associated with breast cancer in premenopausal or postmenopausal women

ACP J Club. 1993 Mar-April;118:55. doi:10.7326/ACPJC-1993-118-2-055


Source Citation

Willett WC, Hunter DJ, Stampfer MJ, et al. Dietary fat and fiber in relation to risk of breast cancer. An 8-year follow-up. JAMA. 1992 Oct 21;268:2037-44.


Abstract

Objective

To determine if dietary fat increases the risk for breast cancer and if dietary fiber decreases the risk for breast cancer.

Design

Cohort analytic study with 8 years follow-up.

Setting

11 U.S. states.

Participants

89 494 women participating in the Nurses' Health Study who provided completed dietary questionnaires and had no previous diagnosis of cancer other than non-melanoma skin cancer. The follow-up rate for nonfatal outcomes was 96%, and the ascertainment of mortality was estimated to be 98% complete.

Assessment of risk factors

Dietary fat and fiber intake were assessed using a validated dietary questionnaire at baseline in 1980 and in 1984.

Main outcome measure

Invasive breast cancer identified from follow-up questionnaires every 2 years or from vital records and confirmed by pathology reports in 95% of cases.

Main results

1439 incident cases of breast cancer were diagnosed, including 774 among postmenopausal women. Patients were divided into quintiles of total fat intake with < 58 g/d being the lowest intake and ≥ 82 g/d being the highest. After adjustment for age, established risk factors, alcohol and total energy intake, and leaving the relative risk (RR) for the lowest fat level at 1.0, the RRs of breast cancer among women ranged from 0.85 to 0.96. The 95% confidence intervals included 1.0, indicating that the associations were not statistically significant. A similar pattern was seen for postmenopausal women. Similarly, no positive association was observed without adjustment for total energy; for saturated, monounsaturated, and polyunsaturated fats; after excluding the first 4 years of follow-up; for women consuming ≥ 49% of total energy intake in fat compared with those consuming < 29%; and for tumors < 2 cm and tumors > 2 cm in diameter. For dietary fiber intake, the quintiles ranged from ≤ 11 g/d to ≥ 22 g/d. After adjustment for established risk factors and leaving the RR of the lowest level of fiber intake at 1.0, the RRs of breast cancer among women ranged from 0.93 to 1.02. Once again all confidence intervals included 1.0.

Conclusions

Total fat intake and type of fat were not risk factors for breast cancer among pre- or postmenopausal women. A protective effect of dietary fiber was not detected.

Source of funding: National Institutes of Health.

For article reprint: Dr. W.C. Willett, Harvard School of Public Health, 665 Huntington Ave., Boston, MA 02215 USA. FAX 617-432-0464.


Commentary

The potential association of dietary fat intake with breast cancer risk has received enormous attention in the scientific literature. Many studies have been reported, with conflicting results. In general, studies using weak designs have been positive and those using strong designs have either been negative or have shown a modest effect.

Thus, the results of the recent update of the Nurses' Health Study are not surprising. This study, which is the largest and arguably the best designed and executed to date, not only fails to provide evidence of a significant association of moderate to high levels of fat intake with breast cancer risk; it also suggests that a nonsignificant inverse trend may exist. Additionally, a protective effect of dietary fiber intake could not be identified.

Where does this leave us? Existing public health recommendations call for a reduction in fat intake to moderate levels (approximately 30% of calories) to improve overall health. Current evidence suggests that compliance with these recommendations will not have a major effect on breast cancer risk, and more extreme reductions (to between 15% and 20% of calories) cannot be justified by existing data. Supporting evidence might be obtained from randomized trials; however, given the reductions in fat consumption that are occurring in the general public and the lack of evidence that the association of fat intake with breast cancer risk is nonlinear, the yield from such costly studies may be small. Because dietary changes affect more than 1 disease, these studies would be more readily justified if they measured overall health benefits and if they examined behavioral issues associated with adoption of healthy lifestyles.

Pamela J. Goodwin, MD
University of TorontoToronto, Ontario, Canada