Dietary fat, body weight, and cancer: Contributions of studies in rodents to understanding these cancer risk factors in humans

Understanding diet and energy balance as risk factors for breast, colon, an d other cancers requires information on the contribution of each factor and of interactions among factors to cancer risk. Rodent models for breast can cer provide extensive data on effects of dietary fat and calories, energy b alance, body weight gain, and physical activity on tumor development. Analy ses of the combined data from many studies have shown clearly that quality and quantity of dietary fat and energy balance contribute independently to increased mammary gland tumorigenesis. These findings were seen in female r ats fed diets high in fat (35-40% of calories) compared to rats fed control diets, with approximately 10% of calories as fat (Fay and Freedman, 1997, Breast Cancer Res. Treat. 46, 215-223). The methods used permit comparison of experimental and epidemiological data, and they may be useful in extrapo lating between species and developing public health recommendations. In add ition to the contributions of lifetime-diet composition, intake, energy bal ance, and physical activity to cancer risk, there are questions about the t iming and duration of alterations in these factors and about the "dose-resp onse" characteristics of cancer risk to the factors. Endocrine mechanisms m ay be significant in mammary gland tumor risk, but experimental and epidemi ological data indicate that cancers at other sites, such as colon and Liver , also are influenced by the factors listed. Other diet and lifestyle facto rs that influence energy, or specifically fat, metabolism may also affect r isk for cancers that are promoted by increased intake of fat and calories. Studies of separate and interactive effects of dietary fat, black tea, weig ht gain, and mammary gland tumorigenesis (Rogers, et al, 1998, Carcinogenes is 19, 1269-1273) have been analyzed. Using adjustment of carcinogenesis en dpoints for body weight, tumor burden, and latency, they were found to be r elated to weight gain within treatment groups in 2 of 3 experiments.