Nutritional factors have been prominent among the environmental determinants of breast cancer hypothesized to account for the large variation in breast cancer incidence around the world and the large increases in rates among the offspring of migrants from countries with low incidence to countries with high incidence. The dominant hypothesis has been that high fat intake increases risk. In this section, evidence for this relationship is reviewed, and alternative hypotheses are suggested.

Dietary Fat

Animal Studies
High-fat diets have long been known to increase the occurrence of mammary tumors in rodents. The interpretation of these and other animal data is controversial, however. Fat is the most energy-dense macronutrient (9 kcal/g compared with 4 kcal/g for protein and carbohydrate); thus, high-fat diets tend to be higher in energy intake unless care is taken to keep energy intake constant. Many animal experiments have not done this, so fat consumption is confounded by energy intake. In a meta-analysis of diet and mammary cancer experiments in mice, Albanes observed a weak inverse association with fat composition (adjusted for energy), whereas total energy intake was positively associated with mammary tumor incidence. Freedman et al. conducted a similar meta-analysis of experiments in rats and mice and reported that higher fat intake and higher caloric intake independently increase mammary tumor incidence. In studies specifically designed to determine the independent effects of fat and energy intake, the effect of fat was either weak or nonexistent in relation to that of energy intake. Furthermore, the relevance to human experience of rodent models, in which animals are given high doses of specific carcinogens to which humans are rarely exposed, is questionable. Notably, in a very large study of rats and mice fed substantially different amounts of corn oil without administration of a carcinogen, no effect of fat intake was found on spontaneous mammary cancer incidence. In a case control study in dogs, fat intake, which ranged from 10% to 70% of energy, was not associated with risk of breast cancer. The clearest message from the animal data is the importance of total energy intake and the need to consider energy balance in epidemiologic studies.

International Correlation (Ecologic) Studies
The dietary fat hypothesis is based largely on the observation that national per capita fat consumption is highly correlated with breast cancer mortality rates. A serious problem with ecologic comparisons of diet and breast cancer is the potential for confounding by known and suspected breast cancer risk factors. National fat consumption per capita is highly correlated with level of economic development; thus, any factor that characterizes affluent Western countries would also be correlated with national rates of breast cancer. Prentice et al. have found that the ecologic relation between fat disappearance and breast cancer incidences was still statistically significant after adjustment for gross national product per capita and average age at menarche. Other breast cancer risk factors, however, such as low parity, late age at first birth, greater body fat, and lower levels of physical activity, are more prevalent in Western countries and would be expected to confound the association with dietary fat. Thus, good reason exists to question whether the international correlation between fat intake and breast cancer represents a causal relationship.

Secular Trends
Estimates of per capita fat consumption based on “food-disappearance” data (the food available rather than the amount actually eaten) and breast cancer incidences increased substantially in the United States during the twentieth century. Surveys based on measures of individual intake, rather than food disappearance, however, indicate that consumption of fat as a percentage of energy has actually declined in the last several decades, a time during which breast cancer incidence has increased. Higher dietary fat consumption has been implicated in the increase in breast cancer incidence in Japan since 1950. However, this increase could also be due to the increasing prevalence of reproductive and other risk factors that characterize Western populations.

The famine that occurred in Norway during World War II provided a natural experiment on the effects of nutritional deprivation on breast cancer risk. Women who were adolescents during the famine have subsequently experienced a reduction in breast cancer risk (approximately 13% lower) at all ages. These data on time trends indicate the sensitivity of breast cancer rates to nutritional and lifestyle factors but do not specifically support a role of dietary fat.

Data from special populations with distinct dietary patterns are valuable, because adherence to a particular diet over many years may represent a more stable long-term exposure than that of most free-living adults, whose diet may change substantially over time. Because these populations often have unusual distributions of nondietary potential risk factors, such as alcohol consumption, smoking, and reproductive behavior, care must be taken in attributing differences in cancer rates to diet alone. Seventh-Day Adventists, who consume relatively small amounts of meat and other animal products, have substantially lower rates of colon cancer but only slightly lower breast cancer rates than other U.S. white women of similar socioeconomic status. Breast cancer rates among British nuns who ate no or very little meat were similar to rates among single women from the general population, findings also suggesting that no substantial association exists between animal fat and risk of breast cancer.