Dietary Fat and Breast Cancer

Extract

A recent review (1) and an article (2), both appearing in the Journal, suggest that the misclassification of fat intake as measured by food-frequency questionnaires (FFQs) is the reason that cohort studies consistently have “been unable” to detect the hypothesized relation of dietary fat intake with breast cancer risk. We wish to respond briefly to these arguments and critiques of our pooled analysis of seven cohort studies of the relationship between dietary fat and breast cancer (3).

Both Wynder et al. (1) and Prentice (2) make much of the data from studies that use doubly labeled water to estimate energy expenditure; these data have been interpreted to show that energy expenditure is generally underestimated by individuals and more severely underestimated by obese individuals. Some underestimation of energy intake by FFQs is undisputed, but it has not been established that obese individuals exhibit bias in reporting the fat composition of their diet (i.e., fat intake adjusted for energy intake). Prentice (2) references one analysis in which reporting of protein (not fat) intake was correlated with obesity even after adjustment for reported energy intake (4). Other analyses, such as the analysis by Lissner and Lindross (5), have concluded that the degree of underreporting of energy intake is not associated with the fat composition of the diet.