More About: Saturated Fat Intake and Lung Cancer Risk Among Nonsmoking Women in Missouri

Extract

Swanson et al. (1) report that the results regarding lung cancer risk related to saturated fat intake, reported in full previously (2), were substantially affected by the method of analysis used. In essence, the method of energy adjustment was found to attenuate the estimated risk increase associated with saturated fat intake.

Two obvious additional issues arise. First, it is noted by Swanson et al. that only two previous observational studies of fat and lung cancer have used any method of energy adjustment at all and also that both the rationale and the appropriate methods of energy adjustment remain controversial. Therefore, for the sake of comparison, it would be interesting to see what results would be obtained in the data of Swanson et al. from an analysis with no energy adjustment. Second, not only the final fat-intake analytic model of Swanson et al. is subject to potential influence by the method of energy adjustment. In their original report, the whole analytic strategy used the same method of energy adjustment, and the selection of nutrients and food groups for final models was based on results from energy-adjusted models with only one nutrient or food group at a time [Tables 3 and 4 in Alavanja et al., 1993 (2)]. Since their study is the largest case-control study of diet and lung cancer in nonsmokers to date, it would be of great interest to know how these first steps of their analysis are affected by the method of energy adjustment. In particular, one wonders how the estimates and significance levels for various other nutrients and food groups in Tables 3 and 4 in Alavanja et al. (2) are affected by different or no energy adjustment and how this might affect the conclusions to be drawn from their study.