Low-Energy Reporters: Evaluation of Potential Differential Reporting in Case-Control Studies

Abstract

Errors in measuring dietary intake can threaten validity of data. Low-energy reporters (LER) are individuals who report lower levels of energy intake than deemed feasible given their basal metabolic rate and physical activity level (PAL). The purpose of this study was to determine whether LER differ by case/control status or by extent of disease of cases. Data from a large population-based case-control study of colon cancer were used to identify LER. Dietary data were collected using a diet history questionnaire. Age- and gender-specific basal metabolic rate was estimated, and Goldberg cut points were used to estimate plausible energy intake and adjusted for PAL. On the basis of standard methods that do not take PAL into account, 16.7% of male cases, 19.8% of male controls, 20.9% of female cases, and 22.2% of female controls were considered LER. There were no case-control differences in the proportion of LER in men or women when PAL-adjusted cut points were used, although more individuals were considered LER. Likewise, there were no differences in LER by colon cancer disease stage. Excluding LER from the population and assessing associations between energy intake and colon cancer yielded results similar to those observed for the total population. In this population, LER were significantly more likely to be older, never to have smoked cigarettes, to be more physically active, and to be overweight or obese. LER reported fewer total food items than non-LER. There does not appear to be differential reporting of low energy intake by cases and controls or by disease stage among cases. However, LER appear to differ depending on exposure characteristics that may be importantly associated with cancer.