Copyright © 2004 by the Johns Hopkins Bloomberg School of Public Health
PRACTICE OF EPIDEMIOLOGY |
Intervention-related Bias in Reporting of Food Intake by Fifth-Grade Children Participating in an Obesity Prevention Study
1 Division of Epidemiology, School of Public Health, University of Minnesota, Minneapolis, MN.
2 Department of Nutrition, School of Public Health and Health Sciences, University of Massachusetts, Amherst, MA.
3 Prevention Research Center, Health Sciences Center, University of New Mexico, Albuquerque, NM.
4 Department of International Health, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD.
5 Division of Epidemiology and Clinical Applications, National Heart, Lung, and Blood Institute, Bethesda, MD.
6 Department of Biostatistics, Collaborative Studies Coordinating Center, University of North Carolina, Chapel Hill, NC.
7 Office of Native American Diabetes Programs, Health Sciences Center, University of New Mexico, Albuquerque, NM.
8 Department of Pediatrics/Center for Applied Research and Evaluation, University of Arkansas for Medical Sciences, Little Rock, AR.
Data collected as part of Pathways, a school-based trial for the primary prevention of obesity in American Indian children conducted between 1997 and 2000, were analyzed to examine possible intervention-related bias in food reporting. The authors hypothesized that children in the intervention schools may have systematically underreported their dietary intake relative to children in the control schools. Nutrient intake estimates for lunch derived from record-assisted 24-hour dietary recalls were compared with intake estimates from observed lunch intakes. Reported nutrient intakes were included in regression analyses as the dependent variables; observed intake, intervention condition, and age were included as independent variables. Results indicated that, among females, intervention condition was a significant predictor of reported energy, fat, and saturated fatty acid intakes. Independently of observed intake, reported lunch energy intake among females in the intervention schools was 66.8 calories lower than reported intake among females in the control schools (p = 0.03). These findings suggest that investigators should consider bias in reporting of dietary intake by intervention condition when conducting diet-focused intervention studies. Specifically, enhancing measures that rely on self-reports with objective measures of dietary intake would help investigators to evaluate whether differential reporting by treatment group has occurred.
bias (epidemiology); intervention studies; nutrition assessment