Skip to main content
ARS Home » Research » Publications at this Location » Publication #296143
Title: Effect of race of interviewer on energy underreporting from food frequency questionaires in a predominantly African American population

Author
item HUYE, HOLLY - University Of Southern Mississippi
item RUPP, RENEE - University Of Southern Mississippi
item CONNELL, CAROL - University Of Southern Mississippi

Submitted to: Federation of American Societies for Experimental Biology Conference
Publication Type: Abstract Only
Publication Acceptance Date: 2/2/2013
Publication Date: 4/23/2013
Citation: Huye, H.F., Rupp, R., Connell, C.L. 2013. Effect of race of interviewer on energy underreporting from food frequency questionaires in a predominantly African American population [abstract]. Federation of American Societies for Experimental Biology Conference. 27:617.20.

Interpretive Summary:

Technical Abstract: Dietary recalls are common practice in health assessments measuring nutritional status. Previous research has identified characteristics of people who are more likely to underreport, but little research has been done to determine if ethnic match between interviewer and participant influences underreporting. The purpose of this research was to assess the effect of race of interviewer on underreporting in a predominantly black population. The analysis used 290 food frequency questionnaires collected at a 3-month health assessment for a nutrition education intervention in the Lower Mississippi Delta (LMD). A predicted total energy expenditure formula by Vinken et al. was used to determine underreporting and indicated 67.5% were underreporters. The majority of the sample was black females (89.6%), aged equal to or greater than 41 y (82.8%), had some college or higher (67.3%), had income over $20,000, and classified as obese (69.9%). 41.4% were interviewed by a white interviewer; 58% were interviewed by a black interviewer. A multiple linear regression model using a dichotomous underreporting variable as the dependent variable and race of interviewer, BMI classification, education level, and income level as independent variables was significant (p less than .0001). However, race of interviewer was not a significant predictor in the model, suggesting race does not influence underreporting in this population.