Self-reported intake of fruits, vegetables, and macronutrients were tested for bias by social desirability trait, the tendency to respond in a manner consistent with societal expectations. Dietary intake was reported by 93 ethnically and racially diverse female employees at five Eastern Massachusetts health centers using a 61-item food frequency questionnaire (FFQ), 7-item fruit and vegetable screener, and a single question on combined fruit and vegetable intake. Discrepancies between these self-reports and those obtained from three 24-hour diet recall interviews were tested for bias using multiple linear regression in which the self-report measures were fit as the dependent variables, in separate models stratified by race/ethnicity and education, while controlling for important covariates. Regression coefficients for social desirability obtained from these models, represent the bias from the comparison method relative to 24-hour recall-derived mean intake. Results indicate that in women with at least a college education, FFQ-derived estimates of total energy and macronutrient intake were associated with a social desirability under-reporting bias (e.g., -23.6 kcal/day/point on the 33-point social desirability scale) compared to women not completing college (36.1 kcal/day/point) (p=0.01). Social desirability bias was not evident in self-reported fruit and vegetable intake. We conclude that an under-reporting bias in self-reported macronutrient intake was evident for the FFQ. Education appeared to be a more important effect modifier of this bias than did ethnicity or race. The results of this study may have important implications for epidemiologic studies, in particular those conducted in diverse populations representing a range of educational attainment.
Learning Objectives: "At the conclusion of the session, the participant (learner) in this session will be able to: 1.Recognize social desirability as a bias that could affect dietary self-reports using structured assessment instruments such as the food frequency questionnaire. 2. Identify education as an important effect modifier of the bias, with women having college or more education evincing the largest bias."
Keywords: Dietary Assessment, Women
Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: None
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.