Evaluation of the Healthy Eating Index-2005.

BACKGROUND: The Healthy Eating Index (HEI), a measure of diet quality as specified by federal dietary guidance, was revised to conform to the Dietary Guidelines for Americans 2005. The HEI has several components, the scores of which are totaled. OBJECTIVE: The validity and reliability of the HEI-2005 were evaluated. DESIGN: Validity was assessed by answering four questions: Does the HEI-2005 1) give maximum scores to menus developed by experts; 2) distinguish between groups with known differences in diet quality-smokers and nonsmokers; 3) measure diet quality independently of energy intake, a proxy for diet quantity; and 4) have more than one underlying dimension? The relevant type of reliability, internal consistency, was also assessed. SUBJECTS: Twenty-four-hour recalls from 8,650 participants, aged 2 years and older, in the National Health and Nutrition Examination Survey, 2001-2002 were analyzed to answer questions 2 to 4. Results were weighted to consider sample design and nonresponse. STATISTICAL ANALYSES: T tests determined differences in scores between smokers and nonsmokers. Pearson correlation coefficients determined the relationship between energy intake and scores. Principal components analysis determined the number of factors that comprise the HEI-2005. Cronbach's coefficient alpha tested internal consistency. RESULTS: HEI-2005 scores are at or very near the maximum levels for all sets of exemplary menus with one exception; the Harvard menus scored low on the milk component because these menus intentionally include only small amounts of milk products. Nine of 12 component scores were lower for smokers than nonsmokers. The correlations of component scores were virtually independent of energy intake (< I.22I). Multiple factors underlie the HEI-2005. Coefficient alpha was .43. The alpha value for all tests was .01. CONCLUSIONS: The HEI-2005 is a valid measure of diet quality. Potential uses include population monitoring, evaluation of interventions, and research. The individual component scores provide essential information in addition to that provided by the total score.

Duke Scholars

Author Bryce B. Reeve Population Health Sciences