RESUMO
BACKGROUND/OBJECTIVE: The growing interest in environmental influences on obesity risk has spawned the development of tools for appraising home food availability. These tools reveal good reliability but tend to be limited in scope and burdensome to use. This cross-sectional study investigated the feasibility of using food categories and scoring algorithms from valid food frequency questionnaires for individuals (that is, Block Dietary Fat and Fruit-Vegetable-Fiber Screeners) to estimate nutrient availability in household food supplies. SUBJECTS/METHODS: Screeners were compared with household food inventories from 100 two-parent families with ⩾1 children ⩽12 years of age. Inventoried foods were coded to match Screener food groups, and amounts available were converted to total adult daily equivalent servings to express the greatest possible frequency at which each food group could be eaten/day/household. Scoring algorithms were converted to express all scores on a per-day basis. For the most conservative assessment, the highest point was used for day ranges for the Fruit-Vegetable-Fiber Screener and the lowest range point was used for the Fat Screener. RESULTS: Spearman's rank-order correlations (r⩾0.76) showed that the Fruit-Vegetable-Fiber Screener ranked households well for fruit/vegetable servings, vitamin C and dietary fiber. The Fat Screener and household inventory were positively correlated (r⩾0.58) for total fat, saturated fat and cholesterol. Concordance of household inventories and the Fruit-Vegetable-Fiber Screener, as determined by kappa with quadratic weighting, were strong and significant. Fat Screener concordance was moderate. CONCLUSIONS: Results indicate that it is feasible to use the efficient, valid Block Dietary Screeners for individuals to appraise household food supplies.
