Comparison of three methods of measuring dietary fat consumption by African-American adults.

OBJECTIVE: To determine dietary fat consumption by African-American adults using three methods that capture different domains of fat consumption and assess the degree of agreement between the three instruments.

DESIGN/SUBJECTS: Data were self-reported from a population-based cross-sectional survey of 658 African Americans, aged 18 to 70 years, in North Carolina using an 11-page questionnaire. Fat consumption was based on a household inventory of high-fat and low-fat foods, fat-related dietary behaviors, and total and saturated fat intakes. Demographic, behavioral, and diet-related psychosocial factors were also assessed.

STATISTICAL ANALYSES: Descriptive statistics were computed. Linear regression and Pearson's correlations examined the degree of agreement between the three dietary fat assessment methods. Linear regression was also used to assess associations of participant characteristics with fat consumption.

RESULTS: The participant mean age was 43.9+/-11.6 years, 41% were men, 37% were college graduates, and 75% were overweight/obese. High proportions of respondents had high-fat foods in their homes (78% had full-fat salad dressing and butter/margarine), whereas fewer had reduced-fat food items (only 19% had 1% or skim milk). Similarly, few respondents practiced healthful fat-related dietary behaviors; for example, 48% did not take the skin off chicken. There was good concordance among the three measures of fat consumption; for example, among those with more high-fat foods in the home, adjusted total and saturated fat intakes were 39.6 g and 14.9 g (compared to 24.1 g and 9.2 g, respectively, for those with none/few high-fat foods), P<0.0001. Pearson's correlation coefficients among the three methods ranged from 0.18 to 0.58 (P<0.0001).

CONCLUSIONS: There was good agreement among three instruments capturing different domains of dietary fat consumption by African Americans. The household food inventory is a brief and practical measure that may be a useful alternative dietary assessment tool in this population.