Sources of variance in 24-hour dietary recall data: Implications for nutrition study design and interpretation

• Published in: The American journal of clinical nutrition Vol. 32; no. 12; pp. 2546 - 2559
• Main Authors: Beaton, G.H, Milner, J, Corey, P, McGuire, V, Cousins, M, Stewart, E, de Ramos, M, Hewitt, D, Grambsch, P.V, Kassim, N, Little, J.A
• Format: Journal Article
• Language: English
• Published: United States 01.12.1979
• Subjects: Adult; Cholesterol, Dietary; Diet Surveys; Dietary Carbohydrates; Dietary Fats; Dietary Proteins; Energy Intake; Female; Food Analysis; Humans; Interviews as Topic - methods; Male; Mental Recall; Nutrition Surveys; Nutritional Physiological Phenomena; Research Design; Sex Factors; Statistics as Topic
• ISSN: 0002-9165; 1938-3207
• DOI: 10.1093/ajcn/32.12.2546

An in-depth statistical analysis of data from the National Heart, Lung and Blood Institute's Lipid Research Clinic provides insight into the sources of variance which affect the 24-hour dietary recall method of collecting diet information. Interviewers and training had no effect; data handling had a small effect. There was a distinct sex difference in total nutrient intake and for females, a day of the week effect, both of which became insignificant in light of nutrient concentration/energy ratios. Interindividual coefficients compared to intraindividual coefficients revealed variation from 1.2 for absolute intakes of protein, carbohydrate, fat, etc. to 2.1 for polyunsaturated fatty acids and cholesterol. Intraindividual variation was found to influence correlation coefficients and regression slopes toward 0. This information should be helpful in design and interpretation of disease and nutritional studies