Dietary Assessment in Epidemiology: Comparison of a Food Frequency and a Diet History Questionnaire with a 7-Day Food Record

• Published in: American journal of epidemiology 1996-05, Vol.143 (9), p.953-960
• Main Authors: Jain, Meera, Howe, Geoffrey R., Rohan, Thomas
• Format: Article
• Language: English
• Subjects: Adult; Aged; Biological and medical sciences; Cost-Benefit Analysis; Cross-Over Studies; diet; Diet Records; Diet Surveys; Energy Metabolism; epidemiologic methods; Female; Human physiology applied to population studies and life conditions. Human ecophysiology; Humans; Male; Medical sciences; Middle Aged; Nutritional survey. Food supply and nutritional requirement; Predictive Value of Tests; questionnaires; Reproducibility of Results; Surveys and Questionnaires - economics; Surveys and Questionnaires - standards; Time Factors
• ISSN: 0002-9262; 1476-6256
• DOI: 10.1093/oxfordjournals.aje.a008839

The validity of two types of diet assessment methods, a self-administered food frequency questionnaire and an interviewer-administered detailed diet history, was assessed relative to a 7-day food record on a population-based sample of 95 men and 108 women in Toronto, Canada, between May 1989 and July 1990. Each study subject completed both questionnaire methods, a food frequency questionnaire and an interviewer-administered diet history, as well as a 7-day food record in a crossover design. Data were analyzed for both unadjusted and energy-adjusted nutrients to estimate Pearson's and intraclass correlations and agreement within categories. Mean values for the intake of most nutrients assessed by the two questionnaire methods were similar. Average, energy-adjusted Pearson's correlation coefficients for men between a food frequency questionnaire and a 7-day food record were 0.55 for macronutrients and 0.48 for micronutrients compared with 0.47 for macro- and 0.48 for micronutrients between an interviewer-administered diet history and a 7-day food record. For women, they were 0.48 for macro-and 0.54 for micronutrients between a food frequency questionnaire and a 7-day food record and 0.46 and 0.49, respectively, between an interviewer-administered diet history and a 7-day food record. The energy-adjusted Pearson correlations were generally higher than were the energy-unadjusted Pearson correlations and the intraclass correlations. The present study suggests that a food frequency questionnaire is comparable with an interviewer-administered diet history as a predictor of nutrients as estimated from a 7-day food record. Am J Epidemiol 1996;143:953–60.