Objective: To validate a semi-quantitative food frequency questionnaire (FFQ).
Design: FFQ was evaluated by a 4-day food record (FR), four consecutive 24-h urinary collections, and study-specific cut-off limits for ratio between energy intake (EI) and basal metabolic rate (BMR).
Setting: Department of geriatric medicine, Göteborg University, Vasa Hospital, Gothenburg, Sweden.
Subjects: 76 consecutive 70-year-olds from representative population studies.
Results: FFQ provided consistently higher intakes of energy and nutrients than FR (1.18-2.51 of the FR values). For energy 95% CI estimates were 7167-8083 KJ and 9333-10579 KJ, by FR and FFQ methods, respectively. The correlation coefficients varied between 0.35 (total sugar) and 0.60 (energy and fat). The proportions and distribution of macronutrients were in good agreement between FFQ and FR. 95% CIs for the ratio between protein intake estimated by dietary methods and urinary protein were 1.14-1.36 and 1.00-1.13 for the FFQ and FR methods, respectively. However, the EI/BMR ratio for FR (1.27) was not in conformity with habitual energy expenditure compared to cut-off limits.
Conclusion: This validity study showed that the agreement regarding distribution of energy from macronutrients between dietary methods and the reasonable ratio of FFQ-EI/BMR allow the conclusion that the FFQ method well reflects habitual intake of elderly Swedes.