Evaluation of energy intake estimated by a diet history in three free-living 70 year old populations in Gothenburg, Sweden

Objective: To evaluate the credibility of estimates of energy intake from a Diet History (DH) by cut off limits for the multiple of energy intake and basal metabolic rate (EI/BMR(est) and by physical activity levels (PAL, total energy expenditure = TEE/BMR). Design: Cohort study. Setting: Departments of Geriatric Medicine and Clinical Nutrition, Goteborg University, Gothenburg, Sweden. Subjects: 369 males and 440 females from three representative cohorts of free-living individuals from the gerontological acid geriatric population studies-H70. Results: Mean values for EI/BMR(est) was 1.50 and 1.60 in males and 1.48 and 1.49 in females according to Schofield, Schofield and James and DHSS 41, respectively. A significant trend was seen when the sample was stratified at different levels of EI/BMR(est) with higher body weight, lower EI, higher proportion of energy from protein and lower of proportion energy from fat in the group with the lower values of EI/BMR(est). A significant difference was shown regarding food choice expressed as proportion of energy from ten defined food groups with respect to different EI/BMR(est) values. Lean body mass (LBM) by bioelectric impedance (BIA) correlated well with BMR according to DHSS 41, 0.90 for males and 0.87 for females. Conclusion: Energy intake was underreported with the DH method-especially in over-weight individuals. Reported food choice varied with EI/BMR values. EI/BMR(est) limits are useful for detecting underestimation of habitual energy intake.