Development and Validation of a Semiquantitative Food Frequency Questionnaire to Assess Dietary Intake in 4065-Year-Old Mexican Women

Introduction: Evaluation of total energy expenditure (TEE) and physical activity level (PAL) is important for treatment of patients with type 2 diabetes mellitus (T2DM). However, the validity of accelerometers (ACC) and physical activity questionnaires (PAQ) for estimating TEE and PAL remains unknown in elderly populations with T2DM. We evaluated the accuracy of TEE and PAL results estimated by an ACC (TEEACC and PAL(ACC)) and a PAQ (TEEPAQ and PAL(PAQ)) in elderly patients with T2DM. Methods: Fifty-one elderly patients with T2DM (aged 61-79 years) participated in this study. TEEACC was calculated with PAL(ACC) using a triaxial ACC (Active style Pro HJA-750c) over 2 weeks and predicted basal metabolic rate (BMR) by Ganpule's equation. TEEPAQ was estimated using predicted BMR and the PAL(PAQ) from the -Japan Public Health Center Study-Long questionnaire. We compared the results to TEEDLW measured with the doubly labeled water (DLW) method and PAL(DLW) calculated with BMR using indirect calorimetry. Results: TEEDLW was 2,165 +/- 365 kcal/day, and TEEACC was 2,014 +/- 339 kcal/day; TEEACC was strongly correlated with TEEDLW (r = 0.87, p < 0.01) but significantly underestimated (-150 +/- 183 kcal/day, p < 0.05). There was no significant difference in TEEPAQ and TEEDLW (-49 +/- 284 kcal/day), while the range of difference seemed to be larger than TEEACC. PAL(DLW), PAL(ACC), and PAL(PAQ) were calculated to be 1.71 +/- 0.17, 1.69 +/- 0.16, and 1.78 +/- 0.24, respectively. -PAL(ACC) was strongly correlated with PAL(DLW) (r = 0.71, p < 0.01), and there was no significant difference between the 2 values. PAL(PAQ) was moderately correlated with PAL(DLW) (r = 0.43, p < 0.01) but significantly overestimated. Predicted BMR was significantly lower than the BMR -measured by indirect calorimetry (1,193 +/- 186 vs. 1,262 +/- 155 kcal/day, p < 0.01). Conclusions: The present ACC and questionnaire showed acceptable correlation of TEE and PAL compared with DLW method in elderly patients with T2DM. Systematic errors in estimating TEE may be improved by the better equation for predicting BMR.