Validity of basal metabolic rate prediction equations in elderly women living in an urban tropical city of Brazil

Abstract

To measure basal metabolic rate (BMR) and to compare it with the values obtained from predictive equations in a sample of elderly (≥60y) women. Cross-sectional study. Seventy-nine women living in Niterói, Rio de Janeiro, Brazil enrolled in physical activity programs open to the community. Anthropometric measures were obtained using standard procedures. Percent body fat (PBF) was assessed by DXA. BMR was measured (BMRm) by indirect calorimetry under standardized conditions and compared with BMR estimated by 14 predictive equations that included elderly individuals in their development. Mean (±SD) age, BMI and PBF were 69.7±6.5y, 27.2±4.6 kg/m2 and 42.1±5.9%. BMRm (4188.3±707.2kJ/day) was significantly lower than estimated BMR by all predictive equations, including the equation developed for the Niteroian adult population (4565.6±607.9kJ/day). This population-specific equation provided the largest number of results within ±10% of BMRm and the lowest overestimation (10.6±15.4%), much lower than the results from the internationally recommended Schofield equation (27.2±17.6%). Regression of calf circumference (CC), age and body mass on BMRm provided similar estimates in comparison to models with fat-free mass (FFM). All predictive equations provided biased, inaccurate estimates of BMR values in comparison to BMRm. Anthropometry and body composition explained only approximately 50% of the variability of BMRm. New equations should account for the variability of organ-metabolic rates and underlying undetected health conditions in older individuals living in tropical regions.