Accuracy of basal metabolic rate estimated by predictive equations in Japanese with type 2 diabetes.

Abstract

Estimation of energy demand using basal metabolic rate (BMR) is a rational approach for optimizing glycemic control and weight management in patients with type 2 diabetes mellitus (T2DM). Here, we assessed the accuracy of predictive equations in estimating BMR in Japanese patients with T2DM. BMR was measured indirectly (BMRm) with a portable gas analyzer in the fasting state in 69 Japanese patients with T2DM. BMR was estimated using the Harris-Benedict equation (BMRhb) and Ganpule equation (BMRg). An original predictive equation (BMRdm) was formulated by stepwise multiple regression analysis using subject age, lean soft tissue mass, fat mass and bone mineral content. Mean differences and 95% limits of agreement between measured and three estimated BMRs were evaluated by Bland-Altman plots. In addition, subjects were divided into three BMI groups (normal, BMI <25; overweight, BMI >=25; obese, BMI >=30), and the influence of BMI on the error size between measured and estimated BMRs was assessed. Between BMRm and the three estimated BMRs (BMRhb, BMRg, and BMRdm), there were small systematic errors with large random errors (mean difference±2SD ; -32±365 kcal，26±405 kcal, and -1.6±349 kcal, respectively) and significant proportional errors (r=0.42, 0.44, and 0.30, respectively). BMI subgroup analysis revealed that the obese group showed larger random errors and significant proportional errors compared to the overweight and normal weight groups. Predictive equations provide unacceptably inaccurate estimates of BMR in Japanese patients with T2DM, particularly in obese individuals.