BACKGROUND: Predictive equations have been reported to overestimate resting energy expenditure (REE) for obese persons. The presence of hyperglycemia results in elevated REE in obese persons with type 2 diabetes, and its effect on the validity of these equations is unknown. OBJECTIVE: We tested whether (1) indicators of diabetes control were independent associates of REE in type 2 diabetes and (2) their inclusion would improve predictive equations. DESIGN: A cross-sectional study of 65 (25 men, 40 women) obese type 2 diabetic subjects. Variables measured were: REE by ventilated-hood indirect calorimetry, body composition by bioimpedance analysis, body circumferences, fasting plasma glucose (FPG) and hemoglobin A(1c). Data were analyzed using stepwise multiple linear regression. RESULTS: REE, corrected for weight, fat-free mass, age and gender, was significantly greater with FPG>10 mmol/l (P=0.017) and correlated with FPG (P=0.013) and hemoglobin A(1c) as percentage upper limit of normal (P=0.02). Weight was the main determinant of REE. Together with hip circumference and FPG, it explained 81% of the variation. FPG improved the predictability of the equation by >3%. With poor glycemic control, it can represent an increase in REE of up to 8%. CONCLUSION: Our data indicate that in a population of obese subjects with type 2 diabetes mellitus, REE is better predicted when fasting plasma glucose is included as a variable.
BACKGROUND: Predictive equations have been reported to overestimate resting energy expenditure (REE) for obesepersons. The presence of hyperglycemia results in elevated REE in obesepersons with type 2 diabetes, and its effect on the validity of these equations is unknown. OBJECTIVE: We tested whether (1) indicators of diabetes control were independent associates of REE in type 2 diabetes and (2) their inclusion would improve predictive equations. DESIGN: A cross-sectional study of 65 (25 men, 40 women) obese type 2 diabetic subjects. Variables measured were: REE by ventilated-hood indirect calorimetry, body composition by bioimpedance analysis, body circumferences, fasting plasma glucose (FPG) and hemoglobin A(1c). Data were analyzed using stepwise multiple linear regression. RESULTS: REE, corrected for weight, fat-free mass, age and gender, was significantly greater with FPG>10 mmol/l (P=0.017) and correlated with FPG (P=0.013) and hemoglobin A(1c) as percentage upper limit of normal (P=0.02). Weight was the main determinant of REE. Together with hip circumference and FPG, it explained 81% of the variation. FPG improved the predictability of the equation by >3%. With poor glycemic control, it can represent an increase in REE of up to 8%. CONCLUSION: Our data indicate that in a population of obese subjects with type 2 diabetes mellitus, REE is better predicted when fasting plasma glucose is included as a variable.