Accuracy of Predictive Equations for Energy Expenditure in Mexicans Living With HIV/AIDS With and Without Antiretroviral Therapy

Introduction: Determination of the resting energy expenditure (REE) is essential for planning nutrition therapy in patients with human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS) to help to improve their nutrition status. We aim to evaluate the agreement and accuracy of prediction equations that estimate the REE in a Mexican population with a diagnosis of HIV/AIDS with and without antiretroviral therapy (ART). Methods: A cross‐sectional study in Mexican patients with HIV/AIDS with and without ART. Weight, height, and body composition measured with dual‐energy x‐ray absorptiometry were evaluated. The REE was determined with indirect calorimetry and estimated using the Mifflin–St Jeor (MSJ), Harris‐Benedict (HB), Schofield 1 and 2, Cunningham, Melchior 91, Melchior 93, and Batterham equations. The Bland‐Altman method assessed agreement between the real and estimated values, and the percent difference between these values was used to assess the prediction accuracy. Results: Sixty‐five adults without ART and 102 adults with ART were included. The mean REE (kcal/kg) was 24.8 ± 2.4 and 23.8 ± 3.6 in patients without and with ART, respectively. Good agreement and reliability were observed in the HB (intraclass correlation coefficient [ICC], 0.75; P < .05), Batterham (ICC, 0.79; P < .05), Schofield 1 (ICC, 0.74; P < .05), and Schofield 2 (ICC, 0.78; P < .05) results in individuals without ART. In individuals with ART, good agreement and reliability were observed with the HB equation (ICC, 0.76; P < .05). The MSJ equation showed good agreement with poor reliability (ICC, 0.05; P < .05). Conclusion: The equations with the best agreement and accuracy were Schofield 2, Batterham, and HB in individuals without ART and HB and MSJ in the population with ART.