Purpose For individuals receiving maintenance dialysis, estimating accurate resting energy expenditure (REE) is essential for achieving energy balance, and preventing protein-energy wasting. Dialysis-specific, predictive energy equations (PEEs) offer a practical way to calculate REE. Three PEEs have been formulated via similar methods in different demographic samples; the Maintenance Haemodialysis Equation (MHDE REE), Vilar et al. Equation (Vilar REE) and the Fernandes et al. Equation (Cuppari REE). We compared them in a US cohort and assessed precision relative to measured REE (mREE) from indirect calorimetry. Because of expected imprecision at the extremes of the weight distribution, we also assessed the PEEs stratified by body mass index (BMI) subgroups. Methods This analysis comprised of 113 individuals from the Rutgers Nutrition and Kidney Database. Estimated REE (eREE) was calculated for each PEE, and agreement with mREE was set at > 50% of values within the limits of ±10%. Reliability and accuracy were determined using intraclass correlation (ICC) and a Bland Altman plot, which analysed the percentage difference of eREE form mREE. Results Participants were 58.4% male and 81.4% African American. Mean age was 55.8 ± 12.2 years, and the median BMI was 28.9 (IQR = 25.3 − 34.4) kg/m 2 . The MHDE REE achieved 58.4% of values within ±10% from mREE; Cuppari REE achieved 47.8% and Vilar REE achieved 46.0% agreement. Reliability was good for the MHDE REE (ICC = 0.826) and Cuppari REE (ICC = 0.801), and moderate for the Vilar REE (ICC = 0.642) ( p < .001 for all). The equations performed poorly at the lowest and highest BMI categories. Conclusion Dialysis-specific energy equations showed variable accuracy. When categorized by BMI, the equations performed poorly at the extremes, where individuals are most vulnerable. Innovation is needed to understand these variances and correct the imprecision in PEEs for clinical practice. KEY MESSAGES Potentially impacting over millions of patients worldwide, our long-term goal is to understand energy expenditure (EE) across the spectrum of CKD (stages 1–5) in adults and children being treated with dialysis or transplantation, with the intent of providing tools for the health professional that will improve the delivery of quality care. Our research has identified and focussed on disease-specific factors which account for 60% of the variance in predicting EE in patients on MHD, but significant gaps remain. Thus, our central hypotheses are that (1) there are unique disease-specific determinants of EE and (2) prediction of EE for individuals diagnosed with CKD can be vastly improved with a model that combines these factors with more sophisticated approaches.

Purpose Approximately 700,000 people in the USA have chronic kidney disease requiring dialysis. Protein-energy wasting (PEW), a condition of advanced catabolism, contributes to three-year survival rates of 50%. PEW occurs at all levels of Body Mass Index (BMI) but is devastating for those people at the extremes. Treatment for PEW depends on an accurate understanding of energy expenditure. Previous research established that current methods of identifying PEW and assessing adequate treatments are imprecise. This includes disease-specific equations for estimated resting energy expenditure (eREE). In this study, we applied machine learning (ML) modelling techniques to a clinical database of dialysis patients. We assessed the precision of the ML algorithms relative to the best-performing traditional equation, the MHDE. Methods This was a secondary analysis of the Rutgers Nutrition and Kidney Database. To build the ML models we divided the population into test and validation sets. Eleven ML models were run and optimized, with the best three selected by the lowest root mean squared error (RMSE) from measured REE. Values for eREE were generated for each ML model and for the MHDE. We compared precision using Bland-Altman plots. Results Individuals were 41.4% female and 82.0% African American. The mean age was 56.4 ± 11.1 years, and the median BMI was 28.8 (IQR = 24.8 − 34.0) kg/m 2 . The best ML models were SVR, Linear Regression and Elastic net with RMSE of 103.6 kcal, 119.0 kcal and 121.1 kcal respectively. The SVR demonstrated the greatest precision, with 91.2% of values falling within acceptable limits. This compared to 47.1% for the MHDE. The models using non-linear techniques were precise across extremes of BMI. Conclusion ML improves precision in calculating eREE for dialysis patients, including those most vulnerable for PEW. Further development for clinical use is a priority.