Agreement between prediction equations and indirect calorimetry to estimate resting energy expenditure in elderly patients on hemodialysis

Machine learning models using non-linear techniques improve the prediction of resting energy expenditure in individuals receiving hemodialysis

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/m2. 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.

KDOQI Clinical Practice Guideline for Nutrition in CKD: 2020 Update

The National Kidney Foundation's Kidney Disease Outcomes Quality Initiative (KDOQI) has provided evidence-based guidelines for nutrition in kidney diseases since 1999. Since the publication of the first KDOQI nutrition guideline, there has been a great accumulation of new evidence regarding the management of nutritional aspects of kidney disease and sophistication in the guidelines process. The 2020 update to the KDOQI Clinical Practice Guideline for Nutrition in CKD was developed as a joint effort with the Academy of Nutrition and Dietetics (Academy). It provides comprehensive up-to-date information on the understanding and care of patients with chronic kidney disease (CKD), especially in terms of their metabolic and nutritional milieu for the practicing clinician and allied health care workers. The guideline was expanded to include not only patients with end-stage kidney disease or advanced CKD, but also patients with stages 1-5 CKD who are not receiving dialysis and patients with a functional kidney transplant. The updated guideline statements focus on 6 primary areas: nutritional assessment, medical nutrition therapy (MNT), dietary protein and energy intake, nutritional supplementation, micronutrients, and electrolytes. The guidelines primarily cover dietary management rather than all possible nutritional interventions. The evidence data and guideline statements were evaluated using Grading of Recommendations, Assessment, Development and Evaluation (GRADE) criteria. As applicable, each guideline statement is accompanied by rationale/background information, a detailed justification, monitoring and evaluation guidance, implementation considerations, special discussions, and recommendations for future research.

A Comparison of the Indirect Calorimetry and Different Energy Equations for the Determination of Resting Energy Expenditure of Patients With Renal Transplantation

OBJECTIVE

We aimed to evaluate the agreement between the resting energy expenditure (REE) obtained by indirect calorimetry and eight prediction equations in adult patients with renal transplantation and a newly developed REE prediction equation for use in patients with renal transplantation in the clinic.

METHODS

A total of 51 patients (30 males and 21 females) were involved in the study. The REE was measured by indirect calorimetry and compared with the previous prediction equations. The agreement was assessed by the interclass correlation coefficient and by Bland-Altman plot analysis.

RESULTS

No significant difference was found in terms of age and body mass index between the genders. Differences between the predicted and measured REEs were maximum in the Bernstein equation (-478 kcal) and minimum in the Cunningham equation (-69 kcal). It was found that underprediction values varied from 27.5% (chronic kidney disease equation) to 98.0% (Bernstein equation). The highest overprediction value was found in the Schofield equation (17.7%). The Cunningham equation and the new equation had the lowest root mean square error (265 kcal/day). In this study, fat-free mass (FFM) was found to be the most significant variable in multiple regression analysis (r²: 0.55). The new specific equation based on FFM was generated as 424.2 + 24.7∗FFM (kg). Besides that, it was found that the new equation and Cunningham equation were distributed randomly according to Bland-Altman analysis. A supplementary new equation based on available anthropometric measurements was developed as -1996.8 + 19.1∗height (cm) + 7.2∗body weight (kg).

CONCLUSION

This study showed that most of the predictive equations significantly underestimated REE. In patients with renal transplantation, if the REE is not measurable by indirect calorimetry, the use of the proposed equations will be more accurate.

Assessment of Metabolic and Nutritional Imbalance in Mechanically Ventilated Multiple Trauma Patients: From Molecular to Clinical Outcomes

The critically ill polytrauma patient is characterized by a series of metabolic changes induced by inflammation, oxidative stress, sepsis, and primary trauma, as well as associated secondary injuries associated. Metabolic and nutritional dysfunction in the critically ill patient is a complex series of imbalances of biochemical and genetic pathways, as well as the interconnection between them. Therefore, the equation changes in comparison to other critical patients or to healthy individuals, in which cases, mathematical equations can be successfully used to predict the energy requirements. Recent studies have shown that indirect calorimetry is one of the most accurate methods for determining the energy requirements in intubated and mechanically ventilated patients. Current research is oriented towards an individualized therapy depending on the energy consumption (kcal/day) of each patient that also takes into account the clinical dynamics. By using indirect calorimetry, one can measure, in real time, both oxygen consumption and carbon dioxide production. Energy requirements (kcal/day) and the respiratory quotient (RQ) can be determined in real time by integrating these dynamic parameters into electronic algorithms. In this manner, nutritional therapy becomes personalized and caters to the patients’ individual needs, helping patients receive the energy substrates they need at each clinically specific time of treatment.

Comparison of a Handheld Indirect Calorimetry Device and Predictive Energy Equations Among Individuals on Maintenance Hemodialysis

OBJECTIVES

Practical methods for determining resting energy expenditure (REE) among individuals on maintenance hemodialysis (MHD) are needed because of the limitations of indirect calorimetry. Two disease-specific predictive energy equations (PEEs) have been developed for this metabolically complex population. The aim of this study was to compare estimated REE (eREE) by PEEs to measured REE (mREE) with a handheld indirect calorimetry device (HICD).

METHODS

A prospective pilot study of adults on MHD (N = 40) was conducted at 2 dialysis clinics in Houston and Texas City, Texas. mREE by an HICD was compared with eREE determined by 6 PEEs using Bland-Altman analysis with a band of acceptable agreement of ±10% of the group mean mREE. Paired t-test and the intraclass correlation coefficient were also used to compare the alternate methods of measuring REE. A priori alpha was set at P < .05.

RESULTS

The mean (±standard deviation) age was 56.7 ± 12.9 years, 52.5% (n = 21) were female, and 85% (n = 34) were African American. Body mass index (BMI) ranged from 18.1 to 47.1 kg/m², 67.5% were overweight (BMI ≥25 kg/m²) and 50% were obese (BMI ≥30 kg/m²). The Maintenance Hemodialysis Equation-Creatinine version (MHCD-CR) was the most accurate PEE with 52.5% of values within the band of acceptable agreement, followed by the Mifflin-St. Jeor Equation and the Vilar et al. Equation at 45.0% and 42.5%, respectively.

CONCLUSION

When compared with mREE by the HICD, the MHDE-CR was more accurate and precise than other PEEs evaluated; however, this must be interpreted with caution as mREE was consistently lower than eREE from all PEEs. Further research is needed to validate the MHDE-CR and other practical methods for determining REE among individuals on MHD.