A retrospective study of resting energy expenditure in children hospitalized with different nutritional status

Introduction

Malnutrition in children can take three forms: undernutrition, overnutrition, and micronutrient abnormalities (1). The basis of nutritional intervention lies in ensuring the adequacy of energy intake. Resting energy expenditure (REE) refers to the energy consumption of the body in a resting state without skeletal muscle activity. It is the main part of total energy expenditure in children with low physical activity, and it may vary among children with different nutritional status (2). Indirect calorimetry (IC) which determines energy consumption by measuring pulmonary gas exchange, is a non-invasive technique and is considered the gold standard for determining individualized REE (3). Limited by a lack of equipment and trained personnel for performing the measurement, predictive energy equations based on age, weight, height, and sex are more often used clinically to estimate energy requirements (4). Most REE predictive energy equations were initially developed from healthy European and American populations, and their use across race or for populations with different health status may cause deviation (4). The predicted results may be higher or lower than the actual situation. Furthermore, many existing equations have been in use for years. Given the evolving dietary structure and lifestyle habits of today’s population, it remains unclear whether these equations adequately address REE for individuals residing in modern, affluent societies (5). It is unclear which energy predictive energy equation fits modern children with different nutritional status. In this study, we aimed to compare the measured REE (mREE) of IC in children with varying nutritional statuses, and to assess the accuracy of predicted REE (pREE) using five energy predictive equations published at different times. We present this article in accordance with the STROBE reporting checklist (available at https://tp.amegroups.com/article/view/10.21037/tp-24-168/rc).

Methods

General information

In this retrospective study, data from patients hospitalized in the internal medicine ward of Beijing Children’s Hospital, Capital Medical University, from May 2017 to December 2020, who underwent IC measurement (CCM Express, Med Graphics, USA) after nutrition consultation, were collected. Coefficient of variation is the mean variation rate of carbon dioxide generation (VCQ₂) and oxygen consumption (VO₂). The steady state is determined when the data points vary by no more than plus or minus 5% from the mean value of VCQ₂ and VO₂. Respiratory quotient (RQ) is the ratio of VCQ₂ and VO₂.

Only the results of the first indirect energy measurement were collected. Inclusion criteria included: (I) 3–18 years old; (II) spontaneous breathing with face tent; (III) coefficient of variation no more than plus or minus 5% (4); (IV) RQ ≤1. Exclusion criteria were as follows: (I) patients requiring mechanical ventilation; (II) patients with fever; (III) patients with diseases affecting gas exchange (asthma, severe lung disease, sleep apnea); (IV) patients with abnormal neurological diseases, hematologic tumor, severe hepatic and renal function abnormalities, and metabolic diseases such as diabetes or thyroid dysfunction.

This study was conducted in accordance with the Declaration of Helsinki (as revised in 2013) and approved by the Ethics Committee of Beijing Children’s Hospital (No. [2022]-E-220-R), which waived the requirement of informed consent due to the retrospective nature. The data were anonymized or maintained with confidentiality.

Study methods

Assessment of nutritional status

Nutritional deficiencies, excesses, or imbalances are all considered forms of malnutrition (6). Generally, “malnutrition” typically refers to undernutrition, while overnutrition is classified as “overweight” or “obesity” based on the assessment. We applied the Z-value scoring method with reference to the 2006 World Health Organization (WHO) Child Growth Standards, to diagnose and grade the malnutrition (6): a Z value of −2 to −1 in height-for-age, weight-for-age, or body mass index (BMI)-for-age is considered mild malnutrition, a Z value of −3 to −2 is considered moderate malnutrition, and a Z value less than −3 is considered severe malnutrition. Referring to the growth curve of BMI of children and adolescents aged 0–18 years in China, BMI-for-age ≥95% was assessed as obesity (7).

Indirect energy measurement

IC measurement was performed daily and maintained by a pediatrician from the Department of Clinical Nutrition. Preheating and calibration of volume and gas (calibration gas mixture 12% O₂, 5%CO₂ and 83% N₂) were performed before each measurement at a room temperature of 24–26 °C, humidity of 45–60%, and atmospheric pressure of 101–102.4 kPa. The patients were made to fast (food and water) for more than 2 h and lying flat quietly for more than 30 min (4). The face tent (canopy) mode, instead of the tight-fitting mask mode, of CCM was used for all patients. The canopy mode has low discomfort and high child acceptance. By reducing the patient’s awareness of breathing effort and avoiding the discomfort associated with direct contact with the nose and mouth, face the canopy mode provides optimal gas exchange measurements in spontaneously breathing patients. Parents accompanied their children through the testing process. The canopy and metabolic flow sensors were used for the measurement while the patients were awake. During canopy data collection, room air and expired air were drawn through the canopy collector by a fan. There should be no voluntary muscle activity during the test. After the patients reached a stable state (oxygen consumption and CO₂ exhalation per minute fluctuate no more than plus or minus 5% for 5 min), the measurement was continued for at least 10 min. Patients with muscle activity during the testing process may extend up to 15 minutes. The average value was taken as the final result.

Predictive energy equations

Current predictive energy equations for pediatric populations include Harris-Benedict (H-B) [1918] (8), Schofield [1985] (9), World Health Organization/Food and Agriculture Organization (WHO/FAO) [1985] (10), Mifflin [1990] (11), and Liu [2001] (12) (see Table 1).

Statistical analysis

SPSS 22.0 software (IBM, USA) was used for statistical processing and analysis. The Shapiro-Wilk test for normality and Levene’s test for the Chi-squared test were performed. Continuous normally distributed variables were presented as mean ± standard deviation (SD). Normal distribution and equal variance were compared using one-way analysis of variance (ANOVA). Comparison of measurements was performed using paired t-test. Kruskal-Wallis non-parametric test to compare measurements with abnormal distribution or unequal variance. The correlation of two values was determined by Pearson analysis. The threshold for significance was P<0.05 for all analyses. The relative error rate of the predictive energy equation was calculated using the equation: (pREE − mREE) / mREE × 100%; a value greater than 10% was considered as an overestimation, while less than −10% was considered as an underestimation, an error within ±10% was considered accurate prediction. Bland-Altman method was used for assessing agreement between mREE and pREE (13). When the 95% limits of agreement (bias ± 1.96 SD) are within the clinically acceptable range, it is considered that there is good agreement between the two. Clinically acceptable range is ±10% of the mean value of mREE.

Results

Sample description

We collected general information from 109 patients; there were 56 males and 53 females. Energy expenditure data is typically presented in KJ/min as kJ is describing gas exchange which is what is being measured. The IC measurement device displays results in kilocalories per day (kcal/d). For ease of comparison and standardization, the measured or formula-calculated REE results in this study are also expressed as kcal/d. Due to variations in patient weight, daily energy expenditure calculated per kilogram of body weight (mREE/kg) is represented as kilocalories per kilogram per day [kcal/(kg·d)]. The mREE in males was 1,396.73±401.60 kcal/d and 36.57±15.04 kcal/(kg·d), and the mREE in females was 1,011.83±374.77 kcal/d and 30.48±10.73 kcal/(kg·d). There were differences in mREE between different genders and mREE calculated per kilogram of body weight (P<0.001 and P=0.02). Grouped by nutritional status, there were 15 cases of mild malnutrition, 30 cases of moderate malnutrition, 32 cases of severe malnutrition, and 32 cases of obesity. There were no significant differences in age (F=1.190, P=0.32) and IC test coefficient of variation (F=0.670, P=0.57) between the groups (see Table 2).

Comparison of mREE in children with different nutritional status

A comparison of the mean values of mREE across the four nutritional status groups revealed significant differences (F=38.516, P<0.001), as did the mean values of mREE/kg (F=4.422, P=0.006) and the RQ (F=4.049, P=0.009). There was no significant difference in mREE and mREE/kg observed among the three malnutrition groups. However, significant differences were observed when each of the malnutrition groups compared to the obesity group (see Table 2).

Analysis of results obtained using IC and the predictive energy equation

The pREE calculated using equations were significantly correlated with the mREE across all populations (P<0.001). The highest prediction accuracy was 47.7%, which result from the Schofield equation (see Table 3).

Relative error rates were assessed for each of the four nutritional status groups in Table 4. The highest accuracy percentage in the mild malnutrition group was 46.7% (H-B and WHO/FAO equations), the moderate malnutrition group was 46.7% (Schofield and WHO/FAO equations), the severe malnutrition group was 34.4% (Liu equation), and the obesity group was 53.1% (H-B equation). There was no significant difference between the pREE using the five equations and the mREE using IC in the mild and moderate malnutrition groups (see Table 5). Only the predicted results using the Liu equation were not significantly different from IC in the severe malnutrition group (P=0.30). In the obese group, the predicted results of the H-B, Schofield, and WHO/FAO equations showed no significant difference from the IC (P=0.50, 0.21, 0.07, see Table 5). Further, on applying the Bland-Altman method to perform the agreement test, the 95% agreement limits of the difference between mREE and pREEs using each equation were greater than the clinically acceptable range and bias ± SD (see Table 6).

Discussion

Globally, approximately 45% of children die from malnutrition each year (1), while the obesity rate in children and related complications are increasing annually (4). Zhang et al. confirmed that REE levels in children correlate with their nutritional status (4), clarifying the importance of REE in optimizing nutritional therapy, and that overfeeding and underfeeding may lead to increased mortality (13). Energy provision based on IC measurement of mREE may lead to better clinical outcomes for patients (3,14).

In this study, we collected data from hospitalized children who underwent IC measurement after nutrition consultation, to assess the REE characteristics of children with different nutritional status. The study results showed a gradual decrease in the mean mREE from the mild, moderate, to severe malnutrition groups, but the difference between the groups was not statistically significant. Patients with obesity had higher mREE than patients with malnutrition and it differed significantly from the results of the remaining groups. The mREEs based on body weight in each malnutrition group were higher than in the obesity group, and the differences were statistically significant. García-Contreras et al. confirmed that the REE was significantly lower in children with moderate/severe malnutrition than in healthy children (15); however, the REE based on body weight was higher than that in healthy children. REE could significantly increase in children with malnutrition after the resumption of nutritional intake (15). Zapata et al. found the REE increased in the obese group compared to the overweight and healthy weight groups (16). Another study found that the REE based on body weight decreased (4). Therefore, maybe in children with malnutrition, especially severe malnutrition, the energy supply can be gradually increased from a low dose at the beginning of the nutritional intervention, and energy can be provided based on the REE using IC measurement if the conditions permit, to avoid the refeeding syndrome (17). In addition, when obese children attempting weight loss, avoiding excessive limitation of energy intake is crucial to ensure adequate growth and development.

When IC is not an option, the predictive energy equation is a useful tool for predicting REE. However, one study found that the energy equation can have an inaccuracy rate of 60% (13). The H-B equation, first proposed in 1918, has been demonstrated inaccuracy in multiple studies (4,11,18), even worse predictive accuracy in group with too low BMI (accurate rate was 39.3%) (18). In 1990, Mifflin modified the H-B equation to improve its applicability (11). The WHO/FAO and Schofield equations were published in 1985 and are now commonly used predictive energy equations for hospitalized children. The Liu equation is a predictive energy equation published in 2001 specifically for the Chinese population. The applicability of these equations to REE estimation for modern populations is unclear (19). The majority of predictive energy equations fail to accurately predict a patient’s energy needs clinically, with the best being accurate to within 70% of the true value (20). Thus, IC is still the most reliable method for measuring energy consumption and guiding energy provision (4,13,18,20). Bland-Altman method is now considered the standard approach for assessment of agreement between measurements (21). In this study, five commonly used REE predictive energy equations for children were validated using the Bland-Altman method, with IC results as the criterion; the result showed that 95% agreement limits for the differences between mREE and the corresponding pREE by each equation were all greater than clinically acceptable range, suggesting that the results of the predictive energy equations have deviations and they cannot be directly adopted. The pREE using each predictive energy equation correlated well with the mREE. One study confirmed that whilst it is not wrong to report a correlation, and indeed can be helpful, it is not sufficient for describing how well two measures agree (21). There was no significant difference between each pREE using the five equations and mREE using the IC measurement in the mild, moderate malnutrition groups. In severe malnutrition group, only the Liu equation had a relatively high accuracy, leading to a pREE not differing significantly from the mREE. In the overall population (see Table 3), the Liu equation had predominantly underestimated predictions (44%). The REE is relatively lower in severely-malnourished children probably because apparent body fat loss, skeletal muscle depletion, and atrophy of internal organs (1). Although the predictive energy equations have clear limitations, they are commonly adopted clinically, and specific equations can be suitable for specific patient groups (20). However, the ultimate goal should be to promote IC through technological innovation, easy operation, and cost reduction. Only when the nutritional assessment and prescription results accurately reflect patients’ nutritional status can they obtain more clinical benefits (3).

There are some limitations in this study due to the retrospective nature. IC was only performed during nutritional consultations, possibly leading to selection bias. The correlation analysis did not consider the impact of pubertal development on REE. Furthermore, the study had a limited sample size and did not represent all Chinese children. The differences in REE between patients and subgroups caused by the small sample size, hindered the development of new equations. Some studies have concluded that body composition (mainly fat-free mass) significantly affects REE (22). In this study, we did not analyze the impact of body composition measurements on the accuracy of the prediction equation due to incomplete data on body composition. It is interesting that the European Society for Clinical Nutrition and Metabolism (ESPEN) expert group confirmed that the addition of body composition measurements did not add to the accuracy of predictive equations whatever the clinical conditions (overweight, underweight, chronic diseases) (20). More research is needed to confirm the relationship of REE and body composition.

Conclusions

In conclusion, we found differences in the REE of children with different nutritional status, as well as bias between the results of each energy predictive energy equation and IC. When REE cannot be measured by IC, it is essential to choose an appropriate predictive energy equation based on the nutritional status of the individual. Severely malnourished patients are advised to choose the Liu equation, obese patients can choose the H-B equation, and mildly to moderately malnourished patients can opt for the WHO/FAO equation.

Acknowledgments

Funding: None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tp.amegroups.com/article/view/10.21037/tp-24-168/rc

Data Sharing Statement: Available at https://tp.amegroups.com/article/view/10.21037/tp-24-168/dss

Peer Review File: Available at https://tp.amegroups.com/article/view/10.21037/tp-24-168/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tp.amegroups.com/article/view/10.21037/tp-24-168/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki (as revised in 2013) and approved by the Ethics Committee of Beijing Children’s Hospital (No. [2022]-E-220-R), which waived the requirement of informed consent due to the retrospective nature.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Becker P, Carney LN, Corkins MR, et al. Consensus statement of the Academy of Nutrition and Dietetics/American Society for Parenteral and Enteral Nutrition: indicators recommended for the identification and documentation of pediatric malnutrition (undernutrition). Nutr Clin Pract 2015;30:147-61. [Crossref] [PubMed]
2. Afroze F, Khoshnevisan F, Harawa PP, et al. Trajectories of resting energy expenditure and performance of predictive equations in children hospitalized with an acute illness and malnutrition: a longitudinal study. Sci Rep 2024;14:3613. [Crossref] [PubMed]
3. Oshima T, Berger MM, De Waele E, et al. Indirect calorimetry in nutritional therapy. A position paper by the ICALIC study group. Clin Nutr 2017;36:651-62. [Crossref] [PubMed]
4. Zhang L, Chen R, Li R, et al. Evaluating the predictive factors of resting energy expenditure and validating predictive equations for Chinese obese children. World J Pediatr 2018;14:160-7. [Crossref] [PubMed]
5. Zhou W, Su H, Tong J, et al. Multiple factor assessment for determining resting metabolic rate in young adults. Sci Rep 2024;14:11821. [Crossref] [PubMed]
6. Bouma S. Diagnosing Pediatric Malnutrition. Nutr Clin Pract 2017;32:52-67. [Crossref] [PubMed]
7. Subspecialty Group of Endocrinologic, Hereditary and Metabolic Diseases, the Society of Pediatrics, Chinese Medical Association. Expert consensus on diagnosis, assessment, and management of obesity in Chinese children. Zhonghua Er Ke Za Zhi 2022;60:507-15. [Crossref] [PubMed]
8. Harris JA, Benedict FG. A Biometric Study of Human Basal Metabolism. Proc Natl Acad Sci U S A 1918;4:370-3. [Crossref] [PubMed]
9. Schofield WN. Predicting basal metabolic rate, new standards and review of previous work. Hum Nutr Clin Nutr 1985;39:5-41.
10. Energy and protein requirements. Report of a joint FAO/WHO/UNU Expert Consultation. World Health Organ Tech Rep Ser 1985;724:1-206.
11. Mifflin MD, St Jeor ST, Hill LA, et al. A new predictive equation for resting energy expenditure in healthy individuals. Am J Clin Nutr 1990;51:241-7. [Crossref] [PubMed]
12. Liu B, Woo J, Tang N, et al. Assessment of total energy expenditure in a Chinese population by a physical activity questionnaire: examination of validity. Int J Food Sci Nutr 2001;52:269-82. [Crossref] [PubMed]
13. Zusman O, Kagan I, Bendavid I, et al. Predictive equations versus measured energy expenditure by indirect calorimetry: A retrospective validation. Clin Nutr 2019;38:1206-10. [Crossref] [PubMed]
14. Wischmeyer PE, Bear DE, Berger MM, et al. Personalized nutrition therapy in critical care: 10 expert recommendations. Crit Care 2023;27:261. [Crossref] [PubMed]
15. García-Contreras AA, Vásquez-Garibay EM, Romero-Velarde E, et al. Energy expenditure in children with cerebral palsy and moderate / severe malnutrition during nutritional recovery. Nutr Hosp 2015;31:2062-9. [Crossref] [PubMed]
16. Zapata JK, Catalán V, Rodríguez A, et al. Resting Energy Expenditure Is Not Altered in Children and Adolescents with Obesity. Effect of Age and Gender and Association with Serum Leptin Levels. Nutrients 2021;13:1216. [Crossref] [PubMed]
17. Yang WL, Yan J, Xu XW, et al. Individualized nutrition support for anorexia nervosa children under metabolic monitoring: a case report. Parenteral & Enteral Nutrition 2017;24:127-8.
18. Jésus P, Achamrah N, Grigioni S, et al. Validity of predictive equations for resting energy expenditure according to the body mass index in a population of 1726 patients followed in a Nutrition Unit. Clin Nutr 2015;34:529-35. [Crossref] [PubMed]
19. Heymsfield SB, Smith B, Dahle J, et al. Resting Energy Expenditure: From Cellular to Whole-Body Level, a Mechanistic Historical Perspective. Obesity (Silver Spring) 2021;29:500-11. [Crossref] [PubMed]
20. Bendavid I, Lobo DN, Barazzoni R, et al. The centenary of the Harris-Benedict equations: How to assess energy requirements best? Recommendations from the ESPEN expert group. Clin Nutr 2021;40:690-701. [Crossref] [PubMed]
21. Mansournia MA, Waters R, Nazemipour M, et al. Bland-Altman methods for comparing methods of measurement and response to criticisms. Glob Epidemiol 2021;3:100045. [Crossref] [PubMed]
22. Achamrah N, Delsoglio M, De Waele E, et al. Indirect calorimetry: The 6 main issues. Clin Nutr 2021;40:4-14. [Crossref] [PubMed]