- Email: [email protected]
Resting energy expenditure and prediction equations in young children with failure to thrive
R
Resting energy expenditure and prediction equations
in young children with failure to thrive Timothy A. Sentongo, MD, Andrew M. Tershakovec, MD, Maria R. Mascarenhas, MD, Miriam H. Watson, MS, and Virginia A. Stallings, MD
Objective: To compare predicted and measured resting energy expenditure (REE) in young children (birth to 3 years) with failure to thrive (FTT). Methods: REE (kcal/d) was measured by indirect calorimetry and compared with predicted REE from 3 sex and age group equations: World Health Organization (WHO), Schofield weight-based (SCH-WT), and Schofield weight- and height-based (SCH-WT-HT). The clinical characteristics associated with inaccuracy of predicted REE were examined. Results: Forty-five subjects (47% female) were evaluated. Their clinical characteristics (mean ± SD) included age 1.2 ± 0.7 years, length/height z score –2.1 ± 1.3, weight z score –2.7 ± 1.0, and measured REE 438 ± 111 kcal/d. All prediction equations were within 10% accuracy <50% of the time. However, SCH-WT-HT did not significantly differ from measured REE (450 ± 138 vs 438 ± 111 kcal/d, P = .2) and was least likely to underestimate REE. Younger age and more severe growth failure (based on weight, length/height, or both) were associated with underestimation of REE by prediction equations. Conclusion: REE should be measured in young infants and children with moderate to severe FTT when knowledge of caloric needs is required for optimal clinical care. The SCH-WT-HT equation was least likely to underestimate REE and is therefore preferred when REE cannot be measured in this group of children. (J Pediatr 2000;136:345-50)
Regardless of the cause of failure to thrive, effective nutritional management partly consists of providing adequate calories to achieve positive energy balance and growth. The World Health Organization Expert Consultation on Energy and Protein Requirements
(1981)1 recommended that “wherever possible, energy requirements should be based on measurement of expenditure rather than intake.” Resting energy expenditure is fairly stable in healthy children between birth and 3 years2 and is the largest single con-
From the Division of Gastroenterology and Nutrition, The Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania. Supported in part by the Nutrition Center at The Children’s Hospital of Philadelphia. Submitted for publication Mar 8, 1999; revision received Aug 30, 1999; accepted Oct 13, 1999. Reprint requests: Timothy A. Sentongo, MD, Division of Gastroenterology and Nutrition, The Children’s Hospital of Philadelphia, 34th and Civic Center Blvd, Philadelphia, PA 19104. Copyright © 2000 by Mosby, Inc. 0022-3476/2000/$12.00 + 0 9/21/103852 doi:10.1067/mpd.2000.103852
tributor to total energy expenditure and thus to total energy requirements.1-4 Total energy requirements, which is the clinically required information, may then be estimated as multiples of REE.1,5 FFM FTT HAZ REE SCH-WT
Fat-free mass Failure to thrive Height for age z score Resting energy expenditure, kcal/d Schofield weight-based resting energy expenditure prediction equation SCH-WT-HT Schofield weight- and heightbased resting energy expenditure prediction equation TOBEC Total body electrical conductance WAZ Weight for age z score WHO World Health Organization weight-based resting energy expenditure prediction equation
The WHO, Schofield weight-based resting energy expenditure prediction equation, and Schofield weight- and height-based resting energy expenditure prediction equation are REE prediction equations that were derived from pooled data of measured REE in healthy populations.1,6 REE is largely dependent on the fat-free mass, which in turn is composed of tissues and organs with different rates of cellular metabolism.7-9 In clinical situations with altered body composition such as FTT, the applicability of prediction equations may be limited. Our previous report showed that the SCH-WT-HT equation predicted REE better than WHO, SCH-WT, and Harris-Benedict equations in children over a broad age range (age 0.2 to 20.5 years) and especially in those with FTT or obesity.10 345
SENTONGO ET AL
The aim of this study was to compare measured and predicted REE with the WHO, SCH-WT, and SCH-WT-HT equations in a nutritionally at-risk group of young children under the age of 3 years with the clinical diagnosis of FTT. The pattern of accuracy between measured and predicted REE and the clinical characteristics that influenced the accuracy of predicted REE in this important clinical population were also evaluated.
METHODS Subjects aged birth to 3 years with a clinical diagnosis of FTT who were referred between 1993 and 1997 for REE measurement as part of clinical management were considered for enrollment into the study. Once referred, a standard protocol including growth assessment, measurement of REE, and body composition was followed. Weight was measured with a digital electronic stand-on scale or infant scale (ScaleTronix Inc, Wheaton, IL) to 0.1 and 0.01 kg, respectively. Recumbent length was measured to 0.1 cm with an infant length board (Holtain, Crymych, UK) in all subjects <2 years of age. Standing height was measured to 0.1 cm with a stadiometer (Holtain, Crymych, UK) in those aged 2 to 3 years. The measurements were made in triplicate, and the mean was used in analysis.11 After the REE assessment was done, body composition was estimated with total body electrical conductance (EmScan, Springfield, IL), and the results were reported as FFM in kilograms and percentage body fat. The subjects were swaddled and then laid supine for passage through the electromagnetic field produced by an oscillating radio frequency current of the large solenoid coil.12 Ten scans were performed on each subject, and the average phase value was calculated. The average duration of the study for each subject was 10 minutes. No additional sedation was administered for the TOBEC test. 346
THE JOURNAL OF PEDIATRICS MARCH 2000 REE was measured by open-circuit indirect calorimetry with a computerized metabolic cart (SensorMedic 2900Z, SensorMedics Corp, Yorba Linda, CA). The measurement was carried out for 30 to 60 minutes between 7:30 and 10:00 AM after an age-appropriate fast ranging from 4 to 8 hours. The hospitalized subjects were transported from their hospital bed by stretcher. The outpatient subjects were admitted to the laboratory and rested in a supine position for a minimum of 30 minutes before the REE was determined. The REE procedure took place in a quiet thermo-neutral room while the child rested in a supine position. Chloral hydrate at a dose of 65 mg/kg was used for sedation. A clear ventilated hood was placed over the subject’s head for sampling respiratory gases at 1-minute intervals. The subjects were observed closely for movement by 1 of 2 trained observers. Measurements of REE obtained immediately after periods of movement or coughing, which altered REE, were excluded. The first 10 minutes of each study were also excluded to account for environmental adjustment by the child. The 1-minute REE measurements from up to 45 minutes (minimum of 15 minutes) were averaged and used for overall REE measurements. The modified Weir equation was used to calculate energy equivalency from the oxygen consumption and carbon dioxide production.13
Statistical Analysis SD scores (z scores) for length/ height, weight, and weight-for height were computed with Centers for Disease Control Anthropometric Software Program (version 3.1, 1988, Division of Nutrition, Centers for Disease Control and Prevention, Atlanta, GA).14 The percentage of measured REE was calculated as predicted REE, divided by measured REE, and multiplied by 100. The predicted REE was considered accurate for the purpose of this analysis if it was within 10% of the measured REE (90% to 110% measured REE).15
A percent measured REE value <90% was classified underestimated by the prediction equations, and likewise a percent measured REE of >110% was classified as overestimated REE. Paired t tests were used to compare measured REE and predicted REE by the WHO, SCH-WT, and SCH-WTHT equations. To determine the equation with the least probability for underestimating REE, the frequency of underestimated REE by the different equations was compared with the Pearson chi-squared test. Analysis of variance was used to evaluate the association between clinical characteristics and accuracy of predicted REE. The association between REE as kilocalories per kilograms per day and WAZ was examined with Pearson correlation. Because the WHO equation is most commonly used in clinical practice, a subsequent analysis of the results was completed with this prediction equation. Subjects with body composition data available were subdivided into accurately predicted (group 1) and underestimated REE (group 2) by the WHO equation. The relationship between REE and FFM in the 2 groups of subjects was examined with Pearson correlation. The association between the clinical variables and REE was analyzed with analysis of variance, with REE as the dependent variable and FFM, age, HAZ, WAZ, and weight for height z score as the independent variables. A P value <.05 was defined as statistically significant. Statistical analyses were performed with SYSTAT software (version 5.2., SYSTAT, Inc).
RESULTS Forty-five subjects were enrolled (47% female)(Table I). All subjects had a clinical diagnosis of FTT, and only 11% had no other secondary clinical diagnosis; 64% of the subjects had more than 2 secondary diagnoses in addition to the FTT. Gastrointestinal disorders including gastroesophageal
SENTONGO ET AL
THE JOURNAL OF PEDIATRICS VOLUME 136, NUMBER 3 reflux, malabsorption, and allergic enteropathy formed most (61%) of the secondary clinical diagnoses. Other clinical diagnoses included neurologic (18%), cardiac (7%), and pulmonary disorders (4%). No subject was acutely ill, febrile, or had a disorder or medication known to directly or indirectly alter energy metabolism such as thyroid disease or β-adrenergic bronchodilators. Two subjects did not receive sedation with chloral hydrate because of unacceptable risks associated with the underlying medical conditions. One subject (0.63-year-old girl, WAZ –4.3 and HAZ –4.9) received an intravenous dextrose infusion (7.2 mg/kg/min) during the REE test because of significant risk for hypoglycemia. Her enteral feeds, which provided 63% of the daily caloric intake, were discontinued 6 hours before REE was measured. Her respiratory quotient did not differ from that of the other subjects, who underwent complete fasting before the study. Excluding her did not change the mean WAZ, HAZ, and REE (440 ± 111 vs 438 ± 111 kcal/d) of the group; therefore she was kept in the analysis. Body composition assessment by TOBEC was performed only after the REE measurement was completed. No additional sedation with choral hydrate was given for the TOBEC measurement. Therefore only subjects who remained sedated after the REE test and were cooperative enough for technically good TOBEC studies had their body composition assessment completed (32 subjects, 50% female) (Table I). The predicted REE by SCH-WTHT did not significantly differ from the measured REE (Table II). However, the predicted REE by WHO and SCH-WT significantly differed from measured REE. Pearson chi-squared test showed that of the 3 equations, the SCH-WT-HT underestimated REE significantly less frequently in this study population (P = .01). Younger age and worse growth failure, as indicated by lower WAZ and HAZ, were
Table I. Clinical characteristics of subjects in the study (mean ± SD)
Male patients (24)
Female patients (21)
1.2 ± 0.7 7.3 ± 2.5 –2.7 ± 1.2 71.6 ± 9.7 –2.04 ± 1.2 –1.9 ± 0.8
1.3 ± 0.7 7.2 ± 2.1 –2.6 ± 1.0 70.0 ± 9.1 –2.1 ± 1.3 1.5 ± 0.7
Male patients (16)
Female patients (16)
1.2 ± 0.7 5.8 ± 1.6 –2.9 ± 1.0 –2.2 ± 1.3 –1.9 ± 0.9 18.1 ± 7.9
1.3 ± 0.7 5.6 ± 1.1 –2.5 ± 0.9 –2.0 ± 1.2 –1.5 ± 0.7 20.6 ± 6.0
Age (y) Weight (kg) WAZ Length/height (cm) HAZ WHZ
Subgroup with body composition assessed Age (y) FFM (kg) WAZ HAZ WHZ %Body fat WHZ, Weight for height z score.
Table II. Comparison of measured and predicted REE (mean ± SD) by the WHO, SCH-WT, and SCH-WT-HT prediction equations
Predicted REE
Measured REE REE (kcal/d) REE (kcal/kg/d) P value* %Measured of REE
438 ± 111 62.0 ± 9.5
WHO
SCH-WT
SCH-WT-HT
391 ± 140 52.9 ± 3.2 .0001 87.5 ± 16.0
398 ± 136 54.2 ± 1.9 .0001 89.6 ± 15.0
450 ± 138 62.3 ± 5.4 .2 102.3 ± 15.2
*Refers to comparison between measured and predicted REE.
Table III. Percentages of subjects in the different predicted REE categories and clinical characteristics associated with underestimated REE
Percent of subjects REE category
WHO
SCH-WT
SCH-WT-HT
58 38 4
51 42 7
24 44 32
Underestimated REE* Accurate REE† Overestimated REE‡
P value Characteristics significantly associated with underestimated REE Age WAZ HAZ WHZ
.02 .0001 .001 .2
.006 .0001 .02 .07
.1 .06 .01 1.00
WHZ, Weight for height z score. *Predicted REE <90% of measured value. †Predicted REE = 90% to 110% of measured value. ‡Predicted REE >110% of measured value. Significance = P value < .05.
347
SENTONGO ET AL
THE JOURNAL OF PEDIATRICS MARCH 2000 with underestimated REE were younger (1.0 ± 0.5 vs 1.6 ± 0.7 years, P = .006) and had more severe growth failure (HAZ –2.7 ± 1.2 vs –1.29 ± 0.8, P = .04; WAZ –3.2 ± 0.9 vs –1.9 ± 0.7, P = .02). Analysis of variance with REE as the dependent variable and FFM, group (accurate REE vs underestimated REE), age, WAZ, and HAZ as independent variables showed that the interaction of group and FFM were significant determinants of REE. This finding suggests that the relationship between REE and FFM differed in the 2 groups.
DISCUSSION Fig 1. Correlation between REE (kilocalories per kilograms per day) and WAZ in 45 children with FTT.
Fig 2. Association between measured REE (kilocalories per day) and FFM in 2 groups of children. Group 1 (closed circles): predicted REE by WHO ≥90% of measured REE. Group 2 (open circles): predicted REE by WHO <90% measured REE (ie, underestimated REE by WHO). Group 2 was younger and had lower HAZ and WAZ than group 1 (P < .001 for all comparisons). significantly associated with underestimation of REE by both WHO and SCH-WT equations, whereas only low HAZ resulted in this association with the SCH-WT-HT equation (Table III). A negative correlation was seen between WAZ and measured REE (kcal/kg/day r = –0.716, P < .0001 (Fig 1). 348
Measured REE positively correlated with FFM in the subgroup of children who had body composition data available. For the same FFM the subjects with underestimated predicted REE by WHO had a higher measured REE than those with accurately predicted REE by WHO (Fig 2). The subjects
In this sample of young children with moderate to severe FTT, all prediction equations were within 10% less than half the time. However, SCH-WT-HT compared best with measured REE and underestimated REE less frequently. Being younger (for WHO and SCH-WT), having more severe growth failure (based on WAZ, HAZ, or both), or both were associated with underestimation of energy requirements by prediction equations. The infants and children in this study had poor growth compared with the reference population, which was consistent with the clinical diagnosis of FTT. These findings confirm the inadequacy of prediction equations in altered growth and body composition10 and more specifically demonstrate that younger age even within the range of 0 to 3 years, and moderate to severe growth failure, are indicators that prediction equations are more likely to underestimate REE, thereby making this approach undesirable in clinical practice. Because nutritional intervention is usually central in the treatment of children with FTT, accurate estimation of caloric needs is necessary. If the goals of clinical nutritional management are to achieve a positive energy balance and growth as quickly as is safely possible, it is better to overestimate, not
SENTONGO ET AL
THE JOURNAL OF PEDIATRICS VOLUME 136, NUMBER 3 underestimate, caloric requirements. Therefore assuming that the greater clinical risk for inaccuracy is in underestimating REE, the SCH-WT-HT equation was superior to the WHO and SCH-WT equations for planning care of young infants and children with moderate to severe FTT. The better REE prediction by SCH-WT-HT, which uses weight and height rather than weight alone in the calculation, suggests that body composition may be an important factor contributing to inaccuracy of the prediction equations. The WHO prediction equations published in 1985 were derived from data obtained from pooled REE measurements in >6000 healthy individuals.1,6 They are considered the best estimate available for prediction of REE of healthy people in any population1; however, inaccuracies in very underweight and overweight individuals have been reported.1,6,7 The Schofield prediction equations were subsequently derived with a data pool approximately twice the size of and overlapping by half the WHO data source. They included a few measurements from ill children and also measurements from >3800 subjects who were used as a secondary source for validating the derived equations.6 All the data sources were screened to ensure that appropriate experimental conditions had been followed and that complete data on age, sex, weight, height, and basal metabolic rate had been included.6 The larger database improved REE prediction from weight alone, and the inclusion of height into the equation specifically improved REE prediction in children <3 years old and the elderly aged 60 years and over.6 Of note, the age groups of birth to 3 years, 3 to 10 years, 10 to 18 years, 18 to 30 years, and so on were based on the commonly used clinical divisions of life span and not from patterns observed in the REE data. Because all 3 prediction equations are based on data obtained from healthy populations, they are best used for predicting energy needs in healthy children.1,6
Malnourished children, especially those with low body fat stores, may have an apparently increased rate of metabolism per unit body weight because of altered body composition. This may increase further during the initial phases of nutritional management in children with FTT,16 as REE increases and the child becomes anabolic. We were able to demonstrate that there was a different relationship between REE and FFM that contributed to the underestimated predicted REE by the WHO equation. Younger children with more severe FTT had a higher REE per FFM than older children with less severe FTT. Therefore based on the WHO equation alone, these children may have been considered hypermetabolic. However, much of the difference disappeared when we used the SCHWT-HT equation to predict REE. The apparent increase in REE per unit of body weight (total weight and FFM) during moderate to severe malnutrition may be due to the state of altered body composition. FFM is composed of tissues and organs with various degrees of metabolic activity. Organs are made up of high-energy-using cell masses (brain, liver, heart, and kidney) and account for >60% of REE, whereas muscle tissue accounts for approximately 30% of REE.7,8 In situations of altered body composition such as occurs in moderate to severe FTT, where there is loss of body fat and muscle mass, a greater proportion of FFM is composed of high metabolic rate organs such as brain, liver, kidney, and heart and a correspondingly reduced proportion of lower metabolic rate tissues such as muscle. Therefore the high metabolic rate tissue comprises a greater proportion of body weight, which leads to an increased REE per unit of body weight, whereas in terms of expected weight (for age), the REE tends to be low.16,17 Our observations support the hypothesis that the FFM of young children with FTT has a larger proportion of high metabolic rate tissue, and therefore the energy expenditure per unit body size is
increased. This finding may partly explain why standard prediction equations with only weight in the calculation may underestimate REE more frequently in young children with moderate to severe FTT. The potential limitation of this study was that timing of referral for measuring REE was not standardized in the subjects; therefore the subjects may have been at different phases of metabolic adaptation to malnutrition. Being anabolic after nutritional intervention may also explain increased REE.16 By history, however, none of the subjects were being successfully renourished; therefore this should not have altered the results. When REE cannot be measured in young infants and children with FTT, the SCH-WT-HT is the preferred equation. In addition, our findings are in agreement with the 1981 WHO1 recommendation that whenever possible, energy requirements should be measured.
REFERENCES 1. World Health Organization. Energy and protein requirements. Report of a joint FAO/WHO/UNU Expert Consultation. (WHO Technical Report Series No. 724). Geneva: World Health Organization; 1985. 2. Waterlow JC. Basic concepts in the determination of nutritional requirements of normal infants. In: Tsang RC, Nichols BL, editors. Nutrition during infancy. Philadelphia: Hanley & Belfus, Inc. 1988. p. 1-19. 3. Wells JCK, Davies PSW. The components of energy metabolism in 12 week old infants [abstract]. Ann Hum Biol 1995;22:175. 4. Pencharz PB, Azcue MP. Measuring resting energy expenditure in clinical practice. J Pediatr 1995;127;269-71. 5. National Academy of Sciences. Recommended Dietary Allowances. Washington, DC: National Academy Press; 1989. p. 24-38. 6. Schofield WN, Schofield C, James WPT. Basal metabolic rate—review and prediction, together with annotated bibliography of source material. Hum Nutr: Clin Nutr 1985;39C(Suppl 1):5-41. 349
SENTONGO ET AL
7. Gallagher D, Belmonte D, Deurenberg P, Wang Z, Krasnow N, Pi-Sunyer FX, et al. Organ-tissue mass measurement allows modeling of REE and metabolically active tissue mass. Am J Physiol 1998;275:E249-58. 8. Butte NF, Moon JK, Wong WW, Hopkinson JM, Smith EO. Energy requirements from infancy to adulthood. Am J Clin Nutr 1995;62:1047-52S. 9. Holliday MA. Metabolic rate and organ size during growth from infancy to maturity and during late gestation and early infancy. Pediatrics 1971;47: 169-79. 10. Kaplan AS, Zemel BS, Neiswender KM, Stallings VA. Resting energy expenditure in clinical pediatrics: measured versus prediction equations. J Pediatr 1995;127:200-5. 11. Cameron N. Methods of auxological anthropometry. Hum Growth 1986;3: 3-46. 12. Harrison GG, Van Itallie TB. Estimation of body composition: a new approach based on electromagnetic principles. Am J Clin Nutr 1982;35:1176-9.
THE JOURNAL OF PEDIATRICS MARCH 2000 13. Weir J. New methods for calculating metabolic rate with special reference to protein metabolism. J Physiol 1949; 109:1-9. 14. Jordan MD. CDC Anthropometric Software Program. Based on Centers for Disease Control and Prevention standard deviation-derived growth reference curves for U.S. children, birth to 18 years. Derived from NCHS/ CDC Reference Population 1988. 15. Firouzbakhsh S, Mathis RK, Dorchester WL, Oseas RS, et al. Measuring resting energy expenditure in children. J Pediatr Gastoenterol Nutr 1993;16: 136-42. 16. Montgomery RD. Changes in the basal metabolic rate of the malnourished infant and their relation to body composition. J Clin Invest 1962;41: 1653-63. 17. Holliday MA, Potter D, Jarrah A, Bearg S. The relationship of metabolic rate to body weight and organ size. Pediatr Res 1967;1:185-95. 18. Stallings VA. Resting energy expenditure. In: Altschuler S, Liacouras C, ed-
itors. Clinical pediatric gastroenterology. Philadelphia: WB Saunders; 1998. p. 606-11.
APPENDIX Schofield equations for predicting resting energy expenditure from body weight (W, kg), height (H, m), sex, and age group5,18 Age range (y) Males 0–3 3–10 10–18 18–30 Females 0–3 3–10 10–18 18–30
kcal/d
0.167W + 1517.4H – 617.6 19.59W + 130.3H + 414.9 16.25W + 137.2H + 515.5 15.057W + 10.04H + 705.8 16.252W + 1023.2H – 413.5 16.969W + 161.8H + 371.2 8.365W + 465H + 200.0 13.623W + 283.0H + 98.2
Receive tables of contents by e-mail To receive the tables of contents by e-mail, sign up through our Web site at http://www.mosby.com/jpeds Choose E-mail Notification. Simply type your e-mail address in the box and click the Subscribe button. Alternatively, you may send an e-mail message to [email protected]. Leave the subject line blank and type the following as the body of your message: subscribe jpeds_toc You will receive an e-mail to confirm that you have been added to the mailing list. Note that table of contents e-mails will be sent out when a new issue is posted to the Web site.
350
Resting energy expenditure and prediction equations
in young children with failure to thrive Timothy A. Sentongo, MD, Andrew M. Tershakovec, MD, Maria R. Mascarenhas, MD, Miriam H. Watson, MS, and Virginia A. Stallings, MD
Objective: To compare predicted and measured resting energy expenditure (REE) in young children (birth to 3 years) with failure to thrive (FTT). Methods: REE (kcal/d) was measured by indirect calorimetry and compared with predicted REE from 3 sex and age group equations: World Health Organization (WHO), Schofield weight-based (SCH-WT), and Schofield weight- and height-based (SCH-WT-HT). The clinical characteristics associated with inaccuracy of predicted REE were examined. Results: Forty-five subjects (47% female) were evaluated. Their clinical characteristics (mean ± SD) included age 1.2 ± 0.7 years, length/height z score –2.1 ± 1.3, weight z score –2.7 ± 1.0, and measured REE 438 ± 111 kcal/d. All prediction equations were within 10% accuracy <50% of the time. However, SCH-WT-HT did not significantly differ from measured REE (450 ± 138 vs 438 ± 111 kcal/d, P = .2) and was least likely to underestimate REE. Younger age and more severe growth failure (based on weight, length/height, or both) were associated with underestimation of REE by prediction equations. Conclusion: REE should be measured in young infants and children with moderate to severe FTT when knowledge of caloric needs is required for optimal clinical care. The SCH-WT-HT equation was least likely to underestimate REE and is therefore preferred when REE cannot be measured in this group of children. (J Pediatr 2000;136:345-50)
Regardless of the cause of failure to thrive, effective nutritional management partly consists of providing adequate calories to achieve positive energy balance and growth. The World Health Organization Expert Consultation on Energy and Protein Requirements
(1981)1 recommended that “wherever possible, energy requirements should be based on measurement of expenditure rather than intake.” Resting energy expenditure is fairly stable in healthy children between birth and 3 years2 and is the largest single con-
From the Division of Gastroenterology and Nutrition, The Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania. Supported in part by the Nutrition Center at The Children’s Hospital of Philadelphia. Submitted for publication Mar 8, 1999; revision received Aug 30, 1999; accepted Oct 13, 1999. Reprint requests: Timothy A. Sentongo, MD, Division of Gastroenterology and Nutrition, The Children’s Hospital of Philadelphia, 34th and Civic Center Blvd, Philadelphia, PA 19104. Copyright © 2000 by Mosby, Inc. 0022-3476/2000/$12.00 + 0 9/21/103852 doi:10.1067/mpd.2000.103852
tributor to total energy expenditure and thus to total energy requirements.1-4 Total energy requirements, which is the clinically required information, may then be estimated as multiples of REE.1,5 FFM FTT HAZ REE SCH-WT
Fat-free mass Failure to thrive Height for age z score Resting energy expenditure, kcal/d Schofield weight-based resting energy expenditure prediction equation SCH-WT-HT Schofield weight- and heightbased resting energy expenditure prediction equation TOBEC Total body electrical conductance WAZ Weight for age z score WHO World Health Organization weight-based resting energy expenditure prediction equation
The WHO, Schofield weight-based resting energy expenditure prediction equation, and Schofield weight- and height-based resting energy expenditure prediction equation are REE prediction equations that were derived from pooled data of measured REE in healthy populations.1,6 REE is largely dependent on the fat-free mass, which in turn is composed of tissues and organs with different rates of cellular metabolism.7-9 In clinical situations with altered body composition such as FTT, the applicability of prediction equations may be limited. Our previous report showed that the SCH-WT-HT equation predicted REE better than WHO, SCH-WT, and Harris-Benedict equations in children over a broad age range (age 0.2 to 20.5 years) and especially in those with FTT or obesity.10 345
SENTONGO ET AL
The aim of this study was to compare measured and predicted REE with the WHO, SCH-WT, and SCH-WT-HT equations in a nutritionally at-risk group of young children under the age of 3 years with the clinical diagnosis of FTT. The pattern of accuracy between measured and predicted REE and the clinical characteristics that influenced the accuracy of predicted REE in this important clinical population were also evaluated.
METHODS Subjects aged birth to 3 years with a clinical diagnosis of FTT who were referred between 1993 and 1997 for REE measurement as part of clinical management were considered for enrollment into the study. Once referred, a standard protocol including growth assessment, measurement of REE, and body composition was followed. Weight was measured with a digital electronic stand-on scale or infant scale (ScaleTronix Inc, Wheaton, IL) to 0.1 and 0.01 kg, respectively. Recumbent length was measured to 0.1 cm with an infant length board (Holtain, Crymych, UK) in all subjects <2 years of age. Standing height was measured to 0.1 cm with a stadiometer (Holtain, Crymych, UK) in those aged 2 to 3 years. The measurements were made in triplicate, and the mean was used in analysis.11 After the REE assessment was done, body composition was estimated with total body electrical conductance (EmScan, Springfield, IL), and the results were reported as FFM in kilograms and percentage body fat. The subjects were swaddled and then laid supine for passage through the electromagnetic field produced by an oscillating radio frequency current of the large solenoid coil.12 Ten scans were performed on each subject, and the average phase value was calculated. The average duration of the study for each subject was 10 minutes. No additional sedation was administered for the TOBEC test. 346
THE JOURNAL OF PEDIATRICS MARCH 2000 REE was measured by open-circuit indirect calorimetry with a computerized metabolic cart (SensorMedic 2900Z, SensorMedics Corp, Yorba Linda, CA). The measurement was carried out for 30 to 60 minutes between 7:30 and 10:00 AM after an age-appropriate fast ranging from 4 to 8 hours. The hospitalized subjects were transported from their hospital bed by stretcher. The outpatient subjects were admitted to the laboratory and rested in a supine position for a minimum of 30 minutes before the REE was determined. The REE procedure took place in a quiet thermo-neutral room while the child rested in a supine position. Chloral hydrate at a dose of 65 mg/kg was used for sedation. A clear ventilated hood was placed over the subject’s head for sampling respiratory gases at 1-minute intervals. The subjects were observed closely for movement by 1 of 2 trained observers. Measurements of REE obtained immediately after periods of movement or coughing, which altered REE, were excluded. The first 10 minutes of each study were also excluded to account for environmental adjustment by the child. The 1-minute REE measurements from up to 45 minutes (minimum of 15 minutes) were averaged and used for overall REE measurements. The modified Weir equation was used to calculate energy equivalency from the oxygen consumption and carbon dioxide production.13
Statistical Analysis SD scores (z scores) for length/ height, weight, and weight-for height were computed with Centers for Disease Control Anthropometric Software Program (version 3.1, 1988, Division of Nutrition, Centers for Disease Control and Prevention, Atlanta, GA).14 The percentage of measured REE was calculated as predicted REE, divided by measured REE, and multiplied by 100. The predicted REE was considered accurate for the purpose of this analysis if it was within 10% of the measured REE (90% to 110% measured REE).15
A percent measured REE value <90% was classified underestimated by the prediction equations, and likewise a percent measured REE of >110% was classified as overestimated REE. Paired t tests were used to compare measured REE and predicted REE by the WHO, SCH-WT, and SCH-WTHT equations. To determine the equation with the least probability for underestimating REE, the frequency of underestimated REE by the different equations was compared with the Pearson chi-squared test. Analysis of variance was used to evaluate the association between clinical characteristics and accuracy of predicted REE. The association between REE as kilocalories per kilograms per day and WAZ was examined with Pearson correlation. Because the WHO equation is most commonly used in clinical practice, a subsequent analysis of the results was completed with this prediction equation. Subjects with body composition data available were subdivided into accurately predicted (group 1) and underestimated REE (group 2) by the WHO equation. The relationship between REE and FFM in the 2 groups of subjects was examined with Pearson correlation. The association between the clinical variables and REE was analyzed with analysis of variance, with REE as the dependent variable and FFM, age, HAZ, WAZ, and weight for height z score as the independent variables. A P value <.05 was defined as statistically significant. Statistical analyses were performed with SYSTAT software (version 5.2., SYSTAT, Inc).
RESULTS Forty-five subjects were enrolled (47% female)(Table I). All subjects had a clinical diagnosis of FTT, and only 11% had no other secondary clinical diagnosis; 64% of the subjects had more than 2 secondary diagnoses in addition to the FTT. Gastrointestinal disorders including gastroesophageal
SENTONGO ET AL
THE JOURNAL OF PEDIATRICS VOLUME 136, NUMBER 3 reflux, malabsorption, and allergic enteropathy formed most (61%) of the secondary clinical diagnoses. Other clinical diagnoses included neurologic (18%), cardiac (7%), and pulmonary disorders (4%). No subject was acutely ill, febrile, or had a disorder or medication known to directly or indirectly alter energy metabolism such as thyroid disease or β-adrenergic bronchodilators. Two subjects did not receive sedation with chloral hydrate because of unacceptable risks associated with the underlying medical conditions. One subject (0.63-year-old girl, WAZ –4.3 and HAZ –4.9) received an intravenous dextrose infusion (7.2 mg/kg/min) during the REE test because of significant risk for hypoglycemia. Her enteral feeds, which provided 63% of the daily caloric intake, were discontinued 6 hours before REE was measured. Her respiratory quotient did not differ from that of the other subjects, who underwent complete fasting before the study. Excluding her did not change the mean WAZ, HAZ, and REE (440 ± 111 vs 438 ± 111 kcal/d) of the group; therefore she was kept in the analysis. Body composition assessment by TOBEC was performed only after the REE measurement was completed. No additional sedation with choral hydrate was given for the TOBEC measurement. Therefore only subjects who remained sedated after the REE test and were cooperative enough for technically good TOBEC studies had their body composition assessment completed (32 subjects, 50% female) (Table I). The predicted REE by SCH-WTHT did not significantly differ from the measured REE (Table II). However, the predicted REE by WHO and SCH-WT significantly differed from measured REE. Pearson chi-squared test showed that of the 3 equations, the SCH-WT-HT underestimated REE significantly less frequently in this study population (P = .01). Younger age and worse growth failure, as indicated by lower WAZ and HAZ, were
Table I. Clinical characteristics of subjects in the study (mean ± SD)
Male patients (24)
Female patients (21)
1.2 ± 0.7 7.3 ± 2.5 –2.7 ± 1.2 71.6 ± 9.7 –2.04 ± 1.2 –1.9 ± 0.8
1.3 ± 0.7 7.2 ± 2.1 –2.6 ± 1.0 70.0 ± 9.1 –2.1 ± 1.3 1.5 ± 0.7
Male patients (16)
Female patients (16)
1.2 ± 0.7 5.8 ± 1.6 –2.9 ± 1.0 –2.2 ± 1.3 –1.9 ± 0.9 18.1 ± 7.9
1.3 ± 0.7 5.6 ± 1.1 –2.5 ± 0.9 –2.0 ± 1.2 –1.5 ± 0.7 20.6 ± 6.0
Age (y) Weight (kg) WAZ Length/height (cm) HAZ WHZ
Subgroup with body composition assessed Age (y) FFM (kg) WAZ HAZ WHZ %Body fat WHZ, Weight for height z score.
Table II. Comparison of measured and predicted REE (mean ± SD) by the WHO, SCH-WT, and SCH-WT-HT prediction equations
Predicted REE
Measured REE REE (kcal/d) REE (kcal/kg/d) P value* %Measured of REE
438 ± 111 62.0 ± 9.5
WHO
SCH-WT
SCH-WT-HT
391 ± 140 52.9 ± 3.2 .0001 87.5 ± 16.0
398 ± 136 54.2 ± 1.9 .0001 89.6 ± 15.0
450 ± 138 62.3 ± 5.4 .2 102.3 ± 15.2
*Refers to comparison between measured and predicted REE.
Table III. Percentages of subjects in the different predicted REE categories and clinical characteristics associated with underestimated REE
Percent of subjects REE category
WHO
SCH-WT
SCH-WT-HT
58 38 4
51 42 7
24 44 32
Underestimated REE* Accurate REE† Overestimated REE‡
P value Characteristics significantly associated with underestimated REE Age WAZ HAZ WHZ
.02 .0001 .001 .2
.006 .0001 .02 .07
.1 .06 .01 1.00
WHZ, Weight for height z score. *Predicted REE <90% of measured value. †Predicted REE = 90% to 110% of measured value. ‡Predicted REE >110% of measured value. Significance = P value < .05.
347
SENTONGO ET AL
THE JOURNAL OF PEDIATRICS MARCH 2000 with underestimated REE were younger (1.0 ± 0.5 vs 1.6 ± 0.7 years, P = .006) and had more severe growth failure (HAZ –2.7 ± 1.2 vs –1.29 ± 0.8, P = .04; WAZ –3.2 ± 0.9 vs –1.9 ± 0.7, P = .02). Analysis of variance with REE as the dependent variable and FFM, group (accurate REE vs underestimated REE), age, WAZ, and HAZ as independent variables showed that the interaction of group and FFM were significant determinants of REE. This finding suggests that the relationship between REE and FFM differed in the 2 groups.
DISCUSSION Fig 1. Correlation between REE (kilocalories per kilograms per day) and WAZ in 45 children with FTT.
Fig 2. Association between measured REE (kilocalories per day) and FFM in 2 groups of children. Group 1 (closed circles): predicted REE by WHO ≥90% of measured REE. Group 2 (open circles): predicted REE by WHO <90% measured REE (ie, underestimated REE by WHO). Group 2 was younger and had lower HAZ and WAZ than group 1 (P < .001 for all comparisons). significantly associated with underestimation of REE by both WHO and SCH-WT equations, whereas only low HAZ resulted in this association with the SCH-WT-HT equation (Table III). A negative correlation was seen between WAZ and measured REE (kcal/kg/day r = –0.716, P < .0001 (Fig 1). 348
Measured REE positively correlated with FFM in the subgroup of children who had body composition data available. For the same FFM the subjects with underestimated predicted REE by WHO had a higher measured REE than those with accurately predicted REE by WHO (Fig 2). The subjects
In this sample of young children with moderate to severe FTT, all prediction equations were within 10% less than half the time. However, SCH-WT-HT compared best with measured REE and underestimated REE less frequently. Being younger (for WHO and SCH-WT), having more severe growth failure (based on WAZ, HAZ, or both), or both were associated with underestimation of energy requirements by prediction equations. The infants and children in this study had poor growth compared with the reference population, which was consistent with the clinical diagnosis of FTT. These findings confirm the inadequacy of prediction equations in altered growth and body composition10 and more specifically demonstrate that younger age even within the range of 0 to 3 years, and moderate to severe growth failure, are indicators that prediction equations are more likely to underestimate REE, thereby making this approach undesirable in clinical practice. Because nutritional intervention is usually central in the treatment of children with FTT, accurate estimation of caloric needs is necessary. If the goals of clinical nutritional management are to achieve a positive energy balance and growth as quickly as is safely possible, it is better to overestimate, not
SENTONGO ET AL
THE JOURNAL OF PEDIATRICS VOLUME 136, NUMBER 3 underestimate, caloric requirements. Therefore assuming that the greater clinical risk for inaccuracy is in underestimating REE, the SCH-WT-HT equation was superior to the WHO and SCH-WT equations for planning care of young infants and children with moderate to severe FTT. The better REE prediction by SCH-WT-HT, which uses weight and height rather than weight alone in the calculation, suggests that body composition may be an important factor contributing to inaccuracy of the prediction equations. The WHO prediction equations published in 1985 were derived from data obtained from pooled REE measurements in >6000 healthy individuals.1,6 They are considered the best estimate available for prediction of REE of healthy people in any population1; however, inaccuracies in very underweight and overweight individuals have been reported.1,6,7 The Schofield prediction equations were subsequently derived with a data pool approximately twice the size of and overlapping by half the WHO data source. They included a few measurements from ill children and also measurements from >3800 subjects who were used as a secondary source for validating the derived equations.6 All the data sources were screened to ensure that appropriate experimental conditions had been followed and that complete data on age, sex, weight, height, and basal metabolic rate had been included.6 The larger database improved REE prediction from weight alone, and the inclusion of height into the equation specifically improved REE prediction in children <3 years old and the elderly aged 60 years and over.6 Of note, the age groups of birth to 3 years, 3 to 10 years, 10 to 18 years, 18 to 30 years, and so on were based on the commonly used clinical divisions of life span and not from patterns observed in the REE data. Because all 3 prediction equations are based on data obtained from healthy populations, they are best used for predicting energy needs in healthy children.1,6
Malnourished children, especially those with low body fat stores, may have an apparently increased rate of metabolism per unit body weight because of altered body composition. This may increase further during the initial phases of nutritional management in children with FTT,16 as REE increases and the child becomes anabolic. We were able to demonstrate that there was a different relationship between REE and FFM that contributed to the underestimated predicted REE by the WHO equation. Younger children with more severe FTT had a higher REE per FFM than older children with less severe FTT. Therefore based on the WHO equation alone, these children may have been considered hypermetabolic. However, much of the difference disappeared when we used the SCHWT-HT equation to predict REE. The apparent increase in REE per unit of body weight (total weight and FFM) during moderate to severe malnutrition may be due to the state of altered body composition. FFM is composed of tissues and organs with various degrees of metabolic activity. Organs are made up of high-energy-using cell masses (brain, liver, heart, and kidney) and account for >60% of REE, whereas muscle tissue accounts for approximately 30% of REE.7,8 In situations of altered body composition such as occurs in moderate to severe FTT, where there is loss of body fat and muscle mass, a greater proportion of FFM is composed of high metabolic rate organs such as brain, liver, kidney, and heart and a correspondingly reduced proportion of lower metabolic rate tissues such as muscle. Therefore the high metabolic rate tissue comprises a greater proportion of body weight, which leads to an increased REE per unit of body weight, whereas in terms of expected weight (for age), the REE tends to be low.16,17 Our observations support the hypothesis that the FFM of young children with FTT has a larger proportion of high metabolic rate tissue, and therefore the energy expenditure per unit body size is
increased. This finding may partly explain why standard prediction equations with only weight in the calculation may underestimate REE more frequently in young children with moderate to severe FTT. The potential limitation of this study was that timing of referral for measuring REE was not standardized in the subjects; therefore the subjects may have been at different phases of metabolic adaptation to malnutrition. Being anabolic after nutritional intervention may also explain increased REE.16 By history, however, none of the subjects were being successfully renourished; therefore this should not have altered the results. When REE cannot be measured in young infants and children with FTT, the SCH-WT-HT is the preferred equation. In addition, our findings are in agreement with the 1981 WHO1 recommendation that whenever possible, energy requirements should be measured.
REFERENCES 1. World Health Organization. Energy and protein requirements. Report of a joint FAO/WHO/UNU Expert Consultation. (WHO Technical Report Series No. 724). Geneva: World Health Organization; 1985. 2. Waterlow JC. Basic concepts in the determination of nutritional requirements of normal infants. In: Tsang RC, Nichols BL, editors. Nutrition during infancy. Philadelphia: Hanley & Belfus, Inc. 1988. p. 1-19. 3. Wells JCK, Davies PSW. The components of energy metabolism in 12 week old infants [abstract]. Ann Hum Biol 1995;22:175. 4. Pencharz PB, Azcue MP. Measuring resting energy expenditure in clinical practice. J Pediatr 1995;127;269-71. 5. National Academy of Sciences. Recommended Dietary Allowances. Washington, DC: National Academy Press; 1989. p. 24-38. 6. Schofield WN, Schofield C, James WPT. Basal metabolic rate—review and prediction, together with annotated bibliography of source material. Hum Nutr: Clin Nutr 1985;39C(Suppl 1):5-41. 349
SENTONGO ET AL
7. Gallagher D, Belmonte D, Deurenberg P, Wang Z, Krasnow N, Pi-Sunyer FX, et al. Organ-tissue mass measurement allows modeling of REE and metabolically active tissue mass. Am J Physiol 1998;275:E249-58. 8. Butte NF, Moon JK, Wong WW, Hopkinson JM, Smith EO. Energy requirements from infancy to adulthood. Am J Clin Nutr 1995;62:1047-52S. 9. Holliday MA. Metabolic rate and organ size during growth from infancy to maturity and during late gestation and early infancy. Pediatrics 1971;47: 169-79. 10. Kaplan AS, Zemel BS, Neiswender KM, Stallings VA. Resting energy expenditure in clinical pediatrics: measured versus prediction equations. J Pediatr 1995;127:200-5. 11. Cameron N. Methods of auxological anthropometry. Hum Growth 1986;3: 3-46. 12. Harrison GG, Van Itallie TB. Estimation of body composition: a new approach based on electromagnetic principles. Am J Clin Nutr 1982;35:1176-9.
THE JOURNAL OF PEDIATRICS MARCH 2000 13. Weir J. New methods for calculating metabolic rate with special reference to protein metabolism. J Physiol 1949; 109:1-9. 14. Jordan MD. CDC Anthropometric Software Program. Based on Centers for Disease Control and Prevention standard deviation-derived growth reference curves for U.S. children, birth to 18 years. Derived from NCHS/ CDC Reference Population 1988. 15. Firouzbakhsh S, Mathis RK, Dorchester WL, Oseas RS, et al. Measuring resting energy expenditure in children. J Pediatr Gastoenterol Nutr 1993;16: 136-42. 16. Montgomery RD. Changes in the basal metabolic rate of the malnourished infant and their relation to body composition. J Clin Invest 1962;41: 1653-63. 17. Holliday MA, Potter D, Jarrah A, Bearg S. The relationship of metabolic rate to body weight and organ size. Pediatr Res 1967;1:185-95. 18. Stallings VA. Resting energy expenditure. In: Altschuler S, Liacouras C, ed-
itors. Clinical pediatric gastroenterology. Philadelphia: WB Saunders; 1998. p. 606-11.
APPENDIX Schofield equations for predicting resting energy expenditure from body weight (W, kg), height (H, m), sex, and age group5,18 Age range (y) Males 0–3 3–10 10–18 18–30 Females 0–3 3–10 10–18 18–30
kcal/d
0.167W + 1517.4H – 617.6 19.59W + 130.3H + 414.9 16.25W + 137.2H + 515.5 15.057W + 10.04H + 705.8 16.252W + 1023.2H – 413.5 16.969W + 161.8H + 371.2 8.365W + 465H + 200.0 13.623W + 283.0H + 98.2
Receive tables of contents by e-mail To receive the tables of contents by e-mail, sign up through our Web site at http://www.mosby.com/jpeds Choose E-mail Notification. Simply type your e-mail address in the box and click the Subscribe button. Alternatively, you may send an e-mail message to [email protected]. Leave the subject line blank and type the following as the body of your message: subscribe jpeds_toc You will receive an e-mail to confirm that you have been added to the mailing list. Note that table of contents e-mails will be sent out when a new issue is posted to the Web site.
350