Assessment of energy expenditure in children and adolescents

Assessment of energy expenditure in children and adolescentsManfred J. Muller and Anja Bosy-Westphal

Purpose of the review

This is a review on recent studies regarding methodological

aspects of assessment of energy expenditure in children and

adolescents.

Recent findings

A variety of methods used for assessment of different

components of energy expenditure has been validated and used

in children and adolescents. Reference values derived from

representative groups of healthy children and adolescents are

now available. Variations in the different components of energy

expenditure and physical activity have been proposed to be

associated with weight gain, and the prevalence of overweight

and obesity. However, recent cross-sectional and longitudinal

data in children and adolescents do not provide strong evidence

for this idea. In contrast, hypermetabolism, which is frequently

seen in critically ill children, may contribute to their tissue

catabolism. In this case beta blockade seems to be a way to

increase ‘metabolic economy’ and thus to reduce tissue

catabolism. In chronically ill children and adolescents (e.g.

patients with cystic fibrosis and sickle cell anemia) energy

expenditure is also frequently increased and group specific

algorithms are needed for predicting energy expenditure when

measurement facilities are not available.

Summary

Methods for assessment of the different components of energy

expenditure have been validated in children and adolescents.

The combined use of these methods together with detailed

analyses of body composition is recommended for future

studies. In patients with acute or chronic illness measurements

of energy expenditure are necessary if disease-specific

algorithms are not available.

Keywords

energy metabolism, energy need, doubly labeled water, indirect

calorimetry, pedometer, accelerometry, physical activity, fitness,

critically ill children

Curr Opin Clin Nutr Metab Care 6:519–530. # 2003 Lippincott Williams & Wilkins.

Institute for Human Nutrition and Food Science, Christian Albrechts University at Kiel,Kiel, Germany

Correspondence to Prof. Manfred J. Muller, Institut fur Humanernahrung undLebensmittelkunde, Christian Albrechts Universitat zu Kiel, Dusternbrooker Weg15-17, D 24105 Kiel, GermanyTel: +49 431 880 5670; fax: +49 431 880 5679;e-mail: [email protected]

Current Opinion in Clinical Nutrition and Metabolic Care 2003, 6:519–530

Abbreviations

AEE activity energy expenditureDIT diet-induced thermogenesisDLW doubly labeled waterFAO Food and Agriculture OrganizationFFM fat-free massPAEE physical activity energy expenditureREE resting energy expenditureTEE total energy expenditureWHO World Health Organization

# 2003 Lippincott Williams & Wilkins1363-1950

IntroductionThe regulation of energy expenditure has been inten-

sively investigated for many decades. Traditional com-

ponents of 24-h energy expenditure, or total energy

expenditure (TEE), include resting energy expenditure

(REE), diet-induced thermogenesis (DIT) and physical

activity energy expenditure (PAEE). In an individual

with a sedentary lifestyle about 70% of TEE is due to

REE, 10% to DIT and 20% to PAEE.

From a regulatory point of view the different compo-

nents of daily energy expenditure are considered as

separate entities. Body cell mass, a familial trait, and

thyroid status are major determinants of REE. DIT

varies in response to food intake, the amount of calories

supplied, the calorie-mix of the diet, substrate transport

and processing. The latter is affected by several

hormones such as insulin and the sympathetic nervous

system. In addition to food intake several drugs and

acute cold exposure also increase DIT.

Physical activity is defined as body movement produced

by skeletal muscle which results in an increase of energy

expenditure. PAEE depends on the amount of physical

work and planned activity but is also explained by a

considerable amount of spontaneous activities (e.g.

fidgeting).

Besides physiological factors different cytokines activate

various neuropeptides which again stimulate effector

systems like the sympathetic nervous system and thus

increase energy expenditure and frequently cause

hypermetabolism in patients with acute and chronic

diseases.

Since all experts and also international organizations –

like the Food and Agriculture Organization (FAO) and

the World Health Organization (WHO) – now recom-

mend that energy requirements and thus dietary energy

DOI: 10.1097/01.mco.0000087967.83880.3a 519

recommendations should be based on measurements of

energy expenditure, there is a need to assess free-living

energy expenditure in people of different age groups.

Major and renewed scientific interest in the measure-

ments of energy expenditure comes from different areas

of research.

Disturbances in energy balance have been described in

overweight, obese and underweight patients. Although

controversial some longitudinal studies have shown that

both hypermetabolism [1] and hypometabolism [2,3] are

associated with weight loss and weight gain, respec-

tively. These data suggest that deviations in energy

expenditure predispose to weight changes. In addition

overfeeding as well as underfeeding are followed by

adaptive responses in the different components of

energy expenditure [4–6]. These adaptations limit

weight changes and thus allow the patient to reach a

new steady state (i.e. a stable body weight).

Recent studies using advances in technologies to

measure TEE together with detailed body composition

analyses by magnetic resonance imaging technologies

have provided further insights into organ contribution to

REE and thus allowed the use of new models in

metabolic research [7,8]. These models consider the

heterogenity of fat-free mass (FFM) by dividing it into

organs with different metabolic activities and thus

adding to our understanding of interindividual variance

in REE. Since further advances in molecular biology

improved our understanding of the complex regulation

of energy intake [9], knowledge of energy expenditure

together with energy intake will add to new concepts of

the regulation of body weight. It has become evident

that regulation of body weight should now be considered

as regulation of body composition.

Methods used to predict and to assessenergy expenditureEnergy requirements can be predicted with some

accuracy. Estimations are based on measurements of

energy expenditure in greater reference populations.

Conceptually estimates of energy requirements refer to

the mean of groups and not to individuals. Predictive

equations based on measurements in a considerable

number of individuals have been developed for REE

and also for TEE. Most of these equations are valid for

adults. For children and adolescents a sufficient database

on energy expenditure has only recently been estab-

lished [10,11]. It includes data on REE as well as TEE

(Table 1).

Methods used to measure energy expenditure in hu-

mans include indirect and direct calorimetry; isotope

dilution, mainly the so-called doubly labeled water

technique (DLW); 24-h heart rate measurements; and

activity monitors (e.g. accelerometry) [12,13]. The

different methods can be used to specifically address

the different components of energy expenditure (Table

2 and Table 3). Indirect calorimetry measures heat

production based on respiratory gas exchange and can be

used to measure REE, DIT and energy expended for

individual physical activities. Direct calorimetry mea-

sures heat dissipation. In a steady-state and under resting

conditions heat loss is identical to heat production. Using

indirect calorimetry (respiratory) or direct calorimetry

chambers TEE can be measured with high accuracy.

PAEE is derived by subtracting REE from TEE.

However measurements within a chamber do not reflect

free-living conditions.

In children there are no problems with the use of

ventilated hood systems but measurements within a

chamber are difficult to realize in a greater number of

subjects. Using a ventilated hood system the participants

were allowed to read or to listen to tapes to minimize

fidgeting during measurements. In that case the book

was placed on top of the hood and an assistant turned

the pages. There were no significant differences in REE

among participants who were sitting, reading, or watch-

ing television [14].

Another approach to measure TEE is the DLW

technique. This method was developed about 50 years

ago. DLW is based on the differences in turnover rates

of 2H2O and H218O in body water. After equilibration

both 2H and 18O are lost as water whereas only 18O is lost

by respiration as carbon dioxide. The difference in the

rate of turnover of the two isotopes can be used to

calculate the carbon dioxide-production rate, VCO2.

Assuming a mean respiratory quotient (i.e. VCO2/VO2

) of

0.85, the oxygen consumption rate (VO2) and thus energy

expenditure can then be calculated from VO2and VCO2

.

The DLW technique is validated against indirect

calorimetry and is now considered to be a gold standard

for measurements of TEE under free-living conditions.

The DLW method is most convenient to children

because it places low demands on the participant’s

performance (only drinking a glass of water and the

collection of some urine samples). Sources of error are

analytical errors in the mass spectrometric determination

of isotopic enrichment, biological variations in the

isotope enrichment, isotopic fractionation during forma-

tion of carbon dioxide and during vaporization of water,

the calculation of total body water and the assumption or

calculation of the 24-h respiratory quotient.

Measurement of physical activity is difficult because it is

a multidimensional variable which includes type, fre-

quency, duration and intensity of movements. Physical

activity is a variable and unstable behavior with habitual

Assessment of nutritional status and analytical methods520

levels of activity varying during the day, with different

days of the week and different times of the year.

Twenty-four hour heart rate and activity monitors have

been widely used to assess physical activity and TEE.

Calculations of TEE from 24-h heart rate and activity

monitors are based on intraindividual calibrations of VO2

against heart rate using open circuit indirect calorimetry

and ergometry. Heart rate monitors are robust and

function well under field conditions. They save 24-h

heart rate data. Free-living energy expenditure is

derived from the minute-by-minute recordings of heart

rate using the individual regression line for VO2versus

heart rate. The individual nature of the heart rate versus

VO2relationship makes it necessary to establish a

regression equation for each participant at several levels

and intensities of activity. One has to keep in mind that

factors other than VO2(e.g. emotions, body position,

ambient temperature, individual muscle groups exer-

cised) also have an impact on heart rate. The method has

been validated against DLW and indirect calorimetry.

The major shortcoming of the 24-h heart rate method is

its inaccuracy at low levels of physical activity. Variations

Table 1. Energy expenditure in different age and sex groups of children and adolescents from affluent societies

Age group (years) Sex n Age (years) BMI (kg/m2) TEE (MJ/day) REE (MJ/day) PAEE (MJ/day) PAL

1–6 M 29 4.7+1.0 15.9+1.7 6.1+1.0 3.8+0.8 2.2+1.1 1.64+0.39F 21 4.9+0.7 14.9+1.0 5.5+0.9 3.6+0.7 1.9+0.8 1.57+0.30

7–12 M 32 9.8+1.8 19.8+4.1 9.8+1.6 5.7+1.0 4.1+1.1 1.74+0.22F 24 9.2+2.0 17.6+2.0 8.0+1.8 4.8+1.0 3.2+0.9 1.68+0.16

13–17 M 31 14.5+1.1 25.3+8.8 14.1+2.7 8.1+1.5 6.0+1.7 1.75+0.19F 26 14.8+1.1 26.6+8.6 11.4+2.6 6.7+1.5 4.7+1.7 1.73+0.24

Data are based on measurements of total energy expenditure (TEE) with doubly labeled water. Resting energy expenditure (REE) was assessed byindirect calorimetry. BMI, body masss index; PAEE, physical activity energy expenditure; PAL, physical activity level; M, male; F, female. Data fromBlack et al. [11].

Table 2. Methodological approaches to the assessment of energy expenditure

Methods used for assessment

Total energy expenditure (TEE) Isotope dilution (doubly labeled water)Indirect calorimetry (respiratory chamber)Direct calorimetry24-h heart rate monitoring

Resting energy expenditure (REE) Indirect calorimetry (ventilated hood)Diet-induced thermogenesis (DIT) Indirect calorimetry (ventilated hood)Physical activity energy expenditure

Activity energy expenditure (AEE) 24-h heart rate monitoringActivity monitor

Activities of daily life Calculation (TEE7(REE+DIT))24-h heart rate monitoringActivity monitor

Spontaneous activities (fidgeting) Calculation (TEE7(REE+DIT+AEE)Motion sensorsRadar method (within a chamber)

Table 3. Use and accuracy of different methods to assess energy expenditure

Indirect calorimetryIsotope dilution 24-h heart rate

Component of energy expenditure Direct calorimetry Chamber Ventilated hood (DLW) monitoring Activity monitor

Minimal or sleeping energy expenditure + +Resting energy expenditure + + +Diet-induced thermogenesis ((+)) +Activity-induced energy expenditure (+) +

Physical activity (+) + +Heat production +Substrate oxidation rate + +

24-h or total energy expenditure + + + +Duration of measurement 1–2 days 1–2 days -6 h 3–14 days -7 days -7 daysCoefficient of variation 52% Approximately

3–5%Approximately

3–5%5–10% Approximately

10%Approximately

10%

DLW, doubly labeled water; + measurable; (+) measurable with limitations; ((+)) weak method for assessment.

Energy expenditure in children Muller and Bosy-Westphal 521

at the lower end of the calibration curve may result in

large errors in predicted TEE. The prediction error may

reach about 15%. The FLEX method is now used to

overcome this shortcoming. FLEX heart rate is defined

as the mean of the highest heart rate under sedentary

conditions (e.g. sitting on the bicycle ergometer without

pedaling) and the lowest heart rate recorded when

performing light work on the ergometer. Thus the

FLEX heart rate point (or the inflection point) is used as

an improvement to discriminate between resting and

exercise heart rate. The method is inexpensive, objec-

tive, simple, and popular with patients. Heart rate

monitoring is the most commonly employed method

for estimating TEE in children.

Movement counters have also been widely applied in

physical activity research. The most frequently used is

the pedometer: a mechanical step counter, which records

movement in one direction. A pedometer does not

record non-step activities, such as during bicycling. It

also cannot measure energy expenditure at different

speeds of walking or running. Accelerometry techniques

(e.g. triaxial accelerometry) have been developed to

assess body acceleration in two or three planes of space.

During the last years the TriTrac-R3D (Stayhealthy.Inc,

Monrovia, USA) accelerometer has been most frequently

used as an activity monitor. The method integrates

acceleration by calculating the square root of the sum of

squared activity counts in each vector (i.e. the so-called

vector magnitude). The vector magnitude can then be

converted to energy expenditure by the use of the

manufacturer’s algorithm. Activity monitoring is used to

distinguish differences in activity levels between and

within individuals. When compared with measurements

of TEE or activity energy expenditure (AEE) acceler-

ometers provide direct measurements of physical activ-

ity. However, physical activity is not equivalent to the

energy cost of activity. Therefore activity monitors have

limitations in quantifying TEE.

Accelerometry has been validated for predicting the

energy cost of children’s activities [15]. When compared

with heart rate and pedometry, triaxial accelerometry

provided the best prediction of oxygen consumption.

PAEE can be subdivided into AEE, that is energy

expenditure for planned or structured activities (e.g.

walking or running) and spontaneous activities (e.g.

fidgeting). AEE is measured by 24-h heart rate or activity

monitoring. PAEE can be calculated from the difference

between TEE and the sum of REE plus DIT. If AEE is

known, energy spent for spontaneous activities can be

derived indirectly from the difference between TEE and

the sum of REE plus DIT plus AEE. Alternatively

spontaneous activities can be measured bymotion sensors

or by the use of radar systems within an indirect or direct

calorimetry chamber. This measurement is irrespective of

work intensity. When compared with motion sensors the

results from a radar system are more reliable and show a

better correlation with energy expenditure.

Methodological aspects of the assessment ofenergy expenditure in childrenThe reproducibility of REE measurements by indirect

calorimetry was tested in a group of 6–11-year-old

children [16]. The investigators used a short and

standardized protocol: no physical activity and 10 min

rest before measurement, measurements between 08.00

and 08.30, in the morning after an overnight fast,

constant room temperature of 22–238C, ventilated hood

system, repeated calibrations, all measurements made by

the same trained observer, during the measurements

children were instructed to lie quietly and motionless

and listened to music or story tapes. The mean

intraindividual coefficient of variation was 2.6+1.7%.

Calculating reproducibility (i.e. variance in REE be-

tween children / variance in REE between children plus

variance in REE within children) was 95% indicating

excellent reliability. This is in line with a previous study

on 19 prepubertal girls aged 6.0–10.1 years [17]. In this

study REE was measured on three consecutive mornings

during two periods 6 weeks apart. There were no

significant differences between the individual REE

measurements (mean CV, 5.8%). Spending the night

before testing at home compared with in a clinical

setting had negligible influence on REE. Therefore,

admission to a clinical research setting is not necessary

for a reliable determination of REE.

Using indirect calorimetry as a clinical tool in pediatric

intensive care units the intraindividual coefficient of

variation of energy expenditure measurements was

7.2+4.5% [18]. There were no significant differences

between means for 30-min and 24-h energy expenditure

measurement. The mean percentage of between day

variations in energy expenditure was 21+16%. The

range of energy expenditure data was high (1–69%).

Thus, in critically ill, ventilated children energy ex-

penditure can be measured with acceptable accuracy but

daily measurements are necessary because of huge

between-day variations.

Heart rate monitoring was calibrated against indirect

calorimetry to assess sleeping energy expenditure in

children [19]. After calibration at different levels of

physical exercise heart rate, respiratory exchange and

also body movements (by the use of a TriTrac-R3D

accelerometer) were measured for at least 2 h between

22.00 and 04.00 hours. Measurements were performed at

home in the bedroom and care was taken to keep the

child asleep during measurements. Using linear or

polynomial regressions resulted in low between method


differences when the heart rate was below FLEX heart

rate (1.0+5.4%). When heart rate exceeded FLEX rate

during sleeping periods the difference between the

methods reached significance (13+5.9%). There were

close correlations between the results of indirect

calorimetry and heart rate monitoring. The authors

conclude that heart rate monitoring can be used to

compute sleeping energy expenditure.

In another study different accelerometer-based physical

activity monitors were validated and calibrated against

6-h indirect calorimetry (respiratory chamber) and heart

rate measurements in 26 children aged 6–16 years [20].

The authors found very close associations between 6 h

energy expenditure or PAEE (as assessed by the

difference between TEE and REE) and the counts of

the activity monitors (r-values between 0.66 and 0.80).

In addition, they could define threshold counts for

sedentary, light, moderate and vigorous activities. They

certified the use of activity monitors as valid and useful

devices for the assessment of physical activity in

children, at least under the restricted possibilities within

a respiratory chamber.

The TriTrac-R3D accelerometer was also validated

against 24-h heart rate monitoring in 20 non-obese

children and adolescents aged 5.5–16 years [21]. TEE

assessed by the TriTrac-R3D accelerometer or 24 h

heart rate monitoring did not differ (6.4+1.0 versus

6.0+1.9 MJ/day). Activity monitors showed good agree-

ment with heart rate monitoring during physical activity

periods, but underestimated energy expenditure during

sedentary periods. There was a high correlation between

vector magnitude and TEE derived from heart rate

monitoring (r = 0.96) which decreased (to r = 0.74) for thecorrelation between physical activity and energy ex-

penditure during sedentary periods. The authors con-

clude that when compared with the 24-h heart rate

method the TriTrac-R3D accelerometer is not system-

atically accurate. The results must be interpreted with

caution when assessing energy expenditure in children.

Prediction of resting energy expenditure inchildren and adolescentsThe agreement between indirect calorimetry and

equations used to predict REE was investigated in 59

non-obese and 57 obese children and adolescents [22].

The formulae frequently used in practice (e.g. the FAO/

WHO/UNU prediction equations or Schofield’s formu-

lae) are based on measurements in considerable numbers

of participants (i.e. more than 7500 3–18-year-old

children in the case of the FAO/WHO/UNU prediction

equations [23]), which were investigated between 1910

and 1980 in different areas all over the world. The

results of the present study showed only minor devia-

tions of predicted REE versus measured REE. The

coefficient of correlation varied between 0.68 and 0.91.

REE calculated according to Schofield’s equation using

height as well as weight showed a mean difference to

measured REE of only 3.7 kcal/day. However the limits

of agreement were between 7293 and + 300 kcal/day.

Group by group comparisons showed considerable

differences in prediction errors when comparing girls,

boys, obese and non-obese participants. The authors

conclude that one prediction formula cannot be used to

calculate REE in all races and different geographic

regions. This is also true for populations differing with

respect to sex and nutritional status. The use of more

complex equations (e.g. FFM specific equations) does

not seem to improve the prediction of REE in children.

Henry et al. [24] developed new equations to estimate

REE in children aged 10–15 years. These equations

were based on measurements of REE by indirect

calorimetry in 195 school children (40% boys, 60% girls).

The authors considered anthropometric data and also

pubertal development. Stepwise multiple regression

analysis was used to develop REE prediction formulae.

In this study body weight was the most important factor

which explained 61% of the variance in REE in boys and

52% in girls, respectively. The regression formulae

for estimating REE (in kilojoules per day) were

weight666.9 + 2876 in boys and weight647.9 + 3230 in

girls, respectively. Sex, weight, height, puberty stage

and skinfolds were used to develop more specific regres-

sion equations. In pre-menarche girls menarche status

improved REE prediction. By contrast, in boys pubic

hair and gonadal stage did not improve REE estimation.

Taken together inclusion of pubertal stage provided only

minor improvements in the estimation of REE in

children.

Effect of puberty and sex on energyexpenditureSince puberty and adolescence are characterized by

rapid anatomic and physiologic developments the

determinants of energy expenditure, and thus TEE,

are also changing. In a cross-sectional study TEE and its

main components were investigated by indirect calori-

metry (respiratory chamber) in 83 children at different

stages of puberty (i.e. Tanner stages 1–6 [25]). The

mean group values in boys and girls were 8.22 and 7.60

MJ (prepubertal), 11.35 and 9.10 MJ (at puberty) and

11.73 and 9.68 MJ (postpubertal), respectively. Sex,

body composition and season but not the stage of

puberty were the main determinants of TEE and REE,

respectively. TEE and also REE could be predicted

from FFM, sex and season. These results are in contrast

to another study [26] in which a significant maturation

effect on TEE (as measured by DLW) and REE (as

assessed by indirect calorimetry) was observed in boys as

well as in girls. Regarding REE, an interaction between


maturation and sex was observed. After adjusting REE

for FFM by ANCOVA the effect of maturation on

resting but not on total expenditure remained significant.

In this group of prepubertal and pubertal children REE

per kilogram body weight showed an inverse association

with age suggesting that the composition of body weight

(and thus FFM) changes during puberty. Both studies

illustrated that there is no or only a small body mass-

independent effect of puberty on energy expenditure.

There are limited and controversial data on the influence

of sex on REE and TEE in prepubertal children. In

adults sex differences in resting expenditure are mainly

explained by differences in FFM. However, when

compared with female participants, TEE as well as

REE remained higher in male patients even after

adjustment for FFM [27]. Previous data have suggested

that there are no or only negligible sex differences in

TEE and REE in children but other authors found an

increased TEE or REE in boys when compared with

girls [11,28–30]. In one study on 371 prepubertal and

postpubertal obese and non-obese children sex added

only 1.1% to the predictability of REE [29]. In a recent

study [31] REE and TEE were measured by indirect

calorimetry and 24-h heart rate monitoring in 40

prepubertal children. In addition to energy expenditure,

sex differences in the intensities of physical activity

during daytime activities were analyzed. In this study

FFM explained 70% of the variance in REE, which again

explained 46% of the variance in TEE. After adjusting

resting or total energy expenditure for FFM no sex

differences could be observed. The authors concluded

that at least in prepubertal children there is no effect of

sex on REE and TEE. Sex also had no significant effect

on FLEX heart rate. Boys and girls showed a similar

percentage of time spent at different physical activity

levels (varying from 1 to above 2). Using a stepwise

regression analysis with REE as the dependent variable

and body weight instead of FFM as determinant, both

body weight (r 2 = 0.72) and sex (r 2 = 0.07) reached

significance. These results again illustrate the effect of

body composition on sex differences in different

components in energy expenditure. The authors specu-

lated that sex differences in TEE and REE might

become apparent under some conditions (e.g. during

winter season when outdoor activities are reduced).

Extending our previous study added some new aspects

to this question. In Fig. 1(a) different relationships

between REE and FFM are apparent for (1) prepubertal

children and (2) a ‘combined’ group of postpubertal

children with young adults. Adjusting REE for FFM

resulted in significant differences between sexes within

each age group as well as between the three groups (i.e.

prepubertal versus postpubertal versus young adults; Fig.

1b). Differences in adjusted REE between sexes were

43.0 kcal for prepubertal children, 190.5 kcal for post-

pubertal children and 95.1 kcal for young adults,

respectively. Altogether these data suggest that there

are mass independent effects of sex and puberty on

REE. It is unclear whether the effects of gender or

puberty will be further reduced or even disappear after

correction for detailed composition of FFM (i.e. non-

muscular and muscular components of FFM).

Energy expenditure, weight gain and obesityin childrenDisturbances in energy expenditure have been consid-

ered as metabolic risk factors for overweight and obesity.

Because overweight and obesity are more prevalent in

African Americans when compared with white Amer-

icans the components of energy expenditure were

studied in racial subgroups. To assess the effect of

energy expenditure on obesity the authors of the Baton

Rouge Children’s Study [32 .] investigated different

components of energy expenditure (TEE by DLW,

REE and DIT by indirect calorimetry; DIT was

measured in response to a standard meal over a period

of 3 h; PAEE was calculated from the difference

between TEE and the sum of REE plus DIT) and

body composition (as assessed by dual energy X-ray

absorptiometry) in a total of 131 preadolescent African

American and white children (mean ages between 10

and 11 years, all children were below Tanner stage 3,

with 101 at stage 1 and 30 at stage 2). The former group

of children had lower TEE and REE. African American

girls spent less time on physical activity, whereas their

male counterparts had a lower REE. In white children

there was a sex difference in REE. Girls had lower TEE

and physical activities than boys. After adjustment for

FFM there was still a significant effect of race (TEE of

9.32 versus 9.76 MJ/day, REE of 5.53 versus 6.01 MJ/

day, in African American versus white children, respec-

tively) and sex (TEE of 9.24 versus 9.84 MJ/day in girls

and boys, respectively). However the effect of sex on

REE disappeared after adjustment for FFM. When

compared with normal weight children energy expended

in physical activity was lower in the obese children. By

contrast obese participants had normal REE and DIT.

After adjustments for body composition, an effect of

obesity was seen for the FFM-adjusted REE (increased

in the obese) and PAEE (decreased in obese children).

However, after adjustment for FFM and fat mass no

difference in REE of lean and obese children was seen.

The authors conclude that racial differences in nutri-

tional status as well as differences between obese and

normal weight children are more likely explained by

differences in physical activity rather than differences in

REE or DIT. The findings of this very detailed study

are contrary to previous data in infants [3] and adults [2].

The effect of parental overweight on energy expenditure

was investigated in a cross-sectional study among non-


obese girls [33 .]. This group may be considered as ‘pre-

obese’, that is, a metabolic risk of becoming overweight.

Body composition (as assessed by total body water),

TEE (measured by DLW) and REE (measured by

indirect calorimetry) were investigated in 196 non-obese

premenarcheal girls aged 8–12 years. When compared

with girls with two normal weight parents REE was

higher among girls with at least one overweight parent.

TEE was also higher among girls with two overweight

parents, but these results were of borderline significance.

Pubertal stage did not affect this effect. By contrast

differences in non-resting energy expenditure (as calcu-

lated from the difference between TEE and REE) were

associated with pubertal stage and race-ethnicity but not

with parental overweight. The authors conferred to three

previous studies in which either no effect of parental

overweight [34,35] or even a higher (rather than a lower)

REE was observed in normal weight children with

overweight parents when compared with normal weight

children with two normal weight parents [36]. Altogether

these results suggest that alterations in REE do not

predispose to weight gain in pre-obese girls. By contrast

decreases in non-resting energy expenditure with

maturation are associated with weight gain.

Familial predisposition to obesity was also assessed by

investigating skeletal muscle energetics by the use of31P-nuclear magnetic resonance [37]. Skeletal muscle

phosphate, phosphocreatine, the ratio of low and high

energy phosphates, intracellular pH, and adenosine

triphosphate were measured using a 3-min rest–exer-

cise–recovery plantar flexion protocol in 22 normal

weight girls with two lean parents, 18 girls with one lean

and one obese parent and 15 girls with two obese parents,

respectively. There were no between-group differences

in skeletal muscle energetics, questioning its role as a

metabolic precursor of childhood overweight. The results

of these two recent studies on familial predisposition to

obesity [33 .,37] are contrary to one aspect of a previous

study [34] in which a total of 74 prepubertal children

(mean age 5.0+0.9 years) were divided into different

groups according to the obesity state of their parents.

After adjustment for FFM there were no between-group

differences in TEE and AEE, but REE was about 6%

lower in children with either an obese mother or an obese

father than in children who had two lean or two obese

parents. These cross-sectional data were extended by a

longitudinal study (over a 4-year period) in prepubertal

children [38]. The average rate of change in absolute fat

Figure 1. (a) Relationships between resting energy expenditure and fat free mass in pre- and postpubertal children. (b) Effect of adjustment ofREE for FFM

REE(kcal/day)

REEadjustedfor FFM(kcal/day)

3500

3000

2500

2000

1500

1000

500

0

2500

2000

1500

1000

500

0

****

*

Female Male Female Male Female MalePrepubertal

children(5–12 years)

Postpubertalchildren

(13–17 years)

Youngadults

(18–35 years)

Postpubertal childrenand young adults

y = 20.879x + 482.16R2 = 0.4998

Prepubertal childreny = 24.907x + 580.26

R2 = 0.7177

0 20 40 60 80 100 120 140FFMBIA (kg)

Prepubertal girlsPrepubertal boysFemalesMales

(a) (b)

(a) Regression lines and respective equations are given for the relationships between resting energy expenditure (REE) and fat-free mass (FFM) in 150prepubertal children (81 boys, 69 girls; mean age 8.5 years; mean body mass index (BMI) 21.0 kg/m2), represented by triangle symbols, and in 24postpubertal children (12 boys, 12 girls; mean age 14.6 years; mean BMI 22.0 kg/m2) together with 432 young adults (184 males, 248 females; meanage 25.7 years; mean BMI 22.7 kg/m2), both represented by circular symbols. (b) Means and standard deviations for FFM-adjusted REE are plottedfor male and female participants of the three age groups. *P50.05; ***P50.001 for differences between sexes and by Mann Whitney U-test.


mass was 0.89+1.08 kg/year. Adjusting the change in fat

mass for FFM resulted in a rate of change of

0.08+0.64 kg/year. Similar results were observed in

children of two non-obese parents as well as in children

with one obese and one non-obese parent. However a

higher rate of change was seen in children with two obese

parents. Unfortunately none of the components of energy

expenditure was inversely related to changes in nutri-

tional state. In a further longitudinal study over a 5-year

period in markedly overweight Pima Indian children

changes in body size, energy expenditure and activity

were measured between the ages 5 and 10 years [39].

Although cross-sectionally the authors found (1) an

inverse association between body fat and sport participa-

tion and (2) a positive correlation between body weight

and television viewing. Prospectively none of the

variables measured was a predictor of body fat at age 10

years. Altogether energy expenditure does not appear to

be a major risk factor for the development of obesity

during prepubertal growth.

REE (as assessed by indirect calorimetry) and whole

body as well as regional body composition (as measured

by dual energy X-ray absorptiometry) were investigated

in 203 5–17-year-old obese African American and white

children and adolescents [40 .]. REE was lower in the

African Americans than in the white children. Ethnic

differences decreased after adjustment for between-

group differences in age, sex and FFM. Further

adjustment for trunk lean tissue mass partially explained

the lower REE of obese African American girls. This is

the most important result of this study. The data add to

the results of a previous study showing that among

premenopausal non-obese women, African American

women had significantly more limb lean tissue, less

trunk lean tissue, and a lower REE [41]. In this study

ethnic differences in REE disappeared after adjustment

for regional body composition. However another group

of authors showed that REE of African American

children remained lower after adjustments for body

composition including trunk lean tissue mass [41,42].

Although controversial, one may conclude that the lower

trunk tissue mass in African American children results in

less metabolically active mass and thus in a lower specific

REE. Although regional assessment of body composition

could explain a greater proportion of the ethnic group

difference in energy expenditure, the ethnic group

difference in REE remains. A further finding of the

above mentioned study [40 .] was that age may have a

significant effect on REE. Using different models the

authors calculated that the resting expenditure of an 18-

year-old obese adolescent would be 728 kJ/day, lower

than that for an 8-year-old obese child. The lower

relative REE in older children may explain limited

success rates in weight management interventions for

pediatric obesity.

Energy expenditure and physical activity inchildrenInactivity has been identified as an emerging serious

concern in children, leading to increasing incidence of

childhood overweight and obesity. Monitoring physical

activities in children is most frequently based on

questionnaires, which require adult mediators, removing

the direct involvement of the child in the process of data

acquisition and thus adding confounding factors. Mea-

surements of physical activity pose problems in children.

Recall of activities is inaccurate.

TEE (as measured by DLW) and physical activity (as

calculated from the TEE/REE ratio where REE was

predicted) were investigated in 106 healthy children

aged 7.8+0.9 years [43]. TEE in both boys and girls was

13 and 9% lower than the FAO/WHO/UNU-recommen-

dation [23]. In boys but not in girls the authors found a

negative association between physical activity and fat

mass (as indirectly assessed by the 18O-dilution space).

These cross-sectional data suggest that a low level of

physical activity is associated with a high fat mass and

thus may be a causal factor for obesity, at least in boys.

The relationship between physical activity and body

composition (in particular body fatness) was examined in

a number of studies. For example, TEE (as assessed by

24-h heart rate) and physical activity levels (as calculated

from the TEE/REE ratio where REE was measured by

indirect calorimetry) did not show an association with fat

mass in a group of 48 prepubertal children [44]. By

contrast, in this cross-sectional study REE was the most

important determinant of fat mass.

There were no differences in TEE (as assessed by

DLW), REE (by analyzing expired air collected in

Douglas bags) or AEE (as calculated by 0.96TEE

minus REE) between obese and normal weight adoles-

cents [45 ..]. However the physical activity level (as

calculated by the ratio between TEE and REE) was

lower in the obese group. The concomitant use of

accelerometry showed significant between-group differ-

ences in total physical activity (as expressed in counts

per minute per day), accumulated and continuous

duration of activity as well as continuous 10-min periods

of physical activity of moderate intensity which were all

reduced in obese boys and girls. Taken together obese

adolescents were less physically active but when

compared with controls PAEE was not reduced. This

finding leads to conceptual issues of physical activity

research, which becomes obvious when two different

methods for the assessment of physical activity are used.

The differences in physical activity between groups with

concomitant similar TEE and AEE may be explained by

the increased cost of moving a larger body mass in the

obese children. In this study the simultaneous use of

energy expenditure and activity monitor data provided


further insight into the influence of different activities on

energy balance. The between-group differences were

not due to sport and fitness-related activities of high

intensities. By contrast, everyday activities of moderate

intensities were significantly lower in the obese children

and had a strong influence on total physical activity. This

finding supports a concept that low to moderate activities

seem to influence TEE to a greater extent than does

high intensity physical activity [46].

TEE and activity patterns of 12–16-year-old adolescents

were assessed by diary and heart rate measurements

validated by whole body calorimetry [47]. Measurements

were performed to determine the inter-individual as well

as the day-to-day variations in energy expenditure. TEE

was higher in spring than in autumn. In active subjects it

was also higher at the weekend than during school days.

By contrast free day activity was lower in sedentary

subjects. Time and energy spent for sport activities

showed great inter-individual variability. The increase in

energy expenditure resulting from sport activities was

moderate and showed a great variance ranging from 0.2

to 2.7 MJ/day. This study clearly demonstrated the

limited value of selected measurements of energy

expenditure and physical activity.

The data of the above mentioned studies indicate the

limits of estimating physical activity from measurements

of energy expenditure. Physical activity is a behavior

that characterizes lifestyle. In contrast physical fitness (as

assessed by testing cardiorespiratory or aerobic fitness as

well as muscular-skeletal function, i.e. muscle strength)

is an attribute. Physical fitness results from physical

activity and training over longer time periods. In

addition, skill-related components (e.g. agility, coordina-

tion) have to be considered. Therefore, future studies on

the relationships between activity, fitness and fatness in

children should be extended beyond the measurements

of energy expenditure and movement. There is a need

for repeated measurements considering different days of

the week and seasons of the year. The possible

differences between the results of the different studies

cited might be explained by the use of different

parameters and methods, by ‘selected’ measurements

as well as by different adjustments of data (e.g. for body

composition).

Metabolic costs of activities in childrenThe metabolic cost of locomotion at any speed given,

expressed as per kilogram body weight, is greater for

children than for adults. This finding suggests that there

is an improvement of metabolic economy with age.

Metabolic economy was measured in children of

different ages using metabolic (i.e. oxygen consump-

tion), kinematic (total body mechanical power, energy

transfer rates) and electromyographic measurements in

groups of 7–8-, 10–12- and 15–16-year-old children [48].

Seventy-seven percent of the age-related variance in VO2

and 62% of the variation in metabolic efficiency was

explained by the variables tested. Age was the best

single predictor for VO2and metabolic efficiency.

Mechanical work and energy transfer were not different

between the age groups tested. By contrast muscular co-

contraction was an important predictor. It is speculated

that growth-related factors affect the metabolic costs of

locomotion as children grow up.

The determinants of energy cost of front-crawl swim-

ming were investigated in 11 children with a mean age

of 12.4 years over 200 m at three speeds [49]. Oxygen

consumption (by Douglas bag method), stroke frequency

and stroke length together with body composition (as

assessed by anthropometry) were measured. The authors

found no association between energy costs of crawling

and the stroke frequency or length and anthropometric

parameters.

The results of both studies [48,49] suggest that

metabolic economy and energy costs of physical activity

cannot be explained by standard measures of energy

expenditure, activity or movement. There is need to

consider further parameters such as detailed body

composition analysis, constitutional factors like propor-

tion of extremities and trunk, or co-ordination. With

respect to factors mentioned above we feel that most of

the unexplained differences are explained by the non-

homogeneity of the groups of subjects investigated.

Meal-induced thermogenesis in childrenDisturbances in DIT have been suggested to contribute

to a positive energy balance, weight gain and thus

overweight [50]. With respect to energy balance, fat

intake is of considerable interest. The underlying causes

are (1) the high energy content of fat and (2) the fact that

fat oxidation is not or is only slightly increased in

response to fat intake [51]. Thus a high fat intake

increases fat storage. To address this problem DIT was

measured by indirect calorimetry in children after a low

fat (20% of energy intake) and a high fat meal (48% of

energy intake) [52]. Measurements were performed in

six obese and six non-obese girls aged about 10 years.

The meals were given in a randomly assigned fashion.

DIT (as expressed as a percentage of energy intake) was

significantly higher after the low fat (6.5+0.7%) than

after the high fat meal (4.3+0.4%). The postprandial fat

storage was eight-fold higher after the high fat meal.

These data suggest that when compared with carbohy-

drates (i.e. a low fat meal) fat intake induces a lower DIT

and a higher fat storage. A high fat intake may thus

contribute to childhood overweight. However there were

no significant differences in DIT between obese and

non-obese children. Therefore, the effect seems to be


independent of the adiposity level of the subjects. This

study showed that there is no clear impairment of

thermogenesis in obese or pre-obese children. The

study, therefore, pointed to the current recommendation

for fat intake in children, which states that 30% of energy

from dietary fat for children aged over 2 years is

sufficient for adequate growth and development [53].

Children can meet their energy requirements from a diet

providing 30% energy from fat. However, if one

recalculates the data of the above mentioned study

[54] there are only minor between-group differences

observed after a high fat versus a low fat meal. DIT

accounted for 26 kcal (108 kJ) versus 39 kcal (163 kJ)

after high fat versus low fat meals, respectively. If we

assume three identical meals per day and no metabolic

adaptations (i.e. an identical postprandial metabolism)

the between-group differences will reach a total of

39 kcal (163 kJ) per day. These data should be re-

considered in face of the relatively huge between-group

differences in fat intake (32 versus 13 g per meal). The

majority of children will consume mixed instead of low

or high fat meals (and thus will have a fat intake of about

23 g/meal, i.e. fat intake is about 35% of energy intake).

It is tempting to speculate that DIT will be somewhere

between 4.3 and 6.5% in response to a mixed meal and

the between-group differences observed after a high fat

versus a low fat meal will thus be further reduced.

Nevertheless these minor effects of fat intake may add

to a positive energy balance and overweight if one

considers longer time periods (i.e. months, years or

decades).

Energy expenditure in ill childrenCritically ill patients frequently show hypermetabolism

and tissue catabolism, suggesting altered nutritional

needs, and thus require a special metabolic support.

REE (by indirect calorimetry) and nitrogen balance

(from nitrogen intake as calculated from parenteral

supply and by measuring urinary nitrogen by the

Kjeldahl method) were investigated in 33 critically ill

children aged about 5 years [54]. The study population

received parenteral nutrition and was mechanically

ventilated. The average energy supplied was over 25%

of their measured energy needs. When compared with

predicted energy requirements mean group REE was

increased by about 20%. Concomitantly nitrogen balance

was negative (789+166 mg per kg body weight and

day). Comparing hypermetabolic patients (as defined by

a measured REE410% of the predicted value) with

‘normo-metabolic’ patients (i.e. measured REE was

within the +10% prediction range) showed a higher

fat oxidation in the former group suggesting that fat is

preferentially oxidized. A high carbohydrate intake was

associated with lipogenesis and thus with an increase in

thermogenesis. Seventy-five percent of the participants

consuming high amounts of carbohydrates had a non-

protein respiratory quotient value above 1.0. The non-

protein respiratory quotient is calculated from the

measured respiratory quotient (i.e. VCO2/VO2

) after

correction for protein oxidation as measured by urinary

nitrogen excretion after correction for changes in serum

urea levels. Thus, the non-protein respiratory quotient

reflects the ‘net’ metabolism of glucose and lipids. A

further finding of this study was that a negative nitrogen

balance was associated with a high protein oxidation. A

high protein intake (about 2.8 g/kg per day) resulted in a

positive nitrogen balance.

How can we tackle hypermetabolism in clinical practice?

This question was addressed in a landmark study by the

group of Robert Wolfe from Shriners Hospital for

children in Galveston, Texas [55]. The authors investi-

gated the effect of a 2-week beta blockade (by

propranolol) on REE and muscle-protein balance in 25

children with acute and severe burns (more than 40%

of body surface area was involved in burn injury). The

study was a randomized trial. REE (by indirect

calorimetry), protein kinetics (by the use of labeled

amino acids) and body composition (by the use of two

independent methods, a whole-body potassium-40

scintillation counter and dual-energy X-ray absorptiome-

try) were assessed before and after 2 weeks of beta

blockade or no treatment (control). Propranolol was

given in doses that decreased heart rate by approxi-

mately 20% from baseline levels. REE increased in

controls – +140+67 kcal/day (583+279 kJ/day) – but

decreased after beta blockade: 7422+197 kcal/day

(1759+821 kJ/day). This effect was associated with an

improvement of muscle protein balance as well as a

better conservation of body cell mass. The authors also

found that propranolol induced an increase in intracel-

lular recycling of free amino acids. These data suggest

that catecholamines are primary mediators of elevated

energy expenditure and tissue catabolism in patients

with severe burns. Long term beta blockade was capable

of decreasing REE and tissue catabolism. The authors

therefore suggested that the therapy may benefit

critically ill children with hypermetabolism. They also

mentioned the side effects of high doses of propranolol

(i.e. a decrease in cardiac output and severe broncho-

spasm). Thus hypermetabolic patients treated by beta

blockade have to undergo continuous thermodynamic

and respiratory monitoring.

Children with chronic diseases such as cystic fibrosis are

a special challenge for clinical nutritionists because of a

need for an adequate energy and nutrient supply for

optimal growth and development. Measurements of

REE are difficult in this group of patients, due to

respiratory distress (i.e. reduced pulmonary function)

and variations in disease activity, resulting in a low

reproducibility. The variations in measurements of REE


(by indirect calorimetry) were addressed in a recent

study [56]. REE was measured for 20 min in 31 children

with cystic fibrosis and 32 healthy children. Anthropo-

metric measurements were used to assess body composi-

tion. No short-term differences between repeated

measurements in children with cystic fibrosis and

healthy children were observed. The measurement

errors were 29 kcal (119 kJ) and 42 kcal (177 kJ) in

cystic fibrosis and healthy children respectively. There

were close associations between REE and FFM in both

groups. Seventy to 80% of the variance in REE was

explained by FFM. Long-term stability of REE was

assessed in another 14 children with cystic fibrosis with

subsequent measurements 1 or 2 years later. There were

no significant differences in REE after adjustment for

body size (i.e. the mean difference between the two

measurements was about 4%). Concomitantly the

children showed no differences in pulmonary function.

This study showed that in children with cystic fibrosis

REE-measurements are highly stable over short as well

as long periods of time. These data were contrary to

results from a previous study in adult patients with cystic

fibrosis showing a higher variability of REE in patients

when compared with age- and sex-matched healthy

controls [57]. The authors of that study explained the

higher variations in cystic fibrosis patients by changes in

systemic inflammatory activity. We feel that the lower

variance observed in children is most likely explained by

their stable clinical condition at mild to moderate disease

activity as well as by the fact that all children were very

familiar with the study protocol (i.e. physiological

measurements). However, as the authors stated them-

selves care must be taken when extrapolating these data

to children with severe pulmonary disease.

In another study REE of children with sickle cell anemia

was investigated [58]. These children have decreased

height and weight and less fat and muscle mass when

compared with their healthy peers. Concomitantly

patients with sickle cell anemia have REE measure-

ments that are about 20% higher than in healthy controls

suggesting that increased REE contributes to their poor

nutritional status. Increased REE is most probably

explained by anemia (low hemoglobin). Because hemo-

globin is the oxygen carrier, low hemoglobin levels lead

to increased cardiac output as a compensatory mechan-

ism to provide oxygen to tissues of the body. REE was

measured by indirect calorimetry in 18 patients and

compared with predicted values. In children with sickle

cell anemia measured REE values were 12–15% higher

than the predicted values. Based on these measurements

the authors developed a prediction formula applying

multiple linear regression analysis. Besides weight,

height, age and sex hemoglobin concentrations were

also used to predict REE. The formula was subse-

quently validated in a different group of 20 patients with

sickle cell anemia. It seems to be useful when

measurements of REE are not available.

ConclusionMethods for assessment of the different components of

energy expenditure have been validated in children and

adolescents. The combined use of these methods

together with detailed analyses of body composition is

recommended for future studies. In patients with acute

or chronic illness measurements of energy expenditure

are necessary if disease-specific algorithms are not

available.

AcknowledgementOur own data presented within this review were supported by grantsfrom Deutsche Forschungsgemeinschaft (DFG Mue-714-8.1) andPrecon GmbH, Bickenbach, Germany.

References and recommended readingPapers of particular interest, published within the annual period of review, havebeen highlighted as:. of special interest.. of outstanding interest

1 Suettmann U, Ockenga J, Hoogestraat L, et al. Resting energy expenditureand weight loss in human immunodeficiency virus-infected patients.Metabolism 1993; 42:1173–1179.

2 Ravussin E, Lillioja S, Knowler WC, et al. Reduced rate of energyexpenditure as a risk factor for body-weight gain. N Engl J Med 1988;318:467–472.

3 Roberts SB, Savage S, Coward WA, et al. Energy expenditure and intake ininfants born to lean and overweight mothers. N Engl J Med 1988; 318:461–466.

4 Jequier E, Tappy L. Regulation of body weight in humans. Physiol Rev 1999;79:451–480.

5 Luke A, Schoeller DA. Basal metabolic rate, fat-free mass, and body cellmass during energy restriction. Metabolism 1992; 41:450–456.

6 Votruba S, Blanc S, Schoeller DA. Pattern and cost of weight gain inpreviously obese women. Am J Physiol Endocrinol Metab 2002; 282:E923–E930.

7 Mueller MJ, Bosy-Westphal A, Kutzner D, Heller M. Metabolically activecomponents of fat-free mass and resting energy expenditure in humans:recent lessons from imaging technologies. Obes Rev 2002; 3:113–122.

8 Wang Z, O’Connor TP, Heshka S, Heymsfield SB. The reconstruction ofKleiber’s Law at the Organ-Tissue level. J Nutr 2001; 131:2967–2970.

9 Spiegelman BM, Flier JS. Obesity and the regulation of energy balance. Cell2001; 104:531–543,

10 Torun B, Davies PSW, Livingstone MBE, et al. Energy requirements anddietary energy recommendations for children and adolescents. Eur J Clin Nutr1996; 50:S37–S81.

11 Black AE, Coward WA, Cole TJ, Prentice AM. Human energy expenditure inaffluent societies: an analysis of 574 doubly-labeled water measurements.Eur J Clin Nutr 1996; 50:72–92.

12 Murgatroyd PR, Shetty PS, Prentice AM. Techniques for measurement ofhuman energy expenditure: a practical guide. Int J Obes 1993; 17:549–568.

13 Schutz Y. The basis of direct and indirect calorimetry and their potentials.Diabetes Metab Rev 1995; 4:383–408.

14 Dietz WH, Bandini LG, Morelli JA, et al. Effect of sedentary activities onresting metabolic rate. Am J Clin Nutr 1994; 59:556–559.

15 Eston RG, Rowlands AV, Ingledew DK. Validity of heart rate, pedometry, andaccelerometry for predicting the energy cost of children’s activities. J ApplPhysiol 1998; 84:362–371.

16 Ventham JC, Reilly JJ. Reproducibility of resting metabolic rate measurementin children. Br J Nutr 1999; 81:435–437.


17 Figueroa-Colon R, Franklin FA, Goran MJ, et al. Reproducibility ofmeasurements of resting energy expenditure in prepubertal girls. Am J ClinNutr 1996; 64:533–536.

18 White MS, Shepherd RW, McEniery JA. Energy expenditure measurementsin ventilated critically ill children: within- and between-day variability. JPENParenter Enteral Nutr 1999; 23:300–304.

19 Beghin L, Michaud L, Guimber D, et al. Assessing sleeping energyexpenditure in children using heart-rate monitoring calibrated against open-circuit indirect calorimetry: a pilot study. Br J Nutr 2002; 88:533–543.

20 Puyau MR, Adolph AL, Vohra FA, Butte NF. Validation and calibration ofphysical activity monitors in children. Obes Res 2002; 10:150–157.

21 Rodriguez G, Beghin L, Michaud L, et al. Comparison of the TriTrac-R3Daccelerometer and a self report activity diary with heart-rate monitoring for theassessment of energy expenditure in children. Br J Nutr 2002; 87:623–631.

22 Rodriguez G, Moreno LA, Sarria A, et al. Resting energy expenditure inchildren and adolescents: agreement between calorimetry and predictionequations. Clin Nutr 2000; 21:255–260.

23 World Health Organization. Energy and protein requirements. Report of ajoint FAO/WHO/UNU expert consultation. WHO Technical Report SeriesNo. 724. Geneva: World Health Organization; 1985.

24 Henry CJK, Dyer S, Ghusein-Choueiri A. New equations to estimate basalmetabolic rate in children aged 10–15 years. Eur J Clin Nutr 1999; 53:134–142.

25 Bitar A, Fellmann N, Vernet J, et al. Variations and determinants of energyexpenditure as measured by whole-body indirect calorimetry during pubertyand adolescence. Am J Clin Nutr 1999; 69:1209–1216.

26 Roemmich JN, Clark PA, Walter K, et al. Pubertal alterations in growth andbody composition. V. energy expenditure, adiposity, and fat distribution. Am JPhysiol 2000; 279:E1426–E1436.

27 Ferraro R, Llioja S, Fontvielle AM, et al. Lower sedentary metabolic rate inwomen compared with men. J Clin Invest 1992; 90:780–784.

28 Maffeis C, Schutz Y, Micciolo R, et al. Resting metabolic rate in six- to ten-year-old obese and non-obese children. J Pediatr 1993; 122:556–562.

29 Molnar D, Schutz Y. The effect of obesity, age, puberty and gender on restingmetabolic rate in children and adolescents. Eur J Pediatr 1997; 156:376–381.

30 Goran MI, Nagy TR, Gower BA, et al. Influence of sex, seasonality, ethnicityand geographic location on the components of total energy expenditure inyoung children: implications for energy requirements. Am J Clin Nutr 1998;68:675–682.

31 Grund A, Vollbrecht H, Frandsen W, et al. No effect of gender on differentcomponents of daily energy expenditure in free living prepubertal children. IntJ Obes 2000; 24:299–305.

32.

DeLany JP, Bray GA, Harsha DW, Volaufova J. Energy expenditure inpreadolescent African American and white boys and girls: The Baton RougeChildren’s Study. Am J Clin Nutr 2002; 75:705–713.

This study uses state of the art methods for assessment of different components ofenergy expenditure and body composition in a considerable number of normalweight and obese children. It could be shown that between group differences inenergy expenditure are due to physical activity rather than resting metabolic rate.

33.

Bandini LG, Must A, Spadano JL, Dietz WH. Relation of body composition,parental overweight, pubertal stage and race-ethnicity to energy expenditureamong premenarcheal girls. Am J Clin Nutr 2002; 76:1040–1047.

Parental weight had an influence on differences in REE and TEE of prepubertalgirls.

34 Goran MI, Carpenter WH, McGloin A, et al. Energy expenditure in children oflean and obese parents. Am J Physiol Endocrinol Metab 1995; 268:E917–E924.

35 Treuth MS, Butte NF, Wong WW. Effects of familial predisposition to obesityon energy expenditure in multiethnic prepubertal girls. Am J Clin Nutr 2000;71:893–900.

36 Wurmser H, Laessle R, Jacob K, et al. Resting metabolic rate in pre-adolescent girls at high risk of obesity. Int J Obes 1998; 22:793–799.

37 Treuth MS, Butte NF, Herrick R. Skeletal muscle energetics assessed by31P-NMR in prepubertal girls with a familial predisposition to obesity. Int JObes 2001; 25:1300–1308.

38 Goran MI, Shewchuk R, Gower BA, et al. Longitudinal changes in fatness inwhite children: no effect of childhood energy expenditure. Am J Clin Nutr1998; 67:309–316.

39 Salbe AD, Weyer C, Harper I, et al. Assessing risk factors for obesitybetween childhood and adolescence: II Energy metabolism and physicalactivity. Pediatrics 2002; 110:307–314.

40.

Tershakovec AM, Kuppler KM, Zemel B, Stallings VA. Age, sex, ethnicity,body composition, and resting energy expenditure of obese African Americanand white children and adolescents. Am J Clin Nutr 2002; 75:867–871.

This study adresses the impact of regional body composition (as assessed byDEXA) on REE. Adjustment for trunk lean tissue partially explained the lower REEof obese African American children. The data clearly show that whole body as wellas trunk FFM are both determinants of REE.

41 Hunter GR, Weinsier RL, Darnell BE, et al. Racial differences in energyexpenditure and aerobic fitness in premenopausal women. Am J Clin Nutr2000; 71:500–506.

42 Sun M, Gower BA, Nagy TR, et al. Total, resting, and activity-related energyexpenditures are similar in Caucasian and African-American children. Am JPhysiol Endocrinol Metab 1998; 274:E96–E101.

43 Ball EJ, O’Connor J, Abbott R, et al. Total energy expenditure, body fatnessand physical activity in children aged 6–9y. Am J Clin Nutr 2001; 74:524–528.

44 Mueller MJ, Grund A, Krause H, et al. Determinants of fat mass in prepubertalchildren. Br J Nutr 2002; 88:545–555.

45. .

Ekelund U, Aman J, Yngve A, et al. Physical activity but not energyexpenditure is reduced in obese adolescents : a case control study. Am J ClinNutr 2002; 76:933–941.

Obese adolescents are less physically active but energy costs of activity arenormal

46 Westerterp KR. Pattern and intensity of physical activity. Nature 2001;410:539.

47 Vermorel M, Vernet J, Bitar A, et al. Daily energy expenditure, activitypatterns, and energy costs of the various activities in French 12–16-y-oldadolescents in free living conditions. Eur J Clin Nutr 2002; 56:819–829.

48 Frost G, Bar-Or O, Dowling J, Dyson K. Explaining differences in themetabolic cost and efficiency of treadmill locomotion in children. J Sports Sci2002; 20:451–461.

49 Ponjade B, Hautier CA, Rouard A. Determinants of the energy cost of front-crawl swimming in children. Eur J Appl Physiol 2002; 87:1–6.

50 Jequier E. Energy expenditure in obesity. Clin Endocrinol Metab 1984;13:563–580.

51 Schutz Y, Flatt JP, Jequier E. Failure of dietary fat intake to promote fatoxidation: a factor favoring the development of obesity. Am J Clin Nutr 1989;50:307–314.

52 Maffeis C, Schutz Y, Grezzani A, et al. Meal-induced thermogenesis andobesity: Is a fat meal a risk factor for fat gain in children. J Clin EndocrinMetab 2001; 86:214–219.

53 Butte NF. Fat intake of children in relation to energy requirements. Am J ClinNutr 2002; 72 (Suppl):1246S–1252S.

54 Coos-Bu JA, Klish WJ, Walding D, et al. Energy metabolism, nitrogenbalance, and substrate utilization in critically ill children. Am J Clin Nutr 2001;74:664–669.

55 Herndon DN, Hart DW, Wolf SE et al. Reversal of catabolism by beta-blockade after severe burns. N Engl J Med 2001; 345:1223–1229.

56 Ashley MA, Broomhead L, Allen JR, Gaskin KJ. Variations in themeasurement of resting energy expenditure in children with cystic fibrosis.Eur J Clin Nutr 2001; 55:896–901.

57 Bell SC, Elborn JS, Nixon LE, et al. Repeatability and methodology of restingenergy expenditure in patients with cystic fibrosis. Respir Physiol 1999;115:301–307.

58 Williams R, Olivi S, Mackert P, et al. Comparison of energy expenditureprediction equations with measured resting energy expenditure in childrenwith sickle cell anemia. J Am Diet Assoc 2002; 102:956–961.


Assessment of energy expenditure in children and adolescents

Documents