Cover Image
close this bookEnergy and Protein requirements, Proceedings of an IDECG workshop, November 1994, London, UK, Supplement of the European Journal of Clinical Nutrition (International Dietary Energy Consultative Group - IDECG, 1994, 198 pages)
close this folderEnergy requirements and dietary energy recommendations for children and adolescents 1 to 18 years old
View the document(introductory text...)
View the documentIntroduction
View the documentTotal daily energy expenditure (TEE)
View the documentEstimates of basal metabolic rate to calculate total energy expenditure
View the documentTime allocation to different activities
View the documentPhysical activity levels of children and adolescents
View the documentDietary energy intake
View the documentGeneral conclusions and recommendations
View the documentReferences
View the documentDiscussion

Total daily energy expenditure (TEE)

Three types of methods have been used to calculate total daily energy expenditure of free-living children and adolescents. Their advantages and limitations have been reviewed by several authors (e.g. Torun, 1984; Durnin, 1990).

(1) Doubly-labeled water. This method has two components: (a) Administration of a marker dose of 2H and 18O, and measurement of the disappearance of the isotope from the body after several days and (b) Calculation of the food quotient or estimation of the average respiratory quotient during that period of time.

The doubly-labeled water is the most accurate of the three methods. However, there are still some doubts about the appropriateness of the assumptions used for the calculation of energy expenditure. Moreover, the number of children so far studied is very small and restricted to few geographical areas due to the high cost of the isotopes and their analysis. Furthermore, it does not provide information on the patterns of physical activity throughout the day.

(2) Heart rate monitoring. This method has three components. (a) Measurement of heart rate while resting and measurement or estimation of the resting and basal metabolic rates. (b) Determination of the relationship between heart rate and oxygen consumption (or energy expenditure) with light, moderate and moderately heavy workloads. This relationship varies among individuals and must be established for every person who will be studied. (c) Minute-by-minute recording of heart rate.

Table 1 Comparison of total daily energy expenditures (MJ/d) measured with two different heart rate monitoring techniques (three-way analysis of variance with unweighted averages)

Minute-by-minute heart rate method (Spurr and Reina, 1988a)

Age (y)

n

Control children
mean

s.d.

n

Mildly malnourished
mean

s.d.

6-8

24

6.6

1.6

21

5.1

1.0

10-12

18

8.4

2.3

23

7.8

2.1

14-16

20

12.1

2.7

26

10.6

2.5

Daytime heart rate accumulation method (Spurr et al., 1986)

Age (y)

n

Control children
mean

s.d.

n

Mildly malnourished
mean

s.d.

6-8

12

6.4

1.1

9

6.3

0.7

10-12

20

9.2

1.8

19

8.0

1.8

14-16

12

11.5

2.0

16

10.2

2.1

Source

F ratio

Probability

A (Age)

109.283

<0.001

B (Nutritional status)

14.463

<0.001

C (Method)

0.413

0.521

A*B (Age*Nutrition)

0.403

0.668

A*C (Age*Method)

1.495

0.226

B*C (Method*Nutrition)

0.462

0.498

A*B*C (Age*Nutrition*Method)

1.054

0.350

Earlier studies used recorders that accumulated all heart beats over some period of time. Average heart rate over 24 h gave unacceptable results due to the poor relationship between heart rate and oxygen consumption at resting and sedentary levels of energy expenditure. However, the method yields acceptable results when the average heart rate is calculated for the period of time when children are awake, and energy expenditure calculated for the remainder of the 24 h from the resting and basal metabolic rates. As shown in Table 1, an analysis of studies by Spurr and collaborators (Spurr et al, 1986; Spurr and Reina, 1988a) indicated that the results with this heart rate accumulation method did not differ from those obtained in similar children with the minute-by-minute rate recording method.

The heart rate monitoring method has been validated with whole body calorimetry and doubly-labeled water. Comparisons varied on an individual basis, but the mean values for groups of individuals were similar to the other methods (Spurr et al, 1988; Ceesay et al, 1989; Livingstone et al, 1990a, 1992a; Emons et al, 1992). Thus, heart rate monitoring can be used to estimate the energy expenditure of groups of children. Minute-by-minute recording also allows, examining the time allocated to different intensities of physical effort.

(3) Time-motion or activity diary techniques. These have two components: (a) assessment of time allocation, which has been explored by direct observations with different timing techniques, by activity records or diaries kept by the subjects or caretakers of young children, and by recall interviews with subjects or caretakers and (b) energy costs of the activities that are observed or recorded, measured by indirect calorimetry or estimated from published values. It should be borne in mind that energy costs of activities published for adults do not apply to children under 15 (Town, 1983, 1990a). Many of the results published with the time motion or diary techniques are questionable due to the inaccuracies inherent in methods based on reporting and in the application of energy cost of activities of adults to calculate energy expenditure of children.

We have based this review and our conclusions on TEE on studies with doubly-labeled water or with appropriate techniques of heart rate monitoring. Some estimates of energy expenditure with time motion/diary techniques were selected as examples to examine the conclusions based on the other two methods.

Another relatively simple way to estimate total daily energy expenditure (TEE), and therefore requirements, of adults was proposed by the 1985 FAO/WHO/UNU Expert Consultation. Sleeping, occupational, discretional, health-promoting and other miscellaneous activities were assigned an energy cost, expressed as multiples of basal metabolic rate ( × BMR) or physical activity level (PAL)1. A factorial calculation accounting for the time allotted to each of those activities allowed the estimation of the mean PAL in 24 h. For populations engaged in occupational activities of different intensities. TEEs of 1.55, 1.78 and 2.10 × BMR were proposed for men with light, moderate and heavy occupational activities, respectively (FAO/WHO/UNU, 1985). The corresponding factors suggested for women were 1.56, 1.64 and 1.82.

We suggest that a similar approach be used to estimate TEE of groups of children and adolescents with different lifestyles. PAL factors are proposed for those estimates in another section of this document ('Physical activity levels of children and adolescents').

Studies with doubly-labeled water

The doubly-labeled water technique is almost 40 years old, but there are still relatively few data on total energy expenditure in children, due primarily to cost. Until the mid 1970s the cost of the stable isotopes (2H and 18O) involved in the technique was restrictive. Since that time advances in technology, notably the development of highly precise isotope mass ratio spectrometers, made it possible to administer significantly less isotope, thus reducing the cost to a more manageable figure. Unfortunately, the cost of isotopes began to rise steeply in 1990 and once again fewer studies are being undertaken.

Studies that allow consideration of TEE and dietary recommendations have been done with well-nourished children and adolescents in urban centers of the United Kingdom (Prentice et al, 1988; Davies et al, 1991, 1994), Holland (Saris et al, 1989; Emons et al, 1992) and the United States (Bandini et al, 1990b; Goran et al, 1993; Fontvieille et al, 1993; Wong, 1994). Table 2 shows their age span and TEE expressed per day, per unit of body weight and PAL.

Figure 1 compares the data from Table 2, expressed as kcal/kg/day, with the FAO/WHO/UNU 1985 recommendations, with and without the allowance for growth. The values for energy expenditure shown in the table and figure do not include the small proportion of energy that should be retained for growth (between 1 and 3%, depending on age).

Current dietary energy recommendations are about 20% higher than energy expenditure of children under 7 years of age in industrialized societies. From 7 years onwards, current recommendations coincide reasonably well with the data from doubly-labeled water studies, although boys throughout adolescence and girls around puberty seem to require 5-15% more dietary energy.

1 Total energy expenditure expressed as × BMR has been considered to reflect an individual's or population's physical activity level (PAL). This term has appeared with increasing frequency in the scientific literature. Thus, we will use it as synonym of × BMR.


FIGURE 1a
Total energy expenditure estimated with doubly labeled water: boys.


FIGURE 1b
Total energy expenditure estimated with doubly labeled water: girls.

Within each sex, the PALs in Table 2 show a trend towards uniformity among children between 1 and 5, 6 and 13, and 14 and more years of age. Therefore, the mean PAL values for those three age groups were calculated (Table 3). Since the studies had similar sample sizes within each age group, calculations of the mean PAL weighted for the number of children in each study gave similar results. Measured BMRs were used for calculations in three studies and BMR estimates with Schofield's equations in all others.

On the average, there were no gender differences at 1-5 and 6-13 years. Boys seemed to have higher PAL than girls after that age, but this observation was based on only three data sets from two studies (Bandini et al, 1990b; Davies et al, 1991). It should also be noted that if PALs were calculated with the equations published by FAO/WHO/UNU (1985) instead of those modified later by Schofield (1985), they would be somewhat lower for girls 1-5 and 6-13 years old than for boys of the same age ranges (1.39 vs 1.47 and 1.74 vs 1.81, respectively).

It should be kept in mind that the values shown in Tables 2 and 3 correspond to studies in a small number of well nourished children with adequate growth patterns, living in societies where food and health services are continuously and readily available.

Table 2 Groups of children, classified by sex and age, whose total daily energy expenditure has been estimated by the doubly labeled water method (does not include 1-3%, depending on age, that should be retained for growth)



Total energy expenditure


Agea (y)

n

Weight (kg)

(kcal/d)

(kcal/kg/d)

PALb

Reference

Boys

1-1.9

8c



83.0d


Prentice et al (1988)

1.5-2.49

11

12.6 ± 1.4e

1075 ± 305

85.8 ± 26.0

1.49

Davies et al (1994)

2-2.9

6c



81.0d


Prentice et al (1988)

2.5-3.49

15

15.0 ± 1.7

1207 ± 181

81.5 ± 15.2

1.41

Davies et al (1994)

3-4

13

15.5

1300

83.9 ± 11.5

1.52

Davies et al (1991)

3-4.49

16

16.9 ± 2.3

1301 ± 211

78.2 ± 14.5

1.47

Davies et al (1994)

4-6

16

20.3 ± 4.3

1438 ± 271

71.5 ± 8.0

1.49

Goran et al, (1993)

5.4 ± 0.3

15

21.1 ± 3.9

1415 ± 252

67.1

1.44

Fontvieille et al (1993)






(1.36)


5-6

12

18.9

1654

87.5 ± 10.0

1.77

Davies et al (1991)

7-8

10

24.6

1958

79.6 ± 9.1

1.84

Davies et al (1991)

9.3 ± 1.4

9

30.9 ± 4.3

2151

69.6

1.78

Saris et al (1989)






(1 77)


9-10

14

29.5

2180

73.9 ± 12.2

1.86

Davies et al (1991)

12-13

8

39.7

2334

58.8 ± 9.8

1.71

Davies et al (1991)

14.5 ± 1.5

13

56.4 ± 10.2

3109 ± 506

56.3 ± 6.4

1.88

Bandini et al (1990b)






(1.79)


15-16

12

60.1

3233

53.8 ± 7.6

1.88

Davies et al (1991)

18-19

12

71.6

3437

48.0 ± 4.3

1.86

Davies et al (1991)

Girls

1-1.9

7c



83.0d


Prentice et al (1988)

1.5-2.49

12

13.0 ± 1.9

1062 ± 212

83.0 ± 19.5

1.46

Davies et al (1994)

2-2.9

6c



81.0d


Prentice et al (1988)

2.5-3.49

16

14.9 ± 1.1

1125 ± 211

75.8 ± 15.0

1.38

Davies et al (1994)

3-4

18

14.8

1150

77.7 ± 10.3

1.46

Davies et al (1991)

3.5-4.49

11

17 ± 2.0

1263 ± 237

74.2 ± 11.0

1.52

Davies et al (1994)

4 6

14

21.0 ± 4.7

1344 ± 314

63.5 ± 5.6

1.47

Goran et al (1993)

5.5 ± 0.4

13

18.9 ± 2.5

1318 ± 189

69.7

1.51

Fontvieille et al (1993)






(1.37)


5-6

16

18.5

1473

79.6 ± 10.5

1.71

Davies et al (1991)

7-8

15

26.0

1989

76.5 ± 17.7

1.96

Davies et al (1991)

8.1 ± 1.3

10

28.2 ± 2.6

1926

68.3

1.82

Saris et al (1989)






(1.69)


9-10

15

29.1

1816

62.4 ± 10.5

1.69

Davies et al (1991)

12-13

10

49.3

2569

52.1 ± 7.9

1.90

Davies et al (1991)

13.2 ± 1.8

9

43.3 ± 8.9

2321 ± 281

53.6

1.82

Wong, (1994)

14.3 ± 1.0

12

55.7 ± 9.4

2385 ± 446

43.9 ± 7.7

1.66

Bandini et al (1990b)






(1.69)


15-16

11

58.0

2453

42.3 ± 6.0

1.67

Davies et al (1991)

18-19

11

62.4

2533

40.6 ± 7.6

1.72

Davies et al (1991)

a Range or mean ± standard deviation.
b Physical Activity Level calculated using basal metabolic rates estimated with Schofield's equations (1985) (or, in parenthesis, measured experimentally).
c Assuming 50% of the children studied were boys and 50% girls.
d Assuming the same values for boys and girls.
e Mean ± s.d.

Table 3 Mean physical activity levels of children in Table 2 grouped by age and sex. (Total energy expenditure measured with doubly labeled water; BMR's were measured or estimated with Schofield's equations)

Age (y)

Boys

Girls

1-5

1.46 ± 0.06 (6)a

1.44 ± 0.06 (6)

6-13

1.79 ± 0.06 (5)

1.80 ± 0.12 (6)

14+

1.84 ± 0.05 (3)

1.69 ± 0.03 (3)

a Mean ± s.d. of mean values in Table 2. Number of data sets in parenthesis. Means weighted by the number of children in each study gave similar values.

Studies with heart rate monitoring

Studies to calculate TEE of children and adolescents through heart rate monitoring have also been done only in a few countries, but they include industrialized and developing societies. These studies were done with either daytime heart rate accumulation (Spady, 1980; Torun & Viteri, 1981a,b; Spurr et al, 1986; Spurr & Reina, 1987) or the minute-by-minute heart rate method (Spurr & Reina, 1988a,b, 1989a,b; Livingstone et al, 1992a; Emons et al, 1992; Torun et al, 1993; Ramirez & Torun, 1994). Table 4 shows their age span and results. All studies involved children living in urban centers. Those in Northern Ireland and Holland involved only between 3 and 6 children in each sex-and-age group, and those in Canada were clone with 11 boys and 10 girls. Most studies in Colombia and Guatemala involved 16-34 boys or girls in each age group (median sample size = 20).

Table 4 Groups of children, classified by sex and age, whose total daily energy expenditure was estimated by heart rate monitoring methods (does not include 1-3% energy, depending on age, that should be retained for growth)




Total energy expenditure






Age (y)

n

Weight
(kg)

(kcal/d)

(kcal/kg/d)

PALa

Country

Conditionc

Methodd

Source

Boys

2.5 ± 0.7

6

11.9 ± 1.0

1060

89.1 ± 9.0

1.56b

Guatemala

Stunted

Accum

Torun & Viteri (1981b)

3.1 ± 0.3

11

12.0 ± 0.8

901

74.8 ± 7.6

1.34

Guatemala

Stunted

Accum

Torun & Viteri (1981a)

6.8 ± 0.5

24

21.9 ± 1.6

1581 ± 374

72.3 ± 16.8

1.60

Colombia


M-M

Spurr & Reina (1988a)

7.0 ± 0.5

12

21.8 ± 1.4

1541 ± 255

70.2 ± 8.4

1.54b

Colombia


Accum

Spurr et al (1986)

7.0 ± 0.5

21

19.3 ± 1.7

1207 ± 243

62.7 ± 12.5

1.46

Colombia

Underweight

M-M

Spurr & Reina (1988a)

7.4 ± 0.7

9

19.4 ± 2.3

1502 ± 176

81.0 ± 9.7

1.59b

Colombia

Underweight

Accum

Spurr et al(1986)

7.5 ± 0.3

6

25.4 ± 6.6

1859 ± 388

74.4 ± 12.2

1.64

UK


M-M

Livingstone et al (1992a)

8.4

5

27.8

2414 ± 394

86.8 ± 14.2

2.13b

Holland


M-M

Emons et al(1992)

9.3 ± 0.2

5

30.2 ± 9.4

2119 ± 182

74.5 ± 17.7

1.88

UK


M-M

Livingstone et al (1992a)

9.4 ± 1.0

11

32.1 ± 4.4

2164 ± 199

66.4 ± 9.8

1.86

Canada


Accum

Spady (1980)

10.8 ± 0.5

34

33.3 ± 2.8

2051 ± 400

61.7 ± 13.0

1.75

Guatemala


M-M

Ramirez & Torun (1994)

10.9 ± 0.6

19

25.9 ± 2.6

1918 ± 425

73.6 ± 16.6

1.72b

Colombia

Underweight

Accum

Spurr et al (1986)

11.0 ± 0.6

20

32.4 ± 3.3

2209 ± 419

68.1 ± 12.7

1.79b

Colombia


Accum

Spurr et al (1986)

11.1 ± 0.6

14

33.1 ± 2.3

2009 ± 421

60.7 ± 12.7

1.67

Colombia


M-M

Spurr & Reina (1988b)

11.1 ± 0.6

23

27.2 ± 2.8

1823 ± 513

67.5 ± 17.2

1.74

Colombia

Underweight

M-M

Spurr & Reina (1988a)

11.1 ± 0.6

19

26.6 ± 3.2

1828 ± 378

68.7 ± 14.2

1.77

Colombia

Underweight

M-M

Spurr & Reina (1988b)

11.1 ± 0.5

34

28.8 ± 3.1

2015 ± 379

70.1 ± 11.5

1.83

Guatemala

Stunted

M-M

Ramirez & Torun (1994)

11.2 ± 0.5

18

33.3 ± 2.5

2020 ± 542

60.5 ± 15.2

1.74

Colombia


M-M

Spurr & Reina (1988a)

12.7 ± 0.3

5

43.8 ± 7.3

2624 ± 315

61.4 ± 12.7

1.76

UK


M-M

Livingstone et al (1992a)

14.6 ± 0.6

16

34.8 ± 5.1

2445 ± 493

71.4 ± 12.5

1.92

Colombia

Underweight

Accum

Spurr et al (1986)

14.7 ± 0.5

12

46.7 ± 3.5

2762 ± 480

58.4 ± 9.0

1.84

Colombia


Accum

Spurr et al (1986)

14.8 ± 0.6

20

49.9 ± 3.2

2896 ± 650

58.4 ± 14.4

1.94

Colombia


M-M

Spurr & Reina (1988a)

14.8 ± 0.4

26

38.9 ± 5.3

2556 ± 580

65.6 ± 13.7

1.93

Colombia

Underweight

M-M

Spurr & Reina (1988a)

15.4 ± 0.4

3

50.7 ± 6.4

2745 ± 33

54.7 ± 6.9

1.71

UK


M-M

Livingstone et al (1992a)

Girls

6.6 ± 0.5

21

21.4 ± 1.1

1386 ± 304

63.0 ± 11.5

1.53

Colombia


M-M

Spurr & Reina (1988a)

7.0 ± 0.5

16

18.2 ± 1.7

1244 ± 254

67.6 ± 13.5

1.40

Colombia

Underweight

M-M

Spurr & Reina (1988a)

7.8 ± 0.3

5

23.5 ± 2.5

1609 ± 260

68.3 ± 5.0

1.55

UK


M-M

Livingstone et al (1992a)

8.4

5

28.3

2079 ± 191

73.5 ± 6.8

1.96b

Holland


M-M

Emons et al (1992)

9.4 ± 0.5

4

33.4 ± 3.8

1729 ± 174

52.0 ± 5.2

1.63

UK


M-M

Livingstone et al (1992a)

9.4 ± 1.2

24

28.3 ± 3.4

1537 ± 340

55.2 ± 13.6

1.43

Colombia


Accum

Spurr & Reina (1987)

9.5 ± 0.8

10

31.6 ± 3 7

1716 ± 243

55.1 ± 11.6

1.52b

Canada


Accum

Spady (1980)

9.8 ± 1.0

20

23.7 ± 2.3

1640 ± 284

69.4 ± 10.3

1.70b

Colombia

Underweight

Accum

Spurr & Reina (1987)

10.8 ± 0.6

21

27.3 ± 4.0

1584 ± 369

55.1 ± 12.8

1.57

Colombia

Underweight

M-M

Spurr & Reina (1988a)

10.9 ± 0.7

11

34.2 ± 3.7

1611 ± 319

46.8 ± 8.9

1.45

Colombia


M-M

Spurr & Reina (1988a)

11.4 ± 0.5

23

29.2 ± 3.3

1867 ± 338

63.6 ± 11.6

1.72b

Guatemala

Stunted

M-M

Torun et al (1993)

11.8 ± 0.6

88e

31.1 ± 4.0

2013 ± 400

64.3 ± 11.8

1.81b

Guatemala

Stunted

M-M

Torun et al (1993)

12.2 ± 0.5

21

33.7 ± 4.4

2170 ± 441

64.5 ± 11.5

1.90b

Guatemala

Stunted

M-M

Torun et al (1993)

12.5 ± 0.4

5

45.1 ± 4.7

2232 ± 234

49.7 ± 5.4

1.60

UK


M-M

Livingstone et al (1992a)

14.9 ± 0.6

19

49.3 ± 2.7

1982 ± 452

41.7 ± 9.6

1.61

Colombia


M-M

Spurr & Reina (1988a)

15.2 ± 0.5

22

42.0 ± 4.1

1950 ± 585

48.6 ± 14.9

1.61

Colombia

Underweight

M-M

Spurr & Reina (1988a)

15.6 ± 0.4

3

55.4 ± 13.2

2365 ± 811

42.9 ± 12.3

1.88

UK


M-M

Livingstone et al (1992a)

a Physical Activity Level calculated using BMR measured by the investigators or estimated mathematically (b).
b PAL calculated using BMR estimated with Schofield's equations (1985).
c Stunted: > 1.5 s.d. below the NCHS median of height-for-age. Underweight: < 95% of weight-for-age and weight-for-height in comparison to Colombian children of upper socioeconomic groups (Rueda-Williamson et al, 1969). All others: adequate height and weight for age.
d Accum = heart rate accumulation during daytime, and BMR while sleeping; M - M = minute-by-minute recording.
e 22 girls measured longitudinally four times at 3-month intervals.

Table 5 Mean physical activity levels of children in Table 4 grouped by age, sex and height or weight. (Total energy expenditure estimated by heart rate monitoring; BMR's were measured or estimated with Schofield's equations)a


Boys

Girls

Age (years)

Adequate
wt and ht

Stunted or underweight

All

Adequate
wt and ht

Stunted or underweight

All

(A) Means of mean values in each studyb

2-3

-

1.45 (2)

-

-

-

-

6-13

1.72 ± 0.11 (10)b

1.68 ± 0.14 (6)

1.71 ± 0.12 (16)

1.53 ± 0.07 (7)

1.66 ± 0.19 (5)

1.58 ± 0.14 (12)

14+

1.83 ± 0.12 (3)

1.92 (2)

1.87 ± 0.10 (5)

1.74 (2)

1.61 (1)

1.70 ± 0.16 (3)

(B) Weighted meansc

2-3

-

1.42 (17)

-

-

-

-

6-13

1.65 (149)

1.71 (125)

1.68 (274)

1.50 (80)

1.67 (101)

1.60 (181)

14+

1.89 (35)

1.93 (42)

1.91 (77)

1.65 (22)

1.61 (22)

1.63 (44)

a Data of Emons et al (1992) excluded due to their unusually high PAL's.
b Mean ± s.d. of mean values in Table 4. Number of data sets in parenthesis.
c Weighted by the number of children in each study (in parenthesis)

PALs in Table 4 were calculated using measured BMR in most studies; estimates with Schofield's equations (1985) were used in only six of them. Table 5 shows the mean PALs for the same age groups as in Table 3. Although there were large differences in sample sizes (3-34), the means of the mean values in each study were within 5% of the mean values weighted for the number of children in every sex-and-age group.

All Canadian, Dutch and Irish children apparently had adequate weight and height. The Colombian children were from low and low-middle socioeconomic groups of Cali. They were classified as well nourished or as marginally malnourished or underweight when their weight-for-age and weight-for-height was above or below 95% of the Colombian standards for children of upper socioeconomic groups, respectively (Rueda-Williamson et al, 1969). Most Guatemalan children were from the lower socioeconomic groups of Guatemala City. While presently well nourished, they were stunted by more than 1.5 s.d. below the NCHS/WHO median of height-for-age. One group of Guatemalan boys (Ramirez and Torun, 1994) was from the middle socioeconomic class and they had adequate height and weight.

Figure 2 compares the data in Table 4, expressed as kcal/kg/day, with the FAO/WHO/UNU 1985 recommendations. Total energy expenditure per unit of body weight was greater among the stunted and underweight children. Since the FAO/WHO/UNU values were derived from data of well nourished, non-stunted children, Figure 3 shows only the values described in Table 4 for such children. They are combined with data from doubly-labeled water in Figure 4.

The higher energy expenditure per unit of body weight often observed in stunted and mildly malnourished children, compared with those of adequate height and weight (Tables 1 and 4), could be partly due to differences in body composition. If so, the differences in TEE would be expected to decrease or disappear when expressed as multiples of BMR (i.e. PAL units). Table 6 shows the PALs of 'normal' and stunted or mildly underweight individuals within the same community. In contrast with TEE per unit of body weight, there was no consistent difference in the PAL of children and adolescents with adequate height and weight, compared with their stunted or slightly underweight counterparts (Tables 5 and 6). This supports the explanation attributing differences to body composition.

However, the differences in TEE could also be related to the children's physical activity patterns. An examination of the minute-by-minute heart rate and its energy equivalence in Guatemalan school-boys of different height and socioeconomic status, showed that during the active hours of the day, the stunted (low income) group spent less time than the taller (middle income) group in 'sedentary' activities (434 ± 160 vs 566 ± 159 min. P < 0.01) and more time in 'light' activities that demanded some degree of physical effort (213 ± 136 vs 103 ± 94 minutes/day, P < 0.01) (Ramirez and Torun, 1994). This was probably due to the different lifestyles imposed by the different socioeconomic conditions of the two groups of children.


Figure 2a Total energy expenditure estimated by heart rate monitoring: boys.


Figure 2a Total energy expenditure estimated by heart rate monitoring: girls.

Table 6 Total daily energy expenditure of well-nourished and of stunted or marginally malnourished children, measured with heart-rate monitoring techniques and expressed as multiples of basal metabolic rate

Age

n

Energy expenditure or PAL

Condition

Reference

Boys

6.9 ± 0.5

41

1.60 ± 0.35

Normal

Spurr & Reina (1989)

7.1 ± 0.6

42

1.46 ± 0.29

Underweight


11.1 ± 0.6

54

1.74 ± 0.45

Normal

Spurr & Reina (1989)

11.0 ± 0.6

82

1.77 ± 0.47

Underweight


11.1 ± 0.5

34

1.75 ± 0.35

Normal

Ramirez & Torun (1994)

10.8 ± 0.5

34

1.83 ± 0.31

Stunted


14.8 ± 0.5

34

1.84 ± 0.50

Normal

Spurr & Reina (1989)

14.8 ± 0.6

47

1.92 ± 0.43

Underweight


Girls

7.0 ± 0.7

29

1.53 ± 0.38

Normal

Spurr & Reina (1989)

7.3 ± 0.7

20

1.43 ± 0.16

Underweight


10.8 ± 0 7

24

1.45 ± 0.21

Normal

Spurr & Reina (1989)

10.8 ± 0.5

32

1.57 ± 0.38

Underweight


14.9 ± 0.6

19

1.61 ± 0.31

Normal

Spurr & Reina (1989)

15.2 ± 0.5

22

1.61 ± 0.43

Underweight



Figure 3a
Total energy expenditure estimated by heart rate monitoring, excluding stunted and underweight boys.


Figure 3b
Total energy expenditure estimated by heart rate monitoring, excluding stunted and underweight girls.


Figure 4a
Total energy expenditure estimated with doubly labeled water or by heart rate monitoring, excluding stunted and underweight boys. Solid symbols: doubly labeled water.


Figure 4b
Total energy expenditure estimated with doubly labeled water or by heart rate monitoring, excluding stunted and underweight girls. Solid symbols: doubly labeled water.

Table 7 Groups of children, classified by sex and age, whose total daily energy expenditure was estimated from time-motion observations or activity diariesa




Total energy expenditure






Age (y)

n

Weight
(kg)

(kcal/d)

(kcal/kg/d)

PALb

Country

Condition

Methodc

Source

Boys

1.5

12d

9.3

725e,f

78.5e,f

1.39

Gambia

Mild malnutrition

O-Estimated EC

Lawrence et al (1991)

2-6

26

12.7 ± 3.2

1026f

81f

1.35

Guatemala

Stunted

O-Estimated EC

Torun (1990b)

4-6

25

17 ± 2

1130

66.5

1.27

Philippines

Normal

O-Estimated EC

Guzman et al (1991)

7-9

26

24 ± 3

1499

62.5

1.43

Philippines

Normal

O-Estimated EC

Guzman et al (1991)

10-12

25

32 ± 4

1971

61.6

1.61

Philippines

Normal

O-Estimated EC

Guzman et al (1991)

12-14

16

31.3 ± 5.6

1810

58.0

1.49

Singapore

Normal

D-Measured EC

Banerjee & Saha (1972)

13-15

24

47 ± 6

2043

43.5

1.37

Philippines

Normal

O-Estimated EC

Guzman et al (1991)

14.5 ± 0.4

102

51 ± 10

2626

51.6

1.68

UK

Normal

D-Adult EC

Durnin (1971)

14.6 ± 2.9

75d

49.3 ± 12.8

2222 ± 572f

45.1f

1.45

Canada

Normal

D-Estimated EC

Bouchard et al (1983)

16-17

65

69.4 ± 9.5

2766 ± 247

39.9

1.47

Australia

Normal, Students

D-Adult EC

McNaughton et al (1970a,b)

16-17

9

65.0 ± 9.6

2886 ± 235

444

1.60

Australia

Normal, Workers

D-Adult EC

McNaughton et al (1970a)

16-19

32

56 ± 5

2726

48.7

1.71

Philippines

Normal

O-Estimated EC

Guzman et al (1991)

18-19

12

72.3 ± 8.1

2714 ± 276

37.4

1.46

Australia

Normal, Students

D-Adult EC

McNaughton et al (1970a)

18-19

9

68.4 ± 8.4

2740 ± 268

40.1

1.52

Australia

Normal, Workers

D-Adult EC

McNaughton et al (1970a)

Girls

1.5

12d

9.3

725e,f

78.5e,f

1.43

Gambia

Mild malnutrition

O-Estimated EC

Lawrence et al (1991)

2-6

22

12.7 ± 3.2

1026f

81f

1.41

Guatemala

Stunted

O-Estimated EC

Torun (1990b)

4-6

27

17 ± 2

1058

62.2

1.28

Philippines

Normal

O-Estimated EC

Guzman et al (1991)

7-9

24

24 ± 2

1528

63.7

1.57

Philippines

Normal

O-Estimated EC

Guzman et al (1991)

13-15

24

46 ± 3

1744

37.9

1.33

Philippines

Normal

O-Estimated EC

Guzman et al (1991)

14.5 ± 0.5

90

52 ± 8

2211

42.5

1.59

UK

Normal

D-Adult EC

Durnin (1971)

14.6 ± 2.9

75d

49.3 ± 12.8

2222 ± 572f

45.1f

1.64

Canada

Normal

D-Estimated EC

Bouchard et al (1983)

16-17

6

50.9 ± 5.3

1893 ± 195

37.2

1.38

USA

Normal

D-Estimated EC

Bradfield et al (1971)

16-17

113

58.3 ± 5.4

2025 ± 167

34.7

1.37

Australia

Normal, Students

D-Adult EC

McNaughton et al (1970a)

16-17

32

54.8 ± 7.2

2139 ± 237

39.2

1.50

Australia

Normal, Workers

D-Adult EC

McNaughton et al (1970a)

16-19

32

50 ± 3

1922

38.4

1.49

Philippines

Normal

O-Estimated EC

Guzman et al (1991)

18-19

21

58.7 ± 5.4

1949 ± 195

33.2

1.38

Australia

Normal, Students

D-Adult EC

McNaughton et al (1970a,b)

18-19

24

54.3 ± 5.6

2073 ± 159

38.2

1.53

Australia

Normal, Workers

D-Adult EC

McNaughton et al (1970a,b)

a Energy expenditure data published by the authors or calculated from their data by B. Torun.
b Physical Activity Level calculated using BMR estimated with Schofield's equations (1985).
c O: Observations during daytime and diary or recall interview at night. D: Activity diary. EC: Energy cost of activities.
d Assume 50% boys and 50% girls.
e Mean of wet (76 kal/kg) and dry (81 kcal/kg) seasons.
f Using the same mean values for boys and girls.

Time-motion data and comparison of methods

PALs calculated from heart rate studies coincided within 5% with those calculated from doubly-labeled water studies, except for girls 6-13 years old (Tables 3 and 5). Figure 4 also indicates that the estimates of daily energy expenditure per unit of body weight calculated by heart rate monitoring coincide quite well with those based on doubly-labeled water, at least among non stunted, well nourished boys and girls.

A review of the literature allowed us to identify several studies that estimated total daily energy expenditure of children from time-motion observations or activity diaries recorded for several days, combined with indirect calorimetry measurements or estimates of the energy cost of the recorded activities. The results of those studies, listed in Table 7, were published as such by the authors or calculated by us from their data.


Figure 5a
Selected values of total energy expenditure estimated with time-motion/diary methods, compared with doubly labeled water (DLW) and heart rate monitoring (HR)*: boys. *B; Bouchard et al 1983; b: Banerjee & Saha 1972; D: Durnin 1971; G: Guzman et al 1991; L Lawrence et al 1988; M: MaNaughton et al 1970a,b; T: Torun 1990


Figure 5b
Selected values of total energy expenditure estimated with time-motion/diary methods, compared with doubly labeled water (DEW) and heart rate monitoring (HR)*: girls. *B; Bouchard et al 1983; b: Banerjee & Saha 1972; D: Durnin 1971; G: Guzman et al 1991; L Lawrence et al 1988; M: MaNaughton et al 1970a,b; T: Torun 1990

Figure 5 shows all the experimental data from the studies described in Tables 2, 4 and 7. Most time-motion/diary results agree reasonably well with the results from doubly-labeled water and heart rate studies, but there is a tendency to underestimate the energy expenditure of older adolescents, especially boys, with the diary method.

Conclusions

Total daily energy expenditure of free-living children has been measured by a limited number of investigators using doubly-labeled water or adequate heart rate monitoring techniques. Most of those studies have been done in industrialized countries, and none in school aged children or adolescents in rural areas of developing countries.

The experimental results suggest that current FAO/ WHO/UNU recommendations for dietary energy are too high for children under 5, and possibly under 7, years of age. By contrast, current dietary recommendations for adolescent boys and for girls around puberty seem somewhat low.

Energy expenditure per unit of body weight of stunted or mildly underweight, but otherwise healthy, school-children and adolescents in developing countries tends to be higher than among those with adequate height and weight. The causes for this must be explored further. In the meantime it seems convenient to make dietary recommendations based on the ideal weights or PALs of the general population.

The validity of these conclusions must be confirmed by other studies, as they are based on research carried out within a very narrow range of geographic and social environments, and most investigations with doubly labeled water or heart rate monitoring in industrialized countries involved small numbers of children in each age and sex group. Studies with heart-rate monitoring in developing countries included larger series of children, but they were done mainly among low income urban groups.

Studies are especially needed in rural areas of the developing world and among middle and upper socioeconomic groups of children in developing and industrialized cities. The minute-by-minute heart rate monitoring technique seems promising for this purpose, provided that the samples of children studied are of appropriate size. If finances allow it, they should be validated in the field with the doubly-labeled water method.

Time-motion/diary techniques can be useful to confirm the accuracy of the recommendations if the values used for the energy cost of activities are appropriate for children and adolescents (Town, 1983, 1990a). They also provide important information on activity patterns that will allow better estimates of the 24-h PAL, and an understanding of the behavioral determinants of physical activity in children and adolescents.