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 of adults: an update on basal metabolic rates (BMRs) and physical activity levels (PALs)
View the document(introductory text...)
View the documentIntroduction
View the documentVariability in adult BMRs
View the documentPredictive equations to estimate bmrs of adults
View the documentEthnic differences in BMR
View the documentEffects of migration from tropical to temperate climate on BMR
View the documentAdaptation and energy requirements
View the documentTotal energy expenditure (TEE) and physical activity levels (PAL) in adults: doubly-labelled water data
View the documentReferences

(introductory text...)

PS Shetty1, CJK Henry2, AE Black3 and AM Prentice3

1 Human Nutrition Unit, Department of Public Health & Policy, London School of Hygiene and Tropical Medicine, 2 Taviton Street, London WC1H OBT; 2 School of Biological & Molecular Sciences, Oxford Brookes University, Headington, Oxford OX3 OBP; 3 Dunn Clinical Nutrition Centre, Hills Road, Cambridge CB2 2DH.

Descriptors: Human energy requirements, basal metabolic rate, adaptation, total energy expenditure, physical activity level

Introduction

The FAO/WHO/UNU Expert Consultation (1985) adopted the principle of relying on estimates of energy expenditure, rather than energy intake from dietary surveys, to estimate the energy requirements of adults. Obtaining data on the energy expenditure of adult males and females has thus gained importance. Since the available reliable data on habitual energy expenditure in free-living adults has been limited hitherto, measures and/or predictions of basal metabolic rate (BMR) have also attracted more attention. The BMR of an individual can simply be defined as the minimal rate of energy expenditure compatible with life. It is measured under standard conditions of immobility, in the fasted state (12-14 h after a meal), and in an ambient environmental temperature between 26 and 30º C, to prevent activation of heat generating processes such as shivering. It can be quantified by direct or indirect calorimetric techniques, the former measuring the heat output directly, while the latter measure oxygen consumption (and carbon dioxide production), which are then appropriately converted to their energy equivalents. The BMR of adults can also be predicted with reasonable accuracy (i.e. with a coefficient of variation of 8%) from predictive equations. Since BMR constitutes between 60 and 70% of the total energy expenditure, it now forms the basis of the factorial approach for the assessment of energy requirements of adults. In this paper, several issues related to BMRs of adults, their relationship to total energy expenditure and physical activity levels and how they influence the estimation of adult energy requirements, are discussed. These include the methodology of BMR measurement, variability of BMRs and total energy expenditure (TEE), and the constancy of BMR over time in adults. A discussion on the usefulness and limitations of equations for the prediction of BMRs from anthropometric parameters, such as body weight. and the likelihood of ethnic variations in BMR and the effects of migration from one climatic zone to another are also included. The chapter also deals with metabolic adaptation and discusses the relevance of this to estimating adult energy requirements. It concludes with a discussion on the factorial approach to assessing total energy expenditure by the use of physical activity levels (PALs) and provides a review of the data to date on doubly labelled water measurements of total energy expenditure and PALs of free-living healthy adults.

Methodological implications of basal metabolic rate measurement in adults

Whether the methods or techniques used to measure the BMR of adults contribute to variability in BMRs is a question that needs to be addressed if BMRs are the basis for estimating total energy expenditure by the factorial method. BMR measurements involve, in the first instance, an estimation of the oxygen consumption of the individual which is then converted into units of heat or energy output. Comparisons of techniques using different equipment, such as Douglas bags, oxylogs, metabolators, and ventilated hoods, show no significant differences between estimates of oxygen consumption of adults obtained by two or more of these techniques in the same individual at the same time (Segal, 1987; Soares et al, 1989).

During the subsequent conversion of oxygen consumption (in ml or litres of oxygen) to energy output or expenditure (in kcals, kJ or MJ) many assumptions are made that can introduce errors into estimates of the BMR. The most important of these potential sources of error are:

(1) the value attributed to the non-protein respiratory quotient (NPRQ or RQ) when the method used does not actually measure the NPRQ from the CO2 production;
(2) the equations used in the calculations to convert O2 consumption and CO2 production when measured (whether or not nitrogen excretion in the urine is estimated) into units of energy output and
(3) the corrections that are made for differences in the volumes of inspired and expired air when both CO2 output and O2 consumption are measured.

It is often believed that these factors do not influence the final results over the physiological range of observed RQs. This is not correct. The difference between the true NPRQ of the subject and the assumed RQ (be it 0.82 or 1.0) in the calculation can introduce an error of over 5% for the same measure of O2 consumption in a subject, if the true RQ is as low as 0.77 (Shetty et al, 1986) (Table 1). For instance, when an oxylog which assumes an RQ = 1 is used and the true RQ of the subject is less than 0.8, the difference in the estimate of energy expenditure is reported to be of the order of 4.6% (Garlick et al, 1987).

Brockway (1987) conclusively demonstrated that differences in the final estimate of BMR due to the various formulae used in the conversion of O2 consumption values, i.e. those of Weir (1949), Consolazio et al (1963), Brouwer (1965) and Passmore & Eastwood (1986), may extend over a range of about 3%.

McLean (1984) has argued that the not uncommon assumption that O2 consumption = outlet ventilation × O2 concentration difference, can introduce an error in the estimate of O2 consumption of the order of ± 6%. This large error emphasises the importance of correcting for differences in volume flow of inspired and expired air when measuring BMR. McLean (1984) states that this large error is fortunately cancelled out by an error in the calorific value of O2 consumed as the RQ of the subject varies. However, this is not necessarily true. For example, the calculation of energy output from values for an O2 deficit of, say, 5%, for a fixed volume of 5 1/min at an assumed RQ of 0.82 using Weir's formula introduces a net error of between +4.5 and -4.0% for a range of physiological RQs of between 0.71 and 1.0, respectively (Shetty et al, 1986).

The errors that arise as a result of the assumptions made in the calculation of BMR from measures of O2 consumption, even with the tacit assumption that the measures of O2 consumption are fairly accurate in themselves, imply that differences in BMR between individuals or groups of individuals of the order of 5% do not have biological significance unless the methodology, the assumptions made and the calculations used to arrive at the BMR values are comparable. It is, of course, assumed that certain stipulated minimal experimental prerequisites, such as absence of gross muscular activity, a post-absorptive state, thermoneutral environment, etc. are strictly met in order to ensure basal levels of metabolism, so that measurements made in different individuals, or in the same individual over time, are comparable and that biological significance can be imparted to differences that are observed under these conditions.

Table 1 Sources of error in conversion of oxygen consumption to energy output between assumed and true respiratory quotients (RQs)a


Error in volume of O2 consumedb


Net error in energy

True RQ

Uncorrected volume

Corrected volumec

Error caused by caloric value of the RQ used in equation (%)

Uncorrected volume

Corrected volume

Assumed RQ=0.82 in equation

(1)RQ=0.71

-7.1

-2.7

+2.6

-4.5

-0.1

(2)RQ=0.79

-5.3

-0.8

+0.7

-4.6

-0.1

(3)RQ=0.82

-4.5

0

0

-4.5

0

(4)RQ=0.94

-1.6

+3.1

-2.7

-4.3

+0.4

(5)RQ=1.00

0

+4.8

-4.0

-4.0

+0.8

Assumed RQ=1.0 in equation

(1)RQ=0.71

-7.1

-

+6.8

-0.3

-

(2)RQ=0.79

-5.3

-

+4.8

-0.5

-

(3)RQ=0.82

-4.5

-

+4.1

-0.4

-

(4)RQ=0.94

-1.6

-

+1.3

-0.3

-

(5)RQ =1.00

0

-

0

0

-

aFor expired air volume = 51/min; O2 deficit = 5%; and Weir's equation [calorific value of 11 of O2 = 3.9+1.1(RQ)].
b [(Assumed-true)/true] × 100.
c Corrected volume indicates the value of O2 consumed that has been corrected for difference in inspired and expired volume at the given RQ.

Variability in adult BMRs

Inter-individual variability

It is generally recognised that in a group of apparently healthy and comparable individuals there is a considerable between-individual or inter-individual variation in habitual, total daily energy expenditure. This, however, is not as large as the inter-individual variation in energy intakes. Edholm (1961) reviewed a number of studies where repeated measurements of total energy expenditure had been made and reported that the coefficient of inter-individual variability was of the order of ± 12.5% on a body weight basis. In recent studies, in which energy expenditure was measured in a respiratory chamber and both the intake and the physical activity levels were controlled, the inter-individual coefficient of variation (CV) ranged between 7.5 and 17.9% (Garby et al, 1984; De Boer, 1985). lt appeared that the CV depended upon the variations in body size; the larger the variation in body weight among subjects, the larger the CV of total energy expenditure.

Comparisons of subjects of similar body weight and body composition showed an inter-individual CV of BMR of 13% (Jéquier & Schutz, 1981). Other reports suggest that the inter-individual CV of BMR varies between 7.9 and 12.0% in both male and female subjects when measurements are made under conditions of controlled intake and physical activity (Schulz, 1984; De Boer, 1985; Daly et al, 1985). The inter-individual CV of BMR was 9.2% when intake was controlled at two levels of physical activity in males (Dallosso & James, 1984) and of the order of 11.7% in free-living males who had a CV of body weights of 15.2% (Shetty et al, 1986). In the few instances where the CVs of inter-individual variation in BMR and TEE have been simultaneously computed (in male subjects who maintained body weight) they were of the order of 10.2% and 10.3%, respectively (Dallosso et al, 1982). This last report emphasises that the inter-individual CV of TEE is reflected in the CV of BMR, since the latter makes a substantial contribution to the total energy output of an individual.

Intra-individual variability in BMR

Sukhatme & Margen (1982) argued that within individual variations in intakes are more important than between-individual variations, and that the observed inter-individual variations can largely be explained in terms of the intra-individual variations. These investigators contend that the well-documented variation in intakes observed among apparently healthy individuals indulging in similar levels of activity is evidence that different individuals operate at different levels within what they consider to be the intra-individual range of 'costless' adaptation. This has resulted in an unsubstantiated claim that intra-individual variations in energy expenditure are also large, with a high coefficient of variation even in subjects accustomed to similar levels of physical activity every day, and that this wide variation needs to be considered when assessing the energy requirements of an adult (Sukhatme & Narain, 1983). Table 2 summarises some recent data on intra-individual variations in BMR obtained from repeated measurements in the same individual when: (1) energy intake and physical activity were controlled while in a respiration chamber Jéquier & Schutz, 1981); (2) energy intake alone was controlled and BMR measurements were made on two levels of physical activity over a 24-h period (Dallosso & James, 1984); (3) physical activity was kept constant over 24 h but the energy intake was varied at two different levels (Dallosso et al, 1982) and (4) when BMR measurements were made in free-living subjects in whom neither intake nor activity were regulated (Soares & Shetty, 1986). In these studies, the CV of the measured BMR has never exceeded 5% and is frequently below 3%.

Estimation of the CV of 24-h energy expenditure measurements by whole body calorimetry also leads to similar conclusions (Table 2). Several studies (Dallosso et al, 1982; Webb & Abrams, 1983; Webb & Annis, 1983; Garby et al, 1984; De Boer, 1985) have confirmed the low CV of intra-individual measurements of 24-h energy output when both energy intake and physical activity are tightly regulated, as is usual in a calorimetry protocol. Even when energy intakes are varied at two different levels, but the activity patterns when inside the calorimeter are maintained constant, the within individual CVs do not vary by more than 2.4 or 2.6% (De Boer, 1985). When energy intakes are unaltered, but 24-h energy expenditure is varied at two different levels of activity in the same subject, a large CV (of the order of 9.8%) is seen. This is to be expected, since the 24-h energy output has been deliberately altered in these subjects. However, even in these experimental situations the CV of the measured BMR in the same subjects at the two different levels of activity while in the calorimeter is no more than 2.2% (Dallosso & James, 1984).

Table 2 Intra-individual variations in basal metabolic rate and total energy expenditure


Sex

CV (%)

Basal metabolic rate

1. Energy intake and physical activity controlled Jéquier & Schutz (1981)

F

2.0

2. Energy intake controlled; physical activity varied Dallosso & James (1984)

M

2.2

3. Energy intake varied: physical activity controlled Dallosso et al (1982)

M

2.8

4. Energy intake and physical activity uncontrolled Soares & Shetty (1986)

M

2.9

Total energy expenditure (24 hour)

1. Energy intake and physical activity controlled

Dallosso et al (1982)

M

1.5

Webb & Abrams (1983)

F

3.3

Webb & Annis (1983)

F

6.0

Garby et al (1984)

M

2.2

De Boer (1985)

F

1.9

2. Energy intake varied; physical activity controlled

De Boer (1985)

F

2.4

De Boer (1985)

F

2.6

The intra-individual variation of total energy expenditure obtained from repeated measurements based on doubly-labelled water studies, where body weight, activity and physiological status remained unaltered, have also been recently compiled. Data from nine such studies are summarised in Table 3. They confirm that the CV is reasonably small, despite measurements being made with doubly-labelled water in the free-living state (Black et al, 1995). The mean within-subject CV of 79 individuals, in whom more than one doubly-labelled water measurement was made, was 8.9%. This includes both the methodological error and the variation in activity levels.

Recent evidence thus supports the conclusion that within-subject variations in BMR are small and insignificant, even when energy intake and physical activity are uncontrolled, (Shetty & Soares, 1988). This effectively refutes the Sukhatme-Margen hypothesis.

The constancy of BMR of adults over time

A critical analysis of the historical data on variations in BMR over long periods of time indicates that the BMR of an individual is constant over time. (Shetty & Soares, 1988; Henry, Hayter & Rees, 1989). More recent data, shown in Table 4, confirm this. BMR in 14 subjects (controls and obese), each tested on five consecutive days, had a CV of about 2% (Jéquier & Schutz, 1981); the BMRs of 166 male subjects studied on two separate occasions had a CV of less than 3% (Dallosso et al 1982). Other studies (Garby et al, 1984; Lammert et al, 1987; Soares & Shetty, 1986; 1987; Henry, Hayter & Rees,1989) support the view that the intra-individual, variations in BMR, measured over a period of days, weeks or even months or years, are small and probably not significant.

Table 3 Within-subject coefficient of variation in doubly labelled water measurement of total energy expenditure where activity, weight and physiological status are unchanged

Subjects

No. of subjects

No. of measurements

CV (%)

Adolescents confined to a metabolic facility during two periods of experimental diet. No control on activity

9

2

6.8

Twice in the calorimeter with the same imposed exercise

4

2

9.1

Mothers measured pre-pregnant and at 16 weeks of pregnancy

9

2

7.4

Mothers in weeks 4, 8 and 12 of lactation

10

3

7.9

Males living in a metabolic facility but following normal occupation. First and last measurements at same weight and activity

8

2

8.1

Males living in metabolic facility but pursuing usual sedentary occupation

7

3

7.1

Physiotherapy students. No apparent change in activity

5

2

10.5

Free-living men

17

2 or 3

11.0

Free-living men during two experimental diets

10

2

10.9

Mean of 9 studies



8.9

CV = within-subject coefficient of variation.

Table 4 Intra-individual variations in BMR (MJ/d) with time




Coefficient of variation (%)


Sex

n

Days

Weeks

Months

Years

Jéquier & Schutz (1981)

F

(14?)

2




Garby & Lammert (1984)

M

(22)

2.4





M

(23)


2.2



Lammert et al (1987)

M

(7)

3.5

4.3




M

(7)


4.8



Soares & Shetty (1986)

M

(5)


3.1



Soares & Shetty (1987)

M

(5)


2.9




M

(10)



2.5


Henry et al (1989)

M

(9)




4.0

Table 5 Intra-individual variations in energy expenditure and body weight over time





CV (%)


Group

n

Time intervala (months)

EE

Body weight

BMRb Males

Entire

(10)

18.2 ± 2.3

2.5

2.5




(7.0 33.0)




Weight stabled

(5)

14.4 ± 2.9

3.2

0.6




(7.0-21.0)




Weight change

(5)

22.0 ± 3.0

1.8

4.3




(15.0-33.0)



24-h EEc Females

Entire

(10)

9.5 ± 2.0

2.4

2.4




(2.0-24.0)




Weight stabled

(5)

7.2 ± 2.0

2.0

1.1




(2.0-13.0)




Weight change

(5)

11.8 ± 3.3

2.7

4.1




(5 0-24.0)



a Mean ± s.e.m.; figures in parentheses = range.
b Soares & Shetty (1987).
c De Boer (1985) (recalculated).
d Considered stable if change < 2.0% of initial body weight.

A critical analysis of the variations seen in BMRs or in 24-h energy output over a period of up to 2 years, when intakes and activity patterns were not controlled over this length of time, is presented in Table 5. The BMRs of 10 male subjects measured on at least three occasions over a period of 6-36 months showed a mean CV of intra-individual differences (separated from measurement error) of the order of 2.5% (Soares & Shetty, 1987). Five of the 10 individuals who had body weight changes of > 2% had even smaller CVs (1.8%) than those who had smaller changes in body weight over the period of time. Measurements of 24-h expenditure by calorimetry in 10 females over a period of 24 months also showed small CVs of 2.4%; however, smaller CVs were seen in those five women who had <2% body weight change over this period (De Boer 1985). In these females, neither intakes nor activity patterns were controlled, except during calorimetry. These recent data confirm the conclusion that BMRs of individuals are relatively constant over a period of several years, despite reasonable fluctuations in body weight, when no attempt is made to regulate either energy intake or physical activity patterns.

Predictive equations to estimate bmrs of adults

BMR can be accurately measured by direct or indirect calorimetry, but it is simpler, in practice, to use predictive equations. By 1951, a plethora of equations to predict BMR were in existence, some easier to use than others. The predictive equations of Aub & Du Bois (1917) tended to over-estimate BMR, as the subjects measured by these authors were under thermal stress and anxious. In contrast, Robertson & Reid's equations (1952) underestimated BMR, as they were based on the lowest recorded values of metabolic rate. Finally, while Quenouille's analysis (Quenouille et al, 1951) was comprehensive, the equations were too complicated to be of routine practical use. Recently Schofield (1985), published predictive equations that were used for the FAO/WHO/UNU report (1985) and thereby became the basis for estimating energy requirements in man.

The Schofield analysis and equations, based on a database of 114 published studies of BMR, representing 7173 data points, is the largest and most comprehensive analysis of BMR to date. While the Schofield equations predict BMR accurately in many individuals from temperate climates, they seem to be less accurate in predicting BMR in populations in the tropics (Henry & Rees, 1991; Piers & Shetty, 1993) and North America (Clark & Hoffer, 1991) and appear to over-estimate BMR in many populations (Piers & Shetty, 1993; Soares, Francis & Shetty, 1993; Hayter & Henry, 1993).

Table 6 List of Italian subjects used in the database of Schofield

Study

n

Sex

Age

Subject details

Pepe & Rinaldi (1936)

217

M

6-16

None provided


143

F

5-12

None provided

Pepe & Perrelli (1937)

257

M

5-16

None provided


235

F

5-12

None provided

Felloni (1936)

532

M

19-25

Students of the Royal Fascist Academy

Lafralla (1937)

213

M

14-20

Students of Naples Royal Military College

Lenti (1937)

525

M

20-25

Military Servicemen

Pepe (1938)

252

M

18-24

Students of Royal Naval Academy

Occhiolo & Pepe (1939)

247

F

20-67

Various social groups

Occhiulo & Pepe (1940)

571

M

22-54

Police officers

Granall & Busca (1941-1942)

186

M

16-55

Labourers and miners

Total

3370




All references in Schofield, Schofield & James (1985).

There are several reasons to suggest that there is a need to re-analyse the worldwide data on BMR using stringent inclusion criteria in order to generate more valid equations to predict BMR in humans worldwide.

(1) Schofield collected data for his study a decade ago. Since then several laboratories have produced a large number of good quality BMR data for different age, sex and ethnic groups that also need to be included.
(2) Henry & Rees (1988) have identified over 1500 data points for Caucasian subjects in the old literature that meet all the strict criteria used by Schofield, but were not included; these values also need to be incorporated.
(3) Certain age groups (children and adults over 60) are under-represented, and these parts of the database need to be expanded in order to generate more reliable predictive equations for those age ranges.
(4) Close examination of the Schofield database (Table 6) shows that of the approximately 6000 BMR values for males between 10-60 years, over 3000 (50%) come from Italian military subjects. The validity of including such a disproportionate number of Italian military subjects may need to be queried, firstly, because the Italian group appears to have a higher BMR per kg than any other Caucasian group (Hayter & Henry, 1993), and secondly, because they may not be representative of the general Italian population. In fact Schofield (1985) noted that when Italians were isolated from the rest of the sample and compared with the derived BMR predictive equation there was a significant lack of fit. The inclusion of this disproportionately large Italian group with a higher BMR per kg may have artificially elevated the predictive equations generated by Schofield.
(5) If an analysis of the BMRs of people from the tropics and sub-tropics (Henry & Rees, 1991) points to a lower BMR than predicted by the Schofield equations, this may be due mainly to the bias imposed by the dominance of the Italian data. More recent data in fact support the view that BMRs of people in the tropics are not different from those in temperate regions (North America and Europe), provided the subjects are well nourished (Henry et al, 1987; Hayter, 1992; Soares, Francis & Shetty, 1993; Piers & Shetty, 1993).

There is thus mounting evidence to suggest that the Schofield equations may be overestimating BMR in many populations, leading to an over-estimation of their energy requirements. This has both practical and political implications. In the light of this, a critical reassessment of the worldwide data on BMR is required.


Figure 1 Linear least square regression lines of basal metabolic rates (MJ/d) on body weight of males (A) and females (B) of Indian, Chinese, North European and North American and Italian groups.


Figure 1 Linear least square regression lines of basal metabolic rates (MJ/d) on body weight of males (A) and females (B) of Indian, Chinese, North European and North American and Italian groups.

Ethnic differences in BMR

In addition to the observation that the Italian data revealed a higher BMR/kg, Quenouille et al (1951) and subsequently Schofield et al (1985) noted that Asiatic subjects (Indians and Chinese) had a BMR 10-12% lower than Europeans. Indeed such a claim for a lower BMR in Indians had first been reported by Mukerjee & Gupta (1931) and Krishnan & Vareed (1932). Extending the observations reported by Schofield et al (1985), Henry & Rees (1988; 1991) showed that the BMR was 8-10% lower in a range of other tropical residents (Filipino, Indian, Japanese, Brazilian, Chinese, Malay and Javanese) as well. In contrast to these earlier reports of low BMR in tropical peoples, recent studies have shown no difference in BMR between Indians and Europeans (Henry, Piggott & Emery 1987; Hayter 1992; Soares, Francis & Shetty, 1993).

Table 7 Recent studies of basal metabolic rate in male migrants from tropical to temperate climate

Study

Subjects

n

Age

Height (m)

Weight (kg)

kJ/kg/d

DeBoer et al (1988)

African

8

31

1.71

69.9

91

SMR


European

7

30

1.84

78.4

87

SMR


Chinese

7

33

1.67*

62.5*

98*

SMR


Indian

8

26

1.72*

58.9*

98*

SMR

Henry et al (1987)

Asian

11

21

1.63

56.2

115

RMR


British

11

25

1.68

57.4

108

RMR

Ulijaszek & Strickland (1991)

Gurkhas

17

25

1.67**

67.1

105

BMR


British

17

23

1.73

66.8

110

BMR

Geissler & Aldouri (1985)

British

15

25

1.74

68.1

117

RMR


Asian

15

27

1.68

63.9

107***

RMR


African

15

28

1.71

67.1

101***

RMR

Blackwell et al (1985)

American

8

31

1.75

75.0

93

SMR


Asian

8

25

1.66*

53.0*

108

SMR

Dieng et al (1980)

W. African

10

34

?

73.0

115

RMR


French

10

36

?

75.0

111

RMR

Hayter & Henry (1993)

Trop 1

9

23

1.70

63.8

113

RMR


Temp 1

9

25

1.76

67.5

114

RMR


Trop 2

21

25

1.69

58.2

118

RMR


Temp 2

20

23

1.77

68.3

114

RMR

Significantly lower with *P < 0.05, **P < 0.01, ***P < 0.005, by the statistical test used in the referred papers. SMR = sleeping metabolic rate, RMR = resting metabolic rate, as specified in the cited papers, expressed in kJ/kg/day. Trop (1 & 2) = two groups of tropical migrants, Temp (1 & 2) = two groups of temperate climate residents. All references in Hayter & Henry (1993).

One approach to studying this problem is to compare the BMR in different population subgroups at similar body weights, thus eliminating a major source of variance in BMR associated with body weight. BMR predictive equations, generated for different population groups over defined body weight ranges, can then be used for comparison. To achieve this, the databases of Schofield et al (1985) and Henry & Rees (1988), were combined and used for analysis (Hayter & Henry, 1993). The resultant data set contained 7737 individual measurements of body weight, height, sex, age and the geographical origin of subjects. The 18-30 year age group was considered most suitable for detailed analysis, as it had a BMR database of 2999 males and 874 females. This age range also showed a negligible effect of age on BMR. Sub samples of Indians (210 males and 137 females), Chinese (200 males and 122 females), North Americans/North Europeans (478 males and 372 females) and Italians (169 males and 135 females) were available for analysis.

The linear least square regression lines of BMR on body weight of those from ethnic subsamples for both sexes are presented in Figure 1. Italian males and females, who comprised 45% of the Schofield database, turn out to be again the most divergent group. The apparently elevated BMR of Italians and their numerical dominance in the sample appear to bias the predictive equations. This may explain why the BMR of Indians and other tropical populations is overestimated by the Schofield equations.

Effects of migration from tropical to temperate climate on BMR

An analysis of the available literature on the effects of recent migration (over 2-4 weeks) from the tropics to temperate zones, and of follow-ups till 9 months later, shows that there are no differences of any significance between the BMRs per kg body weight of tropical migrants and those of their peers born and resident in temperate zones, provided the subjects are from privileged backgrounds and well-nourished (Table 7). There is thus no reason to believe the BMRs of well-nourished tropical individuals are lower than those of European or North American subjects.

Adaptation and energy requirements

A working definition adopted by the FAO/WHO/UNU Expert Consultation (1985) on adaptation states that it is 'a process by which a new or different steady state is reached in response to a change or difference in the intake of food and nutrients'. This definition attempts to deal with both short-term and long-term adaptation; the word 'new' having relevance to short-term responses to acute changes in a subject who is in balance, while the word 'different' refers to long-term changes in individuals or groups exposed habitually to different environmental or nutritional conditions. Three general points were made by this Report in relation to both types of adaptation:

(1) The concept of a 'steady state' is relative, and the time-scale over which a state may be considered steady or stable varies for different functions.
(2) Adaptations are of different kinds: metabolic, biological/genetic, and social/behavioural.
(3) It follows from the above that adaptation must imply a range of steady states, and hence it is impossible to define a single point within the range that represents the 'norm'. Implicit in this is the understanding that different adapted states may have advantages and/or disadvantages.

The concept of a range of adapted states, each with possible advantages and disadvantages, while implying a respect and understanding for different biological and cultural situations, can also serve to condone the acceptance of double standards and the endorsement of the status quo.

An adaptive response is an inevitable consequence of a sustained perturbation in the environment and may be genetic, physiological and/or behavioural. These different types of adaptation are not mutually exclusive; they interact with each other at several levels. Every adaptation has a potential cost. Reduced physical activity in a child may reduce the interaction with the environment needed for normal development. Reduced physical activity in adults, with no apparent biological cost, may have serious economic and social consequences. At some point, adaptation will begin to have both biological and social costs. The processes and costs involved may be overt or covert, reversible or irreversible, and transient or permanent. Adaptation, both in the short-term and in the long-term, is a relatively slow process and should be distinguished from the rapid regulatory role of homeostatic mechanisms. A homeostatic response in a biological system may neither have additional costs to the organism nor lead to compromise in its function, capability, or performance, in contrast to an adaptive response which may do both in order to further the survival of the individual.

Metabolic adaptation

The suggestion that the energy metabolism of individuals is variable and adaptable, and that allowances need to be made for this when making estimates of human energy requirements, has been based on several important publications that have drawn attention to the possibility of such physiological variability in energy utilisation between individuals (Durnin et al, 1973; Edmundson, 1980) and within individuals (Sukhatme & Margen, 1982; Sukhatme & Narain, 1983). Healy (1989) criticized the validity of Sukhatme's approach based on autocorrelation. Norgan (1983) critically evaluated the following four types of evidence that have been adduced to illustrate this variation which is purported to result in adaptation in human energy metabolism:

(1) in any group of 20 or more similar individuals, energy intake can vary as much as two fold (Widdowson, 1962);
(2) large numbers of apparently healthy active adults exist on energy intakes that are lower than their estimated energy requirement (Durnin, 1979);
(3) the efficiency of work and work output is variable per unit energy intake (Edmundson, 1979); and
(4) observations based on studies of experimental or therapeutic semistarvation (Benedict et al, 1919; Keys et al, 1950; Grande, 1964; Apfelbaum, 1978) and overfeeding of humans (Sims, 1976; Norgan & Durnin, 1980).

Differences in body size, levels of physical activity and systematic errors in the estimation of energy intakes may provide explanations for most of these observations (Norgan,1983). However, what is implied or explicitly stated by the proponents of metabolic adaptation is that metabolic efficiency and mechanical work efficiency of the individual are variable and show an adjustment to variations in the levels of energy intake. The decrease in oxygen utilisation of the residual active tissue mass of an individual seen during experimental (or therapeutic) semistarvation (Keys et al, 1950) constitutes the most important biological argument for metabolic adaptation. On the basis of these observations it has been assumed that an enhanced metabolic efficiency is also a characteristic of individuals who are habitually on diets that are low in energy content.

Several physiological mechanisms, chiefly hormonal, operate to account for the changes in the metabolic activity of the tissues to enhance their metabolic efficiency, when a well nourished individual's energy intake is restricted (Shetty, 1990). The activity of the sympathetic nervous system is toned down, signalled by a decrease in energy flux, while the energy deficit lowers insulin secretion and initiates changes in peripheral thyroid metabolism. The latter are characterised by a reduction in the biologically active T3 and an increase in the inactive reverse T3. The reduction in the activities of these key thermogenic hormones acts possibly in a concerted manner to lower cellular metabolic rate. Changes in other hormones such as glucagon, growth hormone and glucocorticoids may influence these changes and at the same time, in association with the insulin deficiency, promote endogenous substrate mobilization which will lead to an increase in circulating free fatty acids and ketone bodies. The elevated free fatty acid levels, alterations in substrate recycling and protein catabolism will also influence this process. These hormonal and metabolic changes that accompany energy restriction aid the survival of the organism during restricted availability of exogenous calories. Hence these physiological changes that are associated with body weight and body composition changes have been considered as an indication of metabolic adaptation which is seen in a previously well nourished individual and which are aimed at increasing the 'metabolic efficiency' of the residual tissues in the body at a time of energy deficit.

It has been assumed that the physiological and metabolic responses of an adult on a low plane of habitual intake are similar to, and can be explained on the basis of, the sort of physiological changes that occur during experimental or therapeutic semi-starvation in previously well nourished adults. Ferro-Luzzi (1985) summarised the thinking at that time on ways in which an individual on habitually low intakes could respond to a sustained energy imbalance by metabolic adaptation. Metabolic adaptation was represented as a series of complex integrations of several different processes that occur during energy deficiency. These processes were expected to occur in phases which could be distinguished, and it was suggested that a new equilibrium could be established at a lower plane of energy intake. At this stage, individuals who had gone through the adaptive processes that occur during long-term energy deficiency, were expected to exhibit more or less permanent sequelae or costs of adaptation, which included a smaller stature and body mass, an altered body composition, a lower BMR, a diminished level of physical activity and possibly a modified or enhanced metabolic efficiency of energy handling by the residual tissues of the body. However, a large number of measurements made during the last decade, in subjects in environments that predispose to low energy intakes, do not confirm the existence of an enhanced metabolic efficiency (Srikantia, 1985; McNeill et al, 1987; Soares & Shetty, 1991). It would therefore appear that an increase in metabolic efficiency in the BMR component of energy expenditure, which has been hitherto considered to be the cornerstone of the beneficial, metabolic adaptation to long-term energy inadequacy, is of doubtful existence. It is more likely that a lower BMR per kg body mass in the chronically undernurished is an arte-fact attributable to the changes in body composition, more specifically the disproportionate reduction in muscle tissue with a normal or even increased non muscle or visceral organ size (Shetty, 19933, possibly enhanced by an increase in number of infective episodes in individuals living in such environments. Hence it is highly unlikely that metabolic: adaptation is of any relevance in chronic energy deficiency, as opposed to a situation where normal individuals are energy restricted.

Behavioural adaptation

The behavioural adaptations in physical activity patterns that accompany low energy intake are related to the individual's allocation of time and energy to different productive and leisure activities and to the biological as well as the economic consequences of these altered behavioural patterns. When there is both a fall in energy intakes and an increased demand for energy expenditure at work, for instance during seasonal agricultural activities, individuals adjust the time they allocate to different tasks; more time is given to work activities and less time and energy to productive tasks at home or socially desirable or pleasurable activities (Immink, 1987).

Lower energy intakes and stunting in preschool children were associated with lower levels of physical activity (Rutishauser & Whitehead, 1972). An analysis of physical activity patterns during voluntary reduction in food intake showed that the behavioural response to a deficient intake and associated weight loss was a change in the pattern of activity: lower effort discretionary activities replaced those which needed greater effort, while obligatory activities were not affected (Gorsky & Calloway, 1983). Rural Guatemalan men were able to carry out the specific agricultural task allocated to them, but took a longer time doing it (Torun et al, 1989); they also took a longer time to walk home and spent nearly 3 h resting or taking a nap or indulging in very sedentary activities during the rest of the day. Rural women in India and Africa with marginal energy intakes and low BMIs have been observed to spend fewer hours working per day and more time resting than better-off individuals in the same socio-economic milieu (Ferro Luzzi et al, 1992, Shetty & James, 1994). Waterlow (1990) computed the saving in energy that may result from doing a task (e.g. walking a certain distance) slowly rather than quickly, at the cost of having less time for other activities. He drew attention to the fact, which could be relevant, that slow muscle fibres are more efficient than fast ones in terms of ATP used per unit force developed.

Appreciable increases in both activity at work and in discretionary activities without concurrent changes in body weight were reported in male agricultural workers whose diet was supplemented (Viteri & Torun, 1975). There was also an improvement in their sense of wellbeing. Similar improvements in subjective well-being with very small body weight increases have been seen in lactating Gambian women when provided with supplementary food (Whitehead et al, 1978). All these studies support the existence of behavioural adaptation in the spontaneous, free-living physical activity of adults which may limit their work output, economic productivity and income-generating ability, at the same time restricting their socially desirable and discretionary or even their obligatory physical activity. This latter behavioural adaptation becomes an important survival strategy. Recommendations for energy requirements have to take into consideration the energy needs to cope with the cost of behavioural adaptation in adults.

Total energy expenditure (TEE) and physical activity levels (PAL) in adults: doubly-labelled water data

Over the past decade a new technique using stable isotopes has revolutionised the study of human energy expenditure. The doubly-labelled water (DLW) method permits determination of energy expenditure of free living individuals integrated over a period of, usually, between 7 and 20 days. The first data from humans were published in 1982 (Schoeller & van Santen, 1982). Since then sufficient data have accumulated to form a basis for establishing energy requirements. A database of 1614 measurements in 1123 individuals aged 2-90 years has been comprehensively analysed by Black et al (1996). Details of the methodologies employed, the database, studies included and excluded, and full references can be found in their paper.

Usage, validity and variability of the physical activity level (PAL) index

Total energy expenditure (TEE) is expressed as a multiple of BMR to determine the requirements of adults as recommended by the last FAO/WHO/UNU Expert Consultation Report (1985) on energy and protein requirements. These multiples of BMR are referred to as physical activity levels (PALs) and calculated by dividing TEE by BMR. The expression of energy expenditure (or requirements) of adults as PALs provides a convenient way of controlling for age, sex, weight and body composition and for expressing the energy needs of a wide range of people in shorthand form. The figures derived by the 1985 Consultation were based on theoretical factorial calculations, making assumptions about the energy cost and duration of day-to-day activities. The data in Table 8 on PAL values in adults are derived from actual measurements using the DLW technique. PAL provides a useful means of categorising energy requirements in a single number, taking into account differences in body size, as represented by BMR. However, the value of PAL depends both on BMR and TEE, and both have errors of measurement, so that PAL is only imprecisely estimated. The CV of BMRs, when actually measured, is very small, as described earlier, while the CV of BMRs predicted using the Schofield equations for given body weights is of the order of about 8% (Schofield, 1985). For TEE, the within-subject CV can be obtained from studies with repeated DLW measurements in persons with stable weight, activity and physiological state. Data from nine such studies, collated by Black et al (1996), have shown that the mean within-individual CV for 79 subjects was 8.9%; this includes methodological error as well as variations in activity levels. Thus, the 95% confidence limits on PALs at the individual level, assuming a measured BMR and no change in body weight or physical activity, is of the order of ± 18.5% representing about ± 0.3 PAL units on a mean PAL value of 1.65.

Table 8 also presents TEE, BMR and energy expenditure for activity (AEE) derived as TEE minus BMR. The latter expression has hitherto not been used, although a related expression, i.e. Physical Activity Ratio (PAR), has been in vogue. PAR is used as an abbreviation for multiple of BMR for an activity, and is used to provide an energy cost for a specific activity, such as sitting down, walking, etc. On the other hand, AEE represents the energy expended by an individual over and above BMR, and includes requirements for thermogenesis, including diet-induced termogenesis (DIT), and for physical activity. The usage of PAL treats it as an index of TEE adjusted for BMR. Since the PAL is a multiple of the BMR and BMR is related to body weight, it is implied that the component of TEE that represents physical activity must also be related to body weight. In theory this would only be true for those activities that involve movement of the body. The relationship of TEE to body weight suggests that most human activities do fall into this category, and the occasion on which significant physical work is done without much movement of the body, e.g. lifting heavy sacks, are relatively rare.

Table 8 Subject characteristics and energy expenditure (obtained by DLW) in different age and sex groups



Age (y)

Height (m)

Weight (kg)

BMI (kg/m2)

Age group (y)

n

mean

s.d.

mean

s.d.

mean

s.d.

mean

s.d.

Females

18-29

89

24.4

(3.7)

1.66

(0.06)

69.2

(22.3)

25.3

(8.1)

30-39

76

33.8

(3.0)

1.64

(0.07)

67.9

(13.9)

25.2

(4.9)

40-64

47

51.6

(8.3)

1.65

(0.07)

70.0

(13.3)

25.9

(4.6)

Males

18-29

56

22.5

(3.5)

1.77

(0.07)

75.6

(18.4)

24.0

(5.3)

30-39

36

34.3

(3.3)

1.79

(0.06)

86.1

(31.4)

26.8

(8.8)

40-64

15

50.6

(8.8)

1.76

(0.06)

77.0

(10.0)

24.9

(3.0)



TEE (MJ/d)

BMR (MJ/d)

AEE (MJ/d)

PAL

Age group (y)

n

mean

s.d.

mean

s.d.

mean

s.d.

mean

s.d.

Females

18-29

89

10.4

(2.2)

6.2

(1.1)

4.2

(1.7)

1.70

(0.28)

30-39

76

10.0

(1.7)

6.0

(0.6)

4.1

(1.5)

1.68

(0.25)

40-64

47

9.8

(1.7)

5.8

(0.7)

4.0

(1.4)

1.69

(0.23)

Males

18-29

56

13.8

(3.0)

7.5

(1.2)

6.3

(2.5)

1.85

(0.33)

30-39

36

14.3

(3.1)

8.2

(1.8)

6.1

(2.5)

1.77

(0.31)

40-64

15

11.5

(1.7)

7.0

(0.8)

4.5

(1.3)

1.64

(0.17)

The limits of human energy expenditure

Studies carried out under special conditions provide information on energy expenditure at the extremes of physical activity levels in adults and thus provide a frame of reference for evaluating values of TEE and PAL from the general population. These studies of TEE measurements using the DLW technique have been summarised by Black et al (1996). At the lower limit of physical activity, studies in non-ambulatory, chair bound subjects and in individuals confined to a calorimeter and apparently not exercising, provide a mean PAL of 1.21. This is slightly lower than the value of 1.27 suggested by FAO/WHO/UNU (1985) as the survival requirement. At the upper limit of physical activity there is a distinction to be drawn between the maximum achievable over a limited period of time and the maximum sustainable as a long-term way of life, given physical fitness and adequate food. The maximum achieved over limited periods of time was a PAL of >4.0 and TEE of 33 MJ/d in a bicycle race and a polar exploration. The maximum for a sustainable way of life may be that represented by soldiers on active service, with a mean PAL of 2.4 and TEE of 18 MJ/d. In support of this, energy intakes of 19.5 MJ/d have been recorded in colliers (Moss, 1923) and in lumberjacks (Karvonen et al, 1961). Among athletes in training, mean PALs of 2-3.5 were found, with TEE ranging from 11 to 18 MJ/d in women, and from 15 to 30 MJ/d in men. PALs greater than 2.4 were obtained in periods of 'rigorous training', which is unlikely to be a sustained lifestyle. The lower values for PAL, 2.0-2.3, were obtained in periods of apparently routine training and may well be sustained for extended periods of time. Similar values have been observed in Gambian women during the farming season (Singh et al, 1989). These data suggest a PAL range of 1.2-2.5 for sustainable lifestyles, where 1.2 is indicative of a non-ambulant life style and 2.5 represents a very physically active lifestyle (Table 9).

Energy expenditure of free-living adults with normally active daily life

A total of 319 adults (212 non-pregnant non-lactating (NPNL) females and 107 males aged 18-64 years) were identified as healthy, free-living, leading a normal daily life, not recruited as having specific and special circumstances, occupations or activities, and in whom BMR had been measured. Table 8 summarises the anthropometric characteristics of the sample, TEE, BMR, AEE and PAL by age and by sex. The data fully encompassed the PAL range from :1.2-2.5, established above as the likely range of sustainable energy expenditures (Table 9). The wide range of expenditures at any age was notable. Regression analysis of the entire data set accumulated by Black et al (1996), which included a total of 574 subjects aged 2-90 years on whom DLW data and BMR measurements were available indicated that equations based on weight, height, age and sex can explain 77% and 86% of the variance in TEE and BMR, and 41% of the variance in AEE. The latter, i.e AEE, was found to be much more sensitive to individual behavioural choices and therefore less definable using purely physiological measures. Age was a negative predictor of energy expenditure, particularly of the activity component (AEE), and remained so when TEE was expressed as PAL. Taken together with the regression analysis, the following key features seem to emerge from the analysis of Black et al (1996):

Table 9 Physical activity levels (PALs) based on doubly-labelled water (DLW) studies

Life style and level of activity

PAL

Chair-bound or bed-bound

1.2

Seated work with no option of moving around and little or no strenuous leisure activity

1.4-1.5

Seated work with discretion and requirement to move around but little or no strenuous leisure activity

1.6-1.7

Standing work (e.g. housework, shop assistant)

1.8-1.9

Significant amounts of sport or strenuous leisure activity (30-60 min four to five times per week)

+0.3 (increment)

Strenuous work or highly active leisure

2.0-2.4

(1) In early life, absolute levels of energy expenditure, whether expressed as TEE, BMR or AEE, rise with increasing body size, peak in the young adult years and decline thereafter. Adjusted for body size, TEE declines with age throughout life.
(2) Adjusted for body size, males have 11 % greater TEE than females.
(3) Expressed as PAL, differences with age remain significant. For females PAL is fairly constant during the adult years, and lower at younger and older ages. For males PAL rises to a peak at 18-29 years and declines thereafter.
(4) Differences in expenditure between the sexes are not completely removed by adjusting for body size using PAL, although the sex effect is confounded with height to some extent.
(5) As expected, mean TEE in the free-living population, however expressed, is well below that of the athletes in training and soldiers on exercises. It is important to note that this sample of 319 adults is from affluent societies and contains very few manual workers with no data from developing country individuals with a high level of physical activity.

PAL values from DLW data on free-living adults in developed societies

Figure 2 (A-D) shows the distribution of energy expenditure of adults aged 18-64 years. The distributions of PAL for both men and women have a modal value at 1.6 (encompassing 1.55-1.65). The distribution for men has a shoulder to the right suggesting the existence of two populations: active and inactive. This could be either real or an artifact of the sample. Many of the authors gave no information about the subjects beyond sex, age and body weight. Subjects designated as free living were typically recruited from among colleagues, from employees in research centres, universities or hospitals, or were volunteers responding to advertising in the local media. Occupations were typically student, housewife, white collar or professional occupation, unemployed or retired. Only three individuals were specifically identified as manual workers. This suggests a predominantly 'sedentary' population. However, some individuals had PALs at a level associated with athletes or soldiers in training, and the limited information on occupation or activity usually suggested plausible reasons for these high values. Among the twenty highest values were three manual workers, six out of thirteen 'university students and laboratory technicians' with an average of 34 min 'strenuous activity' per day and with several active sports specifically mentioned, while five 18 year old college students and two professionals were known to cycle or walk as a primary means of transport. Women were not well represented in the data set at higher PAL levels. Whether this reflects an absence of subjects recruited from more active groups or a general tendency for women to be less involved in strenuous activities is not known.

Figure 2 Data on (A) total energy expenditure (TEE) by doubly labelled water, (B) BMR, (C) energy expenditure for activity and thermogenesis (AEE) and (D) physical activity levels (PAL) of male and female subjects compared with extreme levels of physical activity.


A


B


C


D

Western lifestyle is commonly referred to as 'sedentary', and the recommendation of FAO/WHO/UNU (1985) for light activity (1.55 × BMR) is frequently interpreted as 'sedentary' and taken as applying to this whole population. However, many desk jobs involve frequent moving around. Other occupations, not necessarily strenuous, require persons to be on their feet all day (e.g. housewives, shop assistants, nurses, storekeepers). Thus a PAL of 1.55-1.65 appears to represent the average for the so-called sedentary lifestyle. There are also data to suggest that activities do not have to be obviously strenuous for relatively high PAL values to be achieved. Calorimetry studies allowing 'free activity' provide mean PALs ranging from 1.50-1.75, and individual PALs from 1.39-2.04. A factorial calculation based on 8 h sleep (PAL 0.95), 4 h sitting (PAL 1.2) and 12 h walking around (PAL 2.5) might represent the lifestyle of a housewife and yields a PAL of 1.8 × BMR.

A PAL of 1.35 has been suggested as the lowest PAL compatible with long-term weight maintenance in persons other than the completely chair- or bed-bound; this was the mean PAL in nine calorimeter studies (n = 207) with controlled, limited activity (Goldberg et al, 1991). Figure 2D shows 7.5% of men and 10.9% of women below a PAL of 1.35. However these may not represent the true long-term energy expenditure due to inaccuracies in the methods. As mentioned before, the coefficient of variation on repeat DLW measurements was 8.9% from nine studies (n=79) on subjects with no change in activity, weight or physiological status; while the CV on measured BMR can be as low as 2.5% under the rigorously controlled conditions of a calorimeter, many studies employed less rigorous conditions. The combined error for PAL is equal to at least ± 9.2%; while the FAO/WHO/UNU Report of 1985 suggested that the inter-individual variability in TEE in a specified group of individuals in whom energy expenditure measurements have been made over a week has a coefficient of variation of ± 12.5% on a body weight basis (Edholm, 1973).

The effect of moderate sport on energy expenditure can be gauged from three studies (n = 28) that imposed a programme of exercise on free-living people normally undertaking very little strenuous activity. The mean sedentary and exercising PALs were 1.63 (s.d. 0.16) and 1.99 (s.d. 0.19), respectively. The mean sedentary and exercising TEEs were 10.53 MJ (s.d. 1.67) and 12.54 MJ (s.d. 2.14) respectively. These figures lend support to the mode of 1.6 for 'sedentary' lifestyles, and show that 30-60 min of active sport, 4-5 times per week, can raise PAL by 0.3 units, but need not necessarily be reflected in a PAL above 2.0.

The relationships between lifestyle, activity and PAL suggested by a careful analysis of the measurements by DLW in adults in developed countries are summarised by Black et al (1996). The data provide little evidence to quantify the energy cost of manual occupations with fairly strenuous physical activity levels which are occupation-related, or to make recommendations about PALs. The range of PAL values which are considered as the maximum for a sustainable lifestyle appears to be between 2.0 and 2.4. The higher energy expenditures, seen in adults in the analysis by Black et al (1996), appear to be due to recourse to active means of transportation such as that resulting from cycling or walking, or to regular participation in active sports. This emphasises the importance of sport or active leisure pursuits in raising energy expenditure in sedentary Western populations, which may provide both for socially desirable activities and for increasing physical fitness and the promotion of health.

The FAO/WHO/UNU Expert Consultation (1985) suggested the average daily energy requirement of adults whose occupational work is classified as light, moderate, or heavy, expressed as a multiple of BMR, to be as follows:


Light

Moderate

Heavy

Men

1.55

1.78

2.10

Women

1.56

1.64

1.82

It is obviously difficult to relate these categories to the data in the analysis of DLW studies (Black et al, 1996), as the information on occupations was limited and the categories do not take active leisure into account. The modal value of 1.55-1.65 for adults in the analysis falls between the light and moderate categories. The suggested range for strenuous occupation of 2.0-2.4 is compatible with the recommendation of 2.10 for heavy occupations. DLW data from adults do not appear seriously at variance with the recommendations made by the Expert Consultation.

References

Apfelbaum M (1978): Adaptation to changes in caloric intake. Prog. Food Nutr. Sci. 2, 543-559.

Aub JC & Du Bois EF (1917): The basal metabolism of old men. Arch. Intern. Med. 19, 823-834.

Benedict FG, Miles WR, Roth P & Smith HM (1919):Human vitality and efficiency under prolonged restricted diet. Carnegie Institute of Washington Publication No. 280. Carnegie Institute of Washington, D.C.

Black AE, Coward WA, Cole TJ & Prentice AM (1996): Human energy expenditure in affluent societies: analysis of 574 doubly labelled water measurements. Eur. J. Clin. Nutr. 50, 72-92.

Brockway JM (1987): Derivation of formulae used to calculate energy expenditure in man. Hum. Nutr. Clin. Nutr. 41C, 463-472.

Brouwer E (1965): Report of subcommittee on constants and factors. In Third Symposium on energy metabolism, ed. KL Blaxter, pp 441-443. European Association for Animal Production, publ. no 11. London Academic Press.

Clark HD & Hoffer LF (1991): Reappraisal of the resting metabolic rate of normal young men. Am. J. Clin. Nutr. 53, 21-26.

Consolazio CF, Johnson RE & Pecora E (1963):Physiological methods of metabolic functions in man, pp 313-317. New York: McGraw-Hill.

Dallosso HM, Murgatroyd PR & James WPT (1982): Feeding frequency and energy balance in adult males. Hum. Nutr. Clin. Nutr. 36C, 25-39.

Dallosso HM, & James WPT (1984): The effect of fat over-feeding on 24 h energy expenditure. Br. J. Nutr. 52, 49-64.

Daly JM, Heymsfield SB, Head CA, Harvey LP, Nixon DW, Katzeff H & Grossman GD (1985): Human energy requirements: overestimation by widely-used prediction equation. Am. J. Clin. Nutr. 42, 1170-1174.

De Boer JO (1985): Energy requirements of lean and obese women, assessed by indirect calorimetry. PhD Thesis. Agricultural University, Wageningen, Netherlands.

Durnin JVGA, Edholm OG, Miller DS & Waterlow JC (1973): How much food does man require? Nature 242, 418.

Durnin JVGA (1979): Energy balance in man with particular reference to low energy intakes. Bibl. Nutr. Dieta 27, 1-10.

Edholm OG (1973): Energy expenditure and food intake. In Energy balance in man, ed. M Apfelbaum, pp 51-60. Paris: Masson.

Edholm OG (1961): Energy expenditure and calorie intake in young men. Proc. Nutr. Soc. 20, 71-76.

Edmundson W (1979): Individual variations in basal metabolic rate and mechanical work efficiency in East Java. Ecology of Food and Nutrition 8, 189-195.

Edmundson W (1980): Adaptation to undernutrition: how much food does man need? Soc. Sci. Med. 14D, 19-126.

FAO/WHO/UNU Expert Consultation (1985):Energy and protein requirements. WHO Technical Report Series 724. Geneva: World Health Organisation.

Ferro-Luzzi A (1985): Range of variation in energy expenditure and scope of regulation: In Proc. of XIIIth Int. Congr. Nutr., eds TG Taylor & NK Jenkins, pp 393-399. London: John Libbey.

Ferro-Luzzi A, Sette S, Franklin M & James WPT ( 1992): A simplified approach to assessing adult chronic energy deficiency. Eur. J. Clin. Nutr. 46, 173-186.

Garby L, Lammert O & Nielsen E (1984): Within subjects, between weeks variation in 24 h energy expenditure for fixed physical activity. Hum. Nutr. Clin. Nutr. 38C, 391-394.

Garlick PJ, McNurlan MA, McHardy KC, Calder AG, Milne E, Fearns LM & Broom J. (1987): Rates of nutrient utilisation in man measures by combined respiratory gas analysis and stable isotope labelling: effect of food intake. Hum. Nutr.: Clin. Nutr. 41C, 177191.

Goldberg GR, Black AK, Jebb SA, Cole TJ, Murgatroyd PR, Coward WA & Prentice AM (1991): Critical evaluation of energy intake data using fundamental principles of energy physiology 1: derivation of cut-off values to identify under-recording. Eur. J. Clin. Nutr. 45, 569-581.

Gorsky RD & Calloway DH (1983): Activity pattern changes with decrease in food energy intakes. Hum. Biol. 55, 577-586.

Grande F (1964): Man under caloric deficiency. In Handbook of physiology, adaptation to the environment, pp 911-937. American Physiological Society, Washington.

Hayter JE (1992): The variability of energy expenditure in populations of different geographic origins. PhD Thesis. Oxford Brookes University.

Hayter JE & Henry CJK (1993): Basal metabolic rate in human subjects migrating between tropical and temperate regions: a longitudinal study and review of previous work. Eur. J. Clin. Nutr. 47, 724-734.

Healy MJR (1989): Nutritional adaptation and variability. Eur. J. Clin. Nutr. 43, 209-210.

Henry CJ, Hayter J. & Rees DG (1989): The constancy of basal metabolic rate in free-living male subjects. Eur. J. Clin. Nutr. 43, 727731.

Henry CJK & Rees DG (1991): New prediction equations for the estimation of basal metabolic rate in tropical peoples. Eur. J. Clin. Nutr. 45, 177-185.

Henry CJK, Piggott SM & Emery B (1987): Basal metabolic rate and dietary thermogenesis in Asians living in Britain. Hum. Nutr. Clin. Nutr. 41C, 397-402.

Henry CJK & Rees DG (1988): A preliminary analysis of basal metabolic rate and race. In Comparitive nutrition, eds Blaxter K & Macdonald 1, pp 149-159. London: John Libbey.

Immink MDC (1987): Economic effects of chronic energy deficiency. In Chronic energy deficiency: consequences and related issues, eds B Schurch & NS Scrimshaw, pp 153-174. Lausanne, Switzerland: IDECG.

Jéquier E & Schutz Y (1981): The contribution of BMR and physical activity to energy expenditure. In Body weight regulatory system: normal and disturbed mechanisms, eds Cioffi LA, James WPT & Van Itallie TB, pp 89-96. New York: Raven Press.

Karvonen MJ, Pekkarinen M, Metsala P & Rautenen Y (1961): Diet and serum cholesterol of lumberjacks. Br. J. Nutr. 15, 157-163.

Keys A, Brozeck J. Henschel A, Mickelson O & Taylor HL (1950): The biology of human starvation. Minneapolis: University of Minneapolis Press.

Krishnan BT & Vareed C (1932): Basal metabolism of young college students, men and women in Madras. Indian J. Med. Res. 12, 831858.

Lammert O. Garby L, Maron K, Mork G. Thein M, Flindt-Egebak P & Krogh-Hansen J (1987): Effect of the preceeding days energy intake on the energy cost of rest, arm and leg exercise. Hum. Nutr. Clin. Nutr. 41C, 141147.

McLean JA (1984): Heat production or oxygen consumption? In Human energy metabolism, ed. AJH Van Es, EURO-NUT report 5, pp 187-189. Wageningen, Netherlands.

McNeill G. Rivers JPW, Payne PR, deBritto JJ & Abel R (1987): Basal metabolic rate of Indian men: no evidence of metabolic adaptation to a low plan of nutrition. Hum. Nutr. Clin. Nutr. 41C, 473484.

Moss KN (1923): Some effects of high air temperatures and muscular exertion upon colliers. Proc. R. Soc. B. 95, 181-200.

Mukerjee HN & Gupta PC (1931): The basal metabolism of Indians (Bengalis). Indian J. Med. Res. 11, 807-811.

Norgan NG & Durnin JVGA (1980): The effect of weeks of overfeeding on the body weight, body composition, and energy metabolism of young men. Am. J. Clin. Nutr. 33, 978-988.

Norgan NG (1983): Adaptation of energy metabolism to level of energy intake. In Energy expenditure under field conditions, ed. J Parizkova, pp 56-64. Prague: Charles University.

Passmore R & Eastwood MA (1986): Davidson and Passmore: Human nutrition and dietetics, 8th edn, pp. 18-19. Edinburgh: Churchill Livingstone.

Piers LS & Shetty PS (1993): Basal metabolic rates of Indian women. Eur. J. Clin. Nutr. 47, 586-591.

Quenouille MH, Boyne AW, Fisher WB & Leitch I (1951): Statistical studies of recorded energy expenditure of man. Part I. Basal metabolism related to sex, stature, age, climate and race. Commonwealth Bureau of Animal Nutrition Technical communication No: 17. Aberdeen: Commonwealth Agricultural Bureau.

Robertson JD & Reid DD (1952): Standards for basal metabolism of normal people in Britain. Lancet 1, 940-943.

Rutishauser IH & Whitehead RG (1972): Energy intake and expenditure in 13 year old Ugandan children living in a rural environment. Br. J. Nutr. 28,145-152.

Schoeller DA & van Santen E (1982): Measurement of energy expenditure in humans by doubly labelled water method. Am. J. Physiol. 53, 955-959.

Schofield WN, Schofield C & James WPT (1985): Basal metabolic rate review and prediction, together with an annotated bibliography of source material. Hum. Nutr. Clin. Nutr. 39C (Suppl I).

Schofield WN (1985): Predicting basal metabolic rate, new standards and review of previous work. Hum. Nutr. Clin. Nutr. 39C (Suppl I), 5-41.

Schutz Y (1984): Terminology, factors and constants in studies on energy metabolism of humans. In Human energy metabolism, ed. AJH Van Es, EURO-NUT Report 5, pp 153-168. Wageningen, Netherlands.

Segal KR (1987): Comparison of indirect calorimetric measurements of resting energy expenditure with a ventilated hood, face mask and mouthpiece. Am. J. Clin. Nutr. 45, 1420-1423.

Shetty PS, Soares MJ & Sheela ML (1986): Basal metabolic rates of South Indian males. Report to FAO, Rome.

Shetty PS (1993): Chronic undernutrition and metabolic adaptation. Proc. Nutr. Soc. 52, 267-284.

Shetty PS (1990): Physiological mechanisms in the adaptive response of metabolic rates to energy restriction. Nutr. Res. Rev. 3, 49-74.

Shetty PS, & James WPT (1994): Body mass index: a measure of chronic energy deficiency in adults. FAO Food & Nutrition Paper 56. Rome: FAO.

Shetty PS & Soares MJ (1988): Variability in basal metabolic rates of man. In Comparitive nutrition, eds Blaxter K & Macdonald 1, pp 141-148. London: John Libbey.

Sims EAH (1976): Experimental obesity, dietary induced thermogenesis and their clinical implications. Clinics Endocr. Metab 5, 377395.

Singh J, Prentice AM, Diaz E, Coward WA, Ashford J, Sawyer M & Whitehead RG (1989): Energy expenditure of Gambian women during peak agricultural activity measured by the doubly-labelled water method. Br. J. Nutr. 62, 315-329.

Soares MJ, Francis DG & Shetty PS (1993): Predictive equations for basal metabolic rates of Indian males. Eur. J. Clin. Nutr. 47, 389394.

Soares MJ & Shetty PS (1986): Intra-individual variations in resting metabolic rates of human subjects. Hum. Nutr. Clin. Nutr. 40C, 365-369.

Soares MJ & Shetty PS (1987): Long term stability of metabolic rates in young adult males. Hum. Nutr. Clin. Nutr. 41C, 287-290.

Soares MJ, Sheela ML, Kurpad AV, Kulkarni RN & Shetty PS (1989): The influence of different methods on basal metabolic rate measurements in humans. Am. J. Clin. Nutr. 50, 731-736.

Soares MJ & Shetty PS (1991): Basal metabolic rates and metabolic efficiency in chronic undernutrition. Eur. J. Clin. Nutr. 45, 363-373.

Srikantia SG (1985): Nutritional adaptation in man. Proc. Nutr. Soc. of India 31, 1-16.

Sukhatme PV & Narain P (1983): Intra-individual variation in energy requirements and its implication. Indian J. Med. Res. 78, 857-865.

Sukhatme PV & Margen S (1982): Autoregulatory homeostatic nature of energy balance. Am. J. Clin. Nutr. 35, 355-365.

Torun B, Flores R, Vitteri FE, Immink MDC & Diaz E (1989): Energy supplementation and work performance: summary of lNCAP studies. Proc. XIV Int. Congr. Nutr., pp 306-309. Seoul, Korea: Korean Nutrition Society.

Viteri FE & Torun B (1975): Energy intake and physical work in Guatemalan farmers. Bol. Oficina Sanit. Panam. (in Spanish) 78, 58-74.

Waterlow JC (1990): Mechanisms of adaptation to low energy intakes. In Diet and disease in traditional and developing societies, eds GA Harrison and JC Waterlow, pp 5-23. Cambridge: Cambridge University Press.

Webb P & Abrams T (1983): Loss of fat stores and reduction in sedentary energy expenditure from undereating. Hum. Nutr. Clin. Nutr. 37C, 271282.

Webb P & Annis JF (1983): Adaptation to overeating in lean and overweight men and women. Hum. Nutr. Clin. Nutr. 37C, 117-131.

Weir JB (1949): New method for calculating metabolic rate with special reference to protein metabolism. J. Physiol. 109, 1-9.

Whitehead RG, Rowland MG, Hutton M, Prentice AM, Muller E & Paul A (1978): Factors influencing lactation performance in rural Gambian mothers. Lancet 2, 178-181.

Widdowson EM (1962): Nutritional individuality. Proc. Nutr. Soc. 21, 121128.