The energy requirement was defined (section 2) as the amount needed to maintain health, growth, and an “appropriate” level of physical activity. Using this definition it is impossible entirely to avoid value judgements on what is meant by health and appropriate activity. Values, and consequently decisions, may change under different conditions. The aim of this report is to provide the information on which decisions can be based.
Energy needs are determined by energy expenditure. Therefore, in principle, as was recognized in the report of the 1971 Committee (1), estimates of requirements should be based on measurements of energy expenditure. This kind of information is difficult to obtain, and sometimes the only feasible approach is to estimate requirements from measurements of intake. If people are, on average, in a steady state, with appropriate body composition and levels of activity, measurements of their mean habitual intake will provide an estimate of their mean expenditure. The intention of the word “habitual” is to even out short-term fluctuations in intake, but it is not possible to define it with precision.
As in all previous reports, the requirements derived in this report are intended to apply to people who are healthy, and in general the effects of disease should be considered separately. However, it is recognized that in many populations this condition is unrealistic. In children in particular, repeated infections are so common that the effects are discussed separately in section 9.
It cannot be assumed that observed levels of expenditure or intake always represent what is desirable for the maintenance of health. In developing countries actual intakes may be too low to allow for what was described in the report of the 1971 Committee (1) as “leisure time” activity. Again, in an affluent society some people may be less physically active than is thought desirable to ensure cardiovascular health. These are matters on which value judgements have to be made; in the opinion of the Consultation, estimates of energy requirement should allow for extra activities of this kind, as discussed in section 4.3.
In the great majority of cases the largest component of energy expenditure is the basal metabolic rate (BMR), which can be measured with accuracy under standardized conditions. In this report, therefore, the principle of calculating all components of total energy expenditure as multiples of the BMR has been adopted.
It is recognized that this principle, used for the sake of simplicity, is likely to involve some inconsistencies. The relationship of the energy cost of a given level of physical activity to BMR will be affected by the nature of that activity, whether static or dynamic; by the body weight, because of the different values of BMR per kg at different body weights; and by age, because of age-related changes in body composition and BMR (2–5). As an example, values for the energy cost of walking are shown in Annex 3, expressed as multiples of BMR. For each of the rates of walking, the values for the two sexes, in two age ranges and with different body weights, do not differ by more than ± 10% from the mean. On the other hand, Seliger et al. and Pařízková et al. show age to have more effect in younger subjects. The energy cost of a given activity, expressed as a multiple of BMR, was 45% greater in 35-year-old males than in children aged 12 years. It is clear that much more work is needed on the relative energy costs of different tasks in relation to age, sex, and body weight.
4.2.1 Basal metabolic rate (BMR)
In any individual the BMR is determined principally by body size, body composition, and age. The relationships are complex; the BMR per unit weight varies with age, being higher in children and lower in the elderly. The BMR per unit weight also varies with weight: within a given age range, BMR per kg is higher in short and light individuals and lower in tall and heavy ones.
For practical purposes the most useful index of BMR is the body weight. In the report of the 1971 Committee (1), a table of weights and associated BMRs was given, taken from a paper by Talbot, based on measurements on 2200 children. For the present report, a more extensive set of measurements was compiled from the literature; these are representative of BMR in developed countries and of some data from developing countries (6).1
The data base covers some 11 000 technically acceptable measurements on individuals of both sexes and all ages, who were considered to be healthy. It includes adults of different stature and of different weight for height, as well as individuals who fell within the designated range of “acceptable” weight for height. The data also include some children and adults who may have been on limited energy intakes before the BMR measurement. This may partly explain the lower BMRs in some groups.
1 The figures in the tables of reference 6 differ slightly from those in the present report because additional data were included by the authors of that analysis after the present report was compiled.
A number of studies have attempted to assess the possibility of ethnic differences in BMR but these have failed to identify any differences that could not be related to the nutritional state or possibly to climatic conditions. Therefore, all studies have been incorporated into a single data base for developing the equations shown in later tables (Table 5 and Annex 1). In the opinion of the Consultation, these equations can be regarded as the best estimates at present available for predicting the BMR of healthy people in any population. They are, of course, no substitute for direct measurements when these can be made.
There are many ways in which equations can be developed for predicting the BMR from the collected data. Firstly, it was decided to formulate separate equations for each of the 6 age ranges defined in section 3.5.2.
Secondly, it was found that within each age range the most useful index of the BMR was body weight. The conventional use of surface area, or the inclusion of height, made no significant difference to the accuracy of prediction. Thirdly, many different types of equation were tested—linear, quadratic, logarithmic, etc. The more complex functions again added nothing to the accuracy of prediction. Therefore, for this report, in each age-sex group the BMR has been estimated from the body weight by the simple linear equations shown in Table 5 (see page 71). For the sake of completeness, equations have been derived for children below the age of 10 years, although in practice the BMR has not been used for estimating the energy requirements in this age group (section 6.3.1).
Equations that include height, and some examples of the effect of including it in calculations, are given in Annex 1. It is perhaps surprising that in most groups not only does height have no effect on improving the fit of the regression equations, but it also has little or no effect on the predicted value of the BMR independently of weight. This observation implies that for most age and sex groups the relationship between weight and height (body mass index; BMI) is not an important determinant of the BMR. For example, in a young adult male weighing 70 kg, the difference between the predicted BMR at a height of 1.6 m (BMI 27) and at 2 m (BMI 18) is less than 1%. Further examples are given in Annex 1. The effect of height is somewhat greater in adult women, and becomes significant in young children (0–3 years) and the elderly.
In adults and older children the calculation of the total energy requirement proceeds in two steps:
These factors are discussed in more detail below and evaluated in section 6. The approach is based on the recognition that a substantial proportion of total energy requirement is accounted for by the BMR, and that the cost of physical activity depends in part on body weight. Since the new equations for BMR indicate that the BMR per kg is higher in shorter and lighter individuals, the expression of physical activity in terms of BMR will increase the apparent cost of movement per kg for such individuals. Although there is little factual evidence for this, or for considering that the metabolic efficiency of physical activity correlates with metabolic efficiency under basal conditions, the present state of knowledge seems to justify the expression of activity as increments of BMR. As yet there is no convincing evidence that the total energy requirement of small children and small adults is the same per kg body weight as that of their taller and heavier counterparts, as was suggested by the 1971 Committee (1). The Consultation therefore adopted the present approach to maintain simplicity in calculating the energy requirements of individuals of different size.
Finally, it should be recognized that investigators may find the BMR of groups in their country differs from that predicted by the present general equations. This is to be expected, since in the data used for developing the current equations Indian subjects were found to have BMRs approximately 10% below the average and northern Europeans and North Americans tended to have higher values at equivalent age, height, and weight for height. These observations are not in themselves evidence for ethnic differences in BMR, so for the present the general equations have been maintained for all groups.
The energy cost of growth includes two components: the energy value of the tissue or product formed and the energy cost of synthesizing it. The total cost will therefore depend upon the composition of the product. The energy value is the heat of combustion, without the deductions for losses in urine and faeces which are allowed for by the Atwater factors. The average values for protein, fat, and carbohydrate are 5.7, 9.3, and 4.3 kcalth (24, 39, and 18 kJ) per g. Numerous estimates of the costs of synthesizing protein and fat have been derived from work on animals (7,8). The cost for protein is greater than for fat, even when fat is being synthesized from carbohydrate precursors. Except in the case of young infants and during lactation, the estimates of energy cost are not very critical, since human growth is a slow process, taking up a small proportion of the energy requirement. Moreover, since the composition of the tissue formed cannot be known accurately, and even the composition of breast milk is somewhat variable, it is only possible to make approximate estimates of the energy cost of growth.
In young children a rounded-off value of 5 kcalth (21 kJ) per g for the energy cost of growth has been widely accepted (9). Annex 4 summarizes information on which this estimate is based. Since in this report the energy requirements of children up to 10 years are estimated from intakes (section 6.3.1), a factor for the energy cost of growth has been used only in deriving the requirements of adolescents (Annex 7). In theory a single factor is not appropriate because of variations in body composition at this stage of life (section 3.4). However, during the pubertal spurt the growth component still represents a very small fraction of the total energy requirement. In contrast, the cost of growth becomes very important in any consideration of the requirements for catch-up growth in malnourished children (section 9).
In adults a higher figure is obtained for the energy cost of weight gain under different conditions (Annex 4). This may be because relatively more fat is being laid down.
The extra energy requirements for pregnancy and lactation are discussed in detail in section 6.2. The energy stored during pregnancy includes the energy laid down in the fetus, placenta, and uterus as well as the additional protein and fat stored in the mother. The composition of the tissue laid down varies at different stages of pregnancy, but since the overall extra cost is only some 10% of the total energy requirement, detailed computations, appropriate to the different stages of pregnancy, are not justified.
The energy requirement of the lactating woman must include the energy lost in milk. The daily amount should be taken as that produced by healthy well-nourished mothers (section 6.2). An additional allowance has to be made for the energy cost of producing the milk. As in the report of the 1971 Committee (1), the efficiency of conversion of food energy to milk energy has been taken as 80% (10).
In many societies women do not consume the extra amounts that seem to be needed to meet the energy requirements of pregnancy and lactation, and it has been suggested that there may be metabolic adaptations leading to increased efficiency (section 4.7). At present, however, there is no well documented evidence for this and further research is needed.
4.2.3 Physical activity
The level of physical activity must obviously be considered in detail when assessing energy needs. Some activities are essential for the individual and the community, and can be considered as economic activities which are life-sustaining. These are designated as occupational energy needs. Previous reports have given general estimates of the energy costs of light, moderate, and heavy activity. The difficulty of applying such figures in practice is recognized, since in many cases activity varies from day to day and from season to season.
The 1971 Committee (1) included an allowance for “leisure-time” activities. The present Consultation attaches much importance to such activities, which are perhaps more appropriately termed discretional, as they are considered desirable for the wellbeing of the community and the health of the individual and the population.
Because of the wide range of variation in both occupational and discretionary activities, it is only possible to give examples typical of particular groups. The examples worked out in section 6 are designed to serve as patterns, showing the method by which the energy needs may be calculated for any particular situation. The notes that follow provide some guidelines for the application of this approach.
Occupational activities. The traditional classification of work according to occupation is maintained in this report, but care must be taken to ensure that there is an adequate desrciption of the occupation. For example, farmers in affluent societies may be relatively sedentary compared with farmers in developing countries who are involved in very strenuous manual labour. The energy cost of travelling or walking to work should be considered as part of the essential energy needs. It must also be recognized that older children and women in developing countries commonly play a significant role in agriculture, in caring for livestock, and in looking after younger children. This is an important contribution to the economy and viability of the household, and energy should be allowed for these essential tasks.
Discretionary activities. There may be many benefits to societies from additional activities outside working hours. The requirement to cover them should not be considered as dispensable, since it usually contributes to the physical and intellectual wellbeing of the individual, household, or group. Such activities can be divided into three categories:
Optional household tasks. A number of optional tasks, such as working in the garden or repairing and improving the home, are an important part of family life. In estimating requirements, an energy allowance should therefore be made for adults for all these activities.
Socially desirable activities. A variety of socially constructive tasks, for example attending community meetings, games or festivals, or walking to health clinics or places of worship, require additional energy expenditure. In some developing countries, where people's main occupation involves a large expenditure of energy, there may be limits on the ability of members of the community to respond to a demand for activities of this kind. For children, additional energy is important as part of the normal process of development, for activities such as exploration of the surroundings, learning, and behavioural adjustments to other children and adults.
Activity for physical fitness and the promotion of health. At all ages physical fitness and wellbeing may depend on leisure-time exercise, and an allowance should be made for this even if it is recognized that at present many people in affluent societies do not expend enough energy in this way. Rather than reducing the estimate of energy requirement, it should be maintained to allow such people to become more physically active. It is, however, impossible to state precisely the desirable level and duration of extra activity.
In sedentary workers, cardiovascular responses to exercise are often inappropriate and muscular strength is limited (11, 12). A small allowance for short periods of physical exercise at a relatively high rate would therefore be beneficial. Although not all authorities are agreed, there is in our view good evidence that for middle-aged men short regular periods of physical exercise at a relatively high rate may have a beneficial effect on cardiovascular risk (13, 14).
4.2.4 Metabolic response of food
The increased oxygen uptake (so-called “specific dynamic action”) after a meal depends on the nutrient composition of the food consumed, and the amount of energy ingested. The greater the energy demands of a subject, because of his size or physical activity for example, the greater the absolute rate of energy expenditure in digesting, absorbing, and storing the larger amounts of ingested nutrients. However, the measurement of the energy cost of these processes in the individual is not easy. It is difficult to separate the energy expended in excess of the basal rate after eating a meal, from the energy cost of the physical activity involved in sitting, eating, and digesting.
For the purpose of estimating energy expenditure, the practical solution is to measure the metabolic rate of individuals in the post-prandial state without limiting minor physical movement. The rate obtained in this way represents the resting metabolic rate. It is greater than the BMR because it includes the energy cost of metabolizing and digesting a meal, as well as the cost of increased muscle tone and minor physical activity. By combining the results of measurements made in the morning, afternoon, and evening, an average figure can be obtained which is an estimate of the resting metabolic rate.
The most important component of energy expenditure, the basal metabolic rate, depends on the mass of metabolically active tissue in the body, the proportion of each tissue in the body, and the contribution of each tissue to the energy metabolism of the whole body. The changes in body composition with age, discussed in section 3.3, markedly affect energy requirements, since some organs of the body are much more metabolically active than others. Table 1 shows that, in the neonate the brain comprises about 10% of the total body weight and may account for 44% of the total energy needs of the child under basal conditions. On the other hand, the energy needs for muscle metabolism at this time are very low because of the relatively small muscle mass. Table 1 also shows that the liver is much more metabolically active than muscle, so that when the mass of muscle is reduced in the aging adult, the whole-body metabolic rate relative to lean tissue mass will also alter. These changes in body composition in children and adults have to be taken into account when calculating the energy requirements of a particular section of the population.
|Organ||Weighta||MR/day||% of wholebody||Weight||MR/dayb||% of wholebody|
|Liver||1.6||482 (2 017)|
|Brain||1.4||338 (1 414)|
|Muscle||30.00||324 (1 356)|
|Miscellaneous, by difference|
|Total||70.00||1 800 (7 530)||3.5||197 (824)|
a Organ weights taken from Boyd (15).
b Metabolic rates for the neonate estimated by assuming that the metabolic rate of each organ per unitweight is the same as in the adult. The total activities of the tissues listed are expressed as fractions of thetotal basal energy expenditure in the adult and the neonate. The total basal metabolic rate in the neonateapproximates to that measured by Benedict & Talbot (16).
There are also altered activity patterns with age; children become progressively more active once they are able to crawl or walk and the physical activity patterns of adults are usually dominated by the nature of their work. If adults retire from work, this change in habits must be recognized in estimating their energy requirements, but it should not be assumed that the marked decline in activity that often occurs in the elderly is either inevitable or desirable. If energy intake declines with increasing inactivity, then an individual is much more likely to have a diet deficient in one of the essential nutrients. In determining the desirable minimum activities the Consultation suggests that the energy allowance considered appropriate for discretionary activities should be maintained throughout adult life and increased for those who have retired from work.
The basal energy expenditure on a weight basis differs little between pre-adolescent boys and girls, but since there are differences in body weight and composition from the first few months of life, and different physical demands are made on boys and girls, their energy requirements are considered separately.
After maturity, men have a relatively greater muscle mass than women, which would tend to reduce their BMR when expressed in terms of lean body mass, since muscle has a low metabolic rate (Table 1). However, the greater body fat content of women means that the observed BMR per unit total body weight is somewhat lower in women.
The energy demand for physical activity will often depend on the different types of employment for men and women. However, the heavy burden of agricultural work for women in rural communities must be taken into account. In addition, this report suggests that for both sexes there should be a minimum desirable energy requirement to allow the same amount of discretionary activity.
In any assessment of the average requirement, both intra- and inter-individual variability must be recognized. The former results from short-term fluctuations in energy intake and expenditure. In the United Kingdom it has been found that measurements over a period of 2–3 weeks were needed in order to assess correctly the intakes of individuals (17). There are also short-term fluctuations in energy output, and it has been found that even under well controlled conditions, observations may have to be continued for several weeks before input and expenditure are balanced (18).
It is also generally recognized that in a group of apparently comparable people there is much inter-individual variation in habitual energy expenditure, and hence in requirement. In a number of selected studies, the measurement of total energy expenditure over a week indicates that the inter-individual variability of expenditure, in a specified group, has a coefficient of variation (CV) of about ± 12.5% on a body-weight basis (19).
There is almost no information about the intra-individual variation in energy requirements in developing countries. Unpublished data from Papua New Guinea indicate a coefficient of variation of 10–16%.
4.6.1 Basal metabolic rate (BMR)
The data for BMR are extensive and the equations in Table 5 (see page 71) allow reasonably precise estimates for individuals of a given sex and weight. Direct measurement of the BMR demands close attention to detail and imposes artificial conditions on the subject —who should be in the post-absorptive state and at complete rest in a thermoneutral environment. In practice the BMR measured in this way is approximately equal to the energy expenditure of subjects during sleep. It is therefore considered valid to measure the BMR of individuals and to assign this energy cost to the time during which the subject is asleep. Anxiety is often cited as an important cause of increased energy expenditure, but direct estimates do not confirm this (20, 21).
4.6.2 Physical activity
The different types of activity undertaken by an individual can be identified and the time spent in each activity measured. The energy cost of each activity can then be obtained by measuring the subject's oxygen uptake while performing the task, either with a Douglas bag to collect the expired air or a less restricting spirometer such as the Kofranyi-Michaelis apparatus. Since there are no studies on the energy expenditure of free-living subjects that do not rely, directly or indirectly, on such encumbering apparatus, it is not possible to say whether the use of respirometry introduces errors. Numerous studies of the energy cost of different activities have been made by this procedure (22, 23). It does not give the net energy cost of each activity above the BMR, but monitors the total rate of energy expended during the period of exercise. The values tabulated in Annex 5 are used in section 6 to produce approximate estimates of the energy expenditure of individuals under different conditions.
The energy cost is usually expressed per minute rather than per day. The total for 24 hours is then calculated according to the time spent on that activity. The energy cost of a standardized form of physical activity is relatively easy to measure, and it is also possible to estimate the variability in energy cost between individuals in performing the same task. It is much more difficult, however, to obtain accurate values for those tasks that combine a variety of movements, some of which demand the use of heavy parts of the body while others involve only small muscle groups without major weight-bearing or movement of the body.
It is also difficult to generalize on the extent to which differences in body weight affect the energy expenditure for a given type of physical activity (see section 4.2 and Annex 3). A relation between energy cost and body weight is to be expected when the task involves moving the body, but not when it involves work on external objects. Clearly, many tasks will be a mixture of both types. In the absence of data it has been assumed in this report that, regardless of body weight, the same multiple of BMR can be used to express the energy cost of each activity. It is clearly desirable that, wherever possible, investigators should make their own measurements.
A number of workers have estimated the energy expenditure of free-living subjects over periods of several days by monitoring and integrating the number of heartbeats (24–26). This procedure relies on indirect calorimetry, since for each individual a calibration curve has to be made relating heart rate to oxygen uptake. The main drawback of this method is that at low levels of activity many physiological and psychological factors may affect the heart rate without appreciably affecting energy expenditure.
4.6.3 Time-scale of estimates
For practical purposes the requirement is generally expressed as a daily rate, although the estimates refer to levels of habitual energy expenditure. It is recognized that expenditure varies not only from day to day, but also from week to week. The estimates should represent the average requirement over longer periods. The variability in energy requirements within a group can then be attributed to inter-individual variation only and not to intraindividual variation.
The length of the period chosen will vary with the circumstances. It is well known that there can be substantial variations in the energy requirements of the same subject under different circumstances; for example, the demand for physical activity increases during harvesting in developing countries. At this time of the year energy expenditure may exceed intake and, if so, body weight falls. While some weight loss may be tolerated for a short time, it may be necessary to take account of intermittent periods of heavy work in calculating overall energy needs.
4.6.4 Corrections for metabolizable energy
Once the requirements for energy are obtained from measurements of energy expenditure, the dietary intake needed to meet the energy demand must be determined. Intakes of metabolizable energy have to be calculated to allow for the availability of dietary energy from different sources which may present special problems in some cases, for example a diet rich in fibre (see section 7). The traditional Atwater factors were designed to allow for non-available energy in different foods, but the corrections for unabsorbed carbohydrate are made in different ways in different food tables (see section 7.1). In general, the Atwater factors are still the most suitable in the absence of more specific knowledge on the availability of energy in particular foods.
When intakes are being compared with requirements, it is usually preferable to correct for the metabolizable energy of the diet rather than to adjust the estimate of requirement. For example, if a diet provides only 90% of the metabolizable energy predicted from the Atwater factors, for most purposes it is more convenient to correct the estimated energy intake by a 10% reduction, rather than to increase the requirement by 10%. However, for some uses the latter method may be more appropriate.
The general principles of adaptation have been considered in section 3.1. Adaptation to changes in energy intake can affect energy requirements in three ways; by alterations in body size (already discussed in section 3), by metabolic adaptation, and by behavioural adaptation.
4.7.1 Metabolic adaptation
A very substantial adaptation in total energy requirements can occur on submaintenance intakes, but this involves large changes in most if not all of the factors contributing to total energy expenditure. In the classical Minnesota study of Keys et al. (27) normally nourished subjects were able to achieve approximate energy balance after 6 months on half their usual energy intake. There was, however, a profound fall in body weight (part of which was lean tissue), a reduction in physical activity, and some mental changes, as well as metabolic adaptation shown by a fall in the basal metabolic rate. This degree of adaptation was clearly disadvantageous and the problem is to define the range of adaptation in total energy expenditure that can be achieved without any detectable disadvantage.
When there is a substantial fall in energy intake, in addition to the loss of body weight there is a reduction in the BMR, which declines over a 3-week period by up to 15% when expressed per unit of body weight (27). Thereafter, further falls in BMR are achieved primarily by a progressive loss of active tissue mass. It is uncertain whether this degree of metabolic adaptation can occur in the absence of a reduction in body size, nor is it clear whether, under conditions of slight energy restriction, there is a fall in the BMR which can then be maintained to bring the body back into energy balance. Table 2 shows data on the response to semi-starvation in volunteers, compared with the energy expenditure of men and women in Papua New Guinea (28). Results from a group of well-fed British women studied by similar methods are also given to show that the resting metabolic rates, which include the metabolic response to food, are similar in the different ethnic groups. From these data the major differences in energy turnover appear to relate to the energy available for physical activity. However, subtraction of the resting metabolic rate rather than the BMR from total expenditure will tend to underestimate the expenditure on physical activity.
When subjects of normal weight are overfed experimentally there is an increase in both lean tissue and body fat, but some metabolic adaptation can also occur. Many overfeeding experiments have been undertaken, but there is little information on total energy expenditure under these conditions. Short-term overfeeding studies have generally shown that most of the extra energy is stored and not dissipated as heat (29, 30). However, it has been repeatedly confirmed that there are modest increases in the BMR with substantial overfeeding (31). Long-term overfeeding experiments have been claimed to demonstrate remarkable changes in the body's ability to cope with overfeeding (32). Unfortunately these studies did not include direct measurements of energy expenditure.
|USA men (27)||Papua New Guinea (Kaul) (28)||British womena|
|Body weight (kg)||70|
|Fat-free mass (kg)|
|Daily energy output|
|Basal metabolic rate|
|Resting metabolic rate|
|Metabolic rate—basal or resting per kg|
|Estimated physical activityb|
a Unpublished data of James et al.
b Physical activity estimated in the USA men by subtracting the basal metabolic rate plus a theoretical 10% of intake as specific dynamic action from the total output.For the Kaul group and the British women the resting metabolic rate, which includes the specific dynamic action, has been subtracted from total energy output.
c Basal metabolic rate.
d Resting metabolic rate.
Metabolic adaptation in other components of energy expenditure have been sought under conditions of underfeeding and overfeeding. It has been suggested that there is an interaction between the metabolic response to food and exercise (33). This response may increase with overfeeding and decline during semistarvation, though the effect is small. There is little evidence that exercise performed under fasting conditions in semi-starved or overfed individuals is appreciably different in efficiency. Finally, studies on the specific dynamic action of a standard meal without exercise in semi-starved or overfed individuals do not suggest major changes in their metabolic response, although in children recovering rapidly from malnutrition the metabolism surges after a meal (34), and this is related to their rate of growth.
The documented changes in metabolism when energy intake is altered suggest, therefore, that with the present state of knowledge the range of metabolic adaptation must be considered to be small. A variety of mechanisms have been suggested to explain such differences as have been found in the efficiency of energy utilization, including changes in plasma thyroxine and tri-iodothyronine concentrations and in the processes of protein turnover, substrate cycling, and perhaps activation of brown adipose tissue metabolism. Evidence on the quantitative importance of each of these factors is scanty.
4.7.2 Behavioural adaptation
A marked reduction in food intake leads to a profound decrease in physical activity (35). Children in Guatemala were found to decrease their energy expenditure without changing their growth rate when their dietary energy was reduced by 10% (36). On the other hand, in studies in Mexico, supplementation of the diet of children led to an increase in physical activity and exploratory behaviour (37). It has also been shown (38) that when the diet of male agricultural workers in Guatemala was supplemented with additional food, there were appreciable increases in their activity at work and in their discretionary activity without any increase in body weight. There was also an improvement in their sense of wellbeing. These findings suggest that limited food intake makes an appreciable difference to the work capacity of a community and that this happens without, in general, alterations in body weight. Similar improvements in subjective wellbeing with very small weight increases have been found in lactating Gambian women when given supplementary food (39).
This relationship between energy intake and work output deserves serious consideration during the assessment of energy requirements. The need to provide for the energy cost of socially desirable activity in the home and community has already been emphasized.