ENERGY REQUIREMENTS IN THE TROPICS

D.S. Miller
Department of Nutrition
Queen Elizabeth College,
London W.8

Introduction

Biologists inevitably have trouble with terminology since many of the words that they use have both a well-defined meaning and a common usage. This applies particularly to human nutrition. The word energy to the layman implies joie de vivre, and the word efficiency implies capability with minimum effort. In physics, energy has a very precise meaning and in animal nutrition efficiency is concerned with optimum gain of a product useful to man in relation to food consumed, e.g. weight gain/food intake. In man, fattening is not a desirable characteristic, and one objective of industrial society is to reduce human labour. Many city workers would be pleased to dissipate their energy stores and avoid the consequences of over-eating. Indeed, directors in London are encouraged to indulge in isometric work, that is to say to increase muscle tone, by attempting to push the walls of their offices down. It is claimed that this is an efficient way to reduce weight, but since no work is done their efficiency is zero! The definition of the word requirement involves even greater difficulties, indeed controversy. Thus "minimum physiological requirement" (usually reduced to nutrient requirement) is not the same as "recommended daily intake" (usually reduced to nutrient allowance). The former has been given as the lowest ever recorded, the average of all measurements, and as a figure based statistically to cover 97% of the population.

Much depends upon how the standards are going to be used. Many authorities state that their figures are only intended for assessing the nutritional status of groups, but it is open to doubt whether such exercises are of value. One is implicitly advised that if the mean intake of a group is equal to the mean requirement, then everyone is O.K. However, the thing about a mean is that half the group will be above it and half below it. This argument has been used to calculate not only that 50 percent of the British population is below average intelligence, but also to exaggerate the extent of malnutrition in the world. The real trouble with such abuses is not the calculations of amateur statisticians but that the recommended methods are not locical. They simply perpetuate an old fallacy in a new form; we have all heard the difficulties of the tall man in negotiating doors of average height. An examination of the population by methods of studying the averages of groups would reveal no tall men, and insist that if tall men exist they would choose only to go through tall doors. Both nutrient requirements and nutrient intakes of individuals cover a range of values, and as there is no reason to suppose that those with a higher requirement conveniently eat a higher amount, the only sensible thing to do is to state requirements that cover the needs of practically all individuals. This would mean of course that some would get more than necessary and this might lead to obesity. However, there is accumulating much evidence to show that many individuals can dispose of excess calories by lowering their efficiency of food conversion: within limits man can adapt to his food supplies.

If requirements are to cover most individuals in a population it follows that dietary surveys should be made of individuals rather than groups: we know quite a lot about intakes of nations and of households in many parts of the world, but almost nothing about the distribution of food within the family. With this knowledge estimates could be made of the number of under-nourished individuals in a population. Since father frequently gets the lion's share of the family's food it seems probable that measurement of individual food intakes would reveal a higher incidence of malnutrition than previously suspected from the study of groups.

This paper is in two parts; the first examines energy requirements, and the second how these are modified by the tropics. The figures are in calories despite the adoption this year of the joule as the new unit of energy, but these may easily be converted by multiplying by 4.185. A broad view of the tropics has been taken, in which their influence has been interpreted to include factors other than environmental temperature, such as race, body composition, and the effect of altitude since many populated areas are on elevated plateaux. Also included is the influence of economic status since underdevelopment deprives a population of mechanical aids in everyday life, and also because populations in tropical areas have become adapted to undernutrition. The influence of disease on energy requirements is important, especially since these will influence heat production and cause immobilisation, but this subject is to be dealt with by Dr. Bennett.

Energy Requirements

There are two approaches to the measurement of energy requirements. The first is based on the rather dubious assumption that healthy individuals who are neither gaining nor losing weight must be consuming enough energy to meet their needs, and all one has to measure is their food intake. The second method which is theoretically more sound but technically more difficult is to measure energy losses and to assume that these must be replaced to meet energy requirements.

Measurement of Food Intake

In order to understand the drawbacks of the first method it is necessary to examine the factors that influence calorie balance. Equation 1 has been hailed as being fundamental in energy exchanges in man, and although indisputably true is an oversimplification of a very complex situation.

Body Energy = Energy In - Energy Out..... (1)

Clearly man cannot escape the consequences of physical laws, but the first law of thermodynamics which is concerned with the conservation of energy says nothing about the equivalents of energy and body weight. Gains or losses of weight may consist of water or minerals containing no calories, or for example fat containing nine calories per gram. The second law of thermodynamics which is concerned with entropy states explicitly that in any energy transformations there is an inevitable loss in the form of heat. In biochemical terms no reaction can be one hundred percent efficient; certainly the synthesis of fat must be accompanied by losses of heat, and even dynamic equilibria will require energy to maintain them.

In adult man there appears to be a very precise control of body weight. Thus individuals who can maintain body weight between the ages of 20 and 60 years do so despite a throughout of about 50 tons of food. It has always been assumed that this control is effected through appetite, but recent evidence suggests the possibility of a control also through heat production. There are at least four theories concerning the control of appetite which involve factors such as stomach capacity, blood sugar levels, skin temperature, or the size of the fat stores, all of which are effective through the hypothalamus. The experimental evidence for these theories is good in experimental animals, and poor in man whose appetite is influenced also by social and psychological factors which can override the simple physiological ones. Control of body weight by means of burning excess metabolic fuel has been shown to be possible at least in young people; students at Queen Elizabeth College can double their food intake without laying down fat (Miller & Mumford, 1967; Miller, Mumford & Stock, 1967). They could maintain weight over a wide range of caloric intakes and there is much evidence to support this view. Also, whole populations can achieve energy balance at different levels of intake which produce different stable body weights. The undernourished populations of many tropical areas maintain low body weight on a poor diet, and the affluent westerner a higher body weight on an excessive diet. Weight changes accompanying changes in intake are seldom linear, but plateau at new levels. The heights and weights of British school children have been rising since the bad old days at the turn of the century, and there seems little doubt that the same process will be followed by children in the developing countries. However, the assumption that people who overeat inevitably become obese is also open to doubt. The upper socio-economic classes in Britain consume about 25% more than their requirements despite occupations which are less active than those of the rest of the population, but the prevalence of obesity is known to be less in this group. It appears therefore that a lowered efficiency of food conversion must occur amongst the overeating members of a community.

The observed wide range of food intakes of weight-maintaining individuals raises a much more difficult question as to what level promotes the greatest benefit. Mortality statistics do not help us, for although undernutrition may be condemned for lowering resistance to infection, overnutrition appears to be associated with degenerative diseases. There seems little doubt however that moderate intakes together with preventive medicine would lead to a long life if not a merry one.

Measurement of Energy Losses

The direct measurement of energy expenditure has much to commend it. Energy losses from the body may be both kinetic and potential, and may be most accurately measured by placing subjects inside a calorimeter for the measurement of heat production, and this together with a knowledge of the energy content of the excreta and any work done provides a total measure of such losses. Heat production is derived from two sources, that associated with basal metabolism, and thermic energy which arises from the inefficiency of metabolism and work. Thermic energy is analogous to entropy in thermodynamics. The size of calorimeters limits the type of activity that may be performed in them. The large calorimeter may be most theoretically satisfying, but in view of man's peripatetic nature has very little relevance to everyday life.

There is a great need for portable instruments to measure energy expenditure over long periods of time; such instruments should not interfere with ordinary existence and should be socially acceptable. Portable instruments do exist, but rarely approach these specifications. They depend upon the principle that energy can only be expended if a metabolic fuel is oxidised. From a knowledge of gaseous exchange, in particular oxygen consumption, the metabolic rate of the individual can be calculated. The Douglas bag involves the subject wearing a nose clip and mouth piece, together with a voluminous and cumbersome bag for collecting the expired air. Two improvements on this device have been introduced. The Max-Planck Respirometer is essentially a mechanical gas meter with a device for obtaining an integrated sample of expired air, and the IMP is an electronic model performing the same function. Both machines can be used with a face mask and are worn as a small pack on the back. However, whenever our students walk through London wearing these instruments they are treated as men from Mars. Their main disadvantage is that they cannot be worn for twentyfour hours. Nutritionists are interested in energy balance and the minimum meaningful unit of time is a day. Consequently the use of these machines is based on measurements of specific activities over short periods of time, and the subjects are asked then to keep a careful diary throughout the day of how much time they spend on each activity. (Passmore & Durnin, 1955). The assumption that is made here is that the cost of any given activity is always the same, and no allowances can be made for variations in energetic efficiency. Some people take figures from the literature and apply them in a factorial way to all subjects, but it is better where possible to make measurements on all subjects in which the experimenter is interested and thus assess individual variations.

Measurements of twentyfour hour energy expenditure are possible with a device recently introduced which measures heart rate and which can be worn unobtrusively for a number of days. Attempts to show a correlation between heart rate and oxygen consumption have been most encouraging but difficulties do arise over changes in posture. If these teething troubles can be overcome the instrument has a great future.

The Factorial Method

Using the above two approaches, FAO in common with many national committees have computed energy requirements. The FAO method starts with a consideration of a carefully defined Reference Man and all other individuals are related to him. By analysing measured figures for energy expenditure, FAO showed them to be built up from three factors:

(a) The basal metabolic rate (BMR)

It has long been known that the heat production at rest of an individual depends upon his surface area. Small animals with a large surface-to-weight ratio produce relatively more heat than large animals. The weight in kilogrammes (W) raised to the power of three quarters is used as an indicator of surface area.

(b) Activity

On the other hand, the work done by an individual in moving around is from physical considerations directly proportional to weight. For example, the energy cost of walking, in which most of the work done is by raising the centre of gravity at each step, is directly proportional to body weight. One might think that the customary activity of individuals would show an enormous range, but changes in the pattern of life, at least in industrial societies, have narrowed this considerably. In rural societies the actual time spent doing hard physical labour is less than was previously thought.

(c) The thermic effect of food

As mentioned above, when food is consumed by animals it is usually accompanied by an increase in metabolic rate, a rise in skin temperature and a loss of heat. This process is little understood but contributes to energy requirements, and FAO assumed the value of one tenth of the intake lost in this way, which suggests that man is 90% efficient in food conversion. By comparison with farm animals bred for efficiency this figure is very high, and is surely an overestimate.

Summarising, the calculations of FAO energy requirements (E) are given by equation 2:

E = 0.1E + 90W^3/4 + 10W + 200 ......... (2)

which has the great advantage that each of the terms can be checked by further research. It will be seen from equation (2) that there is a poor correlation between body weight and daily energy expenditure, an observation made by many workers in the field.

Reference Man, who is 25 years of age and weighs 65 kg., is physically fit for active work, and defined as having an eight-hour working day in "an occupation which is not sedentary but does not involve more than occasional periods of hard physical labour"; he requires 3200 kilocalories daily. Despite the fact that it is the view of FAO that the degree of physical activity is the most important factor influencing calorie requirements they feel unable to make allowances for this factor. However, Nicol (1959) from his studies in Nigeria points out that there is certainly no reason to assume that the peasant farmers of developing countries have a lower energy expenditure than Reference Man. The range of requirements from the truly sedentary to exceptionally active is thought by FAO to be between 2400 and 4000 kilocalories per day, but in their view the number of persons engaged in work at these extremes is relatively small.

FAO reduce calorie allowances with age such that the requirement at 45 and 65 are 6% and 21% respectively less than at 25 years. These considerations are based on the observation that elderly people eat less food rather than that they expend less energy. However, there is a small fall in basal metabolic rate with age and no doubt a decline in physical activity.

FAO also define a Reference Woman and allowances are made for pregnancy and lactation on the basis of increased metabolic rate, the requirements for growth of the foetus, the placenta and associated maternal tissues, and the quantity of breast milk produced. For pregnancy an allowance of 40,000 kcals per pregnancy is suggested, and an allowance of 1000 kcals per day for six months for lactation. In providing these figures it has been assumed that physical activity of women is reduced during reproduction, but it is admitted that a poor woman who has already several children will be unable to decrease her activity much. Also it has been assumed that there will be a daily milk production on 850 ml. for six months, standards which certainly do not apply to most developing countries where breast feeding is frequently common for the first two years of life, albeit at a meagre flow-rate.

FAO also make allowances for environmental temperature, but the evidence against this procedure is now so strong that it seems unlikely that the new committee now sitting will include them: this is discussed in detail below.

The Tropics

Geographically the tropics are easily defined, but the conditions that could influence calorie requirements present an enormous variety. The temperature and humidity ranges within the two imaginary lines are large due to location and season. The recommendation that in the assessment of calorie requirement a mean annual environmental temperature should be taken is an example of the use of a meaningless mean. Also it is naive to assume that there are no socio-economic differences which could influence human biology; a mud hut in the Congo and a room in the Nairobi Hilton are after all both in the tropics. The following four factors seem to be the most important.

Climate

The FAO committee recommended that calorie requirements should be decreased by 5% for every 10°C of mean annual external temperature above a reference temperature of 10°C, on the basis that BMR is known to be lower in the tropics (Munro, 1950; Quenouille, Boyne, Fisher & Leitch, 1951; Roberts, 1952). Since this statement arouses much controversy the subject is dealt with here in more detail.

According to Rubner (1902) heat loss by an average man in Europe shows that 44% is lost by radiation, 31% by conduction and convection, 21% by the evaporation of water, and only 4% by the heating of expired air and work. The high figure for radiation will surprise some, but it has been known for some time that man, both black and white, behaves as a Newtonian black body. The rate of cooling and heating of such a body by radiant heat is proportional to its surface area, and the fourth power of the absolute temperature of the surfaces which radiate towards each other: it does not depend upon the temperature of the air between them. However, as the environmental temperature increases the importance of radiation becomes less dominant and homeostatic control of body temperature is maintained primarily by the vaporisation of water. As the environment reaches body temperature the vaporisation of water accounts for almost 80% of the heat loss (Hardy & Dubois, 1938).

If the environmental temperature of experimental animals is increased from a low level, heat production falls until it is equal to the basal metabolic rate. Such extra heat was necessary simply to keep the animal warm. Above this critical temperature, heat production is constant and equal to basal metabolic rate, but for environmental temperatures above body temperature homeostasis breaks down and heat production rises again. With man, allowances must be made for his ability to change his clothing. It seems more appropriate to measure a man's microclimate, that is the temperature under his shirt than the mean environmental temperature for a given geographical location. Arctic explorers nowadays live in centrally heated accommodation and inhabitants of the tropics tend to go around lightly dressed. On these considerations alone it seems unlikely that man should need extra food to keep himself warm in temperature and subarctic zones. In fact there is now a wealth of evidence to show that man requires more food in the tropics, not less.

Consolazio, Shapiro, Masterton & McKinzie (1961) conducted an extensive study in the hot desert heat at Yuma, Arizona. The study was divided into three periods; during the first the men were kept outdoors, during the second they were allowed shade, and in the third period they were moved indoors into an airconditioned room. Measurements were made of energy expenditure and balance, together with changes in body composition. There was a highly significant positive correlation between energy requirement and environmental temperature which was not greatly improved with the addition of humidity. The increased requirements were apparently due to a combination of increased action by the blood in heat transport, increased action of the sweat glands, increased calorie losses due to sweat vaporisation, plus an increased metabolic rate due to an elevation in body temperature. Collins & Weiner (1968) have shown that the pattern of endocrine secretion, particularly of the pituitaryadrenal-thyroid system, is involved under heat stress and that this is accompanied by an increase in metabolic activity. Some criticisms of the Yuma study were whether the men in the experiment were fully acclimatised to the heat, and whether the increase in energy requirements were due to insufficient training prior to the beginning of the experiment. A further study was carried out on eight men under strictly controlled levels of temperature and humidity (Consolazio, 1963). The daily activity levels were controlled by use of a bicycle ergometer and values for oxygen consumption were obtained for light, moderate and heavy physical activity at three different environmental temperatures (see Table 1). This study again showed that there was a significant increase in metabolic rate of men living in the heat.

This work carried out at the U.S. Army Medical Research Laboratory appears to be beyond criticism, and the increase in energy requirement for men living and working in the tropics has important biological aspects. Hitherto the low energy intakes of tropical communities have been explained in part by a higher environmental temperature. The effect of Consolazio's work is to emphasise the overnutrition in temperate climates and the undernutrition in tropical climates. It further demonstrates the importance of energy in the world food problem (Miller, 1970; Sukhatme, 1970).

Race

As mentioned above, FAO assumed that there was an inverse relationship between mean annual temperature and BMR. There is however some doubt as to whether this is a climatic or a racial effect. Classical physiological thinking would suggest that individuals of the same age, height, weight and sex would have the same BMR, but variations of body shape and proportions, especially with respect to muscular development and fatness would have an effect on BMR. In a study of Indians and Europeans in Bombay, Mason, Jacob & Balakrishnan (1964) demonstrated that the Europeans had a significantly higher BMR than Indians when measured in the same place. However, a similar study by Mahadeva (1954) of Asiatics and Europeans in Edinburgh showed no significant differences.

However, values for BMR are stated as rates per unit surface area, a convention that obscures the fact that big men have a higher absolute daily metabolic rate than small men. Equation (2) recommended by FAO to calculate energy requirements takes this into account, and thus provides for a greater food intake for Europeans compared with inhabitants of the tropics because of differences in average adult weights. Little consideration appears to have been given to the possibility that differences in body weight are not racial but nutritional in origin. There is much evidence to indicate that poor nutritional status leads to reduced stature. For instance, Coon (1950) has shown that the height of the genetically isolated Icelander shrank from that of their tall Norwegian ancestors during a century of depression but returned again later to make them amongst the tallest of European people. Similarly, the average height of Japanese that have migrated to the USA increases with successive generations. Examples of the influence of nutrition on the body size of tropical peoples is more direct. Forty years ago, Boyd Orr & Gilks (1931) attributed the differences in adult stature between the Masai and the Kikuyu to diet. A recalculation of these diets suggests that the effect was probably due to energy rather than protein intake, since both diets appear adequate in the latter by modern standards (Table 2). McCarrison (1936) also provides data from which a similar comparison can be made between the diets of the tall Sikh and his smaller countryman the Madrassi with similar results. McCarrison fed rats on the two diets and those fed the Sikh diet weighed on average 255 g. compared with 155 g. on the other. Similar examples may be found in Brazil, Ceylon, Puerto Rico and many other countries (Weiner, 1964).

The question that may be asked is, should one make energy allowances based on actual or potential body size? Perhaps it would be more reasonable in the present state of knowledge to omit this factor as well as that for temperature. The two together provide a reduction in calculated energy requirements from 3200 kcals per day for Reference Man (adult weight 65 kg., environmental temperature 10°C), to say 2450 kcals per day for a man living in the tropics (adult weight 50 kg., environmental temperature 25°C) as shown in Table 3. Omission of this manipulation would suggest an increased requirement of food supplies to the tropical countries of 30%, an alarming prospect indeed. More likely, the temperate zone requirements are set too high. Such calculations indicate the need for more precise physiological information and demonstrate that the deliberations of biologists on international committees can have far-reaching effects in political and socio-economic fields.

Altitude

Many of the populated areas in the tropics are at high altitude and this can have an effect on energy requirements. Work carried out by us in Ethiopia (Miller & Stock, 1969) has demonstrated a long suspected phenomenon that the thermic effect of food is much reduced at altitude. In both Europeans and Ethiopians the thermic effect of a standard meal was zero at 10,000 ft., 5% at 5,000 ft. and 10% at sea level. Increased efficiency of food utilisation on high plateaux in the tropics could be a valuable asset in times of famine, but might lead to obesity in times of affluence.

Socio-economic Status

Although it is now some thirty years since the foundation of the United Nations and its agencies, the food intakes of many individuals in tropical countries is still low. The incidence of overt famine is much reduced, but chronic undernutrition is still a feature of many developing countries. Adaptation to low food intakes takes various forms. There is reduced physical activity and body movements are made with maximum economy, basal metabolism is reduced and body weight is lost (Keys, Brozek, Henschel, Mickelson & Taylor, 1959). Such people are not efficient in the colloquial use of the term, but biologically they become very efficient in utilising their meagre food supplies. Typically, resistance to disease, particularly intestinal infection, is lowered and infant mortality rates are high. It is amazing how a community can survive when half the children born to it do not reach maturity, but such communities are not uncommon in S.E. Asia, India, Africa and South America.

Statistics collected by FAO demonstrate that many tropical countries only just reach adequacy insofar as energy is concerned, but their data are average consumptions compared with average requirements. Such meaningless means hide the poor distribution of food supplies by area, social class, and within the family; and do not reveal the number of individuals who are undernourished. A few surveys of individual food intakes have been carried out and these demonstrate low, almost incredibly low values compared with normal standards. Most field workers can describe adult males, particularly agricultural labourers, of fine physique and apparent good health that are consuming less than 2000 kcals per day. It seems that once past the hazards of childhood man can adapt to very low food intakes. However, the body weights of such men fluctuate considerably showing that they are existing at the lower limits of food intake. Fox (1953) provides data from Gambia of body weight changes in the course of the agricultural cycle over a period of three years. Both men and women were heaviest during February immediately after harvest when food supplies were abundant, and lightest during October when agricultural work was extensive and food in short supply. The overall energy turnover was typically less than 2000 kcals per day, a level providing no buffer against minor misfortunes. Such a community can only achieve energetic equilibrium on a yearly rather than a daily basis.

Provision of only a few hundred extra calories daily would probably have removed these fluctuations in body weight. In the villages surveyed by us in the high Simien mountains in Ethiopia, the mean caloric intake was low and this was matched by low skinfold measurements and low heart rates. However, no seasonal fluctuations of body weight were observed. The physical stature of the men was surprisingly good, although they did not engage in sustained hard physical work. Areskog, Selinus & Vahlquist (1969) have shown that such apparently undernourished Ethiopians are surprisingly physically fit. It is tempting to suggest that the calorie allowances for such men are set much too high, and that many fat business men in Addis Ababa and London would do well to eat their diet. But one suspects that as economic development proceeds the reverse will be true. Certainly industrialisation will provide more mechanical aids for the cultivator and this will lower his energy requirements. Claims that increased food supplies would improve productivity are difficult to substantiate, although evidence is available that lowering food supplies results in reduced work output (FAO, 1962). Northcott (1949) made this point in respect of Kenya twenty years ago, but the matter is complicated by the consequences of economic development and the motivations of labour. But this is the subject of Professor Boshoff's paper.

The World Food Problem

Everyone knows there is a world food problem, but there is a growing controversy over its nature and solution. In recent years emphasis has been placed by some authorities on the specific shortage of protein, and this has led to much enthusiasm for the manufacture of high protein supplements from unconventional sources. But FAO (1965) removed the "protein gap" by recognising the ability of man to maintain nitrogen balance on very low protein intakes. An examination of national food supplies reveals a greater shortage of energy than of protein by these standards. That is not to say that protein deficiency does not exist, since if the diet has a low energy value protein will be used as a fuel rather than for the maintenance of nitrogen balance. About 30% of the basal metabolic rate is utilised for protein synthesis, and Miller & Payne (1964) have shown that this requires approximately 24 kilocalories per gram. Using this information Sukhatme (1970) has demonstrated that only in about one out of ten cases does kwashiorkor occur as a direct result of inadequate protein intake. In the vast majority of cases protein deficiency is the result of an inadequate intake of total energy. Similarly, Miller & Payne (1969) have demonstrated that most diets in the world based on most staples are adequate with respect to protein if consumed in sufficient amount. However it should be realised that whereas the protein requirements used in this debate are based on physiological minima, the energy requirements derive primarily from the average intakes of healthy individuals in Europe and North America. As argued above, the latter may be set too high, and this has the effect of over-emphasising the importance of protein in the diet.

These interrelationships between protein and energy metabolism demonstrate the need to consider both concomitantly when devising nutritional standards. The new FAO committee is doing just this, and we await its report with interest.

Summary and Conclusions

The factors that influence energy requirements and some of the instruments used to measure them are described. Doubt is expressed whether requirements in the tropics should be modified by climate or race. Food intakes in developing countries are low and this is associated with a high infant and child mortality rate, but those adults that survive appear to be adapted to low energy intakes. The standards laid down by FAO (now under revision) are criticised for providing a lower energy allowance for tropical countries compared with that for the affluent nations. The major difficulties in stating energy requirements are (i) man's ability to maintain weight over a wide range of intakes, and (ii) the interrelationships between protein and energy metabolism.

REFERENCES

Areskog, N.H., Selinus, R. & Vahlquist, B. Physical work capacity and nutritional status 1969 in Ethiopian male children and young adults. Amer. J. clin. Nutr., 22, 471

Coon, S.C., Garn, S.M. & Birdsell, J.B. Races, A Study of the Problems of Race Formation 1950 in Man. Charles C. Thomas, Springfield, Illinois

Collins, K.J. & Weiner, J.S. Endocrinological aspects of exposure to high environmental 1968 temperatures. Physiol. Rev., 48, 785

Consolazio, C.F., Shapiro, R., Masterton, J.E. & McKinzie, P.S.L. Energy requirements of 1961 men in extreme heat. J. Nutr., 73, 126

Consolazio, C.F. The energy requirements of men living under extreme environmental 1963 conditions. Wld. Rev. Nutr. Dietet., 4, 53

FAO Calorie Requirements. FAO nutr. Stud. No. 15 1957

FAO Nutrition and Working Efficiency. FFHC Basic Study No. 5 1962

FAO Protein Requirements. FAO nutr. Mtg. Rep. Ser. No. 37 1965

Fox, R.H. A study of energy expenditure of Africans engaged in various rural 1953 activities. Ph.D. Thesis, London University.

Hardy, J.D. & Dubois, E.F., quoted in Brody, S. (1964), Bioenergetics and Growth. New 1938 York, Hafner Publishing Co. Inc. Chapter 11

Keys, A., Brozek, J., Henschel, A., Mickelson, O., & Taylor, H.L. The Biology of Human 1950 Starvation, Vols. I & II. Minneapolis University of Minnesota Press.

McArthur, M. Some factors involved in estimating calorie requirements, with special 1964 reference to persons engaged in agricultural labour in Asian countries. J. Roy. Statist. Soc. (A), 127, 392

McCarrison, R. Nutrition and Health. London, Faber & Faber Ltd. 1936

Mahadeva, K. The energy expenditure at rest of Southern Asiatics in Britain. Indian 1954 J. med. Res., 42, 181

Mason, E.D., Jacob, M. & Balakrishnan, V. Racial group differences in the basal metabolism of Indian and European women in Bombay. Human Biol. 36, 374

Miller, D.S. Dietary assessment of nutritional status. Proc. nutr. Soc. (in press) 1970

Miller, D.S. & Mumford, P. Gluttony. 1. An experimental study of overeating 1967 low- or high-protein diets. Amer. J. clin. Nutr., 20, 1212

Miller, D.S., Mumford, P. & Stock, M.J. Gluttony. 2. Thermogenesis in overeating man. 1967 ibid, 20, 1223

Miller, D.S. & Payne, P.R. Dietary factors influencing nitrogen balance. Proc. nutr. 1964 Soc., 23, 11.

Miller, D.S. & Payne, P.R. Assessment of protein requirements by nitrogen balance. 1969 Proc. nutr. Soc., 28, 225

Miller, D.S. & Stock, M.J. The effect of altitude on thermogenesis in man. Proc. nutr. 1969 Soc., 28, 74A

Munro, A.F. Basal metabolic rates and physical fitness scores of British and Indian males 1950 in the tropics. J. Physiol. 110, 356

Nicol, B.M. The calorie requirements of Nigerian peasant farmers. Brit. J. nutr., 1959 13, 293

Northcott, C.H. African Labour Efficiency Survey, London, H.M.S.O. 1949

Orr, J.B. & Gilks, J.L. The physique and health of two African tribes. M.R.C. Spec. 1931 Rep. Ser. 155. London, H.M.S.O.

Passmore, R. & Durnin, J.V.G.A. Human energy expenditure. Physiol. Rev., 37, 801 1955

Quenouille, M.H., Boyne, A.W., Fisher, W.B. & Leitch, I. Statistical studies of recorded 1951 energy expenditure of man. Comwth. Bur. Anim. Nutr. Tech. Comm. No. 17

Roberts, D.F. Basal metabolism, race and climate. J.R. Anthrop. Inst., 82, 169 1952

Rubner, M. Die Gesetze des Energieverbrauchs bei der Ernährung. Leipzig und Wien, 1902 Deutiche

Sukhatme, P.V. Incidence of protein deficiency in relation to different diets in India. 1970 Br. J. Nutr., 24, 477

Weiner, J.S. In Human Biology, p. 401 et seq. ed. Harrison, G.A., Weiner, J.S., 1964 Tanner, J.M. & Barnicot, N.A. Oxford, Clarendon Press

Table 1

Table 2

Table 3