Measurements of food intake, by techniques which we judge capable of providing reliable data, have been made on more than 6500 boys and girls of preschool age; i.e. up to about 6 y of age. The exact number of children cannot be given because some of the papers are imprecise on this point.
The great majority of these children were in the USA (about 4000) or the UK (1733). Only 638 were in the developing countries; these were from only 3 studies, one of which is suspect as for as the reliability of the data is concerned.
There is surely something seriously wrong with the orientation of nutritional investigation when during the past 10 years only 3 studies have been published concerned with probably the greatest nutritional problem of the present day, i.e. the extent of malnutrition among the most vulnerable group in the developing world. It should be a high priority to investigate this on a wide scale systematic basis.
Of the data published from the U.S., energy intakes of 3327 boys and girls, aged 3m to 6 y, from a random sample of that population in the North-Central States (27) were oblained by a 3 d recording by the mother under partial supervision. The exact procedure is not described and no body weights are given, but they can be calculated from the energy intake data.
A group of 448, 6 m old children (319 Caucasian, 85 Negro and 28 Oriental) were monitored until they reached 6 y of age in Berkeley, California (13). Pooled data for the 2 sexes are given. No difference was present between the intakes of the Caucasian and Negro children at any age, but the Oriental children were consistently slightly lower, although they also had lower body weights.
A semi-longitudinal study was done in Boston (50) on about 80 boys and 80 girls of 3–5 y (the exact longitudinal component of the study is unclear). The technique was a 7 d weighed inventory method but the mother seems to have been unsupervised during the actual survey.
In the UK, a random sample (from the birth register)of 304 infants (aged 3–24 months) in Glasgow, stratified to cover approximately equally the different social groups, was measured by a 5 d weighed inventory method (40).
An approximately random group of 1248 children from 12 to 47 months of age and living in various parts of the UK was studied using a 7 d weighed inventory method (16). There was a significantly greater intake by the boys even at 12 m but unfortunately no information on body weights was given so it is unsure whether this difference simply reflects a larger body mass.
A semi-longitudinal study in Newcastle in the UK was done on 140 children age 8 m to 5 y (4). No data on weights is presented and only pooled results for the sexes are given although the authors state that there was a significant difference between the sexes after 3 y of age. Other unusual findings in this study (e.g. children from the socially poorer families, where the family size was larger, had higher intakes thar those from more privileged and smaller families) have explanations provided by the authors with insufficient evidence to support them. Two surveys, one from France (9) on 108, 5 y old children, and one from Germany (57) on 2–5 y old children, used the “precise” weighing method for periods varying from 7 d to 100 d. Some controversial comments are made in the German study on the necessary duration of measurement of food intake - the authors suggest about 3 weeks, which would effectively reduce to almost zero the amount of available nutritional data in this field.
In the developing countries, 201 boys and girls from 1 to 7 y and living in a coastal and in a highland village in New Guinea (24) had food intakes measured by the individual weighed inventory method during 5–6 d.
In Costa Rica, 151, 1–5 y old children living in a rural environment, had food intakes assessed during one day only using an inexactly described combination of recall and weighing (59). No data on height and weight are given, although many of the children were apparently malnourished. Dietary intakes are highest for the “under ly olds” and then remain almost the same at age 1 y, 2 y, 3 y, 4–5 y.
The combination of the unusual findings, the inadequate information and the questionable methodology makes it difficult to place much reliance on the data.
286 Iranian urban children aged 1–5 y from poor families (29) had food intake weighed during 3 d by “nutritionists” who stayed all day with the families.
A rather unusual survey was made on children in a Thai village (13 bis). The food eaten by each individual child during 3 days was weighed but these results are not reported. Instead, the data was apparently used to construct a typical diet for each age group which was then analysed chemically and by bomb calorimetry.
The results obtained indicate an enormous deficit of intake from requirement and this, combined with the authors' admission of large difficulties in carrying out the study, make it unjustifiable to utilize the data.
The total number of children aged from about 6 y up to about 18 y whose energy intake has been measured and the data have been published since 1970, totals 2931. Of this number, 300 were measured in Boston, USA (50), 2230 in the UK (16, 12, 18, 49), and the remainder comprised small groups in Germany (57), Australia (41), Singapore (2) and India (17).
Only one study, on 281 children and adolescents in New Guinea (24), could be regarded as typical of the situation for the poorer people in a developing country.
It is both remarkable and deplorable that such important population groups as children and adolescents in the developing world should have received so little scientific nutritional attention, and it is to be hoped that some strong positive encouragement will be given to rectify the situation.
In the USA, about 300 children, aged from 6 up to 17 y, were surveyed in the previously mentioned semi-longitudinal study in Boston (50).
In the UK, 321 children more or less randomly selected at age 10–11 y and living in 3 large towns were measured (16).
Another study using the same technique as the above (7 d weighed inventory) was carried out on 312 children aged 8–11 y in Kent (12).
On an older group, 13–15 y, 611 boys and girls living in Glasgow and stratified according to social group were studied using again the 7 d weighed inventory method (18). This investigation was done in two parts, contrasting similar populations of adolescents measured 1) in 1964 and 2) in 1971. There were some apparent trends showing a reduced energy intake by both the boys and the girls in the later study although body weights and heights were similar and the fat mass of the boys in the 1971 survey was actually greater than previously.
In Birmingham, the same method was used on 791, 14 y old boys and girls (16). Again, the methodology was similar in a study in Newcastle on 177 boys and girls of 14 y, half of them from one-parent and half from two-parent families (16); no significant differences were found in the respective intakes of the contrasting groups.
The other studies on Caucasian populations were on 43 German boys and girl aged 6–15 y (57) and 7 Australians aged 16–20 y (41). The Australian investigation was one of a small number where energy expenditure was measured as well as food intake. The implications of having both sets of data available on the same subjects will be dealt with in a separate section giving an overall picture of all of these different studies.
The only other reports on energy intake on school children and adolescents were 1) on 18 Indian boys of 14–17 y living in Coventry in the UK (49) where two groups, one with biochemical signs of rickets and the other a comparable “control” group, were found to have no significant differences in their energy intakes 2) on 16 Indian boys age 12–14 y living in a boys hostel in Singapore where energy expenditure was also measured (2); 3) on 24 Indian girls, some athletes and others not but apparently both groups being relatively socially privileged (17), and 4) on 99 boys and girls, aged 13–18 y, living in a coastal and in a highland village in New Guinea (24).
The total mass of information on this age group is even more restricted in scope than is available on the preschool children, and is very heavily biassed towards only 2 developed countries. Obviously, only a strictly limited interpretation can be made of these results.
Only 3 reports are available on a total of 73 men and they are on such contrasting populations that their implications are problematical. One of these three was on 38 men mostly clerks working in the offices of a shipyard in N. Italy (47). The second was on two small groups of men, aged 20–39 and 40–59, living in a village near Cambridge in England. Their occupations are not described but their intakes are relatively so low that presumably they must have been mostly in the sedentary group (3). The final study was on 10 clerks in Manila (32).
All of these investigations were done using a 7 d weighed inventory method and, as well, the Italian and Filipino men had their energy expenditure measured.
There is a much greater variety of published data on this category, 13 studies on 375 men of whom 260 were farmers. However only 3 developing countries are represented, the Philippines, Upper Volta, and New Guinea.
The occupations vary a little. Thirtyseven shipyard workers in N. Italy (47) were the sole representatives of populations studied in the developed countries, other than 5 groups of farmers. The farmers consisted of 3 groups in Italy, two of them (25) being each of 20 middle-aged men and living in two different parts of Central Italy who were investigated at 5-yearly intervals on two occasions as part of a project related to diet and coronary heart disease; the other Italian study was on 24 middle-aged farmers working in a moderately affluent region in the centre of the country (23), who constituted one of 3 groups with contrasting environments, the others being in Finland (23) and the last in the eastern U.S. (23). The individuals in these 5 groups of farmers were chosen as being reasonably typical of their particular population.
Three very different types of farming were represented in the studies from the developing countries. Nine farmers (33) and 32 sugar cane workers in the Philippines (31), 11 in Upper Volta who were studied during a quiet part of the working year (at the end of the harvest) (8), and 94 subsistence farmers in a coastal and in a highland village in New Guinea (48), make up the total.
Four other occupations were investigated, all in the Philippines. Ten shoemakers (34), 10 “Jeepney” drivers (a “Jeepney” is a type of small bus peculiar to the Philippines) (35), 25 textile-mill workers (36), and 33 fishermen (31) all had their food intake and their energy expenditure measured during a 7 day period. Only in the textile-mill workers was the energy intake measured in a different fashion - all of the other studies on men in the moderately active occupations employed a weighed inventory technique - and this might account for some part of the unusual findings in this particular group.
Many of the investigation in this category of occupation involved energy expenditure measurements as well - all of the studies in the Philippines, that in Upper Volta, in New Guinea, and on the Italian shipyard workers.
Only 3 sets of published data have been usable in this category of occupation covering a total of 152 men, 1 of the groups being in developing countries, the second being in Italy and the third in the UK. Eighteen agricultural workers were studied in Guatemala (60); their food intake was measured by record and weighing for only 1 day. In Italy, 75 men engaged in heavy work in a shipyard (47), and in the UK 59 army recruits (19), constitute the remaining men. All of these 3 investigations assessed both food intake and energy expenditure.
It is obvious that both the number of separate investigations and the total number of men studied in the three divisions of “heaviness” of occupations make it difficult to extrapolate the data in any comprehensive fashion to the assessment of actual intakes relative to requirements. The whole topic of studying energy intakes in a wide-ranging way, so that some general picture can be formulated about the quantities of energy ingested in the food eaten by free-living populations, is one which has received and is receving less attention than is necessary.
Four studies have been reported on elderly men, 2 from the UK, one from Italy, and one from Nigeria. Some of the data is unusual. One of the UK investigations was on a very small number of men aged 60–79 living in a Cambridge-shire village (3); the other measured 158 men aged 62–74 and 54 men aged 75–90 living in Edinburgh (38). The technique of the Edinburgh survey - a diet history plus some 2 d weighed intakes - leaves considerable possibility for error. The Italian men were the farmers mentioned in the section on “moderately active occupations” (25) who were aged 50–69 at the time of the last measurement in the longitudinal study. Small numbers of men were studied in Nigeria, living in urban and rural areas; 10 men were aged 60–69, 9 men aged 70–79 and 4 men aged 80–89 (45). The method was a 2 d weighed intake by “dietitians” and, although the intakes are relatively enormous, the authors write as if their elderly men were malnourished: in the circumstances, assessment of such data is perplexing.
In many ways evaluating the energy requirements of elderly people presents more problems than other age groups: the impact of ill-health and disability and the huge range of variability of the ageing process make it impossible to consider any age span of elderly people as a homogeneous group. However this makes it even more relevant to study varied populations of the elderly, although such studies are of limited value unless adequate description of the subjects is provided.
The information on energy intake of non-pregnant non-lactating women is pathetically scanty.
Studies have been reported on very small groups in the UK, in the Philippines, and in India. Part of the previously mentioned investigation in a Cambridge village (3) involved measuring food intake on 9 women aged 20–39y and 14 women aged 40–59. In the Philippines 10 housewives (34) and 10 typists (32) each had a 7 d weighed measurement of their food intake.
Women in this type of occupation would usually be in one of two categories: they would have one of the relatively small number of jobs in the developed countries which necessitate moderately hard physical labour, or else if they lived in a developing country they would belong to the large mass of the population whose way of life requires frequent physical activity.
In this present review, information is provided on a total of 46 women from 4 countries and 94 women from New Guinea. Six non-lactating women had their diet estimated in Guatemala (54) by a 24 h recall technique carried out on 4 separate days - a method we have usually dismissed as inadequate but which has been accepted here because of the 4 repeated assessments and because data on this sort of group is so scarce. Six young Indian students (17), who were also athletes, made up a second group, but are a considerable contrast to the first group and also to the 14 farmers in Upper Volta (8), measured at the end of the harvest season at a quiet time of the working year, and to the 14 textile-mill workers in the Philippines (36). The energy intake of these Filipino textile-mill workers was very low and it is uncertain how much this result is due to problems which the investigators encountered in the measurement of food intake, which was largely unsupervised. Some of the statistical treatment of the data is also somewhat worrying. Apart from these surveys, the final study of ‘moderately active’ women who were all from developing countries, was on 94 women living in 2 subsistence-farming communities in New Guinea (48).
The total numbers of women included in both the light activity and the moderately active occupational categories comes to 183. Some of them, e.g. all the groups in the Philippines, that in Guatemala, in Upper Volta, and those in New Guinea, also had energy expenditure measured simultaneously. Whilst this considerably enhances the amount of information which can be deduced from the reports, there is still such a very small amount of data available that general implications related to energy requirements can be made only in a very tentative fashion.
Four studies on very dissimilar groups of pregnant women have been reported. A longitudinal study on 12 or 13 or 14 ‘mature’ women living in Berkeley, California (6) measured food intake at 20–28 weeks, again at 29–36 weeks, and at 37–40 weeks. In Leeds in the UK 179 women of unspecified body weight (either pregnant or pre-pregnant) in their first trimester had food intake carefully measured by a supervised 7 d weighing technique (55). In Ethiopia, some very peculiar data were obtained by a 2 d weighed inventory on 20 ‘non-privileged’ and 10 ‘privileged’ young women in their 3rd trimester (28). The two groups of women had similar body weights but the energy intake of the privileged women was twice as high (12.4 MJ/day opposed to 6.4 MJ) as the non-privileged. The implications related to social and physiological circumstances of this finding are so unlikely that it is strange that the authors have not discussed then in some depth. Finally, some low and unusual intakes were reported on 16 women, in the first and second trimester, living in a coastal and in a highland village in New Guinea (48). Intakes were apparently very similar between the pregnant and the non-pregnant women in these village.
One study from California and one from Australia, together with 3 from developing countries, makes up the total reported literature. The Californian study was part of the longitudinal investigation on pregnancy (6) and 12 of the women were re-assessed when they had been lactating for 8–12 weeks. In Perth Australia, 27 women lactating for between 1 to 15 months had energy intakes assessed over 7 d by ‘measured portions’ (51). Of the 3 studies from developing countries, one in Nigeria (1) measured by 24 h recall for 3 days food intakes of 75 rural and 25 urban women who had been lactating for periods of from 3 weeks to 6 months. Their intakes are the highest for any of the groups of lactating women. In Guatemala, 18 women living in a village and lactating for 6 months at least, had intakes assessed as previously described (54). Finally, 32 women living in a coastal village and 20 women in a highland village in New Guinea had food intake and energy expenditures measured (48). Some of these women were lactating for up to 2 y.
It must again be emphasised that there exists a severe paucity of data on groups who are rightly regarded as being among the most important from the nutritional point of view, that is pregnant and lactating women, especially in the developing countries. It might be forgivable for studies to have been done in only 4 developing countries in the past 10 years if a considerable volume of data already existed. It does not exist. And moreover the theoretical energy requirements of pregnancy and lactation usually conflict radically with actual energy intakes, so there exists a biological dilemma as well as a scientific partial vacuum. This was mentioned in part in the previous FAO/WHO report (1973), with apparently no effect.
We consider it of the highest priority that as much influence as possible should be exerted on grant-giving institutions to support this type of research and, more importantly, that it should be emphasized that the problems to be investigated are not only of the highest practical significance but also involve the solution of interesting physiological and biochemical questions, and that therefore they are worthy of the interest of good nutritional scientists.
As parts of the studies already mentioned in the section on elderly men, 8 women aged 60–69 in a Cambridge village in the UK (3), 190 women aged 62–74 and 73 aged 75–90 in Edinburgh (38), and 19 women aged 60–79 and 12 aged 80+ in Nigeria (45) have had food intakes measured by a variety of techniques. The same comments apply as to those described for the elderly men, especially in relation to the Nigeria survey of the 60–79 y women.
Even less material is available in the literature on energy expenditure. The number of studies is small, and the subjects measured in each study are almost invariably very few.
Screening out the papers published in the last ten years, and retaining only those where the methodological section provides sufficient information to allow critical examination of the results, the total number of children below the age of 18, on whom energy expenditure measures were attempted, was 97. Of these, 36 were from developing countries (U ganda and Singapore); the others were British, Italian, Canadian and Australian.
The assessment of energy expenditure in children raises a number of intractable problems. Spontaneous activity of small unrestricted healthy children is often, at least superficially, undirected and sudden changes of tempo, purpose and nature of their activity are quite common. The difficulties encountered in obtaining a usable, sufficiently precise time allocation record is inversely related to the mobility and inventiveness of the child. Actual measurement of the energy cost of activities is also difficult. Available instrumentation has been designed for adults and is unsuitable for a satisfactory adaptation to a small child. Thus it is not surprising that very few studies on this age group are to be found in the literature. Moreover, several of these studies have been conducted with techniques that appear either insufficiently described or of inadequate reliability. So, an even smaller number was eventually retained.
Twenty African and 5 European healthy free-living children, aged between 1.5 and 3.5 years, were studied in Uganda by a simplified activity diary in 10 min. blocks (53). The energy cost of the activities was estimated from published values for older children. The children were observed for two periods of 5 h on different days. No data are given on the body weight and sex distribution; energy expenditure data are reported as KJ per Kg of body weight. African children had a markedly lower energy expenditure than Europeans. Because of the small number of subjects and the doubtful technique we feel that the results should not be given much credence.
Between the ages of 4 and 5 years, only 20 children have been studied. They were a specially selected group, composed of 12 children, with normal-weight parents compared with 8 other children of similar body weight and age whose parents were or had been more than 20 % overweight (30). The study was conducted in the UK. Energy expenditure was assessed by means of heart rates recorded over 4 to 7 d, complemented by individual regression lines constructed from heart rate and O2 uptake. Oxygen uptake was measured by continuous flow diapherometer. Two or more points were used to draw the lines. The children with overweight parents, although being themselves of normal body size, had a significantly lower energy expenditure than the control group, if this notoriously unreliable technique is to be accepted. As resting energy expenditure was reported to be significantly lower in this group, the authors suggest that these children may have had some metabolic peculiarity.
We find the same scarcity of data for the age group spanning 6 to 18 years. At this age range, energy expenditure assessment becomes simpler and there is a progressively increasing ability of the subjects to cooperate. The almost complete absence of direct measurement of energy expenditure in this age group is thus surprising.
Two small groups of 9 year old children were studied in Canada and in Italy. The Canadian children, a small self-selected group of volunteers from public schools, 11 boys and 10 girls, were measured by heart rate (HR) counting during two consecutive days (56). Their energy expenditure was derived, except for sleeping, from the HR/VO2 relationship, individually constructed. Talbot BMR values (58) were used for energy expenditure during sleep. Besides the criticism relevant to the use of heart rate for the evaluation of the energy expenditure, there are some doubts about the unusual procedure used to calibrate the O2 analyzer (by ‘burning alcohol’) and not enough details are provided about the instrumentation used (‘open air circuit’).
The Italian children, 12 nine-year old boys from a rural underprivileged area in southern Italy, were studied by a 7 d dairy of activity, recorded on alternate days by an experienced unobtrusive observer. Energy costs of activities were measured by the modified Max-Planck respirometer (37, 44), hereafter referred to as KM, and portable O2 analyzer. Energy spent during sleep was derived from Fleisch BMR values (26). The children were studied during the school season, and their activity pattern appears to have been strongly influenced by attending daily classes with the associated homework.
Three, 16 y old boys were studied in Australia as part of a small group of 9 subjects on whom both energy expenditure and intake were assessed (41). The evaluation of their daily pattern of activity was done by a 7 d diary, but there was little supervision. The Wolf Integrating Pneumotacograph (IMP) and Schohlander were used to measure the energy cost of the activities. The value for sleeping was derived from the published tables of BMR of Robertson and Reid (52).
The only study on children from developing countries was one conducted in a boy's hostel in Singapore on 16 Dravidian Indian boys, 13 years old (2). Here again the 7 d dairy method was utilized. The subjects apparently kept their own diaries, and the only supervision was that of the hostel warden. KM and Lloyd Haldane for O2 analysis were utilized to measure the energy cost of 10 standard activities. Resting metabolic rates (RMR) were used for computing energy expenditure during sleep. In spite of the overestimation that this procedure may have introduced, the reported daily energy expenditures appear to range on the low side. Some other aspects of the experimental technique (such as measuring very low energy cost activities by KM and some of the results of the anthropometry - an increase of 2.2 kg in body weight and 0.7 cm in height in a two-month period) raise certain doubts about the methodology and the interpretation of the results.
The very sparse and often methodologicaly unsatisfactory material in the literature on the energy expenditure in children is clearly insufficient and unrepresentative; this may well reflect the formidable problems in the assessment, by available methods, of the energy expenditure of unrestrained young children.
Adult man is considered the productive unit in any society and, as production is in most cases associated with an energy cost, the study of the energy spent by man in various types of jobs has attracted a moderate amount of interest and effort by nutritionists, physiologists and socioanthropologists. The higher degree of cooperation that can be obtained from a collaborative adult and the more regular pattern of daily cycles and life styles make the adult a much easier subject to study than the child. As a result there is a larger amount of data on adult individuals. It is difficult to define exactly the overall number of individuals who have been measured, as some reports are very enigmatic about the number of subjects actually involved; other studies have been conducted on numbers which differ in the measure of the energy cost of activities and the diary for time allocation. Some other studies, involving a follow-up, do not state clearly what was done and on whom. In two cases, the observations collected on small number of subjects were extended to the whole community.
A tentative enumeration indicates that daily energy expenditure was measured on approximately 650 adult men, in various countries and in a variety of occupational activities.
For review purposes, the same arrangement adopted for energy intake has been retained.
Only two groups of light activity occupation have been studied. These were 38 Italian office employees and sedentary workers of a shipyard (47) and 10 Filipino clerk typists (32). The energy expenditure of both was measured by a 7 d diary of activity together with measurements by indirect calorimetry (KM and paramagnetic O2 analyzers). Dietary energy intakes were measured as well, and have been reviewed in the appropriate section.
A limited number of occupational groups other than farmers were studied both in developing and in developed countries. Thirtysix shipyard workers were studied in Italy (47). Two young men, a trainee and an engineer, were studied in Australia as part of a small group in part reviewed, for reason of their age, in the previous section (41). Other 6 young men, university students, were measured in Glasgow (46); they took part in an overfeeding experiment. Their spontaneous physical activity was unrestrained, and data on energy expenditure are provided both for the control period and for the 42 days of experimental overfeeding. There was no significant difference in the energy output of the two periods.
Data are available for 33 fishermen (31), 25 textile-mill workers (36), 10 shoemakers (34) and 10 ‘jeepney’ drivers (35), all studied in the Philippines. All the above mentioned studies were performed by indirect calorimetry (KM and Douglas bag plus O2 analysis) and 7 d diary of activity. Dietary intakes of most of these subjects had been measured as well and are discussed in the appropriate section. The body size of the subjects varies greatly between the two ethnic groups; for example, the Italians of the Caucasian group had an average body weight of 78 kg, while the Filipinos were approximately 55 kg. The energy expenditure of the Italian shipyard workers is highest in absolute terms, but the lowest when expressed in terms of body weight.
It is obvious that the available data are not only very scanty but also very heterogeneous.
Farmers and peasants, especially in developing countries, appear to have been the most intensively studied occupational group and account for almost half the total sample of adult men. This interest may be explained on the debatable grounds that the energy cost of physical labour is a critical and limiting factor in determining productivity under the conditions of pre-industrial agriculture. This is far from being demonstrated, and there is, on the contrary, some evidence that self-employed farmers may be successfully productive withouth needing to expend unduly high amounts of energy. For reasons of adherence to real life situations, all farmers have been treated in this section, in spite of the possibility that the existence of a seasonal cycle in their work may be responsible for considerable variation in the level of their energy output.
Two rather comparable groups of subsistence horticulturalists, differing mostly in the degree of involvment in hunting-gathering activities, were studied in Peru (42) and in two contrasting areas in New Guinea (48). They represent the most unaccultured group of farmers that have been studied. Their style of life may be considered typical of subsistence agriculture in tropical areas.
Time allocation for the Peruvian Machiguenga was obtained by a new and interesting method; observations performed during brief random visits to the households, repeated throughout the year, were compounded to provide on overall pattern of activity.
The two New Guinean groups were studied by the classical 5 to 7 day diary of activity by a supervised observer. One group was composed of 42 men living on an island, off the Northern coast of New Guinea; the other, of 40 men in a highland area. The two groups were studied for over one year, thus encompassing the complete seasonal cycle. Both energy intake and expenditure were assessed.
Indirect calorimetry was used both in Peru and in New Guinea to assess the energy cost of various activities. The details given about the relevant technical aspects of the measurements performed on the Machiguenga Indians are insufficient. Their energy expenditure appears to be markedly higher than that of the New Guineans, and also very high in absolute terms. This findings is surprising, considering the similarity of the ecological environment and of the life styles of these two groups. However, not much reliance can be attached to the data from the Machiguenga because of the lack of validation of the method used for assessing time allocation; the absence of any description of the protocol adopted for measuring the energy cost of the activities raises further doubts as to the validity of the results.
African farmers in Upper Volta were also investigated by indirect calorimetry and 2 to 6 d activity diaries (11,8). The study was carried out at different seasons, covering periods of intense agricultural activities and of relative inactivity. Eleven 45y old farmers were assessed in the post-harvest season (8), 23 other subjects were studied in the dry season and 16 in the wet season (11). Total energy expenditure was calculated in a slightly different way in one of the studies, where mean group values were calculated for the energy cost of the activities and were applied to the individual time and motion data for computing individual energy expenditure (8). In one study, the energy cost of sleeping was calculated as a 15% reduction of measured RMR (11); while in the other one published BMR values (26) were used (8). The results indicate the existence of a marked seasonal variation in energy expenditure which underlines the necessity of taking into account seasonality or other yearly cycles when trying to characterize the pattern of energy flow in a community.
A similar seasonality in energy expenditure was also observed in Iranian agricultural workers (10). Four groups, totally 46 men, of paid labourers and low income farmers were studied during 4 seasons by 1 d diary of activity recorded by the team investigators, and by indirect calorimetry. The use of KM for the measure of the BMR, which was equated, after a 10% reduction, to the cost for sleeping, introduces a technical flaw in this otherwise satisfactory methodology. Energy expenditure peaked in summer and declined to a minimum in winter, parallel with the intensity of agricultural work.
Yemenite and Kurdish Jewish farmers were studied in the Negev area of Israel in two different seasons (20). The exact number of subjects is not clear, and it fluctuates according the season and the nature of the observation. A total of approximately 83 subjects seem to have participated. Methodological description is insufficient concerning measure of the energy cost of activities. Only a few measurements (n.50) were performed by indirect calorimetry (KM or Wright flowmeter, and Lloyd Haldane), and heavy recourse to published and unpublished values is aknowledged. One-day time and motion studies were supplemented by 2–6 interview sessions to obtain details of the activity pattern in an attempt to reconstruct the weekly cycle. The two groups had a very similar energy output, and neither exhibited a seasonal cycle. The employed farm workers appear to have been somewhat more active than the self-employed farmers. It is regrettable that such essential information as body weight or age are not given for these subjects.
Two other groups of agricultural workers were studied in the Filipines and another one in Guatemala. The Filipino subjects were 9 rice farmers, aged 17 to 60 years, studied at the peak of the season (33), and 32 sugar cane workers (31). No age nor body weight are given for the latter, but the authors state that a selection for age has been performed, all subjects being over 20 and less than 50 years old. Body weight could be derived by calculation from the tables and appears to be about 54 kg. Although not clearly stated, it seems that both groups were employed agricultural labourers. The Guatemala peasants also were agricultural labourers, working on a milk farm (60). Their age ranged between 17 and 57 and their average body weight was 60 kg. All these three groups of subjects were studied by an activity diary and indirect calorimetry. The lenght of observation for the time allocation study was 7 days for the two Filipino groups and 3 days for the Guatemalan farmers. The energy cost of activities was performed by calibrated KM and O2 analysis. Fleisch values for BMR (26) were used for energy expenditure during sleep for the Filipino rice farmers. Published values of BMR were also used for the Guatemalan subjects, but the source is not quoted. The Guatemalan farmers were the heaviest and had the highest energy expenditure, both in absolute terms or as a function of body weight.
With the exception of the farmers during the high peak of the season, which were reviewed in the previous section, only one group is available in the literature on subjects in the high activity category. It is composed of 75 Italian shipyard workers (47). Their energy expenditure was assessed by the method previously described.
A special occupational category which seems to have received much greater attention is the Army. Three separate studies on the energy expenditure of soldiers are available. All of these appear to have focussed on the intense period of training that soldiers periodically undergo; thus the conclusions that can be drawn about the average cost of being in the Armed Services may be misleadingly high.
Indian soldiers of various ethnic and class extraction were studied during their training in the field (39). The soldiers belonged to infantry, artillery and engineer units. They kept their own activity diary, apparently during 3 cycles of 15 days each over 3 months. No supervision is mentioned. Published and unpublished values of energy costs for all the activities were used. No information is given about the exact number of subjects involved in the study, and the whole of the methodological presentation is very confusing. Sixty-four British troops were also subjects of energy expenditure studies. 35 infantry recruits were studied during their initial training period (19), while another 29 were soldiers exercising in Malaya on a more or less restricted energy intake (14). Their time allocation was studied by activity diary during 24 consecutive days in the Malaya subjects, and for an undefined number of days for the army recruits in the UK. Energy expenditure was measured by IMP worn for most part of the day in the latter, and by KM in the Malaya soldiers. The heaviness of their activity is confirmed by the high values of energy expenditure reported, which appear to be the highest recorded in the available literature.
No studies have been reported on elderly men.
Several of the studies carried out on men covered women as well. In all these cases the methodology for the assessment of energy expenditure has been the same. The total number of adult non-pregnant, non-lactating women studied is approximately 250 with a seasonal repetition on about 50 of the subjects. It is not clear whether these subjects were completely different ones, or were the same ones followed through different seasons.
There are only two groups of women who fall in this category and both have been studied in the Philippines.
The first group is composed of the 10 wives of the Filipino shoemakers, reviewed above (34); the second Filipino group is the female counterpart of the group of clerk typists also reviewed above (32). Both had their energy intakes also measured. Their energy expenditure has been measured by the technique already described for the males. Perusing the results of the time allocation study, it appears that the wives of the shoemakers engaged in some shoemaking activities also, probably in support of their husbands' jobs. Their energy expenditure is higher than that of the typists, but not when expressed on a body weight basis.
The total number of women studied in the light activity category to which most women from the developed countries, and an indefinite proportion of women from the developing countries, are supposed to belong is surprisingly small. This may reflect the small amount of interest given, even in the accademic world, to the work (domestic or otherwise) of women which is associated with the low economic value attributed to it.
The total number of women in this category amounts to approximately 210. Most of these were farmers and peasants. Although women participate in increasing proportions in the production process as employed labour, only one study is to be found that deals with occupational jobs other than farming.
Considering that it is likely that the energy cost of the occupational activity of a women is superimposed on the costs of the housework and that of the reproductive processes, the fact that so little attention has been dedicated to the energy expenditure of women engaged in physically exacting occupations is astonishing and regrettable. one wonders wether the explanation may not be found in neglect, lack of prestige and low economic value attributed in most male-oriented societies to women's labour*.
Four, 19 y old, students and trainess were studied in Australia, by the same method reported above (41). Fourteen textile-mill workers were studied in the Philippines (36). Their energy expenditure was assessed as described for the males textile-mill workers.
* This is the highly biassed view of one of the authors.
As for the farming women, two almost purely subsistence agriculture communities were studied in Peru and in New Guinea. The Peruvian group was composed of Machiguenga Indians, living in the tropical forest of the upper Amazonian tributaries (42). The New Guinean women were studied in a coastal area and in the highlands (48). All seemed to have led a rather similar type of life, engaging in garden work, handicrafts, a certain amount of food-gathering, food preparation and child care.
In the Machiguenga study 20 women, 16–41 y old, had their time-allocation assessed by the already described random visit method (42). Eight women in the same age range, and with an average body weight of 44 kg were measured to assess the energy cost of the recorded activities. The sample of New Guinea women consisted of 23 women, 18 to 29 years old and 17 women, aged 30 to 48, in the coastal group. The highland group consisted of 31 women, aged 18–29, and 7 aged 30–49 years. Body weights for all four groups were similar, ranging between 44 and 51 kg. Also the energy expenditure of the Machiguenga and New Guinean women appears to have been rather similar, as might be expected considering the similarity of the environment. However, as already remarked for the male Machiguenga subjects, the methodology of time allocation used in the Peruvian study needs extensive validation before the results can be fully accepted.
A rather different type of farming, with some cash cropping and marketing activities, was characteristic of the Upper Volta women farmers(7,8). A group of 14 women, aged 14 to 48, was studied in the post harvest season, when agricultural activities are at its lowest. The methodology employed was the same as for the Upper Volta male farmers, reviewed above (8). Another 12 women farmers living in the same area had their energy expenditure assessed during the rainy and the wet season by 2 d diary of activity and by indirected calorimetry (7). Their sleeping energy metabolism was calculated to be 85% of their measured R MR.
There are two other groups of rural women, one in Guatemala, the other in Israel. Since for neither of them is given any indication about their time-allocation, it is impossible to say how similar was their general style of life to that of the groups of female farmers described above.
The Guatemalan group was composed of 6 Ladino paesants who were a control group in a study on energy balance during lactation (54). Their mean age and body weight were comparable to that recorded in the other groups of rural women. The energy expenditure was calculated on the basis of regression lines, individually established between heart rate and O2 uptake. Heart rate was measured during 2 days. This method, as already mentioned, has a limited value because of the incostancy of the relationship under different physiological conditions. The data of this report have been nevertheless retained in this review because of the scarcity of reports on energy expenditure in women.
The last group of peasant women were the wives of the Yemenite and Kurdish Jews reviewed in the ‘male’ section of this paper (20). The women were studied in summer (21 Yemenites and 23 Kurdish) and in winter (24 Yemenites and 28 Kurdish). Their ages and body weights are not given. The methods of assessment of energy expenditure were the same as employed for the men. No seasonal difference in energy output were present.
Pregnancy is an energetically costly condition. The last FAO/WHO Committee evaluated the total cost of pregnancy as about 335 MJ. This quantity also includes the build up of energy stores that meets part of the cost of breast feeding. The very few studies carried out on freeliving pregnant women provide conflicting results and do not allow any firm conclusions to be drawn.
Two groups of women in different stages of pregnancy were measured in California(5, 6). Twelve of them were black adolescents 4 to 9 months pregnant on whom time-allocation was studied by a 1 d diary of activity (5). The energy cost of the recorded activities was measured on 15 other black (with one exception) girls of the same age, 8 months pregnant, confined in a metabolic ward.
A second group of pregnant Californian women was monitored throughout pregnancy and in lactation (6). The original number of partecipating women was 13, their age 26 and their body weight 64 kg. They were 20–28 weeks pregnant. Their energy expenditure, was studied by a 3 d diary of activity repeated twice more during pregnancy at spaced intervals (29–36 weeks and 37–40 weeks). The energy cost of the recorded activities was in part derived from published and unpublished values, and in part actually measured (Douglas bags and O2 analysis). There was a regular increase in energy expenditure as pregnancy progressed reaching a peak immediately before delivery. Energy intakes had also been measured but do not coincide with the energy expenditures.
Other 14 pregnant women were studied in a coastal and a highland area in New Guinea (48). The were part of the study already reviewed under various sections of this paper. The stage of pregnancy is not specified, but most were in their first and second trimester. Their energy expenditure was assessed as described for the rest of the group. The results do not show evidence of any modification in their energy expenditure as compared to non-pregnant women of the same group. The higher energy expenditure of the highland non-pregnant women as compared to the coastal women, is maintained during pregnancy.
Lactation represents an energeticaly very costly process, and again it is therefore surprising and regrettable that only 2 studies of energy expenditure during breast feeding are to be found in the recent literature.
Eighteen rural Guatemalan women were studied while breastfeeding at the 10 th month post partum (54). Their energy expenditure was assessed on the basis of 2 d heart rate count. Other 12 women of mixed ethnic and occupational extraction were studied in California (6). They had been lactating for 10 weeks and it seems that they had been longitudinally followed through pregnancy. Their body weight, 62 kg, was comparable to that of the 6th month of pregnancy, and much higher than that of the above mentioned Guatemalan women, who weighed 49 kg. The energy expenditure of the California women was assessed as described above(6). The Guatemalan women had a markedly higher energy output. possibly indicating that they had gone fully back to their usual farming activities.
The energy expenditure of elderly women has not been studied.