The results are set out in the appropriate tables and in some scatter diagrams. Even though the range of studies on preschool and school children is not wide yet there is a volume of data which allows some general trends to appear and some hypotheses may be formulated. In the case of the male and female adults, there is such a very limited amount of information that very little can be said with any confidence and only the most tentative suggestions are possible.
All the data on energy intakes of children below 18 years of age are graphically presented in Figs. 1–3. The theoretical requirements (21) have also been plotted to serve as a reference. The best fitting curvilinear regression line was obtained by the log transformation of intake and of age.
This presentation is biassed by the fact that each point represents the finding for that age group of any single study, independently of the number and representativeness of the sample.
The first immediate impression emerging from the observation of the plot and of the regression line is that of a gap increasing steadily with age, between the two sets of data. While most of the data on children below 3–4 years of age coincide reasonably with the values supposed to be their energy requirements, those of the older children progressively diverge and are almost always below the reference values. The scatter also is increasingly wider and the slope much less than that of theoretical requirements.
Most of the points in the younger age groups, up to 4, represent data collected on children from developed countries (Fig. 2). They all fall within about ±0.5 MJ of the theoretical requirement by age, and most of them are in even closer agreement. The data for the children of the same age group, from developing countries (Fig. 3), are very scanty (only 7 points), their difference from the requirements exceeds 1 MJ, and the number of investigated children is lamentably low, 391. Of all these children, only three groups seem to have met their theoretical energy requirements. These were all New Guineans.
Above 4 years of age, none of the children from the developing countries have energy intakes that even vaguely approach the theoretical values, and the discrepancy grows with age.
The same observation holds true for the developed countries, although some 7 or 8 groups of children appear to have had intakes that compare satisfactorily with the reference values.
The interpretation of these observations is not easy. There is also the possibility that the smaller body size of the children from the developing countries may have been responsible for their smaller energy intakes. Expressing the same results on a body weight basis disposes of this hypothesis, as the picture (see Fig. 4) remains virtually unchanged. And, at least at the younger age groups, the intakes per kg body weight appear to be very similar in the two sets of data.
An interesting, but isolated, clue for a possible interpretation may be found in the results of the study on the Glasgow children, which reports the findings on energy intakes of two comparable groups at an interval of 8 years (18). The 1964 Glasgow children had intakes of 11.7 MJ, while in 1971 their counterparts of similar body weight and even slightly fatter, had decreased their energy intakes by approximately 0.8 MJ, thus falling about 1.2 MJ short of the theoretical requirement. The meaning of these results might be that the physical activity and thus the energy requirements of these children had declined with time with increasingly sedentary life styles. There is little hard evidence of this to be found in the literature, although plenty anedoctical support is available.
Unluckily this same explanation reveals its limitations when applied to the developing countries. It can be safely assumed that, while spontaneous activity of children from developed countries is rather rigorously organized for at least 5h/d per 5–6 days/week per 9 months/y because of compulsory sitting in a class room, the same does not apply to the child from developing countries. Their school attendance is probably less regular, although it might require covering quite long distances on foot to reach school. Thus spontaneous physical activity is probably less restricted. Besides, the impression gained while visiting developing countries, or even poorer sectors of developed ones, is that very young children may engage on a rather regular schedule in sometimes physically very hard work. This would imply that their energy requirements are comparatively higher than those of their sedentary counterparts in the developed countries. The fact that their observed energy intakes are in fact lower can only indicate either that this impression is not true, or that a very peculiar adaptation to low energy intakes must have taken place. This apparent contradiction needs urgent and qualified attention.
Energy expenditure studies on children, both in the developing and in the developed countries, are extremely rare. Their time allocation has been even less investigated. Of the 6 existing studies, only one has dealt with children from developing countries, Indian boys residing in a Singapore college, and attending regular classes, although this fact is not explicitely described by the authors. For all the energy expenditure studies, the findings indicate levels of energy expenditure which fall well below the theoretical requirements, even when expressed on a weight basis, and close enough to the mean observed values for energy intakes.
While one might postulate that both energy intakes and expenditure may be restricted by externally imposed limitations in the developing countries, this surely does not apply to the developed countries.
There is clearly insufficient material to draw any firm conclusion from these figures, a part from stressing the need for more and better research in this line. The current energy requirements do not seem to fit the existing situation.
Approximately 28 groups of adults males ware studied either for their energy intakes or for their energy expenditure or both. The data, divided in developed and developing countries are summarized in table 3. The same data, expressed as percent of the requirements adjusted for age, body weight and physical activity are presented in Fig. 5. The groups from the developed countries consisted of students, clerks, manual workers and army recruits. Although that number of activities cannot even vaguely cover the variety of possible existing occupational conditions, the three categories of work intensity are represented.
For the developing countries, all but 5 groups were engaged in some form of farming. While this may seem appropriate since agricultural and food production activities engage the majority of the population of developing countries, the diversity of environment and of methods is far from being depicted. The seasonal cycles of the farming activities appears also insufficiently explored.
The overall picture does not suggest any well-defined pattern. Energy intakes and expenditure are scattered both above and below the reference. A slight prevalence of values lower than the requirements can just be seen for the developed countries. However on the average, the values were only 6% below the requirement. For the developing countries the scatter of the results above and below requirement is even wider and thus the fact that the grand mean of all the values, either for intake or expenditure, is almost exactly equal to the requirement (101%) can have only a very limited meaning.
Energy expenditure in the developed countries is either very similar or slightly lower than energy intakes. The picture is rather different for the developing countries, where the energy expenditure values are greater than requirements (+ 5%) as well as than the intakes (+ 10%). Out of the total of 18 groups where both intakes and expenditure had been measured, 4 had a huge discrepancy between the two sets of measurements, 8 were in excellent agreement, and the remaining 6 had a somewhat intermediate position. All the cases of great discrepancy between intake and expenditure occured in the developing countries and were all in one direction, i.e. expenditures were greater than intakes. The meaning of this phenomenon is obscure although, apart from methodological explanations, it could be attributed to the possibility that there is no physiological need for instantaneous or short-range energy equilibrium. The practical implication however would be that the measurement of energy intakes or of energy expenditures need to be carried out in such a way as to become more representative of the long-term situation.
Energy requirements may be quite justifiedly considered as a fluctuating variable, which has never been satisfactory investigated under free-living conditions for a sufficient length of time, and is thus virtually unknown.
What can be made of these results? They appear to be on average in reasonably good agreement with the present requirements and it would appear justified to maintain them. However the scatter of the data appears to be so wide as to weaken any inference.
Energy intake in old age does not appear to have attracted much interest. Apart from a rather large group of 212 men over 62 years of age, living in Edimburgh, Uk, only 3 other groups have been investigated. One consisted of Italian farmers, whose dietary intakes were measured as part of a longitudinal (or follow-up) study on dietary fats and CHD; other serve 60–79 years old men from a small village near Cambridge were studied in UK. The last group was composed of 23 Nigerian men, 60 to 90 years, both urbans and rurals; they were described by the authors as being malnourished, which is in overt conflict with the reported results on energy intakes which are enormously in excess of the actual requirements (147% – 192%). The results of the other group were all reasonably close to the age-adjusted requirement, and more or less equally distributed around it. On none of these was energy expenditure measured, nor was it measured on any other elderly subject. This material is not only totally unrepresentative of old age situations, but is insufficient as well for drawing any sort of conclusions.
Figs. 6 – 8 are scatter diagrams of energy intake and age of 5995 girls between the ages of 3 months and 18 years. This number is made up of 5504 females living in developed countries and 491 in developing countries. Almost the whole of the former comes from data collected in the USA and in the UK. Of the girls living in developing countries, apart from 24 in India, the remainder are from New Guinea, or from either Costa Rica or Iran where neither study differentiate the sexes. It is obvious, therefore, on what tenous grounds any deductions can be made. This is especially true of the children over 5 y in the developing countries (Fig. 8) where all of the data is from New Guinea, apart from two small groups, each of 6 girls, at age 14 y and another two groups of 6 girls at 16.5 y from India. This figure shows also almost complete absence of data on children under 1.5 y, presumably because most children of that age in a developing country are still being breast-fed.
Therefore, most of the conclusions from the data on children and adolescents must be derived from the figure 6 using all the data from whatever source, or else from that dealing only with developed countries. In either case the conclusions will not differ other than in a minor way and the extrapolation to the situation in developing countries can be made without undue concern - any error will be on the safe side.
Fig. 6 shows a picture of actual intakes of energy which are generally appreciably less than the apparent requirement. The situation is less clear in the age range 3 m up to 3–4 y, where there is a considerable scatter of the data, and many of the intakes are close to or slightly above the requirements. Above the age of 3–4 y, all of the measured intakes were less than the ‘requirements’ and only the exceptional result came even near those levels. The slopes of the regression lines of actual intakes and of requirements (Fig.6) demonstrate interesting differences after age 4 y, when they start to diverge progressively. In other words, the ‘requirement’ matches actual intakes very closely below 4 y but becomes more and more in excess, even proportionately, up to age 15–16 y.
The plateau which occurs on the ‘requirement’ curve at age 14-15 and which extends for only one year or so, in fact happens much earlier in real life - at age 10–11 y - and may indeed then remain almost at the same level throughout the whole of the life span until late middle-age. The reasons for this are, of course, complex and are affected by cultural influences so that there is likely to be considerable variation, related to the particular society and certainly to individuals, but there may well be an overall pattern. If a levelling-off process of energy requirements really occurs, it must be related to an interaction affecting both body size and physical activity.
Undoubtedly body mass increases after the age of 20–11 y in females so that, if physical activity were remaining constant, energy expenditures and requirements would increase in some sort of related fashion. However, it seems as if physical activity may, on overage, gradually diminish during early - and mid - adolescence with a consequent reducing effect on energy needs. It would be most useful to have data on energy expenditures or even on activity patterns during this part of the life-span, but unfortunately no adequate data exist.
There would appear to be a plausible case for suggesting a reduction in energy requirements for females, starting at age 5 y, and being of the order of 10–15 % up to age 12 y, then perhaps at least 15 % to age 15 y, and thereafter remaining at a constant level up to middle-age.
An alteration of this degree perhaps seems somewhat substantial but there seems no justification for retaining “requirements” at the present level. None of the data obtained in free-living populations support their retention. The “developed” populations were clearly not suffering from any restriction of food availability, either in quantity or variety. Yet their intakes are uniformly less than “requirements”. It might be argued that girls in developing countries are smaller in size than those in developed countries and so would need more energy at an equivalent body mass. Figure 9 shows that energy intakes related to weight alone, are still much lower than “requirements”, so the general conclusion remains.
If fresh data - which is most urgently needed - demonstrates that the suggested reduction is unrealistic and undesirable, than a reassessment can be made. In the meantime, there is little to justify the retention of the present “requirement” values.
The data on energy intakes and expenditures of females is sub-divided into sections on non-pregnant non-lactating women in different types of occupation, into separate section on pregnancy and on lactation, and finally there are a few data on elderly women.
Table 4 summarizes the information on energy intake and energy expenditure from the various studies in developed and developing countries. As far as those in developed countries are concerned, only two reports are available, the intake and expenditure data are within reach of the “requirement”, and on such very limited information nothing more can be said.
In nine studies on 13 different groups of women in the developing countries 215 women were measured either for energy intake or energy expenditure or both. The gross values for energy are shown in table 4 and the intakes and expenditures, expressed as percentages of “requirements”, are shown in Fig. 10. These combined studies represent a more satisfactory diversity of environment - the Philippines, India, Guatemala, Upper Volta, New Guinea and Peru - but the groups were sometimes very small and the actual areas where they lived might have only a limited degree of validity for much of the relevant country - e.g. 2 groups of 6 college students in India, 20 women living in a village in the Peruvian jungle. Again it is thus dangerous to extrapolate very far from the actual data. Moreover it is difficult to discern any obvious pattern from the data. Many of the energy intakes are higher than “requirements” - the Filipino typists (127 %), both groups of Indian college students (although one of these consisted of 6 non-specified “athletes” and perhaps their requirement should have been set higher), and both groups of women living in a highland village in New Guinea (105 % and 111 %). Does this imply that the requirements are too low?
Several other intakes are much lower than the requirements - the Upper Volta farmers (75%) at the end of harvest when it might be supposed that food was readily available, both groups of women living in a coastal village in New Guinea (73% and 83%) when again food did not appear to be in short supply, the textile-mill workers in the Philippines (77%). Such large differences from supposed requirements when no signs of nutritional inadequacy have been reported, might seem to indicate that the values for “requirements” are too high.
Also, several of the studies where both intakes and expenditures of energy have been simultaneously measured have considerable disagreements between the two estimates - 13% in the Philippine housewives, 15% in the Philippine typists, 21 % in the Upper Volta farmers, 26% and 12% in the New Guinea coastal villagers, and 27% in the Filipino textile-mill workers. In all of these, the expenditures were higher than the intakes except in the case of the Filipino typists. The reasons for these discrepancies are not at clear and their existence complicates any assessment of energy requirements in the light of the present data.
It would appear unjustifiable to suggest any alteration in requirements for non-pregnant non-lactating women, although there is also not much justification for the current values from the data reviewed.
Only 5 studies are available on pregnant women, two of which have very limited usefulness - two otherwise apparently carefully carried-out investigations on 179 women in the first trimester of pregnancy living in Leeds in the UK and 435 women in the 3rd trimester of pregnancy, also in UK, where no body weights, either current or pre-pregnant, are given, and a study with most peculiar results on 20 “non-privileged” and 10 “privileged” women in their third trimester living in Addis Abeba. The third study was on small numbers of pregnant women living in a coastal and in a highland village in New Guinea, where again some unusual results were found (although they at least were fully commented upon by the authors). Lastly, a partly longitudinal study on 12 women in California measured energy intake at 3 periods during the 2nd and 3rd trimesters with little indication that energy intake was markedly increased at these times. There are some peculiarities in fitting the energy expenditure values, which were also measured in this study, to the energy intakes when account is taken of changing body weights. For example, energy expenditures at 29–36 weeks and at 37–40 weeks were 1.5 to 2 MJ/day in excess of energy intakes when the weight gains were such as to require an excess of intake over expenditure of more than 4 MJ/day.
The data from these 5 studies, of course, is completely inadequate in the light of the problem of the energy requirements of pregnancy. However it is clear that none of the findings supports the present recommendations, and the tentative impression is that the energy intakes of pregnant women are probably only slightly greater than those of the non-pregnant state. It would be fitting to consider an appreciable reduction in the extra energy which it was suggested previously was needed in pregnancy.
Much the same can be said about energy needs in lactation as was written above on pregnancy. Only one of the studies on either energy intake or expenditure of lactating women has results which even approach the “requirements”; that was in Nigeria and intakes were assessed by 24 h recall, so there must be some reservation about the validity of the data. The others, again very small in number, have energy levels which are 1–4 MJ/day less than are supposedly needed. It is highly probable that these “requirements” have been over-estimated when applied to normal free-living populations, and they should probably be reduced by at least 1 MJ/day.
Only 3 reports have been reviewed on elderly people and, with one exception (about which there must be reservations in accepting the results), the data more or less support the present “requirements”.
