|Protein-Energy Requirements of Developing Countries: Evaluation of New Data (UNU, 1981, 268 p.)|
|Discussions and recommendations of the task forces|
The working group on protein requirements for children was asked to address the following major questions:
1. An important factor in the utilization of food proteins, especially of plant compared with animal sources, is their digestibility. What differences in faecal nitrogen exist between populations with low protein intakes corresponding to those used to measure obligatory losses of requirement levels, and populations with habitual levels of protein intake from different food sources?
2. What is the significance of zero-balance intercepts; the use of nitrogen retention required for normal growth; and the use of integumentary losses in nitrogen-balance calculations?
3. What are the effects of acute and chronic infections and of other stress situations?
4. How do these affect catch-up growth?
Standardization of Methodology
The drafting group members were conscious of a number of limitations in the investigative approach used in the UNU-sponsored research and agreed that it should be supplemented by research following the guidelines for protein requirement studies on children devised by a 1977 FAO/WHO Expert Consultation (18).
Selection of Subjects
In principle, children should be representative of the general population from which they come, but this was rarely the case. At INCAP, they were all children who had previously been severely malnourished. The children who were studied in the Philippines were mostly mildly undernourished at the time of the study, the majority being below 90 per cent weight/height.
For future investigations it is recommended that the children selected for this type of study be representative of the segment of the population about whom there is concern. Children without obvious morbidity should be selected and, preferably, they should never have been severely malnourished. If, on initial selection, they are below 90 per cent weight/height, they should be rehabilitated for at least four weeks before the commencement of the study. Unpublished INCAP data show that it takes as long as 30 days for the creatinine-height indices to return to normal, even though weight/ height has achieved normality.
While frequently the investigators in the studies carried out so far had little option but to work with the subjects they did, in the future greater efforts might be made to obtain a more representative selection of the target population.
Ideally, the customary energy intake of each individual child should be determined over a representative period of time, perhaps one month, and then the nitrogenbalance procedure should be carried out at that child's established intake. The standardization of energy intake levels used left much to be desired. In the studies carried out using the multiple protein level approach, either the energy intake was fixed at the 1973 WHO/FAD theoretical average requirement for that age or it was ad libitum based on fixed protein-to-energy ratios. In this latter situation, children who had low energy intakes also had low protein intakes.
We must recognize that these inconsistencies represent a shortcoming of experimental design. In the studies by Intengan, for example, the energy intakes were up to 10-20 per cent above the assumed physiological requirement because the ad libitum intakes were measured during recovery from mild to moderate malnutrition. The data must be interpreted in the light of these facts. In spite of this, however, it must be recognized that the INCAP team observed that a 10 per cent energy intake reduction at a fixed protein intake below the WHO/FAD safe level did not affect nitrogen balance in children. The same would appear to be true of the studies carried out by Tontisirin.
Energy Content of Food
Another potential source of error is the assumed energy content of food. It may be inadequate to use food composition tables to define the energy content of the diet, particularly for those based on traditional foods. What is important is the available energy content of food. It is recommended that in future studies the energy content of food and faeces should be measured directly by bomb calorimetry. If necessary, suitably prepared samples should be sent to a regional reference laboratory for this purpose.
General Health Status
The UNU collaborative study represents a mixture of children. Some were infested with intestinal parasites, although usually only mildly or moderately so. The experience of Ju, as described at the meeting, has clearly been that a low level of Ascaris and Trichuris infestation does not affect nitrogen balance. Even though infections with these worms are endemic to most developing countries, it appears unnecessary to deworm such subjects, since this would not significantly affect requirements. The influence of Giardia on nitrogen balance, however, deserves more detailed study. Apart from worm infestations, the UNU research subjects were apparently free of clinically manifest signs of infection that might have influenced the validity of the resuIts.
Protein Characteristics of the Dietary Proteins Tested
In some of the studies presented, information was not given on the amino acid composition of the diets used. Such information should be obtained, and this would facilitate comparisons among the studies from the different research centres. Food table values for amino acid compositions are rarely adequate, and it is recommended that in future studies of this type a reference laboratory should be available for the conduct of amino acid analyses. In the subsequent published accounts of the present studies a concerted attempt should be made to make amino acid data available for comparative purposes, in spite of the difficulties and limitations involved.
Physical activity can also influence the efficiency of dietary nitrogen utilization. Energy output should therefore be standardized. Play facilities should be provided during the investigation and the child should be encouraged to use them. The use of "metabolic beds" should therefore be avoided whenever possible. One or two studies in the UNU project could have been influenced by a lack of attention to this detail.
Number of Levels of Protein to Be Fed for Balance Studies
The drafting group was concerned about the statistical procedures adopted for the analysis of the children's nitrogen-balance regressions. While the pooled regression analysis gives an adequate estimate of the mean nitrogen intake needed for a given level of balance, it does not provide a precise estimate of variation in requirements between individuals. We recommend that the present data be recalculated using regressions for individual children and that future studies be planned in such a way that they can be analysed on this basis.
The use of individual rather than pooled regression analysis necessitates at least four and preferably more levels of nitrogen intake. These should be fed around the estimated level for normal growth plus and minus 20 per cent. Lower or higher intakes would serve no practical purpose.
It was considered premature to make any comments on nitrogen requirements either on the apparent mean or on variance until the data are supplemented with those from additional studies under way and then reanalysed. However, there appears to be a remarkable similarity among most of the data from the different centres.
Integumental Nitrogen Losses
Such information on integumental nitrogen losses is essential for the true interpretation of nitrogenbalance data. The UNU data present, for the first time, nitrogen balance information on children between one and three years of age from two cultures, Taiwan and Guatemala. These studies were done under conditions where there was no overt sweating. With protein intakes of 1.4-3 g/kg/day, integumental losses were virtually the same (6-9 mg N/kg/day). Interestingly, and confirming previous adult findings, lower protein intakes yielded slightly lower integumental losses. With 0.5 9 protein/kg/day, it was 6-7 mg N, and with a nitrogen free diet it was 5 mg N/kg/day. It is recommended, in the interpretation of nitrogen-balance studies on children who are not sweating, that 8 mg N/kg/day be incorporated into the calculation of nitrogen balance. For studies carried out under circumstances where environmental temperatures are such that sweating is more profuse, the extra integumental nitrogen losses are unlikely to produce a significant error in the balance calculations.
Exclusion of Anomalous Data
The decision to use individual regression equations will ensure the detection of individuals producing anomalous patterns of nitrogen retention relative to intake. When the results are biologically unrealistic (for example, a negative regression coefficient) for the relationship between nitrogen intake and balance, the data for that individual should be discarded. In less clear-cut cases, it is recommended that objective decisions be made based on accepted statistical methods. If it is possible to repeat the nitrogen study on a given individual, this should be done.
While the exclusion of data always represents a difficult decision, it is essential to deal with this problem so that the calculation of requirements does not result in unrealistically high values. When data have been excluded in the calculation of nitrogen requirements for the pooled multi-centre studies, this should be clearly stated and the specific reasons given.
Zero-Balance Intercept and Growth
For the non-pregnant, non-lactating adult the zero-balance intercept (Bo), after taking into account the integumental losses, can be used as an indicator of the mean nitrogen requirements. In the case of the growing child, the adequate level of nitrogen intake should be one that allows the child to grow at an acceptable rate.
It was recommended that the present data be analysed by the individual regression approach to produce two values, Bo, the nitrogen for maintenance nitrogen balance, and Bg, the nitrogen intake to allow for daily growth corresponding to the mean annual growth rate at the 50th centile for a given size. Bo is largely of academic interest, but on a per kilogram basis the data would seem to indicate that it is the same for children as adults. For population groups, the relevant value for children is Bg.
At the level of the individual, however, growth, even in healthy children, is by no means a uniform process (7). Variations in growth rate up to five times the 50th centile are commonplace. At these times, more protein would be needed. It is impossible, at the present time, to predict exactly how much more is needed since changes in physiological efficiency are a distinct possibility. Furthermore, this variable growth rate limits the value of protein requirement estimates based on data observed from short-term nitrogen balance studies.
Valuable data would accrue if healthy children in different countries were followed longitudinally to determine precisely the natural fluctuations in growth and the extent to which they are paralleled by fluctuations in food intake. Special attention should also be placed on the voluntary selection of foods and the resulting different nutrient intakes to see whether or not there is a "hunger" for individual nutrients as well as for energy.
Determination of Obligatory Nitrogen Loss in Children
Three studies were presented at the meeting concerning obligatory nitrogen losses on diets containing very low amounts of protein. There was good agreement as to faecal nitrogen losses, but a discrepancy in the urinary nitrogen. The two INCAP studies suggested 34 mg N/kg/day as the obligatory loss, whereas the Taiwan data suggested 54 mg N/kg/day. A possible explanation lies in methodological differences; the diets in the Taiwan study provided 12-28 mg N/kg/day, whereas the INCAP diets were virtually nitrogen-free. Another possible explanation was that the Taiwan data were based on pooled collections covering days four to ten, whereas the daily analyses performed in the INCAP studies suggested that day four was too early to be included in the calculation of obligatory nitrogen losses.
Regardless of this unresolved problem, we do not recommend that new studies be initiated. Obligatory nitrogen losses are of value for calculating protein requirements using the factorial approach and for estimating true digestibility and BV data for protein foods, but are not needed for requirement estimates based on the zero-balance intercept approach. Moreover, ethical problems make further studies of this type undesirable, even though there were no ill effects to the subjects as a consequence of the low-protein intake in the studies presented.
Suggestions for Future Investigative Approaches
Laboratory and metabolic ward approaches are invaluable, but are inevitably of relatively short-term duration and can never accurately reproduce normal living circumstances. To be truly sure that a given level of energy or a nutrient is adequate, long-term studies are essential.
Two approaches suggest themselves. One would be on free-living children for whom food intake would be measured sequentially over an extended period of time in conjunction with measurement of growth, body composition, function, and basic well-being. The aim would be to determine, under prevailing circumstances, the level and quality of food that is compatible with health and with appropriate physiological and psychological performance. This approach cannot identify minimum requirements. In certain circumstances, food intake could be approached by relevant modifications of intake.
The second approach is a natural extension and would apply to circumstances where one had a more complete control over children's food intake, such as in child welfare institutions. Here apparent safe levels for short-term metabolic studies could be introduced and tested on a long-term basis. As an alternative to multilevel studies in a few individuals, a single level of dietary protein intake considered adequate could be given to a large number of individuals to observe the distribution of their nitrogenbalance response. If appropriate for most normal individuals, negative nitrogen balance responses should be rare but might occur occasionally.
The drafting group suggests that relevant indicators of overall health and well-being in studies of dietary protein adequacy should include:
(a) weight and height growth;
(b) assessments of lean body mass and adiposity, using techniques appropriate to technical facilities available; and
(c) the immune response.