7. Gaps in our knowledge

Given the emphasis placed on possible hitherto unmeasured benefits of supplementation, it is appropriate that these should be investigated. Energy expenditure has been a neglected variable because of the difficulty of measurement and the perhaps unjustified disrepute into which the method has fallen. The doubly-labelled water technique will provide much sought-after information on total daily energy expenditure (TDEE) in free-living conditions, but the traditional techniques of diary recording and indirect calorimetry will be retained to identify the components of TDEE that are changing. Identifying changes of habitual physical activity and, in particular, the pace of work may require new approaches in measurement and recording of the social scientist as well as the physiologist. The effect of changes on interpersonal and social interaction of individuals and families can then be investigated.

Supplementation of an individual clearly affects the family and studies of the effects on the whole family are required. Much remains to be answered about leakage, sharing and substitution. Does sharing within the family reflect family need? Apparently not, as it occurs even when the whole family is supplemented. Sharing does not appear to be related to the extent of the food gap. In children, there is no relation between existing intake and net increase in intake (BEATON and GHASSEMI, 1982) although there was a negative correlation in pregnant Bogota women. Does leakage increase purchasing power and how does this affect the family? Do parents control child supplements or is the child satiated? Is the basic disturbance shortage of food or what is it that limits food intake?

Although the advantages of experimental studies were highlighted earlier, BARRETT (1984) has argued a case for non-experimental designs to compare the effects of several independent variables under conditions in which they normally operate and to allow for covariation. Although this approach may not identify the single most important variable, it is more valid ecologically and may be used to improve the design of subsequent interventions. Pathway analysis may help identify intervening and confounding variables, processes and new outcomes (BEATON, 1982). The smaller and less specific effects of indirect multistage processes make greater demands of large sample sizes, better statistical analysis and variables of higher specificity and sensitivity. There are limits to the number of variables that can be studied, drawn by the cost and the quality of data obtained, and hypotheses have to be refined to identify what needs to be measured. The apparent importance of non-nutritional factors of the social, economic and biological environment in determining the extent of the response to supplementation suggest these as rich areas for study. Measurement of the separate effects of nutrition and health care and education stimulation, and whether a synergism exists, is a priority. Is energy supplementation alone worthwhile? Pleas for functional indices of energy nutritional status continue to be made but this has not been a promising area so far. MARTORELL (1984) recommends an outcome of "social competence" - how good an individual is at everyday tasks - which has considerable merit.

The question and investigation of adaptations to chronic energy deficiency is relevant here. Do the low intakes recorded in many studies represent the habitual intakes of the subjects? Is the de-adaptation of supplementation a real phenomenon? What are the effects of discontinuing supplementation? Can individuals readapt?

The policy issues of whether to target the most needy or most responsive and how these might be identified have not been considered here. However, whether to supplement mothers or infants (LECHTIG and KLEIN, 1980) and whether maternal supplementation can affect the offspring without changing maternal nutritional status are questions that need to be answered. Similarly, the factors responsible for individual variations in responsiveness and even participation have yet to be fully described. In each community, different factors may operate and there may be no circumventing the need for a full description of the ecology of the community.