by Dr. G. H. Beaton
In the accompanying graphic are portrayed certain of the concepts of nutrient requirement and the Recommended Intake as applied by the FAO/WHO Expert Group on Requirements for Thiamine, Riboflavin, Niacin and Vitamin A and as further adopted by the FAO/WHO Expert Group on Requirements for Vitamin D, Ascorbic Acid, Folic Acid, Vitamin B12 and Iron. Thiamine is chosen as the example to illustrate these concepts. The Group, after examining the data available to it, concluded that clinical lesions of thiamine deficiency were associated with intakes of about 0.22 mg/1000 kcal or less. However, they also noted that when the relationship between thiamine intake and urinary excretion was examined, there was a "break point" at about 0.33 mg/1000 kcal. This figure was accepted by the Group as the average requirement for thiamine in the population. In the accompanying graphic, these concepts are applied to the interpretation of nutritional status in the average person (the "reference man"?).
However, the Group also recognized that individual requirements vary for reasons that we cannot yet predict, and that the above figure (0.33 mg/1000 kcal) would be sufficiently high to meet the individual requirements of only one-half of the population. They accepted that the variability of thiamine requirement was &+-; 10 per cent. In the middle portion of the graphic is plotted the cumulative distribution of thiamine requirements in a population applying the above figure for variability of requirement. The Group accepted that a desirable objective would be to meet the nutrient requirements of 97.5 per cent of the population (average + 2 S.D.); at the level of the Recommended Intake, the requirements of all but 2.5 per cent of the individuals within the population would be met.
The corollary of this approach is also shown in the accompanying graphic. Accepting the same hypotheses as were accepted by the Group, at the level of the Recommended Intake, the individual has only one chance in 40 of not meeting his own requirement. As his intake drops to the level of the average requirement, he runs a risk of one chance in two that he will not meet his own requirement and at an intake of 0.26 mg/1000 kcal, there is only one chance in 40 that he will meet his own requirement.
The applications arising from this, as applied to the individual are:
1. The implied public health objective is that each person achieve the Recommended Intake (i.e. that the individual be at minimal risk of failing to meet his requirements).
2. The relationship between observed dietary intakes of individuals and their actual nutritional status must be considered in terms of the relative risk of inadequacy of intake to the individual. Thus individuals ingesting the Recommended Intake or greater are believed to have a very low risk of dietary inadequacy while those ingesting, in the case of thiamine, an intake of 0.26 mg/1000 kcal are almost certainly (very high risk) ingesting less than their individual requirements.
A further implication of this interpretation is seen when population data, rather than individual data, are considered. If (1) above is accepted as correct, then it is apparent that a desirable objective in planning food supplies for a population would be to provide an average intake that exceeds the Recommended Intake. How far above the Recommended Intake the average intake should be will depend upon:
This implication can be exemplified by some calculations based on a study of 2700 persons in Canada. In this group, examination of dietary data revealed that the variability of thiamine, riboflavin and niacin intakes 1 were about 30 per cent. (S.D./mean intake); it was noted that the distributions did not fit the normal distribution and this was taken into account in the calculations. By statistical manipulation of the data, the average intakes of these nutrients were adjusted to the Recommended Intakes, keeping the variability of the data proportionate to the mean. Applying the assumptions made by the FAO/WHO Group about the variability of requirement, and having satisfied ourselves that there was no covariance between intake and requirement when both were expressed as mg/1000 kcal, we were able to calculate the proportions of individuals who had intakes below their own requirements.
When the average intake was equal to the Recommended Intake it was predicted that 25-35 per cent (depending upon the nutrient) of the individuals would not have met their personal requirements.
As a public health objective in a nutrition programme, it is doubtful that this could be considered to be a satisfactory situation.
It is important to understand and to apply these concepts and implications in interpreting dietary intake data at either the individual or the population level.
1 These nutrient intake data are based upon the mean daily intakes as derived from 7 day records of food intake (a small proportion were 4 day records). In our experience with Canadian dietaries, extension of the collection period beyond 4 days does not alter the variability of group data (i.e. variability of daily intake by the individual is no longer an important factor).