Previous Page Table of Contents Next Page


METHODOLOGY IN REVIEWED ARTICLES

In the various published articles reviewed here, very few indeed satisfy these criteria, even as far as the bare essentials. The methodology has almost never been described in sufficient detail to allow an adequate assessment to be made of the likely accuracy of the technique. No respectable journal of physiology or biochemistry would accept papers where the “method” section was so imprecise as was almost invariably the case in these nutritional papers. For example, many of the papers reviewed (which represent, in our opinion, the best of the published work, much else having been discarded) gave no information on such a basic requirement as body size — no data on weight or height! This was the case even in many studies on infants and children. Some authors stated that anthropometric data was given elsewhere: it is surely extremely remiss not to quot the basic anthropometry together with the nutritional data, otherwise it is impossible to evaluate the results in any satisfactory fashion. one paper, which we know personally was otherwise carefully carried out (although this was not made clear in the article) stated that heights and weights were reported elsewhere but gave no reference ! Another study on infants and children which specifically mentioned that it was “intended primarily to study the relationship between growth and health” gave no data whatever on the body weights, even though it discussed at length the effects of family size and social class; the interaction between these factors and body size was clearly relevant. This same paper also stated that there was a significant difference between the intakes of boys and girls over 3 y of age and yet presented only pooled data for the two sexes. Many other papers also pooled data, either for sexes or for age ranges, although real differences might thus have been concealed.

Sometimes authors reported really extraordinary findings without making any comment, particularly in relation to whether methodological errors or difficulties might have been responsible for the unusual results. For example, one group of children whose intake was assessed in yearly groups from an initial age of below 1 y up to 5 y of age, showed a very considerable excess in the youngest group relative to supposed requirements, and then a progressive deficit relative to requirements reaching very large levels at age 5, with the intakes of each age group being almost exactly uniform — that is, children less than 1 y old, at age 1, then age 2, age 3, age 4 and 5, were all eating the same quantity of food (59). This remarkable finding did not seem to elicit any surprise in the authors.

Another paper on two groups of pregnant women, one being from the “privileged” and the other from the “unprivileged” sectors of the community of a large city in a developing country, had intakes of 12.4 and 6.4 MJ/day respectively (28). Their body weights were exactly the same. Either the non-privileged women, existing under severe economic restraints apparently, were living a life which required so little physical activity that they could maintain their body weights on intakes of energy only half those of their very energetic privileged “sisters” - an unlikely circumstance - or there must have been some unusual factor, possibly methodological, to account for these curious results. When the responsible authors seem oblivious to the incongruity of their data, it is difficult to avoid a feeling of reluctance to accept the findings at their face-value.

Similar criticisms could be made of several other papers included in this review.

The overall impression strongly generated by our perusal of this literature is of a standard, at the best, indicative of poor scientific descriptive ability and, at the worst, illustrative of severely deficient methodology. It is sad that this state of affairs seems to exist so generally in this area of nutritional research.

We would urge most strongly that the first duty of the editors and referees of nutritional journals is to assist in the attainment of reasonable standards in the presentation of articles which are published. Unless all the appropriate data have been obtained in an acceptable scientific fashion and unless it is clear that this is the case, it is impossible to assess the scientific value of the findings. It is a high responsability on editors and referees to ensure that data published in their journals have been obtained by reputable methods which have been fully and clearly described.


Previous Page Top of Page Next Page