In 1948 it was noted that whatever the standard of dietary control there was little variation in food intake from day to day. In some patients there was a tendency for a reduction of food intake towards the end of the week, and this was assumed to indicate an element of financial difficulty before pay or pension day, especially as it was more obvious in patients with limited means. In complete contrast, in 1968 there were considerable fluctuations in intake from day to day, with no particular pattern discernible during the week. Some variability was noted even in those patients who, on average, were keeping close to the prescribed diet. The extreme example of such variation was in a diet which ranged from 1,265 to 2,850 calories per day within the week, the patient, perhaps surprisingly, was a woman aged 63. Though these fluctuations may have an adverse effect on blood glucose levels they are in keeping with the variations observed in normal people, and are presumably related to the changing energy requirements from day to day. Booyens and McCance (1957) showed that in six normal subjects there was a wide range of daily food intake, the calorific intake varying from an average of 2,114 to 3,007 calories during a one-week or two-week period. There are obviously many social and economic factors which influence a diabetic patient's ability and desire to follow dietary advice. One factor is the cost of the diabetic diet, and this was assessed in both Leeds surveys. In 1948 it was found that the average weekly cost of food for diabetics was 23s., while the average cost of food for the non-diabetic members of the same families was just over 10s. In 1968 the cost of an ideal diabetic diet, of the same calorific value as the average prescribed for the patients in the survey, was found to be 44s. Id. This figure was calculated after a careful survey of the range of local retail prices. The probable average cost is not the arithmetical mean of the lowest and highest possible cost. It is unlikely that many patients would buy food at the highest prices and even more unlikely that they could buy all their food at the cheapest prices, for these were usually limited to one area within the city. We have therefore assumed, after reviewing the patients' records and the range of prices for each item of food, that the probable actual cost of each item was the minimum cost plus two-thirds of the difference between minimum and maximum costs. In assessing the cost of the diets actually consumed it was found that many patients failed to identify their food in sufficient detail for a confident calculation to be made. In those cases where the information supplied was adequate the probable actual cost, on average, was 49s. 3d. Though the results of these calculations must be accepted with caution, they all indicate that the cost of food for diabetics is higher than for non-diabetics (38s. 4d.) (National Food Survey, 1968), and support the clinical finding that patients with low incomes continue to have difficulty in purchasing a correct d...
