With regard to the suddenness with which the diabetic regimen should be adopted by the patient, it is found to be the rule that it is easier for most patients to begin with the fully restricted diet at once than to gradually eliminate one article after another from the menu. The urine should always be thoroughly examined, both qualitatively and quantitatively, for sugar, urea, and albumin, before the restricted diet is commenced, and periodic examinations must be made during the course of the treatment in order to determine the effect upon the disease of withholding different articles of food and drink.

At first it is well to make such examinations daily, and subsequently, if the patient is doing well, once or twice a week is sufficient.

The effect of no dietetic system is immediate, and at least two days may be required for the patient to come fully under the influence of treatment, because the materials already present in the body when the new diet is commenced may serve as a source of sugar for some little time afterwards.

The general principles for the dietetic treatment of diabetic patients are, first, to exclude from the diet, when possible, all those articles which are most likely to result in the formation of sugar; secondly, to devise means for relief of the craving for starchy and saccharine foods which patients kept on a nitrogenous diet are certain to experience before long, and to see that while reducing the amount of sugar in the food, the patient is not suffering to a dangerous extent from lack of nutrition. It is difficult to formulate rules of diet which can be adhered to in all cases. There are many patients confined to an exclusive nitrogenous diet who lose flesh and strength so rapidly that although the advantage of reducing the glycosuria is attained, a new danger arises from anaemia and emaciation.

For convenience of description of dietetic treatment cases of diabetes may be subdivided into three classes: (a) Those patients who pass a considerable quantity of urine containing a large percentage of sugar, but in whom the general health is still good, (b) Cases in which, in addition to the passage of considerable sugar in the urine, there is more or less dyspepsia, emaciation, and debility, (c) Cases in which the constitutional symptoms become rapidly severe after the first appearance of sugar in the urine.

(a) In the first class of cases the dietetic treatment is productive of the greatest benefit, and not rarely the patients begin to gain flesh and strength; they sleep better; the daily quantity of urine falls perhaps from three hundred ounces to seventy, and it approaches the normal composition; the excessive appetite and thirst diminish; the digestion improves, and in one to three weeks the sugar may entirely disappear. Such patients are not to be regarded as cured, however, as soon as the sugar disappears - not, in fact, until they can eat starches in ordinary quantity without exciting the appearance of glycosuria. According to the statement of Dujardin-Beaumetz, patients of this class may be cured who have been eliminating as much as three thousand grains of sugar per diem. Whenever sugar, or amylaceous food absorbed as sugar, passes through the circulation unaltered, it is of no value to the system, and, as Bauer says, it serves merely as "useless ballast".

(b) The second class of cases is also amenable to dietetic treatment, but the benefit is not so immediate and usually not so great as in the first group, and it may be impossible to cause the total disappearance of the sugar, although it may be reduced to two hundred or three hundred grains a day. When the amount of urine is lessened by the changes of menu, the specific gravity remains high, and sugar is abundant, the prognosis is very grave. The patient cannot tolerate the diet, and drugs must be resorted to.

(c) In the third and most severe class of cases dietetic treatment is sometimes of avail, and it should be always undertaken, for it may prevent the patient from becoming worse, although it is unable to accomplish a cure.

But usually the treatment of these cases is unsatisfactory, because the patients, who are often young subjects, continue to form sugar from nitrogenous food after starches and sugars have been proscribed, and it is practically impossible to devise any special combination of food, or to find any single food upon which they can support life and from which they are not able to produce sugar. Their nutrition is extremely poor, and they go rapidly from bad to worse. Not seldom they are thin, neurotic subjects when the disease first attacks them, and therefore have very little capital to draw upon. In extreme cases a careful comparison between the amount of food ingested and sugar eliminated has furnished evidence that they are even capable of manufacturing sugar or glycogen out of the muscular tissues of their own bodies, as well as from proteid foods.

A clinical comparison of the different varieties of diabetes above described, with the theories which have been discussed in regard to the possible method of the production of the disease, emphasises the conclusion that cases of diabetes may not all have exactly the same aetiology.

By different writers all gradations of diet have been recommended, so long as the carbohydrates are restricted, from the absolute meat diet of Cantani to the skim-milk diet of Donkin and the more liberal menu of the majority of authorities.

Some patients who have a very good appetite when allowed a mixed diet lose it altogether when put upon an exclusive nitrogenous regimen. There are those who can live contentedly on an exclusive diet of proteid food and fats for a certain length of time - say ten days or a fortnight - if the principle of the treatment be explained to them and they are anxious of being cured, but sooner or later they almost always find it intolerable to wholly abstain from starchy and saccharine foods, and many prefer to live less long, but in more comfort than such restriction implies. This is probably attributable more to long-continued habit or heredity than actual inability to support life on a purely nitrogenous diet, for, as stated elsewhere (p. 320), the Eskimos thrive upon a diet absolutely free from starches and sugars of every kind. It is believed by Ebstein and others that an exclusive meat diet may be injurious on account of a tendency to produce acetonemia, and it may favour the increase of uric-acid deposits in those having the uric-acid diathesis.

To obtain enough carbon from such a diet for the needs of the system, a very large quantity of proteid food must be consumed.

When chronic nephritis complicates diabetes the difficulty of dieting is much enhanced, for meat is injurious for the nephritis, and starches are prohibited in diabetes. This condition has been aptly described as the "Scylla and Charybdis of the diabetic." As a compromise such patients must usually be put upon a milk diet.

Dujardin-Beaumetz states that when the quantity of sugar eliminated falls to ten grammes a day, less rigorous withholding of starches is needed.

Mild cases of diabetes seldom occur in children, hence the diet for them must be even more closely confined to proteids than for adults.

Many elaborate diet tables have been prepared, and for those whose means permit of indulgence in delicacies considerable variety may be secured without the use of carbohydrates; but for the poor in hospitals and at home it is a difficult problem to furnish inexpensive variety without occasional recourse to starchy foods.