The cause of diabetes can rarely be removed. Occasionally cases are met in which glycosuria has been provoked by the excessive use of sugar and starch. These can be cured radically by limiting the ingestion of carbohydrate foods. The severity of all cases of diabetes is increased by prolonged mental depression or great anxiety, or other strenuous and not exhilarating mental states. These conditions sometimes are the exciting cause of the sickness; more frequently they provoke relapses or aggravate it. The patient's mode of life should be regulated so as to correct and avoid these states. Sciatica and other troubles involving the peripheral nervous system are apparent causes of diabetes in some cases; in others, its result; though often the first symptoms to attract attention. Structural disease of the medulla, brain, or cord may apparently give rise to the symptom-complex. Destructive or functional inactivity of the pancreas also has been demonstrated to be a cause of many cases of it. It is at times associated with hepatic lesions Sometimes, however, no definite structural lesion in nervous system, liver, or pancreas can be found.

Nutritional changes characterize all cases. In well-marked instances all metabolic processes are increased; in the mildest cases, those of the liver only, or at least the most persistently. The daily output of urea is increased, although it can be demonstrated that nitrogenous food is well utilized. The nonnitrogenous foods are imperfectly oxidized. Oxyhemoglobin is increased in the blood. Carbonic acid gas is expired in larger volume than is normal, and the absorption of oxygen is augmented. The combined sulphates and hippuric acid are increased. In some cases the body waste, in spite of the eating of large amounts of food, shows that retrogressive metabolism exceeds reparative changes. In the severest cases sugar is made from protein. Von Noorden believes that the liver is unable to make glycogen and it is certainly almost entirely absent from the liver in severe cases.

For convenience in discussing the diet of diabetics we can group cases under three heads: first, the mildest; second, the moderate; third, the severe cases. Acute cases, which are seen almost exclusively in childhood and early adult life, belong to the last group, as do some of those chronic from the start. Mild cases may gradually grow more severe. Not infrequently a case is from its inception of moderate or of severe type.

In those cases that can be called mildest, the patient will cease to eliminate sugar in the urine when the ingestion of sugar is forbidden and the eating of starches is limited. They usually occur during or past middle life in persons disposed to obesity and goutiness, and readily yield to treatment. The moderate cases are those of patients from whose urine sugar can be removed by withholding all foods containing either sugar or starch. The severe cases are those of patients from whose urine sugar cannot be removed by depriving them of carbohydrate food. Sugar can be made to disappear from the urine of some of these by depriving them entirely of carbohydrates and at the same time limiting the amount of albuminous foods. In others the same thing can be accomplished by careful dieting and medicinal treatment. There are many patients, however, from whose urine the sugar cannot be made to disappear by any treatment.

It is necessary to prescribe carefully and definitely both the character and the quantity of food that diabetics shall take.

First, in order to ascertain to which of the categories described a given case belongs, the patient must be placed on the diet from which sugars and starches are absolutely excluded. However, this strict diabetic diet - that is, one containing no carbohydrates - should be instituted gradually, as in many cases coma has been brought on by a sudden and complete change, while in other cases indigestion has been caused by it. From the beginning all sugar should be forbidden, and the amount of starchy food should be diminished slightly. The latter should then be further lessened day by day during the first week of treatment, so that, if the patient's condition permits, all starch will be excluded by the end of that time. Simultaneously with the diminution of carbohydrate food, albuminoids, and especially fats, should be increased. The total quantity of urine for twenty-four hours and the percentage of sugar that it contains should be ascertained at least every second or third day. Moreover, Gerhardt's test with ferric chlorid must be made frequently. If it gives a positive reaction, or if there is other evidence of acetone or diacetic acid in the urine, a modification of the strict diet should be made at once. Indeed, it is safest to enforce Ebstein's rule that under these conditions the amount of albuminoids eaten should be lessened and the amount of carbohydrates increased. The presence of the acetic compounds in the urine signifies a great liability to diabetic coma. If the reaction is found when severe cases first come under treatment, sugar only should be excluded from the diet and the amount of albuminous food should be limited. At the same time the intestines should be cleansed and albuminous indigestion, if it exists, corrected. Alkaline salts and alkaline mineral waters are also useful under these circumstances. They may have to be given hypodermically if any quick results are required.

Acetone, diacetic acid and oxybutyric acid are due to the breaking up of fatty acids. When they are generated in large quantities they produce or are associated with an acidosis or the phenomena of diabetic coma. Alkalis neutralize the acid and help to prevent or to overcome the acid state. The administration of carbohydrates in some unknown way also prevents the formation of acetone, diacetic acid and oxybutyric acid even when fats are eaten in generous amounts. This is the reason for at once adding carbohydrates to the diet when diacetic acid is found in the urine in large quantity or persistently. It is safest also to restrict very considerably the amount of fat which such a patient is to eat and somewhat also the amount of meat. Van Noorden's oatmeal diet particularly is a good one under these circumstances.