After Long and Lukens (1936) had shown in classical studies that adrenalectomy ameliorates diabetes in laboratory animals, it was found in man that the development of adrenal cortical insufficiency in diabetic patients lessened the severity of the diabetes. It is possible to re-establish the diabetes at the pre-adrenalectomy level by the administration of large doses of adrenal cortical extract or by treatment with the so-called glucocorticoids of the adrenal. In 1941, Ingle induced steroid diabetes in the normal rat and a few years later was able to reproduce this metabolic derangement by the administration of large doses of ACTH. Steroid diabetes is rare in man but it is known to occur either as a symptom of Cushing's disease or as a sign of hypercorticalism caused by overdosage with either a glucocorticoid such as cortisone or with ACTH. It disappears in both man and laboratory animals when the hypercorticalism is corrected unless, as in a few patients, diabetes mellitus co-exists.

It has sometimes been claimed that hypercorticalism can cause diabetes mellitus. Hypercorticalism aggravates diabetes mellitus and may make it overt in a case too mild to be detected by ordinary methods. But the proposition that disturbed adrenal cortical function is a common primary cause of diabetes mellitus is supported only by faulty logic and flimsy evidence. Pancreatic diabetes and steroid diabetes are different diseases. In the former there is a defect in the oxidation of glucose, but no defect in oxidation of glucose is known to exist in steroid diabetes, although there is some change from normal in other phases of carbohydrate utilization, such as conversion of carbohydrate to fat.

Exposure to severe stress activates the pituitary-adrenal axis and causes an increase in secretion of glucocorticoids and it must be asked if this can lead to a diabetic state or exacerbate previously existing diabetes mellitus. Although the putative effect of trauma and infection in precipitating or worsening diabetes mellitus has been debated for several decades without agreement among clinical investigators, there is no evidence that nonspecific stressors worsen either clinical diabetes or experimentally induced diabetes. Indeed, the glycosuria of the partially depan-creatized rat is reduced during severe exercise, by exposure to cold, injection of formalin, by aspirin and in the presence of tumours. Recently it has been shown that when Walker carcinoma 256 is successfully transplanted in rats with severe steroid diabetes the glycosuria disappears (Ingle, 1960). This tumour causes marked hypertrophy of the adrenal cortices which is not completely prevented by cortisone in doses causing compensatory atrophy of the adrenal cortices under non-stress conditions.

Since pancreatic diabetes can be ameliorated by adrenalectomy, it may be claimed that a state of hypercorticalism existed before the adrenal glands were removed. The amount of cortical hormone does have a causal relationship to the severity of glycosuria, but elsewhere we have argued that this is an example of a permissive or supporting cause rather than a primary cause (Ingle, 1954). We know the primary cause of pancreatic diabetes is lack of insulin and yet when the glucocorticoids are withdrawn, the diabetes improves. Another possibility to be considered is that the adrenal cortices become hyperactive following pancreatectomy. Twenty years ago one of us (D.J.I.) did the following experiment. Adrenalectomized tube-fed rats were maintained on a normalizing dose of adrenal cortex extract which was uniform during the experiment. Severe glycosuria developed after extensive pancreatectomy; the secretory activity of the adrenal was not a factor, and when the intake of ACE was decreased, the glycosuria decreased. We believe that the role of the adrenal hormones here is best regarded as secondary, or permissive or supporting. It is reasonably certain that adrenalectomy reduces the severity of diabetes because hepatic gluconeo-genesis is thereby reduced and because the animal now conserves some of the carbohydrate possibly by converting it into fat, which was formerly lost in the urine.