SINCE pancreatectom in dogs resulted in hyperglycemia (1) which was reversed by treatment with the product of the pancreatis islets (2, 3), diabetes was considered to be due to insulin deficiency. This view was questioned almost 50 years ago when Himsworth suggested that diabetic patients could be subdivided into two groups: insulin-sensitive and insulin-insensitive (4). Later, Himsworth pointed out that insulin-sensitive patients tended to be ketosis-prone, while insulin-insensitive patients were usually middle-age diabetics without tendency to ketosis (5).

Himsworth's concept gained support by the findings that plasma insulin activity, determined by biological assays, was absent in diabetics with weight loss and ketosis but was essentially normal in overweight diabetics without ketosis (6). Subsequently, a strong proof for the concept was provided when Yalow and Berson (7), using their specific sensitive radioimmunoassay for insulin, demonstrated that, in response to provocative testing, plasma insulin levels in diabetics ranged from absolute deficiency to markedly elevated values.