Clinical and epidemiological studies of the concentrations of insulin in man.

Abstract

Clinical and epidemiological studies were conducted to identify factors related to the wide range of serum-insulin values encountered in man. A reliable modification to the double antibody immunoassay for insulin was established. Serum-insulin levels were log-normally distributed. Lean healthy subjects showed a ten fold range of insulin values. "Insulin resistance" was identified in consistently high insulin secretors, who showed correspondingly higher blood-sugar values. Normoglycaemic and borderline diabetic patients with vascular disease or obesity had elevated serumrinsulin levels, but more severe diabetes was associated with a delayed and inadequate response to oral glucose. The one hour sample after 50 g. glucose identified most clearly the "insulin resistance" of normal subjects as well as the impaired insulin response of diabetics. In population studies, the independent relationships of various factors with the "one hour" serum-insulin levels were delineated. Blood-sugar levels showed the strongest association with serum-insulin. The latter rose steeply with blood-sugar levels up to 180 mg/100 ml., but beyond this a progressive decline of insulin values occurred. Other major factors associated positively and independently with elevated serum-insulin levels were age, obesity, hyperuricaemia, and hypokalaemia. The insulin response to the standard glucose load varied inversely with height. Females had consistently higher levels than males. Fasting serum-insulin levels were closely correlated with serum triglycerides both in normal subjects and those with coronary heart disease. Presumably insulin may have a physiological role in2 promoting hepatic triglyceride synthesis. Measurement of serum-insulin provided useful insight into possible mechanisms of disease; but was of limited use in discriminating diseased individuals. In the stages of diabetes mellitus, borderline hyperglycaemia was related to "insulin resistance", which may, in part, reflect varying vascular permeability. More severe hyperglycaemia was associated with a progressive impairment of insulin response, of undetermined cause, although hyperosmolality may contribute to this. Hyperinsulinism is possibly a "pre-diabetic" phenomenon. In subjects with vascular disease in the population, the association with elevated serum-insulin levels was confirmed, particularly in the younger subjects with coronary heart disease and in hypertensives generally. Serum-insulin may usefully characterize the metabolic status of population samples, for healthy and active subjects showed low levels that did not rise until late age; whereas obese, hyperglycaemic, or vascular disease subjects had prematurely elevated levels, and an accelerated rise with age. The findings are compatible with the hypothesis that insulin has an aetiological role in the early genesis of atherosclerosis.