The heralded isolation of insulin in Toronto in 1921 was followed immediately by treatment of diabetes. It soon became clear that while insulin was effective in regulating the blood glucose levels in most patients, there were some subjects in whom insulin appeared to be ineffective [1,2]. This lack of insulin effect was termed “insulin resistance” as early as 1925 . In a remarkable early study, Bainbridge observed that rats and mice, fed a carbohydrate-free, excess fat diet, developed a high degree of resistance to insulin. In 1936, Himsworth noted that the effect of insulin injection was less pronounced in obese subjects and subjects with “mild” diabetes . He suggested that obesity and diabetes are related to insulin resistance  and that insulin sensitivity could be increased by high carbohydrate, low fat diets . However, lack of ability to measure insulin directly made the interpretation of Himsworth’s elegant studies difficult in terms of insulin action per se. Introduction of the radioimmunoassay by Berson and Yalow provided the first compelling evidence for the existence of insulin resistance. These authors were aware that higher insulin after oral glucose implied insulin resistance, and suggested insulin resistance in obese patients, and patients with “maturity onset” (type 2) diabetes [7,8]. However, absent a direct assessment of insulin resistance, the putative role of this presumed defect in type 2 diabetes (T2DM) remained controversial. Adding to the controversy was Kipnis’ elegant demonstration of an insulin secretory defect in type 2 patients. Kipnis reported a diminished plasma insulin response to a matched plasma glucose pattern in type 2 diabetic patients compared to nondiabetic patients . Thus as of 1970, supportive evidence existed suggesting a defect in β-cell function was responsible for “maturity onset” diabetes; au contraire, insulin resistance could likewise be implicated.
Opening the loop: the pancreatic suppression test
Controversy regarding insulin resistance in T2DM was due to the “closed-loop” feedback system, which regulates the blood glucose concentration (Figure 15.1). In normal individuals, ingestion of a meal increases glucose and other secretagogues, including GIP and GLP-1, which act in concert to stimulate insulin secretion. Hyperinsulinemia in turn acts to renormalize the plasma glucose level by suppressing glucose production and increasing glucose utilization. In the presence of insulin resistance per se, the elevated insulin will be less able to normalize glucose, therefore resulting in a secondary stimulus to the β cells and relative hyperinsulinemia. In fact, postprandial hyperinsulinemia was considered by Berson and Yalow to be the signature for insulin resistance [7,8]. However, β-cell oversecretion could also account for postprandial hyperinsulinism. In fact, some investigators have suggested that hyperinsulinism itself is the primary event in the development of T2DM . Because of the closed-loop relationship between insulin secretion and insulin action it is problematic to infer the existence of insulin resistance directly from a “closed-loop” procedure such as the oral glucose tolerance test (OGTT), although this approach has become more widely used lately (see later, [11,12]). Equally important, it appeared it was not possible to assign a quantitative value to insulin resistance based solely on the closed-loop response to oral glucose.
Shen and Reaven demonstrated definitively the existence of insulin resistance in patients with T2DM [13–15]. These investigators intervened experimentally to interrupt the closed-loop. They made the β cells relatively unresponsive to increasing ambient blood glucose. Initially this was done with epinephrine [14,15] and later with somatostatin [16,17]. Infusing a compound that renders the β cells unresponsive during glucose infusion, the resultant steady-state glucose concentration “SSPG” is suggestive of insulin resistance. They reported that SSPG was higher in obesity, but highest in T2DM [18,19]. Thus, Reaven and colleagues provided persuasive evidence that insulin resistance existed concomitant with T2DM. However, the existence of insulin resistance did not rule out the possibility that β-cell deficiency could also play a role in the pathogenesis of T2DM. Also, SSPG, while indicative of apparent insulin resistance, does not represent a quantitative in vivo index of insulin sensitivity of the tissues. In addition, SSPG is highly dependent on utilization of glucose at high concentrations independent of the ambient insulin level (“glucose effectiveness,” cf. ). In fact, the debate regarding the relative importance of resistance versus insulin secretory defects in T2DM raged for several decades beyond the introduction of the pancreatic suppression test, which confirmed the insulin resistant state [21–25].
Qualitative versus quantitative
Often it is not sufficient to establish whether insulin resistance exists, relative to a normative population. Much more common is the need to measure insulin resistance, that is, to attach a number to it. A drawback of the qualitative resistance concept is that (as discussed earlier) virtually complete insensitivity to insulin can exist (“infinite insulin resistance”). It is much more practical to think in terms of “insulin sensitivity”, where the latter concept is the inverse of resistance. Thus, insulin sensitivity determined by whatever method may have a minimum value near zero (total resistance), and values may extend to a finite number representing great sensitivity to the hormone. The first quantitative method to measure insulin sensitivity was originated by the late Ruben Andres, and it is still called the “euglycemic glucose clamp” .