The incretin effect in health
The idea that gastrointestinal factors contribute to the control of postprandial glucose regulation dates back to the beginning of the twentieth century, when Moore and colleagues reported reductions in glucosuria after the oral administration of gut extracts xin patients with juvenile diabetes . Even though it is questionable whether these glucose-lowering effects were really attributable to the incretin activity of the extract (which is unlikely, because most gastrointestinal peptide hormones are inactivated by the gastric acid), this report can be considered as the first description of an incretin-like effect. The term “incretin” was coined in 1932 by La Barre to describe a putative substance from the upper gut mucosa, which lowers glucose without activating exocrine secretion [2,3]. At this early time period, various reports about glucose-lowering activities of gut extracts had appeared. However, probably owing to the variations in the purity of these extracts, the results of these experiments were quite heterogeneous, and the idea that gut hormones contribute to glucose regulation was not explored further until the 1960s .
The development of the first radioimmunoassay for insulin in 1962 by Berson and Yallow opened up a new era in metabolic research. Elrick and colleagues  and McIntyre and colleagues independently reported that oral administration of glucose led to significantly higher increments in insulin secretion than intravenous glucose administration , thereby suggesting the existence of gut-derived factors that augmented insulin secretion, that is, an “incretin effect” (Figure 11.1). While these initial studies had therefore already demonstrated some incretin activity in humans, it was still difficult to quantify the respective contribution of this effect to the overall postprandial rise in insulin levels, because the circulating plasma glucose concentrations after i.v. glucose injection significantly exceeded those after oral glucose administration, thereby confounding direct comparisons of the respective insulin concentrations. Therefore, Nauck and colleagues developed the “isoglycemic clamp technique” to quantify the contribution of the incretin effect  (Figure 11.2). In these experiments, glucose was first administered by mouth, and the plasma glucose rises were recorded at 5-min intervals. On a second day, glucose was administered intravenously, and the respective glucose infusion rates were varied to exactly copy the glucose concentration pattern of the first experimental day. By these means, it was possible to compare the rises in insulin and C-peptide concentrations at identical plasma glucose level . The incretin effect was then calculated by the equation based on the integrated incremental responses (AUCs) of insulin or C-peptide concentrations or of insulin secretion rates derived by deconvolution of C-peptide concentration-time curves.
Based on these studies, it was estimated that the incretin effect contributed between 27.6% and 62.9% to the overall increments in C-peptide levels after oral glucose ingestion . It was also demonstrated that the “size” (i.e., the percentage contribution) of the incretin effect increased with greater amounts of glucose being administered . Because insulin is subject to considerable first-pass clearance by the liver, the C-peptide-based calculation was considered to be more accurate than the calculation using insulin concentrations. In fact, hepatic insulin clearance is significantly reduced by oral, but not by intravenous glucose administration, meaning that calculations based on insulin levels might lead to an overestimation of the incretin effect.
While augmentation of insulin secretion is clearly the most prominent action of the incretin hormones, there also appears to be an effect on glucagon secretion (Figure 11.1). Thus, when glucagon levels after oral and isoglycemic intravenous glucose administration were directly compared, the suppression of glucagon levels was found to be greater after intravenous compared to oral glucose administration . This finding is surprising given the glucagonostatic actions of GLP-1 . However, it appears that these glucagon-lowering effects of GLP-1 are outweighed by the glucagonotropic actions of GIP and potentially GLP-2 . Indeed, when GIP was exogenously co-infused along with GLP-1, the GLP-1-induced suppression of glucagon was completely abolished . Therefore, the net incretin effect on glucagon levels appears to be a modest stimulation of α-cell secretion under normal conditions.
Although the quantification of the incretin effect based on the isoglycemic clamp technique has been accepted as a reference standard for most subsequent studies, a couple of methodological caveats need to borne in mind. Thus, matching circulating plasma glucose concentrations with oral and intravenous glucose administration inevitably results in differences in the amount of glucose being administered on both occasions. Thus, typically ∼20 – 25 g of i.v. glucose were required to reach isoglycemia with 50g of glucose ingested orally . One might therefore argue that the study design could overestimate the incretin effect due to the different amounts of glucose being administered . Along this line, another argument to consider is the fact that the oral glucose load engages pancreatic β cells through various mechanisms (direct stimulation by circulating glucose, afferent neutral signals, and humoral stimulation by incretin hormones), whereas i.v. glucose represents a single β-cell stimulus. It seems conceivable that under conditions of a general impairment of β-cell mass or function the relative impairment in insulin secretion might be greater with the large combined stimulus (oral glucose) than with the smaller stimulus (i.v. glucose) . It is also important to bear in mind that the estimation of the incretin effect based on isoglycemic clamp experiments is not necessarily representative of the postprandial situation, because the potential contributions of the fat and protein components of a typical mixed meal and other gastrointestinal hormones with insulinotropic effects released by these food components are not taken into account. In line with this, studies in humans have demonstrated significant incretin effect after oral fat or amino acid ingestion as well [12,13]. Finally, variations in the rate of gastric emptying may alter the patterns of insulin secretion after oral glucose ingestion, independent of incretin activity.
Other groups have therefore applied other techniques, such as a relative comparison of insulin secretion after oral glucose ingestion during fixed hyperglycemia at 8 and 10.5 mmol L−1 in a single experiment . However, such study design does not take into account the typical postprandial changes in plasma glucose concentrations.
Taken together, the quantification of the incretin effect is complex and can be potentially confounded by various factors. Despite various limitations, the isoglycemic clamp technique still seems to represent the most appropriate measure of the incretin effect.