insulin secretion can be measured accurately in humans with intact renal function [34, 35].
The other factor when assessing insulin secretion is the nature of the stimulus – a physiological stimulus mimicking situations similar to everyday life e.g. an oral glucose tolerance test or a mixed meal tolerance test may be better at detecting subtle defects in insulin secretion as compared to supraphysiologic stimuli. The latter use stimuli intended to produce submaximal insulin secretion e.g. glucagon or arginine stimulation tests [36–38]. Similarly, an intravenous bolus of glucose that can change peripheral concentrations of glucose very rapidly (as part of an intravenous glucose tolerance test – IVGTT) is also a supraphysiologic stimulus to β‐cell secretion. Intravenous glucose challenges produce a characteristic biphasic pattern of insulin secretion in healthy subjects that is not observed in people with early type 1 diabetes or in people with type 2 diabetes [39].
While isolated measurement of insulin secretion in response to a standardized stimulus might provide a qualitative measurement of β‐cell function, truly quantitative measures require expressing insulin secretion as a function of insulin action. Such indices generated through mathematical modelling can predict risk of progression to diabetes or quantify response to an intervention. However, at present they have limited utility on an individual basis due to lack of sufficient normative data [6]. Nevertheless, absent or near absent C‐peptide concentrations in the fasting state or in response to stimuli may help demonstrate an absence of endogenous insulin secretion.
Autoantibodies
Several autoantibodies can be used in clinical practice to document the presence of islet autoimmunity. These are antibodies directed against Glutamic Acid Decarboxylase (GAD‐65), insulin, tyrosine phosphatases, insulinoma‐associated protein 2 (IA‐2 and IA‐2β), and the zinc transporter (ZnT8). The greater the number of autoantibodies that are positive and the higher the titer of these antibodies, the more likely that islet autoimmunity is present [40]. However, people with a clinical presentation compatible with type 1 diabetes are often antibody negative. The opposite sometimes applies. Indeed, in a cohort of Belgian patients with type 1 diabetes presenting before age 40, 24% did not have autoantibodies [41]. In the Botnia study, 4.4% of healthy individuals had GAD antibodies [42].
Latent autoimmune diabetes of adults (LADA) likely represents a “forme fruste” of type 1 diabetes where a lower genetic burden of predisposition results in a later age at presentation, decreased amount and degree of autoantibody positivity, and a longer persistence of endogenous insulin secretion. Patients with LADA tend to have a lower BMI than patients with type 2 diabetes. Of note, antibodies against insulin and IA‐2 are rarely present in late‐onset diabetes. Islet autoantibodies are most useful in screening family members of patients with type 1 diabetes for risk of developing the disease [43].
“Atypical” diabetes, glucose toxicity and the honeymoon period
Atypical diabetes or “ketosis‐prone type 2 diabetes” were monikers used to describe obese minority (typically African‐American) patients presenting with diabetic ketoacidosis as their first manifestation of diabetes [44]. Typically, after acute metabolic control is achieved, such patients are able to maintain glycemic control over long periods of time without insulin (often with diet alone). The concept of glucose toxicity where hyperglycemia per se impairs insulin secretion and action has been put forward to explain these observations [45]. Indeed, during recovery, these indices do not differ from those of age‐ and weight‐matched controls [46]. Similar short‐term exposure to lipotoxicity also has little effect on β‐cell function [47].
Short‐term intensive insulin therapy in a cohort of Chinese patients with a first presentation of diabetes produced a “remission” of diabetes with preservation of insulin secretion at one year when compared to conventionally treated patients [48]. Diabetes remission, independent of weight‐loss, is also reported after bariatric surgery but – at least in the short‐term is likely explained by caloric restriction and β‐cell “rest” [49]. This is in keeping with the concept that insulin over‐production may produce endoplasmic reticulum stress and if unchecked lead to β‐cell death [50].
These factors probably explain the resumption of endogenous insulin secretion and decreased or absent exogenous insulin requirements during the honeymoon period in type 1 diabetes. The term refers to a variable interval after presentation with hyperglycemia and subsequent restoration of metabolic and glycemic control. Some patients have a prolonged remission and indeed tight glycemic control seems to prolong the persistence of endogenous insulin secretion in affected patients.
Cystic fibrosis‐related diabetes
Diabetes is the most common extra‐pulmonary complication of cystic fibrosis and rises with age so that 50% of affected patients over age 30 have diabetes [51]. There is a recognized abnormality in insulin secretion which is likely explained by the actions of the Cystic Fibrosis Transmembrane Regulator (CFTR) on β‐cell electrophysiology. Drugs that improve CFTR function also have salutary effects on insulin secretion [52]. Exocrine pancreatic insufficiency may decrease intraluminal release of nutrients that stimulate incretin hormone secretion and delay gastric emptying thereby contributing to postprandial hyperglycemia. Endocrine dysfunction is also strongly correlated with destruction of the exocrine pancreas.
Nutritional deficiency and also the systemic inflammation associated with cystic fibrosis likely impair insulin action which is also a feature of Cystic Fibrosis‐related Diabetes. Macrovascular complications are uncommon and microvascular complications occur at lower rates than they do in “classical” type 1 and type 2 diabetes [53].
Novel approaches to reclassifying diabetes
Ahlqvist et al. undertook a data‐driven cluster analysis in patients with newly diagnosed diabetes from a Swedish cohort. They utilized six variables measured at the time of diagnosis – age at diagnosis, BMI, HbA1c, GAD antibodies, and homoeostatic model assessment (HOMA) estimates of β‐cell function and insulin resistance. The investigators identified five clusters based on these variables – one cluster was antibody positive, and relatively lean with poor metabolic control. The second cluster differed from the first in that antibodies were absent, while the third cluster had an elevated BMI and insulin resistance. Cluster 4 was also obese but defects in insulin action as measured by HOMA were not as severe as in the 3rd cluster. Finally, patients in the 5th cluster were older but, like cluster 4, had less severe metabolic abnormalities. These clusters were subsequently validated in 3 independent cohorts. Over subsequent follow‐up, the incidence of diabetic ketoacidosis and of non‐alcoholic fatty liver disease were highest in the first and third cluster, respectively. This latter cluster also had the highest risk of chronic kidney disease [54]. HOMA measures have significant limitations and poor predictive value for progression to diabetes [55] – although in fairness the investigators used C‐peptide rather than insulin to estimate β‐cell function. Whether this will help guide therapeutic interventions and improve outcomes in individual patients remains to be ascertained.
Another approach to characterizing different sub‐types of diabetes is based on common genetic variation and its influence on quantitative traits such as glucose tolerance in response to a standardized challenge. For example, Dimas et al. identified 5 clusters of genetic risk loci that altered glycemic traits [56]. In this study examining data from ~58 000 non‐diabetic subjects, one cluster altered insulin sensitivity. A second cluster altered fasting glucose while a third cluster altered the ratio of proinsulin to insulin in the fasting state. Another cluster was primarily characterized by defects in post‐challenge insulin secretion. The final cluster of risk loci did not alter glycemic traits. While this exercise has certainly helped understand the effect of genotype on phenotype, the effect size of each risk allele on a given phenotypic trait is so small as be unhelpful in terms of predicting individual clinical behavior.
Pharmacogenomics
