Comment on: Gale EAM (2007) To boldly go—or to go too boldly? The accelerator hypothesis revisited. Diabetologia 50:1571–1575—a reply to the editor

Abstract

To the Editor: The Editor’s erudite commentary [1] on the accelerator hypothesis and its impact on the classification of diabetes [2] points to where the hypothesis may be vulnerable, offers the opportunity for adjustment and sets an intellectual standard by which the hypothesis will stand or fall. The hypothesis is based on the construct of tempo [3], and envisages a hierarchy of genotypes that respond to insulin resistance with corresponding rates of beta cell loss. Accordingly, a rise in insulin resistance across a population would predict a leftward shift, resulting in younger age at onset and an increase in incidence, both fundamental tenets of the hypothesis because they spell acceleration. Constitution predisposes to accelerated apoptosis Gale’s expectation from the accelerator hypothesis is that ‘both forms of diabetes share a common genetic susceptibility to beta cell loss’. But this is to misconstrue the hypothesis (‘Type 1 and type 2 diabetes are the same disorder of insulin resistance set against different genetic backgrounds’). The hypothesis does not propose a common genetic soil, but rather a common environmental risk—insulin resistance—to which different genotypes respond in a hierarchical fashion. I have responded to this point elsewhere [4], but an analogy may help to clarify it. If the fish in a lake were ‘accelerator’ genes encoding susceptibility according to their size, only the widest of meshes (the least insulin resistance) would be needed to net the largest of them, i.e. those (HLA) genes associated with fast tempo diabetes (type 1). Smaller fish would not be selected by the trawl, and would not find themselves partitioned one side of a net designed to catch type 1 fish. These smaller fish would be trapped only with a narrower mesh, i.e. the genes associated with type 2 diabetes would be expressed only when insulin resistance was greater. I would therefore expect to find type 1 genes overexpressed in type 2 diabetes, but not type 2 genes, such as TCF7L2, in type 1 diabetes, as Field et al. observed [5]. FTO is not an accelerator gene, but a determinant of body mass that is over-represented in type 2 diabetes, where BMI surfeits are large, but not in type 1 [6] because, although corresponding surfeits are consistently reported [2], they are smaller. Insulin resistance is the principal cause of non-immune diabetes I recognise that insulin resistance and beta cell performance are inter-dependent functions in a non-linear loop, but the question at issue is whether the same insulin resistance that most would accept underpins type 2 diabetes also underpins type 1. The two examples in the literature which I cited to suggest that it might, both adjusted their measure of insulin resistance for beta cell reserve, as Gale requires. Both reported that those who went on to type 1 diabetes were more insulin resistant relative to, in the first instance, their identical twin who did not [7] and, in the second, those in the same randomised cohort who did not [8]. I think both satisfy Gale’s concern for ‘relative to what?’ Everybody is, by degrees, insulin deficient, and the hypothesis argues that the tempo of progression from health to diabetes moves along a continuum, not from one arbitrary category to another. Insulin sensitivity only becomes an important influence upon progression to hyperglycaemia once critical levels of Diabetologia (2007) 50:2604–2606 DOI 10.1007/s00125-007-0829-3