Accelerator hypothesis

Type 1 diabetes with onset in childhood may influence height and weight trajectories in patients around and after diagnosis. Patients with type 1 diabetes are known to display some degree of insulin resistance, although probably different in nature from that among patients with type 2 diabetes [1]. The role of obesity and insulin resistance as risk factors for subsequent type 1 diabetes is less established. It has been proposed that weight gain explains the increase in incidence of type 1 diabetes [2][3].

The accelerator hypothesis

As originally proposed in 2001[2] (p. 914-915):

“The ‘Accelerator Hypothesis’ argues that Type 1 and Type 2 diabetes are one and the same, distinguished only by their rate of beta cell loss and the accelerators responsible. The first accelerator, a constitutionally (intrinsically) high rate of beta-cell apoptosis, is necessary for diabetes to develop but in itself rarely sufficient to cause it. Insulin resistance, the second accelerator, results from weight gain and physical inactivity which further increases the rate of beta-cell apoptosis and accounts for the rising incidence of Type 1 as well as Type 2 diabetes in industrially developed societies. Finally, a small and genetically defined subset of patients with both intrinsic lesion and insulin resistance develops beta-cell autoimmunity, the third accelerator.”

It is understood that by constitutionally, and intrinsic lesion, Wilkins refers to genetic susceptibility, in particularly carrying at least one of the established HLA class II haplotypes (“DQ8 and /or DQ2”) associated with increased risk of type 1 diabetes. For recent reviews by Wilkin, including also discussions of potential mechanisms involved, see e.g. [4][5]. Not surprisingly, the “accelerator hypothesis” has attracted controversy in the type 1 diabetes community [6][7].

Is type 1 diabetes and type 2 diabetes the same?

It is clear that in some patients, particularly in young adults, it may be difficult to distinguish between type 1 diabetes and type 2 diabetes, but the general statement that type 1- and type 2 diabetes is highly controversial. Discussions of this complex question soon wander into philosophy and subjective judgments. Most clinicians still consider it advantageous to split type 1 diabetes and type 2 diabetes. For the purpose of this review, we will leave it at that, and rather review other more specific aspects of the hypothesis below.

Figure  1
Figure 1
Insulin resistance: role in initiation of islet autoimmunity It has long been considered plausible that insulin resistance may contribute to the precipitation of clinical disease [8] (Figure 1).

According to the accelerator hypothesis [2], islet autoimmunity is viewed as a response to increased beta-cell apoptosis induced by insulin resistance (and other factors such as genetic susceptibility). The mechanisms involved in both views are complex, and here, we will rather examine the available epidemiological evidence supporting the contention that insulin resistance or obesity predicts increased risk of islet autoimmunity (figure). In the German BABYDIAB-study there was no association between estimated insulin resistance or BMI and islet autoimmunity [9]. In an independent, cross-sectional study, Raab reported no association between insulin resistance and islet autoantibodies [10]. Although measures of insulin resistance were not done, BMI or weight gain was not associated with islet autoimmunity in the DAISY study, which rather suggested an association with height growth velocity [11]. An analysis of the smaller Australian BABYDIAB-study, suggested a relation of weight gain in infancy and islet autoimmunity [12]. A randomized trial of an intervention to reduce weight gain or insulin resistance in subjects without detectable islet autoantibodies at baseline to test whether this would influence the risk of islet autoimmunity would be the final test of this hypothesis, but this would be a major undertaking. On the other hand, such a trial carried out in subjects already positive for islet autoantibodies attempting to halt progression to clinical diabetes would not be informative regarding this aspect of the accelerator hypothesis.

Insulin resistance: role in progression from islet autoimmunity

Other published observational studies of the role of insulin resistance as a risk factor for type 1 diabetes have all investigated whether insulin resistance predicted the rate of progression clinical type 1 diabetes in those already positive for islet autoantibodies [13][14][15][16][17]. While insulin resistance predicts risk of type 2 diabetes, a reduced first phase insulin response (FPIR) to an intra-venous glucose tolerance test (IVGTT) is an established risk factor for rapid progression to type 1 diabetes in high risk children [18]. In studies of type 1 diabetes, estimated insulin resistance was in most instances either not different, or only borderline significantly different in progressors compared to non-progressors to type 1 diabetes [13] [14] [16]. In most studies, variation in FPIR was a much better predictor of progression to type 1 diabetes than was insulin resistance (an exception was DPT-1 [15], but later analyses from the same cohort suggested that HOMA-IR added little to the prognostic model on top of FPIR and other measures [19]). Consequently, a higher ratio of insulin resistance to FPIR (consistent with a low FPIR) also predicted progression to type 1 diabetes in these studies. Most of studies did not find any predictive value of BMI in progression from islet autoimmunity to type 1 diabetes [11] [13] [15] [16]. BMI was an independent predictor in the DIPP study [17], but the authors concluded that the magnitude of association was minor.

In some studies, few subjects with type 1 diabetes, sensitivity of the results to adjustment for other variables (in some instances the components of ratios and indices), potential selection bias, and lack of repeated measures of insulin resistance during follow-up should call for some caution in the interpretation of the consistency of the results. Variability in age and stage of disease progression may also complicate the interpretation [18].

Childhood weight, growth or obesity and subsequent risk of type 1 diabetes

According to a systematic review of literature up to February 2010, children developing type 1 diabetes were on average taller, heavier or gained more weight or height prior to diagnosis compared to their peers [20], although there was substantial variation in age at anthropometric measurement and other methodological problems. From a pooled analysis of four partially comparable studies, it was estimated that childhood obesity was associated with an approximately doubling in the risk of type 1 diabetes. Svensson et al, who found one of the strongest relations, concluded that this could not explain the changing incidence of type 1 diabetes in Denmark [21]. In a nested case-control study within the DiPiS-study (not included in the systematic review), postnatal growth in body size in 58 children who developed type 1 diabetes was compared with that in 51 HLA matched controls and 104 non-HLA matched controls. BMI-SDS in the first 18 months of life was not significantly different, but linear growth was higher in children who developed type 1 diabetes [22]. A few studies have reported a relation between BMI and younger age at diagnosis of type 1 diabetes, but these are susceptible to bias (or else, why would anyone bother to collect data from controls, let alone to run large cohort studies?).

While most of the case-control studies were able to minimise the risk of information bias (recall bias) by using growth charts, the necessary inclusion of population controls bears a high risk of selection bias. These risks are minimised by the case-sibling control design [23], or by the utilising cohort design. In a cohort analysis of nearly 40,000 Norwegians aged 18 or above, BMI in 1984-6 among was not a significant predictor of type 1 diabetes approximately 11 years later, but there were only 18 cases of type 1 diabetes [24]. In the British 1970 birth cohort, BMI at age 10 years was associated with a higher risk of self-reported type 1 diabetes (n=47) at age 10-30 years [25]. In a Tasmanian study of infants at high risk for sudden infant death syndrome (selected, for instance to have a large proportion of low birth weight), BMI, weight gain or skin folds at age five weeks were not predictive of type 1 diabetes (n=26), although mid upper arm circumference was [26].

Two recent, large scale cohort studies found that for each standard deviation (approximately 1 kg) increase in weight 12 months or weight gain from birth to 12 months of age, the risk of type 1 diabetes [27] or multiple islet autoantibodies [28] increased by approximately 20 percent. Although apparently consistent, the relatively weak association is unlikely to have direct clinical relevance at the individual level. It is still possible that infant or early life weight or growth is a surrogate marker for some other factor or process that is more directly related to type 1 diabetes pathogenesis.

Finally, a covariation in time trends in the prevalence of childhood obesity and incidence of type 1 diabetes is of some interest [29], particularly when this occurs both during periods of rise and fall. Some recent data suggest a plateau in both BMI or obesity and incidence of type 1 diabetes. However, questions remains regarding representativeness of the obesity studies and potential time-lags in effects, and the fact that obesity and type 1 diabetes is measured in different individuals makes such observations fundamentally susceptible to ecological bias.


In summary, there currently is little or no support for the contention that insulin resistance is a primary driver of islet autoimmunity or has a major role in progression from islet autoimmunity to clinical type 1 diabetes. Several studies suggest a role of childhood obesity or growth as a risk factor for later type 1 diabetes, but the Association seems weak and it remains unclear if it is causal. Because of the potential importance for prevention, more high quality studies in the field are warranted.


  1. ^ Cleland SJ et al. Insulin resistance in type 1 what is 'double diabetes' and what are the risks? Diabetologia 2013;1462-1470.

  2. ^ Wilkin TJ. The accelerator hypothesis: weight gain as the missing link between Type I and Type II diabetes. Diabetologia 2001;44:914-922.

  3. ^ Dahlquist G. Can we slow the rising incidence of childhood-onset autoimmune diabetes? The overload hypothesis. Diabetologia 2006;49:20-24.

  4. ^ Wilkin TJ. The convergence of type 1 and type 2 diabetes in childhood: the accelerator hypothesis. Pediatr Diabetes 2012;13:334-339.

  5. ^ Wilkin TJ. Is autoimmunity or insulin resistance the primary driver of type 1 diabetes? Curr Diab Rep 2013;13:651-656.

  6. ^ Gale EA. To boldly go-or to go too boldly? The accelerator hypothesis revisited. Diabetologia 2007;50:1571-1575.

  7. ^ Rewers M. The fallacy of reduction. Pediatr Diabetes 2012;13:340-343.

  8. ^ Leslie RD, Taylor R, Pozzilli P. The role of insulin resistance in the natural history of type 1 diabetes. Diabet Med 1997;14:327-331.

  9. ^ Winkler C et al. Is islet autoimmunity related to insulin sensitivity or body weight in children of parents with type 1 diabetes? Diabetologia 2009;2072-2078.

  10. ^ Raab J et al. Continuous rise of insulin resistance before and after the onset of puberty in children at increased risk for type 1 diabetes - a cross-sectional analysis. Diabetes Metab Res Rev 2013;631-635.

  11. ^ Lamb MM et al. Height growth velocity, islet autoimmunity and type 1 diabetes the Diabetes Autoimmunity Study in the Young. Diabetologia 2009;2064-2071.

  12. ^ Couper JJ et al. Weight gain in early life predicts risk of islet autoimmunity in children with a first-degree relative with type 1 diabetes. Diabetes Care 2009;94-99.

  13. ^ Fourlanos S et al. Insulin resistance is a risk factor for progression to type 1 diabetes. Diabetologia 2004;1661-1667.

  14. ^ Mrena S et al. Models for predicting type 1 diabetes in siblings of affected children. Diabetes Care 2006;662-667.

  15. ^ Xu P et al. Role of insulin resistance in predicting progression to type 1 diabetes. Diabetes Care 2007;2314-2320.

  16. ^ Bingley PJ et al. Insulin resistance and progression to type 1 diabetes in the European Nicotinamide Diabetes Intervention Trial (ENDIT). Diabetes Care 2008;146-150.

  17. ^ Siljander HT et al. Insulin secretion and sensitivity in the prediction of type 1 diabetes in children with advanced beta-cell autoimmunity. Eur J Endocrinol 2013;479-485.

  18. ^ Williams AJ, Long AE. Following the fate of the failing b-cell: new insights from first-phase insulin responses. Diabetes 2013;62:3990-3992.

  19. ^ Xu P et al. Prognostic accuracy of immunologic and metabolic markers for type 1 diabetes in a high-risk receiver operating characteristic analysis. Diabetes Care 2012;1975-1980.

  20. ^ Verbeeten KC et al. Association between childhood obesity and subsequent Type 1 a systematic review and meta-analysis. Diabet Med 2011;10-18.

  21. ^ Svensson J et al. Growth in the first year of life and the risk of type 1 diabetes in a Danish population. Paediatr Perinat Epidemiol 2007;44-48.

  22. ^ Larsson HE et al. Children developing type 1 diabetes before 6 years of age have increased linear growth independent of HLA genotypes. Diabetologia 2008;1623-1630.

  23. ^ Kharagjitsingh AV et al. Revisiting infant growth prior to childhood onset type 1 diabetes. Clin Endocrinol (Oxf) 2010;620-624.

  24. ^ Carlsson S et al. Age, overweight and physical inactivity increase the risk of latent autoimmune diabetes in results from the Nord-Trøndelag health study. Diabetologia 2007;55-58.

  25. ^ Viner RM et al. Childhood body mass index (BMI), breastfeeding and risk of Type 1 findings from a longitudinal national birth cohort. Diabet Med 2008;1056-1061.

  26. ^ Ponsonby AL et al. Infant anthropometry, early life infection, and subsequent risk of type 1 diabetes a prospective birth cohort study. Pediatr Diabetes 2011;313-321.

  27. ^ Magnus Maria C.. Infant Growth and Risk of Childhood-Onset Type 1 Diabetes in Children From 2 Scandinavian Birth Cohorts. JAMA Pediatrics, JAMA Pediatr. American Medical Association (AMA); 169(12):e153759. Available from:

  28. ^ Elding Larsson Helena. Growth and risk for islet autoimmunity and progression to type 1 diabetes in early childhood: The Environmental Determinants of Diabetes in the Young Study. Diabetes. American Diabetes Association; :db151180. Available from:

  29. ^ Knip M et al. Does the secular increase in body mass in children contribute to the increasing incidence of type 1 diabetes? Pediatr Diabetes 2008;46-49.


Nobody has commented on this article

Commenting is only available for registered Diapedia users. Please log in or register first.