In This Issue of Diabetes Care

The effect of conventional lifestyle interventions on type 2 diabetes incidence varies according to prediabetes phenotype, according to Sathish et al. (p. 1903). Specifically, they found that there were significant risk reductions in individuals with impaired glucose tolerance (IGT) and in those with both IGT and impaired fasting glucose (IFG) but not in those with isolated IFG. The findings come from a reanalysis (meta-analysis) of individual participant data from four randomized controlled trials that recruited individuals with isolated IFG, with isolated IGT, or with both IFG and IGT. The authors looked at subsequent incidence of type 2 diabetes following exposure to either conventional lifestyle interventions for type 2 diabetes, usual care, or minimal intervention. Just under 2,800 participants were included in the analysis. The authors found that after a median of 2.5 years of follow-up, individuals with IFG alone had a hazard ratio (95% CI) for diabetes incidence of 0.97 (0.66, 1.44). The equivalent ratio in individuals with IGT alone was 0.65 (0.44, 0.96), and those with both had a ratio of 0.51 (0.38, 0.68). Based on this analysis, they conclude conventional lifestyle interventions reduce diabetes incidence in individuals with IGT (irrespective of the presence of IFG) but not in those with IFG alone. As for why there are differences in risk reduction between the phenotypes, the authors suggest it is due to variations in pathophysiological abnormalities. They point to clear differences in the phases of insulin secretion and the types of insulin resistance the groups have. On this basis, they suggest that different therapeutic interventions are needed for the groups to prevent progression to full type 2 diabetes. Commenting further, author Thirunavukkarasu Sathish said, “The novel findings of this study are likely to advance the field of precision prevention of type 2 diabetes, subsequently informing the diabetes prevention guidelines to consider recommending different lifestyle intervention strategies according to the prediabetes phenotypes.”

Diets with a high glycemic index (GI) and glycemic load (GL) appear to be associated with numerous changes in DNA methylation sites, with many occurring on genes known to affect metabolism, according to Ott et al. (p. 2067). Their work focused on children and adolescents and found that the level of association was much higher in individuals with overweight and potentially also explains the known health benefits of diets with low GI and GL. The findings come from an epigenome-wide association study and meta-analysis of six cohorts of children and adolescents who provided blood and dietary data (to calculate dietary GI and GL). The authors also used a stratified analysis of participants with either normal weight or obesity/overweight to investigate whether there were different epigenetic patterns according to weight. Just under 1,200 individuals were included in the analysis, with about 800 having normal weight and 400 having obesity. Additionally, they looked at gene expression in subcutaneous adipose tissue samples from more than 200 children. Overall, they found one methylation site associated with GL in the overall group and well over 500 sites in the BMI-stratified groups, with most being in the overweight/obesity group. In many cases, the sites were in genetic regions of regulatory factors and genes that may promote metabolic issues. The authors could only speculate as to why there were vastly more associations when the participants were stratified by BMI. However, they suggest that genetic and environmental selection factors are enriched in the stratified groups, leading to stronger associations between dietary GI and GL and DNA methylation. Commenting further, author Sandra Hummel said, “Our findings indicate that the molecular response to a high-GI, high-GL diet differs between children and adolescents with and without overweight or obesity, supporting the need for personalized nutritional intervention to improve long-term metabolic health.”

Hydroxychloroquine does not appear to prevent progression from stage 1 to stage 2 of type 1 diabetes, according to Libman et al. (p. 2035). However, the drug does show evidence of effects on the acquisition of autoantibodies and possibly on glucose levels. “This clinical study has provided some very provocative data about the pathologic mechanisms of early type 1 diabetes,” said author Kevan C. Herold. “Importantly, the hydroxychloroquine treatment showed biologic effects, but these immune activities were not sufficient to prevent disease progression.” The findings come from a randomized, placebo-controlled trial that included 273 individuals with type 1 diabetes at stage 1 (i.e., two autoantibodies present) in treatment with hydroxychloroquine (n = 183) or control (n = 90). The authors then assessed whether the drug could delay or prevent progression to stage 2 of the disease. The trial was stopped early, after 23.3 months, by the data safety monitoring board because of futility. Specifically, enrollment was terminated after a planned interim analysis indicated a lack of efficacy for the treatment in terms of the clinical outcomes. As an indication of the trial’s lack of efficacy, they found that 34 individuals in the hydroxychloroquine group and 17 in the placebo group had confirmed abnormal oral glucose tolerance test results. Additionally, the unadjusted hazard ratio for the time to confirmed abnormal oral glucose tolerance test or type 1 diabetes was 0.95 (95% CI 0.56–1.61). The authors note that the study coincided with the start of the coronavirus disease 2019 (COVID-19) pandemic, which they say heavily affected recruitment and study visits and likely affected the assessment of time to abnormal glucose tolerance. Nevertheless, the authors did find secondary effects that were in line with the postulated mechanisms of action. “The findings raise questions about the role of autoantibody-producing cells in the early stages,” added Herold. “These and other questions should be explored further as we develop effective and lasting preventative treatments for type 1 diabetes.”

A feature observed in aging, called clonal hematopoiesis of indeterminate potential (CHIP), may explain why type 2 diabetes rates increase with age, according to Tobias et al. (p. 1978). Significantly, they also identify key gene sets that link diabetes, coronary heart disease, and atherosclerosis with the suggestion that mechanistic studies can uncover a way to target CHIP as a therapeutic target for diabetes and cardiometabolic risks. CHIP involves fundamental changes to progenitor blood stem cells, with mutations appearing to significantly increase for individuals above the age of ∼65 years. Using individual participant data from six cohorts, the authors examined the presence of CHIP and incident type 2 diabetes in ∼17,600 participants over just under 10 years of follow-up. They also looked at the specific associations between incident type 2 diabetes and a series of variants, all previously linked to coronary heart disease. They found that CHIP prevalence was 6.0% (n = 1,055) in the overall set of participants at baseline, and just under 2,500 were diagnosed with type 2 diabetes over the study period. They then found that participants with CHIP had a 23% higher risk (hazard ratio 1.23 [95% CI 1.04, 1.45]) for type 2 diabetes compared with individuals without CHIP. They conclude that CHIP is likely a mediator of type 2 diabetes risk via atherosclerosis-related pathways and coronary heart disease–related mutations. As such, they call for mechanistic studies to identify the pathways, since there is a high likelihood that they will determine whether any pathways exist for therapy via CHIP. “It is unclear what causes some people to accumulate CHIP as they age, or how this might in turn contribute to higher risk of type 2 diabetes,” said author Deirdre K. Tobias. “Inflammation is a plausible pathway, but further research is needed to uncover other important mechanisms that link CHIP to aging-related deteriorations in glucose tolerance and diabetes risk.”