In This Issue of Diabetes

The interaction between cell division regulator MAD2, or mitosis arrest deficiency 2, and insulin receptors appears to control glucose metabolism in vivo, according to Park et al. (p. 1781). It also appears to establish a physiological function for insulin receptor trafficking in insulin and glucose homeostasis and demonstrates the importance of insulin receptor endocytosis in metabolic regulation. The authors show that genetic ablation of the interaction in mice delays insulin receptor endocytosis, increases insulin receptor signaling, and prolongs the action of insulin at plasma membranes. These effects resulted in a defect in insulin clearance with increasing levels of insulin and glucagon, modestly altering glucose metabolism. Further experiments showed that disruption of the insulin receptor–MAD2 interaction results in increases in serum fatty acid concentrations and hepatic fat accumulation and in distinct changes in metabolic and transcriptomic profiles in liver and adipose tissue samples. There was also evidence that disruption of the interaction enhanced lipogenesis, lipid transport, and lipolysis, all affecting levels of circulating fatty acids and hepatic fat accumulation. Prior to the study, it was far from clear what, if any, role insulin receptor endocytosis has in whole-body insulin sensitivity. Despite the findings, the authors are clear that further research will be needed, particularly on mechanisms of increased levels of glucagon upon disruption of the insulin receptor–MAD2 interaction and tissue-specific function of this interaction. “Our studies demonstrate that targeting insulin receptor endocytosis may be beneficial for treating diabetes,” said author Eunhee Choi. “Due to the difficulties with gene therapy at the present time, direct targeting of the MAD2-interacting motif in insulin receptors will be limited. However, we can design and investigate inhibitors that target the MAD2–insulin receptor interaction or the regulatory module of insulin receptor endocytosis, such as tyrosine phosphatase SHP2, which facilitates active insulin receptor endocytosis.”

A series of noninvasive retinal markers might help identify individuals with type 2 diabetes who are at risk of cognitive impairment, according to Pedersen et al. (p. 1853). Specifically, individuals with cognitive impairment had thinner macular retinal nerve fiber and macular ganglion layers and less nasal quadrant oxygen saturation compared with those without cognitive impairment. The findings come from the use of noninvasive retinal imaging and are based on the notion that the retina and brain share similar biological origins, with the retina potentially acting as a window to the brain. As such, the authors wanted to determine whether there were differences in retina-based markers in individuals with type 2 diabetes with and without mild cognitive impairment. After recruiting 134 individuals with type 2 diabetes, neuropsychological tests revealed that 28% had mild cognitive impairment. Using a series of noninvasive techniques, including fundus photography, optical coherence tomography, and retinal oximetry, the authors used multivariate cluster analysis to identify the changes associated with cognitive impairment. Notably, they also found no differences in a wider series of markers, such as retinal vessel density, and measures from optical coherence tomography–angiography. The authors conclude that the differences they identified might be useful as markers to identify individuals at risk of cognitive impairment to provide an early diagnosis and to manage modifiable risk factors via, for example, better metabolic control. However, in comparing the study findings with those of previous studies it is clear some inconsistencies remain, although most likely these inconsistencies are because of differences in study designs and confounding factors considered. They note that the cross-sectional design precluded them from examining causal relationships and that there is a risk of spurious relationships because of multiple testing. Nevertheless, they suggest that the findings warrant further investigations, including longitudinal studies, which are reportedly planned in a wider European Union–funded research project.

High glucose variability appears to be a key metric in determining the risk of neuropathy in children and adolescents with type 1 diabetes, according to Oberhauser et al. (p. 1835). Specifically, the metric, together with but independently of mean glucose levels, appears to be an unexpectedly strong predictor of slowed nerve conduction velocity. The findings come from a prospective study that included 53 children and adolescents aged 5–23 years with type 1 diabetes and 50 healthy children as control subjects. Various measures of glycemia, including continuous glucose monitoring (CGM), were then collected along with nerve conduction velocity in three motor nerves and one sensory nerve. The authors found that nerve conduction velocity in all four nerves was significantly reduced in an age- and height-matched subgroup of individuals with diabetes aged 10–16 years. The most affected was the peroneal nerve, with reduced nerve conduction velocity seen in all subgroups aged between 5 and 23 years. Height also reduces peroneal nerve conduction velocity, so after adjusting for height, the authors found it correlated negatively with HbA_1c and highly significantly with glucose variability. “We are pleased to have found that glucose variability is such an important preventable risk factor,” said lead author Sarah S. Oberhauser. “In the future, low glucose variability should be targeted, even if a slightly higher HbA_1c is accepted.” Given the findings, the authors advise that greater attention should be paid to glucose variability. Oberhauser added, “We plead for greater attention to parameters such as glucose standard deviation and the coefficient of variation for mean glucose to help contain glucose variability.” Looking to the future, senior author Philip J. Broser said, “We are further developing this project and now make extensive use of other CGM sensor data. While glucose variability appears important, we think the data recorded by the CGM devices will allow us to answer more complex and clinically relevant questions, such as diabetes control at night and the relationship between [physical] activity, glucose control, and neuropathy.”

A post hoc analysis of the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial by Okuno et al. (p. 1864) suggests that long-term, but not short-term, glycemic variability predicts risk for renal disease in type 2 diabetes. Chronic elevation of blood glucose is a recognized driver of renal disease risk in diabetes, but recently glycemic variability has also emerged as an additional contributing risk. Involving just under 4,000 participants, the analysis looked at the relationships between 1,5-anhydroglucitol (1,5-AG) plasma levels (i.e., short-term variability) and visit-to-visit fasting plasma glucose coefficient of variation (CV-FPG) (i.e., long-term variability) and micro- and macroalbuminuria end points. The authors report that high CV-FPG and low 1,5-AG were independently associated with microalbuminuria after adjustment, but only CV-FPG remained associated after additional adjustment for HbA_1c. For macroalbuminuria, CV-FPG was the only significant risk factor. Stratified analysis according to intensive or standard treatment groups (as included in the original trial design) showed that higher CV-FPG was more harmful in those receiving the less intensive standard treatment for glucose control. Specifically, the association between microalbuminuria and CV-FPG was significant in the standard glucose control group after full adjustment but not in the group receiving intensive glucose lowering. Based on the findings, the authors conclude there is potential value in the different measures of glycemic variability for determining renal complication risk. However, it seems that a single measurement of 1,5-AG is not sufficient to capture all renal risks, and that means that the longer test for glycemic variability is still needed. “In conjunction with our prior research on glycemic variability, we are of the belief that the management of diabetes and its complications should focus on avoiding both elevated glucose levels and prolonged glucose fluctuations,” said author Jin J. Zhou. “Although we utilized one-time 1,5-AG as an indicator of short-term glycemic variability, future metrics of glycemic variability—such as those stemming from continuous glucose monitoring—might warrant further evaluation regarding their ability to predict risk for diabetes complications.”