In This Issue of Diabetes Care

Undiagnosed diabetes in the U.S. is less common than previously thought, according to Fang et al. (p. 1994). Although overall rates of diabetes increased in the U.S. between the late 1980s and early 2020, the proportion of undiagnosed cases declined substantially, suggesting there have been major successes with diabetes screening and detection programs. According to the authors, undiagnosed diabetes now affects only between 1% and 2% of U.S. adults, with over 90% of all cases now successfully diagnosed. Nevertheless, they stress that undiagnosed diabetes is not a universal issue across the U.S. population. It mostly affects high-risk, underserved patients with poor access to health care, typically uninsured or underinsured individuals from racial and ethnic minorities, with a mix of overweight or obesity and low income. Complications of diabetes also varied considerably across the different groups affected by undiagnosed diabetes. Consequently, the authors suggest that addressing this remaining burden will require enhanced health care access for affected groups as well as the development of highly targeted screening programs for diabetes. The findings come from further analysis of the National Health and Nutrition Examination Survey (NHANES) and involved ~30,500 U.S. adults without diagnosed diabetes over the period 1988–2020. Specifically, the authors used clinical standards for diagnosing diabetes to detect potential cases of undiagnosed diabetes from the records. This required identification of elevated levels of both fasting plasma glucose and HbA1c. They found that between 1988 and 2020 there was an overall increase in the prevalence of diagnosed diabetes—4.6% in 1988 to 11.7% in 2011. In contrast, there were no real changes in the rates of undiagnosed diabetes according to the two different measures they used (i.e., from 2.23% to 2.53% or 1.10% to 1.23%). Translated, that means the proportion of all undiagnosed diabetes cases declined from either 32.8% to 17.8% or 19.3% to 9.5% over the study period, depending on the measure.

The incidence and prevalence of visual impairment due to diabetic retinopathy in Finland has fallen since the peak years of the mid-1990s, according to Purola et al. (p. 2020). This was despite there being a background of increasing prevalence of diabetes in the country since data were first collected in the early 1980s. The reasons for the improvements are likely multifactorial, according to the authors, but largely due to developments in medical care that have been implemented over the years. While they observed that severity of disease seems to have lessened and that age of onset has increased, they note that patients may still experience vision issues with diabetes. Thus, the study, and its encouraging trends, underlines the importance of timely screening and treatment of diabetic retinopathy and diabetes in general. The study, which is based on the Finnish Register of Visual Impairment, included ~2,500 cases of nonproliferative and ~2,000 cases of proliferative diabetic retinopathy (NPDR and PDR, respectively) over the period 1980–2019. Purola et al. found that incidence of both types of diabetic retinopathy increased in the 1980s, in line with increasing rates of diabetes in the population. However, by the mid-1990s, incidence peaked at 102.3 cases per 100,000 for NPDR and 39.9 for PDR. Rates then fell to 5.5 and 7.4 cases per 100,000, respectively, by 2019. They also note that the severity of disease has fallen and that age of onset has increased, in certain cases substantially so. In general terms, the authors point to the successful development of screening methods and therapies over the years for both diabetes in general and for diabetic retinopathy specifically. While clearly relevant to the population of Finland, the authors add that generalizability is limited, and thus they call for more population-based studies with long-term follow-up in other countries.

Between 1996 and 2015, risk of a first-time ischemic stroke halved in Danish patients with incident type 2 diabetes without atherosclerotic cardiovascular disease (ASCVD), according to Gyldenkerne et al. (p. 2144). The changing risk profile coincided with substantial increases in usage of prophylactic cardiovascular medicines, but as the authors point out, multiple factors were likely involved in bringing down the risk of stroke in patients with diabetes. The findings come from an analysis of several of Denmark’s nationwide health registries that the authors used to identify patients with incident type 2 diabetes but without ASCVD. They were then matched for age and sex with up to three control subjects from the general population and followed for 5 years to identify temporal trends in ischemic stroke and medication use. The authors identified 288,825 individuals with incident type 2 diabetes during the study period. Within this group, they found that 5-year risk for ischemic stroke approximately halved between 1996 (5.2%) and 2015 (2.7%). However, they also found that compared with the control subjects from the general population, individuals with diabetes still had higher risk for stroke, even though the risks declined over time. The authors examine medication usage over the study period and describe striking increases in statin use (5% to 50%) as well as substantial increases in multiple other cardiovascular-related medications. However, the reductions in stroke risk most likely reflect a combination of medical advances, according to the authors. In addition to improved prophylactic treatment, they point to improvements in awareness and screening for diabetes and wider use of HbA1c to diagnose diabetes. Nevertheless, they add that there is still likely room for further reductions in stroke incidence in type 2 diabetes, given that, for example, “only” 50% of patients received statins in the later study period.

The gut microbiome of individuals with long-standing type 1 diabetes (T1D) differs from that in those without diabetes, according to analysis by van Heck et al. (p. 2084). Specifically, they found that numerous bacterial taxa were significantly depleted or enriched in individuals with T1D compared with healthy control subjects. The authors stress that the study cannot assign causality because of its design, but many parameters of glycemic control and disease-related complications correlated with dynamics in microbial diversity and metabolism. The findings are based on an analysis of metagenomic shotgun sequencing data of fecal samples obtained from 238 individuals with an average of nearly 30 years of T1D and ~3,000 age-, sex-, and BMI-matched healthy individuals. Crucially, the authors also collected disease-related characteristics and any complications in a bid to look for correlations with risk factors and end points. While a measure of microbiome diversity per se did not show any difference, there was a slight increase in the diversity of microbial biochemical pathways in the group with diabetes. More significantly, the authors report 43 bacterial taxa that were significantly depleted in T1D, while 37 were significantly enriched. For context, the microbiome in the human gut is thought to consist of many thousands of microbial species that are divided among many other taxa and families of bacteria. Average blood glucose levels in the form of HbA1c and disease duration explained a significant amount of the variation in the microbiome data. HbA1c levels also explained the abundance of several microbial species and pathways in the two groups that previous studies identified with health or disease status. Also, macrovascular and microvascular complications explained part of the variation in the microbiome, in particular nephropathy and macrovascular complications. The authors write, “Overall, our study provides evidence for a difference in gut microbiome composition in participants with long-standing type 1 diabetes compared with healthy control subjects that associates with glycemic control and disease-related complications.”