Data-driven Cluster Analysis Reveals Distinct Diabetes Subtypes in Black/African Americans in the United States

Despite considerable advances in understanding the pathophysiology and drug development, prevention and treatment of type 2 diabetes mellitus (T2DM) remain suboptimal, with substantial variations in individual responses to lifestyle and pharmacological interventions. T2DM is characterized by 2 primary pathophysiological factors: defective insulin secretion by pancreatic β cells and the inability of insulin-sensitive tissues to respond appropriately to insulin. Increasing evidence, however, indicates that T2DM is phenotypically heterogeneous, including several distinct subtypes of diabetes, each with unique pathophysiological characteristics and clinical presentations driven by genetic variation, lifestyle, and other environmental exposures (1). Understanding these subtypes is crucial, as it opens the door to more personalized treatment approaches that can improve patient outcomes and reduce the burden of complications associated with diabetes. Recent studies using data-driven cluster analysis have identified several subtypes of diabetes based on multiple etiological pathways such as age at diagnosis, β-cell function, body mass index, β-cell mass, insulin action, insulin sensitivity, fat distribution, and the presence or absence of glutamic acid decarboxylase (GAD) antibodies (2-4). Major subtypes include severe autoimmune diabetes, severe insulin-deficient diabetes (SIDD), severe insulin-resistant diabetes, mild obesity-related diabetes, and mild age-related diabetes (MARD) (2-5). These studies reported that individuals allocated to these clusters have different characteristics. The severe insulin-deficient diabetes cluster progresses quickly to insulin treatment and is more prone to retinopathy, while the severe insulin-resistant diabetes group is more susceptible to nephropathy. Although most studies used Scandinavian and North European patient cohorts or predominantly White populations in North America (2, 4, 6, 7), similar diabetes subtypes have been reported in studies from China, Mexico, India, Qatar, Thailand, Africa, Japan, and Korea. It should be noted that most studies have been carried out in relatively homogeneous populations, which often lack the racial and ethnic diversity needed to apply the findings to the increasing global population with diabetes. Lu et al (8) have reported a retrospective study using data-driven cluster analysis to identify distinct subtypes of diabetes and examine how racial background affects subtype distribution and characteristics among Black/African Americans in the Deep South of the United States. They included the 6 parameters reported in previous studies and found that the diabetes clusters in Black/African Americans are significantly different from that in White individuals. Among the identified diabetes subtypes, the study highlighted a higher prevalence of the more severe impairment of β-cell function (SIDD), a subtype with metabolic perturbations (higher glycated hemoglobin A1c [HbA1c] and lipid abnormalities), and a higher risk for complications, despite not being more obese or insulin resistant than other subtypes. The study also found that the risk of cardiovascular disease, chronic kidney disease, and metabolic-associated steatotic (fatty) liver disease was higher in Black/African American individuals with SIDD as compared to Whites. In contrast to SIDD, the MARD subgroup was characterized by milder disease and a lower odds ratio of developing complications compared to other clusters. By identifying these diabetes subtypes, the study underscores the need for targeted screening and intervention strategies that consider the specific metabolic profiles of Black/African Americans. This study is subject to certain limitations, including its retrospective design based on International Classification of Diseases codes. Of the 89 875 patients identified with diabetes over the study period, only 1207 (1.3% of the screened population) had comprehensive data covering all 6 required clinical parameters. Additionally, anti-GAD antibodies were tested in only a minority of cases. Other autoantibodies such as anti-insulin, anti–islet cell autoantigen, and anti-zinc, as well as indicators of insulin deficiency—such as reduced C-peptide levels, lower homeostatic model assessment of β-cell function (HOMA2-β) values, decreased triglyceride/high-density lipoprotein ratios, elevated HbA1c, increased ketone propensity, and reduced instances of obesity and insulin resistance—were not systematically documented. Traditional approaches to diabetes management apply a one-size-fits-all model, which focuses on broad categories of diabetes without accounting for the underlying heterogeneity within these groups. Identifying specific subtypes in racially and ethnically diverse populations by clustering underscores the importance of personalized approaches considering metabolic and genetic factors affecting diabetes subtypes. For example, patients with severe insulin-resistant diabetes may benefit more from interventions targeting insulin resistance, such as lifestyle modifications, weight loss, and insulin-sensitizing medications. Meanwhile, those with severe insulin-deficient diabetes may need more intensive insulin therapy and close monitoring of β-cell function. This information can help to guide efforts to tailor treatment to each subtype, leading to improved glycemic control, reduced risk of complications, and ultimately, enhanced quality of life for patients with diabetes. This is particularly important in addressing health disparities, as certain populations may be more predisposed to specific subtypes and their associated complications. Diabetes heterogeneity has considerable implications for diagnosing, managing, and preventing acute and chronic diabetic complications. Identifying distinct subtypes within diverse populations highlights the need for personalized treatment approaches that address each subtype's unique characteristics. As research continues to evolve, health-care providers and researchers must remain vigilant in refining diabetes classification and treatment, ensuring that all patients receive the most effective care possible based on their specific needs. P.K. is supported by the Association of British Clinical Diabetologists and the National Institute of Health and Care Research (research grant No. NIHR303671). G.E.U. is supported by the National Institutes of Health (NIH; research grant NATS UL 3UL1TR002378-05S2) from the Clinical and Translational Science Award program and the NIH and National Centre for Research Resources (NIH/National Institute of Diabetes and Digestive and Kidney Diseases No. 2P30DK111024-06) and has served as a member of advisory boards for Dexcom and GlyCare. I.C.R. declares no conflicts. glycated hemoglobin A1c mild age-related diabetes severe insulin-deficient diabetes type 2 diabetes mellitus