Beyond HbA1c: How Diabetes Subtyping Predicts Comorbidities and Mortality Over 14 Years

Highlights

• Severe Insulin-Resistant Diabetes (SIRD) and Severe Insulin-Deficient Diabetes (SIDD) subgroups face a nearly three-fold increase in the risk of kidney failure compared to mild age-related diabetes.
• The SIRD subgroup presents with the highest baseline prevalence of hypertension, dyslipidemia, and steatotic liver disease, despite having lower HbA1c levels than autoimmune groups.
• Mortality risk is significantly elevated in SIDD, SIRD, and Mild Obesity-related Diabetes (MOD) subgroups, largely driven by cardiovascular events, persisting even after adjustment for traditional risk factors.
• Subgroup-based classification at the time of diagnosis provides a superior prognostic framework for predicting retinopathy, neuropathy, and heart failure than current glucose-centric models.

Introduction: The Shift Toward Precision Diabetology

For decades, the clinical management of adult-onset diabetes has been dominated by a glucose-centric paradigm. While hemoglobin A1c (HbA1c) remains the gold standard for monitoring glycemic control, it often fails to capture the underlying pathophysiological heterogeneity of the disease. This heterogeneity explains why two patients with identical HbA1c levels may have vastly different trajectories regarding renal failure, cardiovascular events, and mortality. Recent advancements in machine learning and clinical phenotyping have led to the proposal of five distinct diabetes subgroups: Severe Autoimmune Diabetes (SAID), Severe Insulin-Deficient Diabetes (SIDD), Severe Insulin-Resistant Diabetes (SIRD), Mild Obesity-Related Diabetes (MOD), and Mild Age-Related Diabetes (MARD). A new prospective cohort study from Sweden, tracking over 19,000 individuals for up to 14 years, provides definitive evidence that these subgroups are not merely academic constructs but essential tools for predicting long-term outcomes and guiding early intervention.

The Swedish ANDIS Cohort: Study Design and Methodology

The study utilized data from the All New Diabetics in Scania (ANDIS) project, a comprehensive registry of individuals newly diagnosed with diabetes in southern Sweden. The researchers analyzed two distinct subsets: ANDIS1 (enrolled 2008–2016) and ANDIS2 (enrolled 2016–2022). The inclusion criteria were rigorous: participants had to be 18 years or older, diagnosed within one year of enrollment, and possess complete data for the six variables used for clustering.

Clustering Methodology and Participant Characteristics

The classification system utilized six key clinical variables collected at diagnosis: glutamate decarboxylase antibodies (GADA), age at diagnosis, BMI, HbA1c, and the Homeostasis Model Assessment 2 (HOMA2) indices for beta-cell function (B) and insulin resistance (S). Based on these, participants were assigned to one of the five clusters:

• SAID: Characterized by the presence of GADA (essentially Type 1 and LADA).
• SIDD: GADA-negative, characterized by low insulin secretion and high HbA1c.
• SIRD: Characterized by severe insulin resistance and high BMI.
• MOD: Characterized by obesity but relatively mild metabolic derangement.
• MARD: The largest group, characterized by older age at onset and a milder clinical course.

Out of 25,590 individuals screened, 19,076 were included in the final analysis. The median follow-up in the ANDIS1 cohort was nearly a decade, providing a robust window into the natural history of these disease subtypes.

Long-Term Outcomes: A Deep Dive into the Findings

The core of the study’s findings lies in the divergence of comorbidity risks across clusters. Using MARD as the reference comparator, the researchers identified specific patterns of organ damage associated with each phenotype.

Microvascular Complications: The Predominance of SAID and SIDD

As expected, the subgroups with the highest glycemic burden—SAID and SIDD—exhibited the highest risk for microvascular complications. The risk of diabetic retinopathy was significantly higher in SAID (adjusted HR 1.35) and SIDD (adjusted HR 2.11). Similarly, the risk of neuropathy was more than doubled in these two groups. These findings reinforce the traditional understanding that chronic hyperglycemia is the primary driver of small-vessel damage. However, the study also revealed that SIDD patients had a higher risk of stroke (adjusted HR 1.32), a macrovascular complication, suggesting that insulin deficiency itself may contribute to cerebrovascular fragility.

Macrovascular and Renal Risks: The Heavy Burden of SIRD

Perhaps the most clinically significant findings involve the SIRD subgroup. At the time of diagnosis, individuals in the SIRD group already carried a heavy burden of comorbidities. Nearly 70% had hypertension, and 44% had dyslipidemia. Furthermore, SIRD was strongly associated with steatotic liver disease and early-stage kidney disease. Over the follow-up period, SIRD patients faced a dramatically increased risk of incident kidney failure (adjusted HR 3.41) and heart failure (adjusted HR 1.55). Crucially, these risks were present despite the fact that SIRD patients often maintain lower HbA1c levels than those in the SAID or SIDD groups. This suggests that the metabolic syndrome components—hyperinsulinemia and lipotoxicity—drive organ damage independently of blood glucose levels.

Mortality Trends: A Sobering Outlook

The study found that SIDD, SIRD, and MOD were all associated with an increased risk of total mortality (adjusted HRs ranging from 1.44 to 1.52). Even after adjusting for traditional risk factors such as smoking, LDL cholesterol, and blood pressure, these subgroups remained at higher risk. The primary driver of this excess mortality was cardiovascular disease. Interestingly, while the SAID group has a high risk of microvascular issues, their mortality risk was not significantly higher than the MARD group, potentially due to closer monitoring or different underlying vascular pathology.

Clinical Implications and Expert Commentary

The results of this study suggest that a “one size fits all” approach to diabetes management is no longer tenable. The identification of SIRD as a high-risk group is particularly vital. Because these patients may not present with extreme hyperglycemia, they may be overlooked in traditional risk assessments that prioritize HbA1c.

Redefining Risk: The Case for SIRD Identification

Clinicians should consider that patients with the SIRD phenotype require aggressive management of non-glycemic factors. Early use of SGLT2 inhibitors and GLP-1 receptor agonists—which have proven benefits in renal protection and heart failure—should likely be prioritized in this group, even if their HbA1c is near target. Conversely, patients in the SIDD group need early and intensive insulin therapy to prevent the rapid onset of retinopathy and neuropathy.

Limitations and Generalizability

While the ANDIS cohort is one of the largest and most well-characterized in the world, the population is predominantly of Northern European descent. The distribution of these clusters and their associated risks may vary in other ethnic groups. Furthermore, the use of HOMA2 indices requires C-peptide measurements, which are not currently routine in all primary care settings. However, the study argues that the prognostic value of these measurements justifies their inclusion in standard diagnostic protocols for new-onset diabetes.

Conclusion: Moving Toward Subtype-Specific Care

This 14-year follow-up study confirms that the initial clinical presentation of diabetes carries a profound prognostic signature. By moving beyond HbA1c and adopting a subgroup-based classification, healthcare providers can identify high-risk individuals—particularly those in the SIDD and SIRD categories—much earlier in their disease course. This shift toward precision medicine offers the potential to tailor pharmacological interventions, optimize follow-up schedules, and ultimately reduce the heavy burden of cardiovascular and renal complications in the diabetic population.

Funding and References

This research was supported by the Swedish Research Council, the Swedish Heart and Lung Foundation, the Novo Nordisk Foundation, and several other Swedish governmental and private foundations including AstraZeneca.

Reference: Asplund O, Thangam M, Prasad RB, et al. Comorbidities and mortality in subgroups of adults with diabetes with up to 14 years follow-up: a prospective cohort study in Sweden. Lancet Diabetes Endocrinol. 2026;14(1):29-40. doi:10.1016/S2213-8587(25)00283-9.