Highlight
The accuracy of HbA1c thresholds to predict type 1 diabetes progression in islet autoantibody-positive adults improves when accounting for age-related increases in HbA1c. Traditional thresholds tend to overestimate risk in adults compared to children. Applying an age-adjusted HbA1c or a higher fixed threshold (≥6.0% [42 mmol/mol]) better aligns progression risk across age groups, particularly among those aged 30 and older.
Study Background
Type 1 diabetes (T1D) is an autoimmune disorder characterized by the destruction of pancreatic beta cells, leading to insulin deficiency. The presence of islet autoantibodies is a critical marker for identifying individuals at risk for developing T1D, often before clinical onset. HbA1c, a measure of glycated hemoglobin reflecting average blood glucose levels over 2–3 months, is widely used to define dysglycemia and predict progression risk in autoantibody-positive individuals.
However, HbA1c levels naturally increase with age, potentially confounding risk stratification in adults when using a universal dysglycemia threshold (HbA1c ≥5.7% [39 mmol/mol]). Current clinical practice, which applies static HbA1c cutoffs irrespective of age, may overestimate progression risk in adults, leading to possible misclassification and suboptimal preventive interventions. This study addresses the unmet need to refine HbA1c-based risk prediction across different age strata in islet autoantibody-positive populations.
Study Design
This analysis leverages data from two large cohorts. The primary dataset includes 5,024 first- and second-degree relatives of T1D patients who are positive for one or more islet autoantibodies, enrolled in the TrialNet Pathway to Prevention study. Of these, 3,720 were children and 1,304 adults. To model age effects on HbA1c, 6,273 adults from the population-based Exeter 10000 cohort were analyzed, establishing normative age-related HbA1c trajectories.
Progression risk to T1D was evaluated using three approaches: the conventional dysglycemia threshold of HbA1c ≥5.7% (39 mmol/mol), an age-adjusted HbA1c metric derived from population data, and an alternative fixed threshold of HbA1c ≥6.0% (42 mmol/mol). Risks were compared across age groups (children vs. adults) and autoantibody positivity status (single vs. multiple autoantibodies). Post hoc analyses further explored progression risk in adults younger than 30 years.
Key Findings
At the standard HbA1c threshold (≥5.7%), children with single autoantibody positivity exhibited a markedly higher 1-year progression risk than adults (38% versus 13%, P<0.001), a trend also observed among those with multiple autoantibodies (55% versus 38%, P<0.001). This disparity suggested potential overestimation of risk in adults when age-related HbA1c changes were not considered.
Applying age adjustment to HbA1c reduced these differences substantially. Progression risks in single autoantibody-positive adults increased to 27%, closing the gap with children (38%; P=0.32). For multiple autoantibody-positive individuals, differences were attenuated though still present.
Using a higher fixed HbA1c threshold (≥6.0%) for adults achieved risk parity with children across autoantibody groups. This approach also identified adults over 30 years old with progression risks comparable to pediatric populations, improving the specificity of risk classification.
Interestingly, post hoc analysis indicated that adults younger than 30 years demonstrated progression risks statistically similar to children (P=0.1), suggesting that age-specific thresholds might be particularly beneficial in older adults.
Expert Commentary
This study provides robust population-level evidence that age-related increases in HbA1c must be accounted for to accurately stratify type 1 diabetes risk in islet autoantibody-positive adults. The authors’ use of a large, well-phenotyped cohort and incorporation of normative population data strengthens the generalizability of findings.
The implications extend into clinical trial design and individualized prevention strategies: adopting age-adjusted HbA1c thresholds could prevent unnecessary enrollment of lower-risk adults or misclassification in risk scoring systems, thereby optimizing resource allocation and patient monitoring.
Limitations include reliance on cohort data, which may not fully capture ethnic and geographic variation in HbA1c dynamics. Furthermore, the choice of an alternative threshold (≥6.0%) remains somewhat arbitrary, warranting prospective validation. Mechanistically, age-related red blood cell turnover and glycation rates likely contribute to observed HbA1c variation, but these biological factors require further elucidation.
Conclusion
Accounting for age-related HbA1c variation markedly improves the precision of dysglycemia classification and risk prediction for type 1 diabetes progression in autoantibody-positive adults. Implementing either age-adjusted HbA1c values or a higher threshold of ≥6.0% for adults aged 30 years and older harmonizes progression risk with pediatric cohorts, facilitating better risk stratification in preventive care and clinical research contexts.
Future directions should explore prospective validation of age-specific HbA1c thresholds across diverse populations and investigate integration with other biomarkers to enhance predictive accuracy for type 1 diabetes onset.
Funding and ClinicalTrials.gov
This study was supported by TrialNet and associated foundations. The TrialNet Pathway to Prevention study is registered with ClinicalTrials.gov (NCT00097292).
References
- Templeman EL, Thomas N, Martin S, et al. Accounting for Age-Related Increases in HbA1c More Accurately Quantifies Risk of Type 1 Diabetes Progression in Islet Autoantibody-Positive Adults. Diabetes Care. 2026 Jul 1;49(7):1262-1269. PMID: 42090204.
- Redondo MJ, Steck AK, Pugliese A. Genetics of type 1 diabetes. Pediatr Diabetes. 2018;19(3):346-353.
- Oram RA, Sims EK. Assessing Risk for Type 1 Diabetes in Autoantibody Positive Individuals. Curr Diab Rep. 2017;17(9):64.
- Herman WH. The Use of HbA1c for the Diagnosis of Diabetes Mellitus: Some Considerations. J Diabetes Sci Technol. 2019;13(4):663-666.
