What the research found

Researchers used genetic risk scores for type 1 and type 2 diabetes to analyse 309 people with atypical or unclear diabetes diagnoses (specifically those without islet autoantibodies, which normally mark type 1 diabetes). Surprisingly, these individuals carried elevated genetic risk factors for both type 1 and type 2 diabetes compared to genetically matched controls, despite being selected for having non-standard presentations.

The genetic scores predicted functional differences in insulin secretion measured during oral glucose tolerance tests. A higher type 1 diabetes polygenic score correlated with lower C-peptide levels (indicating weaker pancreatic insulin production), while a higher type 2 diabetes score predicted higher C-peptide levels. Notably, individuals with type 1 diabetes genetic scores above a certain threshold were 13 times more likely to have both poor insulin secretion on testing and to require intensive insulin therapy (both basal and bolus doses).

This suggests some people with atypical diabetes who lack the classic autoimmune markers of type 1 diabetes may actually have a genetically similar form of insulin deficiency—and polygenic scores could flag who needs insulin intensification rather than oral agents alone.

Why it matters for you

If you're tracking fasting glucose, post-prandial glucose response, or C-peptide via OGTT, genetic risk profiling could help explain why your glucose control responds better to certain therapies than others. Someone with "type 2 diabetes genes" but poor insulin secretion in practice may be misclassified—and thus undertreated—if diagnosed by phenotype alone.

For users taking TRT or using peptides that affect insulin sensitivity, knowing your underlying genetic predisposition to insulin deficiency versus resistance reshapes the strategy. A person genetically loaded toward type 1-like pathology shouldn't expect insulin sensitizers alone to solve poor glucose control; they may need to accept insulin as a primary tool earlier. Conversely, someone with type 2 genetic risk may benefit more from lifestyle and sensitivity-enhancing agents.

The practical takeaway: if your blood glucose biomarkers don't fit the textbook type 1 or type 2 picture, or if your response to treatment is atypical, a polygenic score could be worth discussing with your clinician. It's one more data point to personalise your approach—especially if you're deciding between intensifying insulin versus trialling other agents.

Caveats

• Observational, cross-sectional design — associations don't prove causation, and the direction of effect from genetics to treatment choice isn't tested prospectively.
• Small, highly selected population — 309 individuals with rare/atypical diabetes; findings may not generalise to common type 1 or type 2, or to other ancestry groups if representation is limited.
• Polygenic scores are probabilistic, not diagnostic — individual prediction accuracy is modest; scores work best at population level.
• No intervention or outcome trial — doesn't show that using the score to guide treatment actually improves glucose control or clinical outcomes.
• Ancestry matching limited — polygenic scores perform less well outside the ancestry group they were developed in; applicability depends on your genetic background.