Diabetes can hide the mitochondrial signal researchers use to study Alzheimer's

Edited by Samael A. Tejada, Founder & Editor-in-Chief

Diabetes is a major risk factor for Alzheimer's disease, and the two conditions share mitochondrial dysfunction, inflammation and oxidative stress. Kaikini, Malik and colleagues at King's College London asked a measurement question with practical consequences: if a study measures mitochondrial DNA copy number (mtDNA-CN) in brain tissue without accounting for diabetes, what does it actually see?

What did the two diseases do to the same measurement?

Opposite things. Alzheimer's disease was associated with lower mtDNA-CN; when people with diabetes were removed from the analysis, mtDNA-CN was 19 to 30 per cent lower in people with Alzheimer's than in cognitive controls. Diabetes, by contrast, was associated with higher mtDNA-CN across all three brain regions — and surprisingly, all diabetes cases, with and without dementia, had higher mtDNA than non-diabetic controls.

Why does pooling the samples hide the Alzheimer's finding?

Because the two effects cancel out. Senior author Afshan Malik, Professor of Metabolic Disease at King's, put it plainly: grouping diabetic and non-diabetic data for mtDNA-CN together can result in the loss of real changes that may be important in disease mechanisms. The team suggests this masking may help explain why previous studies of mitochondrial dysfunction in Alzheimer's have produced conflicting results.

What should change in testing and trials?

Stratification. The authors argue diabetes status needs to be accounted for when mtDNA-CN is used as a biomarker or a trial readout, and that Alzheimer's can affect people differently at a biological level depending on conditions like diabetes. First author Dr Aakruti Kaikini notes the team does not yet know whether the higher mtDNA in diabetes is a stress response, a compensation, or something else — the next step is multi-omics work to understand what the change does to brain function. This is a post-mortem observational study, not a validated clinical test.

Where the evidence stands

• Established: Diabetes is a major risk factor for Alzheimer's disease; both involve mitochondrial dysfunction, and mtDNA copy number is widely used as a proxy for mitochondrial health in research.
• Preliminary: In post-mortem human hippocampus, amygdala and cerebellum (Alzheimer's & Dementia, 2026), Alzheimer's was associated with lower mtDNA-CN (19–30% lower once diabetes cases were excluded) while diabetes was associated with higher mtDNA-CN (41–64% higher in Alzheimer's cases with diabetes), so pooling masks the Alzheimer's association.
• Absent: A validated clinical mtDNA-CN test for Alzheimer's, evidence on whether the diabetes-associated rise helps or harms brain function, and longitudinal data in living patients.