BAA was associated with the initial onset of stroke, dementia, and mortality, whereas its associations with later progression to comorbidity were not evident, suggesting that BAA may be most informative for identifying early vulnerability before clinical disease onset.
Key Findings
Results
PhenoAge acceleration was significantly associated with increased risk of stroke from baseline.
Each standard deviation increase in PhenoAge acceleration was associated with a hazard ratio of 1.16 (95% CI: 1.14–1.19) for stroke from baseline.
The study included 220,236 UK Biobank participants free of stroke and dementia at baseline.
Over a median follow-up of 15.31 years, 4,436 participants experienced stroke.
Similar trends were observed for KDM-BA acceleration.
Results
PhenoAge acceleration was significantly associated with increased risk of dementia from baseline.
Each standard deviation increase in PhenoAge acceleration was associated with a hazard ratio of 1.10 (95% CI: 1.06–1.13) for dementia from baseline.
Over a median follow-up of 15.31 years, 3,101 participants developed dementia.
Similar trends were observed for KDM-BA acceleration.
A multi-state Cox model was used to evaluate associations across eight disease transitions.
Results
PhenoAge acceleration was significantly associated with increased risk of death from baseline.
Each standard deviation increase in PhenoAge acceleration was associated with a hazard ratio of 1.24 (95% CI: 1.23–1.25) for death.
17,375 participants died over the follow-up period.
Similar trends were observed for KDM-BA acceleration.
This was the strongest association observed among the three initial event outcomes.
Results
No significant associations were observed between biological age acceleration and transitions from stroke or dementia to comorbidity.
291 participants accumulated comorbidities (i.e., transitioned from stroke or dementia to the comorbidity state) over follow-up.
The lack of association applied to both PhenoAge acceleration and KDM-BA acceleration.
This contrasts with the significant associations found for initial disease onset transitions.
The multi-state model evaluated eight disease transitions across five states: baseline, stroke, dementia, comorbidity, and death.
Results
The associations between biological age acceleration and disease outcomes were consistent across multiple time intervals and demographic and clinical subgroups.
Time intervals examined were 0–1 years, 1–3 years, and >3 years after baseline.
There was no evidence of temporal modification in the associations.
There was no evidence of effect modification by demographic or clinical subgroups.
This consistency supports the robustness of the associations between BAA and initial disease onset.
Methods
Biological age acceleration was estimated using two established algorithms applied to UK Biobank data.
The two algorithms used were the Klemera-Doubal Method-Biological Age (KDM-BA) and Phenotypic Age (PhenoAge).
BAA was calculated as the residual difference between biological and chronological age.
The study sample comprised 220,236 participants free of stroke and dementia at baseline.
Median follow-up was 15.31 years.
Methods
A multi-state Cox model framework was used to capture sequential neurovascular disease trajectories.
The model evaluated eight disease transitions across five states: baseline, stroke, dementia, comorbidity, and death.
This approach allowed assessment of both initial disease onset and subsequent disease progression.
Nonlinear exposure-response patterns and time-interval-specific associations were also evaluated.
The framework captured transitions including baseline to stroke, baseline to dementia, stroke to dementia, stroke to comorbidity, dementia to comorbidity, and transitions to death from various states.
Cheng W, Chen L, Kang P, Tang B, Zhang S, Zhang G, et al.. (2026). Accelerated biological aging and the dynamic transitions of stroke and dementia: A multi-state trajectory analysis.. Neurobiology of disease. https://doi.org/10.1016/j.nbd.2026.107332