Biological ageing, diet quality, and stroke: evidence from two population-based studies and an exploratory hospital-based Chinese cross-sectional dataset.

Background: Diet quality and biological ageing have each been linked to stroke, but whether the association between accelerated biological ageing and stroke reflects poorer diet quality, or is largely independent of it, is unknown. Treating diet quality as an upstream determinant of biological ageing, we examined the association of Phenotypic Age (PhenoAge) acceleration with stroke and how far it changed after adjusting for diet quality. Methods: We analysed 400,695 UK Biobank participants, 21,754 NHANES adults and 312 adults from a hospital-based dataset (Second Xiangya Hospital). PhenoAge was derived from chronological age and nine biomarkers, and PhenoAge acceleration was PhenoAge minus chronological age. Diet quality was assessed by a Healthy Diet Score (UK Biobank, hospital) and the Healthy Eating Index-2020 (NHANES). Outcomes were incident stroke (record linkage) in the UK Biobank and prevalent stroke in NHANES (self-reported) and the hospital dataset (physician-documented). All models estimated associations; no risk-prediction model was developed or externally validated. Results: During a median 15.83 years, 11,345 UK Biobank participants developed incident stroke; 729 NHANES and 58 hospital participants had prevalent stroke. Each 1-year increase in PhenoAge acceleration was associated with incident stroke in the UK Biobank (HR 1.060, 95% CI 1.056-1.064) and with prevalent stroke in NHANES (OR 1.058, 1.037-1.080) and the hospital dataset (OR 1.082, 1.002-1.169). In the UK Biobank the association was confined to ischaemic stroke (HR 1.066, 1.061-1.071) but showed no association with haemorrhagic stroke (HR 1.009, 0.998-1.020). Better diet quality was associated with lower PhenoAge acceleration (UK Biobank, high versus low adherence: -0.746 years, -0.811 to -0.681; NHANES: -0.695 years, -0.781 to -0.609). Nevertheless, adding diet quality changed the HR/OR only modestly (11.9, 10.4 and 0.7%; HR 1.060 → 1.053, OR 1.058 → 1.052 and 1.082 → 1.081). Conclusion: Better diet quality was associated with slower biological ageing, yet accelerated biological ageing was associated with stroke, largely independently of measured diet quality. Because outcome definitions differed and these analyses compare adjusted models rather than decompose a causal pathway, the three estimates are complementary rather than replicative and do not support inferences about prevention or clinical risk stratification.