Association of pulse pressure with incidence of dementia independent of established risk factors

Objectives: To investigate the association between pulse pressure and dementia incidence, independent of other blood pressure measurements and established risk factors, and to assess whether this association differs across dementia subtypes. Design: Prospective population-based study. Setting: UK Biobank. Participants: Of the 502,211 participants in the UK Biobank, 470,986 completed at least one blood pressure measurement and were included in the analysis. These participants were recruited between March 2006 and July 2010 and were followed for up to four assessments through to February 2024. Main outcome measures: Incidence of dementia, identified through linked health records using ICD-9 and ICD-10 diagnosis codes, self-reported diagnoses or records of dementia-specific medication use. The association between pulse pressure and risk of dementia was investigated using Cox proportional hazard models. Models were adjusted for age, sex, education, hearing problems, lipid levels, depression, traumatic brain injury, physical activity, diabetes, smoking, hypertension, body mass index, alcohol consumption and mean arterial pressure. Dementia subtype-specific associations were examined using competing risk models, with cause-specific Cox analyses included as supplementary sensitivity analyses. Results: During a median follow-up of 13 years, 9,028 persons developed dementia (Alzheimer's disease: 3,011; Vascular dementia: 1,270; Dementia with Lewy bodies: 234; Frontotemporal dementia: 191; Other/Mixed dementia: 4,322). Each 10mm Hg increase in pulse pressure was associated with a 5.4% higher risk of dementia (95% confidence interval on hazard ratio: 1.036 to 1.071), even after adjustment for age, mean arterial pressure and other established dementia risk factors. The effects were disproportionately stronger for Alzheimer's disease and vascular dementia, with no clear evidence for increased risk for dementia with Lewy bodies or frontotemporal dementia. Results were robust across sensitivity analyses including alternative blood pressure metrics, complete-case models, and alternative dementia classifications. Conclusions: Pulse pressure is independently associated with incidence of dementia beyond conventional blood pressure measures.