Epidemiology
Elevated Lp(a) is not linked to dementia risk in a 151,117-patient propensity-matched real-world cohort (Eur J Intern Med 2026)
Original title: Lipoprotein(a) and risk of dementia: A propensity score-matched real-world analysis from a global federated research network
Retrospective propensity-matched cohort study using the TriNetX global federated research network, comparing 54,841 matched pairs of adults with elevated (50 mg/dL or more) versus normal Lp(a), followed up to 10 years for incident dementia. Elevated Lp(a) was not associated with increased risk of all-cause dementia (hazard ratio 0.99, 95% CI 0.88-1.10), Alzheimer's disease (hazard ratio 1.04, 95% CI 0.85-1.27), vascular dementia (hazard ratio 1.05, 95% CI 0.86-1.27) or frontotemporal dementia (hazard ratio 0.94, 95% CI 0.48-1.83), among 151,117 patients with Lp(a) data overall. Findings held across sensitivity analyses accounting for competing mortality risk and across subgroups by age, sex and prior cardiovascular disease or stroke, arguing against a causal Lp(a)-dementia link.
Original abstract
Background: The association between lipoprotein(a) [Lp(a)] and dementia remains controversial.
Purpose: To assess the relationship between Lp(a) and dementia in a global federated research network.
Methods: A retrospective cohort study using data from the TriNetX network. Adults with at least one Lp(a) measurement from 2010 onward were categorized as having elevated (≥50 mg/dL) or normal (<50 mg/dL) Lp(a). Propensity score matching (1:1) balanced baseline characteristics. The primary outcome was incident all-cause dementia; secondary outcomes included Alzheimer's disease, vascular dementia, and frontotemporal dementia. Follow-up extended up to 10 years from the index Lp(a) measurement. Sensitivity analyses accounted for death as a competing event, a 30-day landmark analysis, and comparison across low, high, and very high Lp(a) levels. Subgroup analyses were performed by age, sex, and history of atherosclerotic cardiovascular disease or stroke.
Results: Of 151,117 patients with available Lp(a) data, 54,929 had elevated Lp(a) (mean age 56.5 ± 15.9 years; 53.2% female), while 96,188 had normal Lp(a) (mean age 56.4 ± 16.5 years; 48.4% female). After matching, each cohort included 54,841 patients. Over a maximum follow-up of 10 years, elevated Lp(a) was not associated with an increased risk of all-cause dementia (HR 0.99, 95% CI 0.88-1.10), Alzheimer's disease (HR 1.04, 95% CI 0.85-1.27), vascular dementia (HR 1.05, 95% CI 0.86-1.27), or frontotemporal dementia (HR 0.94, 95% CI 0.48-1.83). Findings were consistent across sensitivity and subgroup analyses.
Conclusion: In this large real-world cohort, elevated Lp(a) levels were not significantly associated with an increased risk of incident dementia or its major subtypes.
Summary written by lp-a.org from the published abstract; figures as published. Page updated 17 August 2026. Methods.