This was a population-based retrospective cohort study using linked administrative databases in Alberta, Canada. The investigators identified patients aged 40 years or older with incident non-valvular atrial fibrillation (AF) diagnosed between April 1, 2012 and March 31, 2024. Patients with a prior history of cognitive impairment, dementia, stroke or transient ischemic attack, or a contraindication to oral anticoagulation (OAC) were excluded. The final sample included 66,885 individuals.
The study is presented as a clinical preprint and has not been peer reviewed. The University of Alberta research ethics board approved the study (Pro00010852). The ethics panel judged the work a retrospective database review for which individual consent was not required. The authors report that all relevant ethical guidelines were followed and that data used in the manuscript are available on request.
Of the 66,885 patients, 43% were female. Compared with males, females were older (reported ages: 74 vs 68 years), had greater measured frailty (14.7% vs 10.1%), and a higher proportion met a CHA2DS-VA score ≥2 (73% of females vs 65% of males). These baseline differences reflect an older and more comorbid profile among women at the time of incident AF diagnosis in this cohort.
Oral anticoagulation was assessed in the early post-diagnosis period. OAC was initiated within 120 days of AF diagnosis in 60% of patients of both sexes. The paper distinguishes between warfarin (vitamin K antagonist) and non–vitamin K oral anticoagulants (NOACs) when evaluating associations with cognitive outcomes.
Outcomes included incident cognitive impairment, Alzheimer’s disease, vascular dementia, and all-cause mortality. Survival models were used to estimate associations between sex and each outcome, and to evaluate the relationship between OAC therapy (warfarin, NOAC, or no OAC) and outcomes stratified by sex. The abstract reports adjusted hazard ratios (aHRs) with 95% confidence intervals and p values for the principal comparisons.
After multivariable adjustment, females had a higher risk of incident cognitive impairment compared with males (aHR 1.14; 95% CI 1.07–1.20; p<0.0001). In contrast, females had a lower adjusted risk of vascular dementia (aHR 0.75; 95% CI 0.63–0.90; p=0.0016) and lower all-cause mortality (aHR 0.91; 95% CI 0.88–0.94; p<0.001) than males.
There was no statistically significant sex difference in the adjusted risk of Alzheimer’s disease (aHR 1.01; 95% CI 0.88–1.17; p=0.8496).
The authors report that use of NOACs was associated with a lower risk of cognitive impairment compared with warfarin or no OAC in both sexes. Regarding dementia subtypes, NOAC use was associated with a lower risk of Alzheimer’s disease and vascular dementia in males only; this male-specific association for dementia outcomes is reported in the abstract. Details on absolute event rates, duration of follow-up, dose, adherence, or time-updated exposure definitions were not provided in the abstract and would require consulting the full manuscript or supplementary materials.
Strengths evident from the abstract include a large, population-based sample of incident non-valvular AF and exclusion of patients with prior cognitive disorders or stroke/TIA, which reduces certain baseline confounders. The study used administrative data across a long time window (2012–2024) and included assessment of anticoagulation type.
Limitations specifically reported in the source abstract are limited. The abstract does not detail potential residual confounding, measurement methods for cognitive diagnoses in administrative data, follow-up duration, competing risks handling, or sensitivity analyses. Because the paper is a preprint, these methodological details and limitations should be reviewed in the full manuscript before interpreting causal implications.
Ethical oversight was provided by the University of Alberta research ethics board, which approved the retrospective database review; the study authors declare no competing interests.
In this large population-based cohort of patients with newly diagnosed non-valvular AF, females were older and frailer at diagnosis and had a higher adjusted risk of incident cognitive impairment than males, while showing lower adjusted risks of vascular dementia and all-cause mortality. No sex difference was observed for Alzheimer’s disease.
Use of NOACs was associated with a lower risk of cognitive impairment in both sexes compared with warfarin or no OAC, and with lower risks of Alzheimer’s disease and vascular dementia in males only, as reported by the authors. The findings suggest potential sex-specific patterns in cognitive outcomes after incident AF and raise questions about the influence of anticoagulant choice on cognitive trajectories.
These results are reported in a preprint and have not been peer reviewed; clinicians and guideline developers should interpret the findings cautiously and consult the full paper for detailed methodology, absolute risks, duration of follow-up, and subgroup analyses before applying them to practice.