Clinical implications of different types of dementia in patients with atrial fibrillation: Insights from a global federated health network analysis

Atrial fibrillation (AF) associates with higher Alzheimer's disease (AD) and vascular dementia risks but the clinical implications have been scarcely investigated. We examined the association between AD or vascular dementia and adverse outcomes in AF patients.


| INTRODUCTION
An increased risk of dementia has been well recognized in patients with atrial fibrillation (AF). 1,2 Although AF has been predominantly associated with vascular dementia, based on common traditional cardiovascular risk factors, several studies have demonstrated that the risk of dementia persists even in the absence of history of an overt clinical stroke. [1][2][3] Therefore, it has been hypothesized that multiple mechanisms act in a synergistic manner, including silent cerebral ischemia, hypoperfusion, systemic inflammation, cerebral hypoperfusion, and microbleeds all contributing to the development of cognitive decline in patients with AF. 4 Recently, brain imaging studies reporting an increased amyloid deposition in patients with AF suggest that not only vascular dementia but also degenerative subtypes of dementia, and more specifically Alzheimer's disease (AD), are associated with AF. More recently a metaanalysis based on 56 370 patients from six studies reporting on the clinical association between AD and AF, found that patients with AF had a 30% increased chance of developing AD compared with people without AF. 5 Hence, AF may be linked to both vascular and degenerative dementia, which may also manifest in mixed forms. While the epidemiological association between AF and dementia has been well documented, the clinical implications have been scarcely investigated.
Current guidelines on AF recommend the use of oral anticoagulation therapy in patients with dementia according to their CHA 2 DS 2 -VASc score and good control of risk factors, but do not suggest any clinical utility in differentiating between dementia subtypes. 6,7 In this study, using a global federated database of electronic health records, our aim was to investigate the association between subtype of dementia (AD or vascular dementia) and the risk of adverse outcomes in patients with AF. We hypothesized that the specific feature of vascular versus AD, as for example amyloid deposition, may confer a different prognosis in AF patients.

| METHODS
The data source used was TriNetX, a global federated health research network with real-time updates of anonymized electronic medical records (EMRs). The network includes healthcare organizations (HCOs), academic medical centers, specialty physician practices, and community hospitals, with data for >80 million patients predominately based in the United States. To comply with legal frameworks and ethical guidelines guarding against data reidentification, the identity of participating HCOs and their individual contribution to each data set are not disclosed. As a federated research network, studies using the TriNetX health research network do not require ethical approval as no patient identifiable identification is received.

| Statistical analysis
Continuous variables were expressed as mean and standard deviation (SD), and tested for differences with independent-sample t-tests.
Categorical variables were expressed as absolute frequencies and percentages, and tested for differences using χ 2 test.
The TriNetX platform was used to run 1:1 propensity score matching (PSM) using logistic regression. The platform uses "greedy nearest-neighbor matching" with a caliper of 0.1 pooled SDs and PROIETTI ET AL. | 657 difference between propensity scores ≤0.1. Covariate balance between groups was assessed using standardized mean differences (SMDs). Any baseline characteristic with a SMD between cohorts <0.1 is considered well matched. 8 Hazard ratios (HR) and 95% confidence intervals (CI) were calculated following PSM, and Kaplan−Meier survival curves with logrank tests. No imputations were made for missing data. Two-sided p < 0.05 were accepted as statistically significant. Statistical analyses were performed using the TriNetX analytics function in the online research platform.

| Participant characteristics
The search identified 2377 patients with AF and a diagnosis of dementia, which included 1225 patients with vascular dementia (mean age: 66 ± 6.4; 37.3% female) and 1152 with AD (mean age: 67.9 ± 6.1; 43.7% female). Table 1 summarizes the baseline characteristics of patients with AF and vascular or AD, before and after PSM. Patients with AF and AD were significantly older (66 ± 6.4 vs. 67.9 ± 6.1, p < 0.001), mostly of non-Hispanic White ethnicity and presented a lower cardiovascular risk profile characterized by alower prevalence of hypertension, diabetes, ischemic heart disease, and heart failure, among others (p < 0.05 for all). Following PSM, 615 patients were included in each of the two cohorts (i.e., 1:1) and these variables were well-matched and no longer significantly different (Table 1).

| Mortality
Mortality was higher among patients with vascular dementia compared to AD (HR: 1.25, 95% CI: 1.01−1.58). Figure 4 reports the survival curves showing that survival was lower in patients with vascular dementia (log rank p = 0.049).

| DISCUSSION
In this retrospective analysis of patients with AF from a large global federated data set, there was a similar prevalence of patients with AD and vascular dementia. However, the risk profile and prognosis of these two forms of dementia appear to be different, since patients with vascular dementia presented with a higher risk of ischemic stroke/TIA, rehospitalization, and mortality compared to AD, whereas AD patients had a higher risk of ICH. To the best of our knowledge, there is no previous study comparing the risk of adverse clinical outcomes in patients with AF according to different dementia types.
The differences we detected in our analysis, showing a slightly lower prevalence of AD in African American, are consistent with prior data reporting that missed diagnosis of AD was common in this group and a well documented high prevalence of AD among non-Hispanic White patients. 9,10 The higher cardiovascular risk profile of patients with vascular dementia also reflects the higher CHA 2 DS 2 -VASc risk associated with stroke. It has been already reported that hypertension is not a risk factor for AD 11 ; therefore, the significant difference found in the prevalence of hypertension between the two groups is unsurprising. It is worth mentioning that type 2 diabetes has been increasingly recognized as a risk factor for AD through different mechanisms including inflammation, insulin resistance, and hyperglycemia. 9 Our analysis reports that patients with AF and AD resent a higher risk of non-traumatic ICH than in those with vascular dementia.
However, rate of ischemic stroke/TIA in patients with AD in this cohort (~30%) was not neglectable to justify specific adjustment on anticoagulation therapy in this population because of a higher bleeding risk. Indeed, the increased risk of ICH in patients with AD is aligned with the clinical dilemma of anticoagulation in patients with cerebral amyloid angiopathy given that perivascular amyloid deposition decreased cerebral arterial compliance and cerebrovascular resistance leading to increased susceptibility to micro-bleeding, hemosiderosis, and ICH. 12 However, cerebral amyloid angiopathy is a challenging diagnosis mostly linked to magnetic resonance imaging findings. A recent consensus paper did not advise any indication for brain imaging in patients with AF for risk stratification or clinical management. 13 In contrast, our finding directly correlates the increased risk of ICH with the clinical diagnosis of AD.
T A B L E 1 Comparison of clinical characteristics of the study cohort before and after propensity score matching. lesions and the susceptibility to amyloid deposition and neural damage that characterize mixed forms of dementia and common in AF requires further investigation. 9 Cerebrovascular disease has been suggested to contribute to AD neuropathological changes including selective brain atrophy and accumulation of abnormal proteins, such as amyloid-beta. 12 The observed higher risk of death and hospitalization in patients with vascular dementia compared to AD may be linked to the predisposition to ischemic events.
Considering our results, there is a need for a more holistic or integrated care approach to patients with AF, 15 which has been associated with an improved clinical prognosis. 16 Such an integrated management should also consider dementia risk, 17

| CONCLUSION
This retrospective analysis suggests that among patients with AF and dementia, AD was almost as prevalent as vascular dementia. The two forms of dementia in patients with AF are associated with different prognosis, with AD being associated with a higher risk of ICH compared to vascular dementia, and a higher risk of ischemic stroke/ TIA, hospitalization, and mortality among those with vascular dementia.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are not available since derived from the TriNetX research network. These data are not publicly available due to privacy and permission restrictions. Special authorization is required for accessing TriNetX.