Oral anticoagulation: a step toward solving the puzzle of dementia related to atrial fibrillation?

This editorial refers to ‘Less dementia with oral anticoagulation in atrial fibrillation’†, by L. Friberg and M. Rosenqvist, on page 453. As populations age, the rate of dementia increases worldwide. Europe has the highest prevalence of dementia,1 and the rate doubles for every 6.3 years of age.2 Dementia represents a major global challenge to health systems and requires novel and systematic approaches to address or at least slow this epidemic. Atrial fibrillation (AF) has been identified as a potential risk factor of dementia.3 The exact mechanisms behind this association are largely unknown, but it has been proposed that ischaemic strokes, silent brain infarcts, and micro-embolism could help to explain the cases of dementia related to AF.4 Based on this hypothesis, oral anticoagulation (OAC) is seen as a potential approach to prevent this type of dementia. However, a recent meta-analysis failed to demonstrate a significant reduction in the risk of dementia among AF patients treated with OAC.5 These negative results may be because the relationship between AF and dementia is complex and subject to substantial confounding (e.g. shared vascular risk factors). Further intricacy is added by the significant difficulty in distinguishing between degenerative, vascular, and mixed types of dementias. Moreover, >75% of patients aged 70 years or older with a primary neuropathological diagnosis of Alzheimer’s disease have evidence of concomitant vascular pathology, meaning that vascular lesions may contribute to the overall cognitive decline, even in primarily degenerative dementias.6 Based on this, assessing the association between OAC use and risk of dementia may depend on whether the dementias are primarily vascular or degenerative in a given study cohort, and the proportion of degenerative dementias with a cerebrovascular component (e.g. brain infarcts). Another challenge of dementia studies is that, unlike with hard cardiovascular outcomes such as stroke or myocardial infarction, the diagnosis and time of dementia onset are generally unclear and highly dependent on factors such as patients’ awareness of their own cognitive decline and the ability of their family members or caregivers to detect signs or symptoms of new-onset cognitive dysfunction. In the current issue of the journal, Friberg and Rosenqvist report that the use of OAC in patients with recently diagnosed AF was associated with a reduced risk of dementia.7 This study was based on a retrospective analysis of the Swedish Patient Register, which comprises data on hospitalizations in Sweden since 1987 and all specialized outpatient visits since 2001, and provided incidence of dementia data. Propensity score matching was used to create a balanced cohort of patients who were treated with OAC and those who did not receive OAC. The main result showed that anticoagulation was associated with a 29% risk reduction in the incidence of dementia. As acknowledged by the authors, the study has several limitations such as the retrospective and non-randomized design, administrative nature of the data set, imperfect ascertainment of OAC treatment, and lack of systematic adjudication of dementia.7 Also, death could have been analysed as a competing risk (e.g. some patients may have not developed dementia because they died of heart disease, cancer, or stroke). Although an on-treatment-like approach was used to account for treatment adherence, landmark analyses may have helped to estimate the association between OAC use and dementia more accurately during follow-up by considering the crossover between treatment groups. Importantly, there were no data available regarding international normalized ratios or time in the therapeutic range among patients treated with vitamin K antagonists. Furthermore, the authors did not describe the reasons why patients were not treated with OAC. Since the study population had AF and most patients were at increased risk of stroke, the fact that 241 060 (54.3%) AF patients not on an oral anticoagulant were used to match the OAC cohort without further explanation about the reasons behind the lack of OAC is concerning and a major limitation considering the study hypothesis. Finally, patients with lower socio-economic status or educational level may have less access to medical care and lower treatment adherence, resulting in worse outcomes. These factors, which are known to influence the risk of dementia considerably, were not contemplated in the propensity score matching or in the Cox regression analyses. The main findings of Friberg and Rosenqvist’s study support the role of cerebrovascular disease as the main pathophysiological mechanism explaining the association between AF and increased risk of dementia (Figure 1). Despite considerable overlap between vascular and degenerative pathology in most dementias, the effect of AF on the risk of dementia seems to be highly influenced by either clinically evident (e.g. stroke) or covert (e.g. silent brain infarct and microinfarcts) cerebrovascular disease. Although non-vascular mechanisms have been proposed as one of the potential causes of dementia among AF patients (e.g. hypoperfusion, inflammation, or hippocampal atrophy),8 these are still hypothetical and should be considered with caution. Indeed, a recent study evaluating the association between AF and the severity of neurodegeneration among a cohort of 1593 patients with primary neuropathological diagnosis of Alzheimer’s disease from the National Alzheimer’s Coordinating Center database found a significant association between AF and milder neurodegenerative changes.4 AF patients also lived longer, which resulted in more years of exposure to vascular risk factors, leading to a higher prevalence of cerebrovascular pathology. Therefore, either macroscopic or microscopic brain infarcts seem to lower the threshold for the clinical diagnosis of dementia in AF patients with otherwise milder forms of Alzheimer’s disease. This may be one of the reasons why some studies have found an association between AF and higher risk of Alzheimer’s disease. The presence of silent brain infarcts9 may also help explain the association between AF and dementia in patients without history of stroke.3 Pathophysiological model of atrial fibrillation-related dementia and possible role of oral anticoagulants in dementia prevention. Dotted lines imply lack of solid data supporting an association. Solid lines represent consistent data supporting an association. The question mark represents pathophysiological mechanisms or neuropathological findings for which evidence about an association with dementia risk is lacking. Superscript numbers represent reference numbers. AF, atrial fibrillation; OAC, oral anticoagulation; OR, odds ratio; HR, hazard ratio; RR, relative risk3 or risk ratio.10 *Studies in which the association between silent brain infarcts and cerebral microinfarcts was evaluated without specifying whether patients had AF. **The darker grey box represents the risk of dementia in patients without a history of stroke. For the overall summary measure of patients without stroke,3 the association between silent brain infarcts and microinfarcts with the risk of dementia was not formally assessed and cannot be ruled out. The cerebrovascular hypothesis of AF-related dementia aligns nicely with the results of the study by Friberg and Rosenqvist. AF is associated with a five-fold increased risk of stroke.10 Likewise, stroke precedes dementia by ∼10 years and doubles its risk.11 Furthermore, stroke in AF patients is associated with an almost three-fold higher risk of dementia.3 As such, if most of the effect of AF on dementia risk is explained by either symptomatic or asymptomatic cerebrovascular disease, OAC therapy would be expected to prevent a proportion of strokes, silent brain infarcts, or micro-infarcts, which would ultimately result in a reduction in the risk of dementia. In line with this concept, a recent study showed a consistent reduction in the incidence of stroke since 2002 in the province of Ontario, Canada, followed by a subsequent drop in the incidence of dementia (2008 onwards).12 The tandem reduction in the incidence of stroke and dementia was possibly explained by improved preventative strategies (e.g. better control of hypertension and other risk factors) in the province, including a higher use of OAC,13 although this remains hypothetical. In summary, the study by Friberg and Rosenqvist may not have a major impact on current clinical practice, since it is widely accepted that most patients with AF and increased risk of stroke should be treated with OAC. Nonetheless, these results are reassuring about the importance of OAC therapy for patients with AF, not only to prevent strokes but perhaps also to delay dementia onset or to reduce its risk. It also provides additional indirect evidence that AF-related dementia is primarily mediated by cerebrovascular disease. Still, further studies are needed to explore how OAC and other therapies may affect the different types of dementia that may be associated with AF. In the meantime, the prevention of different types of dementias, including AF-related, remains a complex and heterogeneous challenge. L.A.S. is supported by the Kathleen & Dr Henry Barnett Research Chair in Stroke Research, the Edward and Alma Saraydar Neurosciences Fund, and the Opportunities Fund of the Academic Health Sciences Center Alternative Funding Plan of the Academic Medical Organization of Southwestern Ontario (AMOSO). Conflicts of interest: L.A.S. has received research funds from Boehringer Ingelheim and Medtronic, and speaker honoraria from Boehringer Ingelheim and Pfizer. R.D.L. has received research grants from Bristol-Myers Squibb, GlaxoSmithKline, Medtronic, and Pfizer Inc., and consulting fees/honoraria from Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Daiichi Sankyo, GlaxoSmithKline, Medtronic, Merck & Co., Pfizer, and Portola.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.