Atrial fibrillation (AF) is a leading cause of adverse outcomes, including stroke, heart failure and death.^1-3 A growing body of recent evidence suggests a significant association between AF and cognitive impairment.^4,5 In the Swiss‑AF cohort study,^4,5 both overt and silent brain infarcts were independently associated with cognitive decline (β = −0.23; 95% CI, −0.44 to −0.01; P = 0.042 and β = −0.14; 95% CI, −0.21 to −0.06; P <⁠0.001, respectively).

Given that cognitive impairment and dementia are key drivers of health care costs and represent a growing public health challenge in aging societies,⁶ understanding their underlying causes and mechanisms is of major importance.

In this issue of Polish Archives of Internal Medicine, Rewiuk et al⁷ addressed the question whether patients with clinically silent (asymptomatic) AF have a different cognitive performance, as compared with individuals with symptomatic AF. Patients were included from the NOMED‑AF (Non‑invasive Monitoring for Early Detection of Atrial Fibrillation) study, a cross‑sectional observational cohort designed to assess the prevalence of AF and its comorbidities in a sample of community‑dwelling Polish adults aged over 65 years. Baseline AF history and related symptoms were recorded at enrollment. All participants underwent continuous ECG monitoring for up to 30 days to determine the presence of AF. Silent AF (SAF) was defined as electrocardiography (ECG)-confirmed AF in individuals who did not report any associated symptoms. In contrast, symptomatic AF referred to cases where patients experienced and reported symptoms. Additionally, cognitive function was assessed using the Mini‑Mental State Examination (MMSE), with scores under 24 indicating probable dementia.

Among 2203 participants (mean [SD] age, 77.1 years; 48% women) included in the study, 491 (22.3%) had diagnosed AF, of which 203 (41.3%) were classified as SAF. The patients with AF, regardless of symptom presentation, had lower MMSE scores in comparison with those without AF (mean [SD], 25.2 [0.22] vs 26 [0.11]; P <⁠0.001). The patients with SAF had lower MMSE scores than those with symptomatic AF (mean [SD], 24.3 [0.38] vs 25.8 [0.26]; P = 0.001). In multivariable adjusted models, SAF was significantly associated with probable dementia (odds ratio, 1.74; 95% CI, 1.1–2.73; P <⁠0.05).

There are several reasons that may explain these findings. In this cohort, AF patients were older, had more cardiovascular comorbidities, and a higher prevalence of prior stroke. All these factors are well‑established contributors to cognitive decline, raising the possibility of residual confounding. However, the findings are consistent with prior studies demonstrating reduced cognitive function in AF patients.⁵ This relationship was partly driven by a higher burden of silent cerebral infarcts, even though 90% of the patients were on oral anticoagulation therapy. The underlying mechanisms linking AF and cognitive decline remain an area of active investigation. Potential other pathways include hemodynamic alterations affecting cerebral perfusion^8,9 and the contribution of shared vascular risk factors, such as inflammatory processes and coagulation abnormalities.^10,11

A key strength of this study is its focused investigation of the clinically important relationship between AF and cognitive function. However, there are several important limitations. First, the cross‑sectional design of this analysis limits our ability to establish a causal relationship between SAF and cognitive decline, as well as the directionality of the association. Patients may have already experienced cognitive decline before their first detected AF. Second, anticoagulation therapy plays a key role in AF management and may influence cognitive outcomes. However, incomplete data on anticoagulation use precludes adjustment for this important confounder. Third, the study provided very little details about the definition of symptomatic versus asymptomatic AF, and it is not clear whether AF patients had asymptomatic but known AF or whether most of them had subclinical AF, that is, AF detected via long‑term monitoring, but not known clinically. Fourth, although all patients underwent 30‑day rhythm monitoring, the duration of individual AF episodes and the overall AF burden were not reported. Lastly, the manuscript does not specify which symptoms were included and how they aligned with the timing of AF episodes.

These are important points as AF is a more heterogeneous disorder than we previously thought, and emerging evidence supports the concept of a more refined classification into AF subpopulations. For example, symptomatic AF may lead to greater healthcare utilization and earlier detection, whereas asymptomatic AF often goes unrecognized, potentially biasing outcome assessments. Nevertheless, large cohort studies have demonstrated similar risks of stroke and mortality between asymptomatic and symptomatic AF patients,¹² suggesting that symptom burden alone offers limited prognostic value. Another key distinction is between clinical AF, diagnosed on surface ECG, and subclinical AF, which is asymptomatic and detected only via continuous monitoring. Patients with subclinical AF have a lower stroke risk than those with clinical AF, even if the CHA₂DS₂-VASc score is similar.^13,14 The risk of silent brain infarcts in patients with subclinical AF is currently unknown. If Rewiuk et al⁷ describe mostly patients with subclinical AF, then the lower MMSE scores are somewhat counterintuitive and potentially due to bias or residual confounding. A third category represents perioperative AF, an important cause of stroke and mortality.¹⁵ Despite this association, it is currently unclear whether these patients should be treated with oral anticoagulants, and randomized trials are ongoing (eg, ASPIRE‑AF [Anticoagulation for Stroke Prevention In Patients With Recent Episodes of Perioperative Atrial Fibrillation After Noncardiac Surgery]; NCT03968393). Together, these observations emphasize that the description of the specific AF population being studied is of major importance.

In summary, the study by Rewiuk et al⁷ provides relevant insights into the relationship between SAF and cognitive decline in patients with AF. With the aging of the population, addressing the impact of AF on cognitive decline is a critical area in need of further attention.