The purpose of this study is to compare the change of global cognitive performance after 52 weeks of intervention among participants with nonvalvular atrial fibrillation (NVAF) receiving rivaroxaban versus a vitamin K antagonist (warfarin).

The secondary objectives are to compare, among participants with NVFA receiving rivaroxaban versus warfarin :

• the changes of global cognitive performance after 26 weeks of intervention
• the changes of executive functions after 26 and 52 weeks of intervention
• the changes of episodic memory after 26 and 52 weeks of intervention
• the changes of independence and autonomy after 26 and 52 weeks of intervention

Detailed Description:

Vitamin K antagonists (VKAs) are commonly used for their role in hemostasis by interfering with vitamin K cycle decreasing the bioavailability of the vitamin K active form. In addition to a role in blood coagulation, vitamin K participates in brain health and function by regulating the synthesis of sphingolipids, a constituent of the myelins sheath and the neurons membrane, and through the biological activation of vitamin K-dependent proteins (VKDPs) involved in neuron survival. Epidemiological studies have reported a positive association between higher serum vitamin K concentration and better verbal episodic memory performance in older adults, and an inverse association between dietary vitamin K intakes and behavioural disorders and cognitive complaint. The clinical implication is that the use of VKAs, which deplete the active form of vitamin K, may be responsible for Central Nervous System (CNS) disorders.

CNS abnormalities were observed in newborns exposed in utero to VKA. Similarly, the investigators and other researchers reported that the use of VKAs (especially fluindione) was directly associated with cognitive decline (notably executive dysfunction) and hippocampal atrophy in older adults, even while taking into account the history of atrial fibrillation, stroke and vascular brain changes. These cross-sectional and longitudinal studies were yet limited by their observational design. Clinical trials are now warranted to explore the effect on cognition of VKAs against direct oral anticoagulants (DOACs), whose indications are similar but whose mechanism does not interfere with vitamin K. The favorable impact of the use of DOACs compared to VKA in the incidence of dementia was also observed in a US retrospective population-based study of patients managed per routine clinical care.

The investigators hypothesize that VKAs have a deleterious impact on cognition and brain morphology compared to DOACs, due to the decrease in vitamin K bioavailability. A review of the published clinical trials comparing the effects of VKAs and DOACs, especially rivaroxaban, shows that cognition and brain volume were not assessed as outcomes in these trials.

Inclusion Criteria:

• Men or women ≥ 70 years old
• Newly diagnosed hemodynamically stable NVAF longer than 52 hours or of unknown duration, and CHA2DS2-VASc score according to ESC 2016 guidelines for anticoagulation treatment indications
• MMSE score ≥ 20
• Subjects who can give written consent to participate in the study
• Affiliation to a social security scheme.

Exclusion Criteria:

• Known history of stroke and/or a diagnosed condition of dementia (DSM-IV criteria) and/or severe depressive symptomatology (score on the 4-item Geriatric Depression Scale > 3)
• Moderate or severe mitral stenosis
• Conditions other than NVAF that require anticoagulation
• Use of anticoagulant more than 3 days at inclusion or more than 15 days in the preceding 12 months
• Regular use of antiplatelet medications and/or nonsteroidal anti-inflammatory agents and/or azole class of antifungal agents and/or inhibitor of HIV protease
• Acute thromboembolic events or thrombosis (venous/arterial) within the last 14 days prior to randomization
• Known presence of cardiac thombus or myxoma or valvular atrial fibrillation
• Any contraindication to anticoagulation, high risk of bleeding, and any other contraindication listed in the local labeling for the experimental treatment and comparator treatment
• Unstable health, severe hepatic failure, or severe and moderate renal failure (creatinine clearance <50 mL/min), acute coronary syndromes
• Participation in another simultaneous clinical trial
• Inability to understand and speak French
• Refusal to participate from the participant
• Persons deprived of their liberty by administrative or judicial decision, persons under psychiatric care under duress, adults subject to a legal protection measure or unable to express their consent