Mild cognitive impairment in hypertensive patients with atrial fibrillation: the type of anticoagulation makes a difference

Abstract Background In hypertensive patients, new tools have been developed to detect mild cognitive impairment (MCI), which is an early-stage cognitive performance loss. However, cognitive assessment in hypertensive patients affected by atrial fibrillation (AF) is often neglected, and while the anticoagulation therapy represents the standard in prevention of stroke, its effect on mild cognitive impairment is still not fully clarified. Purpose Consequently, our aim was to study a population of hypertensive patients with AF, to search for MCI and its predictive factors, with a particular attention to the potential role of the anticoagulation regimen (VKA or DOAC). Methods We enrolled 179 patients (age 69.71±0.99years;70.39%male) from two departmental hypertension clinics. The patients had a history of paroxysmal or permanent AF and arterial hypertension, in therapy with oral anticoagulants and no history of cerebrovascular and/or neurodegenerative diseases. In particular, we divided the population into two groups: patients with MCI (MCI group) and without MCI (No MCI), on the basis of their cognitive performance as assessed by a specific and sensitive screening test (Qmci total49.4 defined the presence of MCI). Due to the small sample size, the relative low number of MCI patients and the consistent number of potential MCI predictors, in order to avoid multicollinearity and select only the most significant covariates, we employed first a penalized elastic net regression analysis and, successively, an unpenalized multivariable regression model to detect the variables significantly associated with MCI. Results Firstly, we recorded a not negligible rate of MCI in the enrolled population(32.96%). Analyzing the clinical parameters, the two groups (MCI vs No MCI) did not differ in age (70.38±1.61 vs 69.47±0.74years;p0.554), while differed in anticoagulation regimen (DOAC:37.29% vs 81.67%,p0.001), mechanical heart valves (MHV: 0 vs 6.67%, p 0.014), urea (61.84±4.05 vs 47.47±2.67 ml/min;p 0.003), incidence of paroxysmal AF (45.76 vs 63.33 %,p 0.023) and presence of CKD (47.46 vs 30.00 %,p 0.042), presence of left atrium enlargement (LAE:84.48% vs 71.67%;p 0.028) and presence of left ventricular hypertrophy (LVH:73.21% vs 41.67%,p 0.001). After the penalized elastic net framework, only five variables were retained: anticoagulation regimen, Mini Nutritional Assessment (MNA) score, LVH, glycemia and uremia. However, after unpenalized multivariable logistic regression only the anticoagulation regimen, in particular the VKA use, was statically associated with MCI (Fig. 1). Conclusions MCI has a high prevalence in patients with AF, and is more pronounced in patients receiving VKA, than in patients receiving DOACs. Above all, it might be useful to introduce Qmci test in clinical practice to monitor the cognitive profile of hypertensive patients with AF, in order to identify early compromised subjects and intervene with appropriate therapeutic strategies.For image description, please refer to the figure legend and surrounding text.