Subclinical left ventricular dysfunction assessed by global longitudinal strain correlates with mild cognitive impairment in hypertensive patients

Abstract

Prevention of dementia represents a public health priority. Hypertension is a risk factor for mild cognitive impairment (MCI), a precursor to progressive dementia. A great effort is underway to develop accurate and sensitive tools to detect the MCI condition in hypertensive patients. To investigate the potential association of subclinical left ventricular dysfunction expressed by the global longitudinal strain (GLS) with the MCI, defined by the Italian version of the quick mild cognitive impairment (Qmci-I). This multi-centric study included 180 consecutive hypertensive patients without medical diseases and/or drugs with known significant effects on cognition but with a not negligible comorbidity burden to avoid a possible "hyper-normality bias". The study cohort was classified into two main groups concerning the median value of the GLS. A weighted logistic regression model was employed after an inverse probability of treatment weighting (IPTW) analysis to characterize a potential association between GLS and MCI. Almost 41,1% of the whole study population was female. The mean age was 65,6 ± 7,2. 39 patients (21,7%) showed MCI. After IPTW, the GLS was significantly associated with the study endpoint (OR, 1,22; 95% CI: 1,07-1,39, P = 0.003). Our results highlight that the GLS is a potential predictor of MCI and, therefore, a valuable tool for establishing preventive strategies to arrest the progression toward a cognitive decline in hypertensive patients.

Keywords: Brain; Cognitive decline; Global longitudinal strain; Propensity weighting analysis; Vascular risk factors.

Fig. 1

Representation of the correlation, with the relative coefficients, between the GLS* (A) and the EF (B) with the i-Qmci. *We used the GLS absolute value for easier interpretation. GLS global longitudinal strain, EF ejection fraction, i-Qmci Italian version of quick mild cognitive impairment

Fig. 2

A Propensity score distributional overlap before (unadjusted) and after (adjusted) propensity score weighting. B Love plot for standardized mean differences comparing covariate values before (red circle) and after (blue triangle) propensity score weighting. A standardized mean differences <0.1 was considered acceptable to support the assumption of balance between the GLS sub-groups