Magnetic resonance-based assessment of myocardial 2-dimensional strain using feature tracking association with cardiovascular risk factors in a population-based cohort free of cardiovascular disease.
Purpose: Myocardial strain analysis is a promising tool for the detection of subtle but relevant alterations of left ventricular function, also in asymptomatic subjects. Thus, we determined the feasibility of cardiac magnetic resonance-based 2D global strain analysis using feature tracking and its association with cardiovascular risk factors in a sample from the general population. Materials and Methods: Subjects without a history of cardiocerebrovascular disease were enrolled in a substudy of the population-based KORA (Cooperative Health Research in the Region of Augsburg) cohort. In all participants with the absence of late gadolinium enhancement, longitudinal and circumferential global strains were measured on Cine SSFP imaging (TR: 29.97 ms, TE: 1.46 ms, ST: 8 mm), using a semiautomatic segmentation algorithm (CVI42, Circle, Canada). Differences in strain values according to age, sex, body mass index, hypertension, diabetes mellitus, and hyperlipidemia were derived using linear regression analysis. Results: Among 360 subjects (mean age, 56.2 +/- 9.2 y, 57% male), the average global systolic radial strain was 40.1 +/- 8.2%, circumferential 19.9 +/- 2.7%, and longitudinal 19.8 +/- 3.2%. Male sex was associated with decreased global strain values, independent of the strain direction (all P<0.001). Although many cardiovascular risk factors were correlated with strain in univariate analysis, mainly waist-to-hip ratio and HbA1c remained associated with decreased radial and circumferential strains in fully adjusted models. Similarly, higher radial and circumferential strains were observed in older subjects (beta=0.14, P=0.01 and beta=0.11, P=0.04, respectively). Conclusions: Strain analysis using magnetic resonance feature tracking is feasible in population-based cohort studies and shows differences with respect to age and sex as well as an independent association with markers of metabolic syndrome.