Longitudinal Study of Left Ventricular Mass Index Trajectories and Risk of Mortality in Hypertension: A Cohort Study.

Background Left ventricular mass index (LVMI) has been extensively studied for its relationship with mortality but has been typically assessed at a single time point. We, therefore, describe the trajectory of LVMI in a population with hypertension over 6 years to study the subsequent risk of mortality. Methods and Results We assessed LVMI that was collected during annual health examinations in round 1 (2010-2012), round 2 (2013-2014), and round 3 (2015-2016) with 2 allometric scalings, height ^2.7 , and body surface area, in a cohort of participants with hypertension to identify 6-year trajectories of LVMI by latent class trajectory modeling. We followed up with participants for mortality by latent trajectory from the last echocardiographic examination (September 17, 2014-December 8, 2016) to December 31, 2018. We calculated mortality hazard ratios by assigned trajectory using Cox proportional hazards models. We obtained data for LVMI from 2453 participants (mean age, 61.80 [SD, 12.14] years, 1428 [58.2%] female). We identified 3 trajectories of LVM/H ^2.7 , characterized by maintained low stable (1298 [52.9%]); moderate stable (935 [38.1%]); high stable (220 [9.0%]), as well as 3 trajectories by LVM/body surface area. During a median follow-up of 2.15 years, 167 participants developed all-cause mortality, and 71 were cardiovascular mortality. Only the high stable trajectory was associated with a higher risk of all-cause mortality compared with the low stable trajectory by LVM/H ^2.7 or LVM/body surface area (all P <0.05). In Kaplan-Meier survival analysis, the trajectory with high stable LVM/body surface area had significantly lower survival probability. Conclusions In community hypertension, the individuals with high stable LVMI trajectory had the highest risk of all-cause mortality. The individuals in the moderate stable trajectory had a similar risk for mortality as those in the low stable trajectory.