The association between left ventricular mass index and all-cause mortality

Abstract Background Left ventricular hypertrophy (LVH) has been associated with a higher risk of all-cause mortality. However, there is a lack of optimal cut-off values for grading the severity of LVH in the general population. Purpose We aim to investigate the association between the left ventricular mass index (LVMi) estimated by echocardiography and all-cause mortality and determine the optimal cut-off values of LVMi for the general population to predict outcomes. Methods Consecutive participants undergoing echocardiography in a tertiary medical center in Taiwan during 2005-2021 were enrolled. LVMi was calculated using the cube formula from the linear method. The primary outcome was all-cause mortality by linking to the National Death Registry. The association between LVMi and all-cause mortality was assessed using the joinpoint linear regression analysis to derive the optimal cut-off values for LVMi. Using the Kaplan-Meier methods, we validated our results with a community-based cohort with a long-term follow-up period. Results 85,503 participants (age 64.3 ± 16.9 years, 53.7% men) were enrolled. During a mean follow-up of 67.6 ± 48.5 months, there were 14,565 deaths (17%). A U-shaped relationship between LVMi and the risk of all-cause mortality could be observed. The joinpoint linear regression analysis revealed that LVMi could be categorized into the following groups (less than 66 gm/m2, 66-87 gm/m2, 88-132 gm/m2, 133-170 gm/m2, and more than 171 gm/m2), with each group representing a different risk for all-cause mortality. (Figure 1) In the validated cohort categorized by the cut-off values of LVMi, the Kaplan-Meier analysis showed divergent risks for all-cause mortality between different groups in the community-indwelled populations (Figure 2, log-rank P-value< 0.001). Conclusion There was a U-shaped relationship between LVMi and the risk of all-cause mortality. Grading for the severity of LVH using the LVMi is warranted for the risk assessment in the general population.Figure 1Figure 2