Change in left atrial function and volume predicts incident heart failure with preserved and reduced ejection fraction: Multi-Ethnic Study of Atherosclerosis.

ABSTRACT

AIMS: The role of change in left atrial (LA) parameters prior to the onset of heart failure (HF) remains unclear. We used cardiac magnetic resonance (CMR) imaging to investigate the relationship between longitudinal change in LA function and incident HF in a multi-ethnic population with subclinical cardiovascular disease (CVD). METHODS AND RESULTS: In this prospective multi-ethnic cohort study, 2470 participants (60 ± 9 years, 47% males), free at baseline of clinical CVD, had LA volume and function assessed via multimodality tissue tracking on CMR imaging at baseline (2000-02) and a second study 9.4 ± 0.6 years later. Free of HF, 73 participants developed incident HF [HF with preserved ejection fraction (HFpEF), n = 39; reduced ejection fraction (HFrEF), n = 34] 7.1 ± 2.1 years after the second study. An annual decrease of 1 SD unit in peak LA strain (ΔLASmax) was most strongly associated with the risk of HFpEF [subdistribution hazard ratios (HR) = 2.56, 95% confidence interval (CI) (1.34-4.90), P = 0.004] and improved model reclassification and discrimination in predicting HFpEF [C-statistic = 0.84, 95% CI (0.79-0.90); net reclassification index (NRI) = 0.34, P = 0.01; and integrated discrimination index (IDI) = 0.02, P = 0.02], whilst an annual decrease of 1 mL/m2 of pre-atrial indexed LA volumes (ΔLAVipreA) was most strongly associated with the risk of HFrEF [subdistribution HR = 1.88, 95% CI (1.44-2.45), P < 0.001] and improved model reclassification and discrimination in predicting HFrEF [C-statistic = 0.81, 95% CI (0.72-0.90); NRI = 0.31, P = 0.03; and IDI = 0.01, P = 0.50] after adjusting for event-specific risk factors and baseline LA measures. CONCLUSION: ΔLASmax and ΔLAVipreA were associated with and incrementally predictive of HFpEF and HFrEF, after adjusting for risk factors and baseline LA measures in this population of subclinical CVD.