Association between Left Atrial Late Gadolinium Enhancement and Atrial Fibrillation: The Multi-Ethnic Study of Atherosclerosis (MESA).

Purpose: To determine the prevalence and correlates of left atrial (LA) late gadolinium enhancement (LGE) at cardiac MRI and its association with atrial fibrillation (AF) in a population-based sample from the Multi-Ethnic Study of Atherosclerosis (MESA).
Materials and Methods: In this secondary post hoc analysis of the MESA cohort (ClinicalTrials.gov no. NCT00005487), participants without AF underwent LGE cardiac MRI at the fifth examination (2010-2012). LA LGE burden was quantified using the image intensity ratio technique on biplane long-axis two-dimensional (2D) LGE images without fat saturation. Survival analysis was performed with log-rank testing and Cox regression.
Results: Of 1697 participants (mean age, 67 years ± 9 [SD]; 872 men), 1035 (61%) had LA LGE, and 75 (4.4%) developed AF during follow-up (median, 3.95 years). At univariable analysis, LA LGE was associated with age (β = .010 [95% CI: .005, .015], P < .001), diastolic blood pressure (β = .005 [95% CI: .001, .009], P = .02), HbA1c level (β = .06 [95% CI: .02, .11], P = .009), heart failure (β = .60 [95% CI: .11, 1.08], P = .02), LA volume (β = .008 [95% CI: .004, .012], P < .001), and LA function (emptying fraction, LA global longitudinal strain, LA early diastolic peak longitudinal strain rate, and LA late diastolic peak strain rate; all P < .05). After adjusting for the variables in the Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) AF score, LA LGE independently helped predict incident AF (hazard ratio = 1.46 [95% CI: 1.13, 1.88], P = .003). The highest tertile (LGE > 2%) was twice as likely to develop AF.
Conclusion: Although limited by the 2D LGE technique employed, LA LGE was associated with adverse atrial remodeling and helped predict AF in a multiethnic population-based sample.Clinical trial registration no. NCT00005487 Keywords: MR Imaging, Cardiac, Epidemiology Supplemental material is available for this article. © RSNA, 2023.
Competing Interests: Disclosures of conflicts of interest: T.Z. No relevant relationships. T.Q.A.C.S. No relevant relationships. B.A.V. No relevant relationships. E.X. No relevant relationships. M.R.O. No relevant relationships. M.H. No relevant relationships. D.A.B. Editor emeritus for Radiology. E.Z.S. No relevant relationships. C.O.W. No relevant relationships. S.R.H. Support for the present article from the National Institutes of Health (NIH), paid to author's institution. S.N. Research funding from Biosense Webster, Imricor, Siemens, ADAS software, and the NIH; consulting fees from CardioSolv and Circle CVI; payment for lectures from Circle CVI. J.A.C.L. NIH grant support.
(© 2023 by the Radiological Society of North America, Inc.)