Left Atrial Strain as Evaluated by Two-Dimensional Speckle Tracking Predicts Left Atrial Appendage Dysfunction in Chinese Patients with Atrial Fibrillation

Table 3

The linearity between PLAS and LAA dysfunction.

Variables	Incidence, n (%)	Crude model			Multivariate-adjusted model 1			Multivariate-adjusted model 2
		OR (95% CI)	value	for trend	OR (95% CI)	value	for trend	OR (95% CI)	value	for trend
PLAS (continuous, %)	Tertiles (n = 248)	0.90 (0.85, 0.95)	0.0005		0.87 (0.83, 0.92)	<0.0001		0.90 (0.85, 0.95)	0.0004
T1: 5.3–17.3	82 (33.3)	1		<0.0001	1		<0.0001	1		0.0340
T2: 17.4–32.0	83 (33.3)	0.23 (0.11, 0.45)	<0.0001		0.23 (0.10, 0.53)	0.0007		0.58 (0.20, 1.65)	0.3046
T3: 32.1–69.5	83 (33.3)	0.05 (0.02, 0.11)	<0.0001		0.04 (0.01, 0.14)	<0.0001		0.20 (0.04, 0.93)	0.0398

Odds ratios were derived from the multivariate logistic regression analysis. Crude, no adjustment. Model I, adjusted for gender, PMH of ablation, INR, course of AF, age, heart rhythm, heart rate, and CHA2DS2-VASc score. Model II, adjusted for gender, PMH of ablation, INR, course of AF, age, heart rhythm, heart rate, CHA2DS2-VASc score, LAD, LA longitudinal diameter, LA transverse diameter, LAVmax, LAVmin, LAEF, E/e′, LVDd, LVSd, LVDV, LVSV, and Simpson LVEF. PLAS, peak left atrial strain. CI, confidence interval.