U-shaped relationship of left ventricular ejection time index and all-cause mortality.

Background: Previous studies have suggested that systolic and diastolic heart failure is associated with alterations of left ventricular ejection time index (LVETI). We sought to examine the relation of LVETI to mortality in an elderly population.

Methods: We prospectively enrolled 852 patients undergoing cardiac catheterization for suspected coronary artery disease (CAD) in 2001 and 2002. LVETI was measured noninvasively using radial applantation tonometry and pulse waveform analysis. Mortality data were assessed by telephone interviews with general practitioners, hospital records, and the national mortality register.

Results: The mean age was 64.8 years, 60.7% of subjects were men, 70.1% of subjects had significant CAD, and 28.6% of subjects had impaired systolic function. After a mean follow-up of 8.2 ± 2.3 years, 183 deaths occurred. At baseline, LVETI was significantly associated with age, systolic and diastolic blood pressure, pulse pressure, and N-terminal probrain natriuretic peptide. A shorter LVETI was significantly and independently associated with impaired systolic function. Kaplan-Meier analysis revealed that both prolonged and shortened ejection time index (1st and 3rd tertile LVETI) were associated with a decreased survival probability (P <0.05, log-rank-test) compared with normal LVETI (2nd tertile). In multivariable Cox regression analysis, the hazard ratios for all-cause mortality were 1.66 for 1st tertile LVETI (P = 0.01) and 1.75 for 3rd tertile LVETI (P = 0.006). The effect of a shortened LVETI on mortality was partly due to the effect of impaired systolic function on ejection duration.

Conclusions: We observed a U-shaped relation between ejection duration and all-cause mortality.