. Author manuscript; available in PMC: 2020 Jun 13.

Published in final edited form as: Int J Cardiol. 2013 Feb 4;168(2):1496–1501. doi: 10.1016/j.ijcard.2012.12.102

Table 3.

Univariate and multivariate Cox regression analyses for HRR in the overall group and peak RER subgroups.

	HRR as continuous variable		HRR as dichotomous variable (≤/>12 beats)
	Chi-square	Hazard ratio & adjusted hazard ratio (95% confidence interval)	Hazard ratio & adjusted hazard ratio (95% confidence interval)
Overall group (163 events)	35.9	0.95 (0.94–0.97)^***	2.4 (1.7–3.4)^***
Overall group (163 events)	35.9	0.97 (0.95–0.98)^**	1.6 (1.2–2.2)^**
Peak RER< 1.00 subgroup (29 events)	12.8	0.93 (0.89–0.97)^***	4.1 (2.0–8.6)^***
Peak RER< 1.00 subgroup (29 events)	12.8	0.92 (0.88–0.96)^**	4.2 (2.0–8.8)^***
Peak RER 1.00–1.09 subgroup (46 events)	9.1	0.95 (0.91–0.98)^**	2.3 (1.3–4.0)^**
Peak RER 1.00–1.09 subgroup (46 events)	9.1	0.95 (0.91–0.98)^**	2.1 (1.1–3.7)^*
Peak RER> 1.10 subgroup (88 events)	19.0	0.96 (0.94–0.98)^**	2.0 (1.3–3.1)^**
Peak RER> 1.10 subgroup (88 events)		0.98 (0.96–1.00)^*	-

Multivariate Cox regression was performed to obtain the adjusted hazard ratio to examine HRR in relation to other patient characteristics and CPX results. The number of predictor variables included in each model was different because the number of events in each RER subgroup was different. Variables in each model included: overall group (age, heart failure etiology, beta-blocker use, LVEF, resting heart rate, peak VO₂, VE/VCO₂ slope, and HRR), peak RER<1.0 (peak VO₂, VE/VCO₂ slope, and HRR), peak RER 1.0–1.09 (LVEF, peak VO₂, VE/VCO₂ slope, and HRR), peak RER≥1.10 (age, heart failure etiology, beta-blocker use, LVEF, resting heart rate, peak VO₂, VE/VCO₂ slope, and HRR).

p<0.05.

^**

p<0.01.

^***

p<0.001.