Table 2.

Association of serum bicarbonate with the primary CVD outcome, heart failure and all-cause mortality

	Serum bicarbonate <22 mEq/L		Serum bicarbonate >26 mEq/L
Outcome	HR (95% CI)	P	HR (95% CI)	P
Primary CVD outcome
Model 1	2.02 (1.47–2.79)	<0.001	0.97 (0.82–1.15)	0.73
Model 2	2.00 (1.45–2.76)	<0.001	0.95 (0.80–1.12)	0.53
Model 3	1.54 (1.11–2.14)	0.01	1.04 (0.88–1.24)	0.65
Heart failure
Model 1	1.86 (1.04–3.34)	0.04	0.86 (0.62–1.19)	0.35
Model 2	1.70 (0.93–3.09)	0.08	0.83 (0.60–1.16)	0.27
Model 3	0.98 (0.52–1.86)	0.95	0.97 (0.69–1.36)	0.86
All-cause mortality
Model 1	1.57 (1.03–2.38)	0.04	0.83 (0.67–1.03)	0.10
Model 2	1.51 (0.99–2.31)	0.06	0.80 (0.64–1.00)	0.05
Model 3	0.98 (0.63–1.54)	0.94	0.91 (0.73–1.14)	0.40

HRs are based on Cox regression and are relative to the reference group of a serum bicarbonate between 22 and 26 mEq/L. The primary CVD outcome includes nonfatal MI, acute coronary syndrome not resulting in an MI, nonfatal stroke, nonfatal acute decompensated heart failure and death from cardiovascular causes. Heart failure includes nonfatal and fatal acute decompensated heart failure.

Model 1: adjusted for treatment group.

Model 2: Model 1 + adjustment for age, sex and race/ethnicity.

Model 3: Model 2 + adjustment for smoking status, history of CVD, systolic BP, use of diuretics, ACEi or angiotensin II receptor blockers, eGFR, log of UACR and serum potassium.

There were a total of 618 primary CVD events, 173 heart failure events and 378 deaths included in the analyses.