Table 3.

CVD mortality according to quartiles of serum selenium concentrations among patients with CKD^a.

Characteristics	Serum selenium concentration (μg/L)
	Q1<156.1	Q2 156.1–181.7	Q3 181.7–201.5	Q4 201.5–734.8	P -trend
CVD mortality
Number of deaths/total	185/776	74/768	40/762	37/767	336/3,063
Model 1b	1	0.878 (0.670, 1.153)	0.560 (0.394, 0.794)	0.535 (0.371, 0.771)	P < 0.001
Model 2c	1	0.865 (0.657, 1.139)	0.540 (0.380, 0.767)	0.525 (0.363, 0.758)	P < 0.001
Model 3d	1	0.834 (0.633, 1.100)	0.526 (0.370, 0.748)	0.513 (0.356, 0.739)	P < 0.001
Model 4e	0.992 (0.989, 0.996)				P < 0.001

^aCompeting risk models were used to estimate the HRs (95% CIs) of cardiovascular-cause mortality according to quartiles of serum selenium concentrations. Q, quartile.

^bModel 1 was adjusted for age (continuous), sex (male or female), and race (non-Hispanic white or other).

^cModel 2 was adjusted for age (continuous), sex (male or female), race (non-Hispanic white or other), family income–poverty ratio (>3.0, 1.1–3.0, and ≤1), BMI (≥30 or <30 kg/m²), serum triglycerides (≥200 or <200 mg/dL), serum total cholesterol (≥240 or <240 mg/dL), and serum uric acid (≥7 or <7 mg/dL).

^dModel 2 was adjusted for age (continuous), sex (male or female), race (non-Hispanic white or other), family income–poverty ratio (>3.0, 1.1–3.0, and ≤1), BMI (≥30 or <30 kg/m²), serum triglycerides (≥200 or <200 mg/dL), serum total cholesterol (≥240 or <240 mg/dL), serum uric acid (≥7 or <7 mg/dL), diabetes (yes or no), hypertension (yes or no), and smoking status (smoker or never smoker).

^eContinues model (each per 1% increase in serum selenium concentrations) adjusted by variables in model 3.