Skip to main content
. Author manuscript; available in PMC: 2014 Jul 31.
Published in final edited form as: Arch Intern Med. 2012 Jul 23;172(14):1071–1077. doi: 10.1001/archinternmed.2012.3020

Table 3.

Correlates With Mortality in Multivariate Models

Variable HR (95% CI)a P Value HR (95% CI)b P Value HR (95% CI)c P Value
Age, per 10 y 1.39 (1.28–1.51) <.001 1.38 (1.27–1.50) <.001 1.38 (1.27–1.50) <.001
Male sex 1.01 (0.86–1.20) .87 0.93 (0.79–1.10) .40 1.00 (0.85–1.18) .99
White race 1.35 (1.10–1.65) <.004 1.40 (1.15–1.71) .001 1.35 (1.10–1.65) .004
Medicaid vs other payer 1.16 (0.94–1.43) .18 1.23 (1.00–1.51) .05 1.15 (0.93–1.43) .19
Current smoker 1.49 (1.06–2.10) .02 1.47 (1.04–2.07) .03 1.48 (1.05–2.08) .03
Frailty 1.59 (1.27–1.99) <.001 1.57 (1.25–1.97) <.001
eGFR per 5 mL/min/1.73m2 1.12 (1.02–1.23) .02 1.08 (0.98–1.19) .11
Albumin quartiles/missing <.001 for trend .007 for trend <.001 for trend
 Group 1: ≤2.5 g/dL 1.48 (1.14–1.92) .003 1.45 (1.12–1.88) .005 1.47 (1.13–1.90) .004
 Group 2: >2.5–3.0 g/dL 1.21 (0.90–1.62) .21 1.18 (0.88–1.58) .26 1.21 (0.90–1.63) .21
 Group 3: missing 1.16 (0.88–1.54) .28 1.20 (0.91–1.59) .20 1.16 (0.87–1.53) .31
 Group 4: >3.0–3.5 g/dL 1 [Reference] 1 [Reference] 1 [Reference]
 Group 5: >3.5 g/dL 0.89 (0.68–1.17) .40 0.90 (0.69–1.17) .42 0.89 (0.67–1.17) .40
Hemoglobin quartiles/missing .75 for trend .75 for trend .87 for trend
 Group 1: ≤9 g/dL 1 [Reference] 1 [Reference] 1 [Reference]
 Group 2: >9–10 g/dL 1.06 (0.81–1.38) .68 1.06 (0.82–1.37) .68 1.06 (0.81–1.38) .68
 Group 3: missing 1.08 (0.77–1.52) .65 1.06 (0.76–149) .71 1.06 (0.76–1.49) .73
 Group 4: >10–12 g/dL 1.06 (0.86–1.31) .60 1.02 (0.82–1.26) .88 1.03 (0.83–1.28) .78
 Group 5: >12 g/dL 1.02 (0.75–1.40) .88 0.96 (0.70–1.33) .82 0.98 (0.71–1.36) .92
Hemodialysis vs PD 0.74 (0.55–0.99) .04 0.76 (0.57–1.00) .06 0.74 (0.55–1.00) .05
Comorbidity
 Diabetes mellitus 1.02 (0.86–1.23) .77 1.04 (0.87–1.25) .65 1.00 (0.83–1.20) >.99
 Congestive heart failure 1.12 (0.92–1.36) .27 1.12 (0.93–1.36) .24 1.11 (0.91–1.35) .30
 Atherosclerotic heart disease 1.19 (0.99–1.46) .10 1.17 (0.96–1.42) .11 1.19 (0.97–1.46) .09
 CVA/TIA 0.88 (0.63–1.22) .44 0.94 (0.70–1.28) .70 0.87 (0.63–1.21) .41
 Peripheral vascular disease 1.23 (0.97–1.56) .09 1.25 (0.99–1.58) .06 1.27 (0.97–1.55) .13
 COPD 1.35 (0.99–1.84) .06 1.32 (0.98–1.78) .07 1.32 (0.97–1.81) .08
 Cancer 1.04 (0.74–1.46) .83 1.12 (0.80–1.56) .51 1.04 (0.74–1.46) .83
Erythropoietin use .80 for homogeneity .60 for homogeneity .31 for homogeneity
 Nonusers 1 [Reference] 1 [Reference] 1 [Reference]
 Users 1.01 (0.83–1.22) .94 1.06 (0.86–1.29) .59 1.02 (0.84–1.24) .85
 Unknown or missing 0.92 (0.69–1.21) .55 0.92 (0.70–1.21) .56 0.92 (0.70–1.22) .58
Early nephrology referral .58 for homogeneity .35 for homogeneity .62 for homogeneity
 No 1 [Reference] 1 [Reference] 1 [Reference]
 Yes 0.98 (0.79–1.23) .89 0.96 (0.77–1.19) .69 0.98 (0.79–1.23) .87
 Unknown or missing 1.17 (0.80–1.69) .42 1.20 (0.83–1.74) .32 1.15 (0.79–1.67) .45

Abbreviations: COPD, chronic obstructive pulmonary disease; CVA/TIA, cerebrovascular accident/transient ischemic attack; eGFR, estimated glomerular filtration rate; HR, hazards ratio; PD, peritoneal dialysis.

SI conversion factors: To convert albumin to grams per liter, multiply by 10; to convert hemoglobin to grams per liter, multiply by 10.

a

Multivariate model in which frailty but not eGFR was included among predictor variables.

b

Multivariate model in which eGFR but not frailty was included among predictor variables.

c

Multivariate model in which both eGFR and frailty were included among predictor variables.