Skip to main content
. 2023 Jun 19;52(5):1545–1556. doi: 10.1093/ije/dyad082

Table 2.

Cox proportional hazard results for red blood cell distribution width (RDW), C-reactive protein (CRP) and 25 hydroxyvitamin D [25(OH)D] associations with all-cause mortality, adjusted for sex, age, ethnicity, body mass index (BMI), smoking pack-years, drinking and deprivation index

Model All-cause mortality hazard ratio per 1% increase in the RDW (95% CI) N = 324 467 All-cause mortality hazard ratio per 1-mg/L increase in CRP (95% CI) N = 317 917 All-cause mortality hazard ratio per 1-nmol/L increase in 25(OH)D (95% CI) N = 303 858
(i) Uncorrected Cox proportional hazard 1.203 (1.194, 1.212) 1.029 (1.028, 1.031) 0.993 (0.992, 0.994)
(ii) Corrected for regression dilution in main exposure using intraclass correlation coefficient 1.430 (1.407, 1.453) 1.107 (1.099, 1.116) 0.988 (0.986, 0.989)
(iii) Corrected for regression dilution in main exposure using regression calibration 1.482 (1.398, 1.567)a 1.119 (1.095, 1.143)a 0.988 (0.987, 0.990)a
(iv) Corrected for regression dilution in main exposure and both BMI and smoking pack-years (confounders) using regression calibration 1.480 (1.394, 1.565)a 1.118 (1.095, 1.142)a 0.989 (0.987, 0.990)a
a

Bootstrapped 95% CIs with 10 000 replications.