Skip to main content
. 2014 Nov 12;11(11):11597–11615. doi: 10.3390/ijerph111111597

Table 3.

Multivariable cox proportional hazards models for the associations of HOUSES and educational level with two-year post-MI mortality.

Regression Models HOUSES Index (Quartiles) Education Level (4 Categories)
(Hazard Ratio, 95%CI, p-Value) b (Hazard Ratio, 95%CI, p-Value) b
4 (ref.) (highest SES) 3 2 1 (lowest SES) p a 4 (ref) (highest SES) 3 2 1 (lowest SES) P a
Model 1 (unadjusted model) 1 1.52 1.76 2.47 <0.001 1 1.01 1.27 2.53 <0.001
(0.85–2.73) (1.01–3.06) (1.46–4.19) (0.59–1.72) (0.79–2.03) (1.55–4.14)
Model 2 1 1.37 1.27 1.86 0.036 1 0.95 1.16 1.84 0.015
(0.75–2.52) (0.71–2.26) (1.07–3.24) (0.55–1.67) (0.70–1.90) (1.11–3.05)
Model 3 (full model) 1 1.29 1.19 1.45 0.24 1 0.82 0.83 0.93 0.84
(0.68–2.43) (0.65–2.16) (0.82–2.58) (0.46–1.44) (0.51–1.37) (0.55–1.57)

Notes: a: p value is statistical significance for overall test (trend) for the association between socioeconomic status and MI mortality using Cochran-Armitage trend tests; b: Hazard ratio per an increment of strata of SES group; Model 1: univariate model only including SES variable as a predictor variable; Model 2: adjusted for all variables listed in Table 1 including sex, hypertension, hyperlipidemia, smoking, BMI, Killip class, STEMI, Anterior MI and ejection fraction, PTCA, CABG, statin, beta blocker and aspirin; Model 3: adjusted for age and Charlson Comorbidity Index in addition to all variables included in Model 2.