Table 4

Multivariable Cox regression model adjusted for age, arterial disease and cystatin C (208 patients).

OM

CVM

N-VM


P-value

HR (95% CI)

P-value

HR (95% CI)

P-value

HR (95% CI)

Smoking

0.02

3.4 (1.3-9.2)

SLICC>1[25]

0.008

4.1 (1.4-17.3)

0.05

5.6 (1.0-103.6)

2GP1

0.03

3.4 (1.2-9.7)

Any aPL medium titer

0.05

2.8 (1.0-8.2)

Sjogrens syndrome B antibodies

0.02

1.3e-6 (0-0.7)

Warfarin

0.05

3.4 (1.0-10.4)

High sensitivity CRP

0.04

1.3 (1.0-1.6)

0.02

1.6 (1.1-2.3)

Fibrinogen

0.04

3.7 (1.0-13.1)

0.05

6.7 (1.0-45.4)

α-1-antitrypsine

0.007

2.7 (1.3-5.2)

0.004

4.3 (1.6-10.7)

Soluble vascular cell adhesion molecule 1

0.05

2.7 (1.0-6.7)

0.02

5.3 (1.3-19.3)


2GP1, anti-β2 glycoprotein-1; any aPL, any antiphospholipid; positive antibody against cardiolipin IgG/IgM at medium titer, β2GP1 or a positive lupus anticoagulant test; OM, overall mortality; CVM, cardiovascular mortality; HR, hazard ratio; N-VM, non-vascular mortality; SLICC, systemic lupus international collaborating clinics. Age, arterial disease and cystatin C were included in all multivariable models. Each variable that was significant in age-adjusted Cox regression (Table 3) for OM, CVM and N-VM, was included one by one, together with the three above-mentioned variables, creating a model with four variables. The table presents the results of the significant associations. Only variables that have P-values ≤ 0.05 in either group are presented. In this way, five different multivariable models are presented for OM, six for CVM, and four for N-VM, respectively.

Gustafsson et al. Arthritis Research & Therapy 2012 14:R46   doi:10.1186/ar3759

Open Data