. 2015 Dec 14;11(2):287–296. doi: 10.2215/CJN.01240215

Table 3.

Multivariate logistic regression to model plaque progression at 24 months stratified by CKD stages 3, 4–5, and 5D

	CKD Stage 3		CKD Stages 4 and 5		CKD Stage 5D
	OR (95% CI)	P Value	OR (95 %CI)	P Value	OR (95% CI)	P Value
Age, decades	1.65 (1.24 to 2.25)^a	<0.001^a	1.95 (1.43 to 2.75)^a	<0.001^a	1.48 (1.07 to 2.06)^a	0.02^a
Smoking, current and former versus no	1.77 (1.24 to 2.55)^a	<0.001^a	3.95 (1.91 to 8.50)^a	<0.001^a	1.32 (0.64 to 2.74)	0.45
Diabetes, yes versus no	1.59 (1.06 to 2.41)^a	0.03^a	1.45 (0.92 to 2.31)	0.11	2.37 (0.66 to 10.74)	0.22
Systolic BP ≥150 mmHg	1.46 (1.01 to 2.12)^a	0.04^a	2.84 (1.86 to 4.39)^a	<0.001^a	0.64 (0.29 to 1.40)	0.26
Phosphate >3.6 mg/dl	1.62 (1.09 to 2.43)^a	0.02^a	0.84 (0.54 to 1.30)	0.44	0.80 (0.28 to 2.17)	0.66
Phosphate binders use, yes versus no	0.42 (0.19 to 0.88)^a	0.02^a	0.78 (0.51 to 1.19)	0.25	1.38 (0.56 to 3.34)	0.48
25-hydroxy vitamin D, pg/ml	0.62 (0.43 to 0.89)^a	0.01^a	1.34 (0.82 to 2.20)	0.24	0.08 (0.01 to 0.47)^a	0.01^a
Plaque at baseline, yes versus no	1.33 (0.85 to 2.06)	0.21	3.16 (1.68 to 6.02)^a	<0.001^a	1.20 (0.50 to 2.82)	0.67
Age × phosphate	1.70 (1.16 to 2.54)^a	0.01^a	0.95 (0.63 to 1.42)	0.22	1.02 (0.44 to 2.09)	0.95
Age × plaque at baseline	0.66 (0.45 to 0.96)^a	0.03^a	0.62 (0.41 to 0.93)^a	0.02^a	0.58 (0.29 to 1.11)	0.11
Ferritin >220 ng/ml	1.33 (0.82 to 2.18)	0.25	1.37 (0.89 to 2.14)	0.16	2.54 (1.20 to 5.52)^a	0.02^a
Age × ferritin >220 ng/ml	0.91 (0.58 to 1.49)	0.69	0.66 (0.45 to 0.97)^a	0.04^a	1.22 (0.70 to 2.11)	0.95
Smoking × plaque at baseline	0.45 (0.19 to 1.02)	0.06	0.32 (0.13 to 0.78)^a	0.01^a	2.75 (0.60 to 13.12)	0.20
BMI, kg/m²	0.99 (0.96 to 1.03)	0.76	1.01 (0.97 to 1.05)	0.67	0.79 (0.66 to 0.93)^a	0.01^a
Dyslipidemia, yes versus no	1.27 (0.86 to 1.85)	0.23	1.39 (0.89 to 2.15)	0.14	5.14 (2.34 to 12.01)^a	<0.001^a
Cholesterol >180 mg/dl	1.15 (0.81 to 1.63)	0.45	1.06 (0.71 to 1.6)	0.77	0.31 (0.13 to 0.70)^a	0.01^a
Uric acid, mg/dl	1.08 (0.96 to 1.21)	0.22	1.02 (0.91 to 1.16)	0.72	1.38 (1.06 to 1.84)^a	0.02^a
cIMT, mm	1.07 (0.93 to 1.23)	0.37	1.12 (0.95 to 1.33)	0.18	1.66 (1.22 to 2.32)^a	<0.001^a
Ln 25-hydroxy vitamin D × BMI	1.01 (0.94 to 1.09)	0.82	0.95 (0.87 to 1.03)	0.22	1.16 (1.02 to 1.33)^a	0.02^a

Results are expressed as odds ratios (ORs; exponential-β for independent variables with interactions) and 95% confidence intervals (95% CIs). The following variables were introduced to build multivariate models by CKD stages, because they were significant on bivariate testing or potential confounders: sex, CKD stage, age (decades), diabetes, smoking, dyslipidemia, systolic BP ≥150 mmHg (highest tertile in CKD stage 3), pulse pressure, body mass index (BMI), basal plaque, common carotid artery intima media thickness (cIMT), ferritin >220 mg/dl (highest tertile in CKD stages 4 and 5), uric acid, C-reactive protein, total cholesterol >180 mg/dl (the level of 180 was selected on the basis of clinical criteria), LDL-cholesterol, hematocrit, statins, antiplatelet drugs, triglycerides, 25-hydroxy vitamin D (ln 25-hydroxy vitamin D for interactions and CKD stages 4–5 and 5D and 25-hydroxy vitamin D ≥18.1; vitamin D highest tertile in CKD stage 3), phosphate (highest tertile in CKD stage 3 >3.6 mg/dl), and parathyroid hormone. Only significant variables in multivariate analysis in each group of CKD were included in the final model. The exponential-β and P values of the rest of the variables are the values obtained if added to the final model. CKD stage 3: Hosmer Lemeshow = 0.46; AUC=0.72; CKD stages 4 and 5: Hosmer Lemeshow = 0.82; AUC=0.74; and CKD stage 5D: Hosmer Lemeshow = 0.16; AUC=0.83. AUC, area under the curve.

Significant variables included in each model.