Table 2.

Association of apolipoprotein A-IV with history of cancer (615 and 611 out of 5039 patients). Results are given for both eGFR adjustments (once calculated with cystatin C and once with creatinine)

	eGFRcystatin-C adjusted			eGFRcreatinine adjusted
	OR	95% CI	p-value	OR	95% CI	p-value
Calculations for median of apoA-IV concentrationsa
Model 1	0.85	0.71–1.03	0.09	0.81	0.67–0.97	0.02
Model 2	0.84	0.70–1.02	0.07	0.79	0.66–0.96	0.02
Calculations per quartile of ApoA-IV concentrations
Model 1
Quartile 1: 163 casesb	1.00			1.00
Quartile 2: 174 casesb	1.07	0.85–1.35	0.58	1.03	0.82–1.31	0.79
Quartile 3: 147 casesb	0.89	0.69–1.14	0.36	0.84	0.65–1.07	0.16
Quartile 4: 131 casesb	0.88	0.67–1.15	0.35	0.80	0.61–1.06	0.12
Model 2
Quartile 1: 163 casesb	1.00			1.00
Quartile 2: 172 casesb	1.05	0.83–1.33	0.69	1.01	0.80–1.28	0.92
Quartile 3: 144 casesb	0.87	0.68–1.12	0.27	0.81	0.63–1.05	0.11
Quartile 4: 128 casesb	0.86	0.65–1.14	0.29	0.78	0.59–1.04	0.09

Model 1: adjusted for age, sex, ln-urine albumin-creatinine ratio, eGFR_{cystatin C} or eGFR_creatinine

Model 2: as model 1 plus statin use, smoking, BMI, diabetes

Abbreviations: OR odds ratio, CI confidence interval

^aReference category includes apoA-IV values below median. The median apoA-IV concentration is 27.6 mg/dL (referring to the total group of 5039 patients). There are 278 cases above, and 337 cases below the median (model 1), and 272 cases above, and 335 cases below the median (model 2). Small differences in the number of cases between eGFR_{cystatin C} (n = 615) and eGFR_creatinine (n = 611) adjustment are caused due to few more missing variables for eGFR_creatinine

^b“Cases” refers to the number of patients with history of cancer. Differences in number of cases between model 1 and 2 are explained by patients with few missing covariates for model 2