Relation of multi-marker panel to incident chronic kidney disease and rapid kidney function decline in African Americans: the Jackson Heart Study

Table 3 Incremental predictive utility of biomarkers for incident chronic kidney disease (CKD), rapid kidney function decline (RKFD) showing C-statistics and reclassification metrics

	C-Statistics	NRI		IDI		Calibration Statistics^a (χ², P)
	C-Statistics	Events correctly reclassified	Non-Events correctly reclassified	Mean probability for events	Mean probability for non-events	Calibration Statistics^a (χ², P)
Chronic kidney disease (CKD)
Model 1: age-sex-MV^b	0.87	11%	5%	32%	7.4%	12.93 (P = 0.11)
Model 2: age-sex-MV-Biomarker^c	0.88	11%	5%	33%	7.3%	14.26 (P = 0.08)
P-value comparing models 1 vs. 2	0.003	0.08	0.01	0.01
Rapid kidney function decline
Model 1: age-sex-MV	0.76	15%	11%	19.3%	9%	12.19 (P = 0.14)
Model 2: age-sex-MV-Biomarker^d	0.77	15%	11%	20.3%	9%	18.42 (P = 0.02)
P-value comparing models 1 vs. 2	0.10	0.01	<.0001	0.0001

Abbreviations: NRI net reclassification index, IDI integrated discrimination index
^aA Hosmer-Lemeshow goodness-of-fit test indicate poor calibration if P-value < 0.05
^bMV adjusted for age, sex, baseline estimated glomerular filtration rate (eGFR), systolic blood pressure, hypertension, use of hypertension medication, smoking, body mass index (BMI), total cholesterol to high-density lipoprotein cholesterol (HDL) ratio and diabetes
^cIn backward elimination of the biomarker panel, adiponectin and leptin are significant
^dIn backward elimination of the biomarker panel, adiponectin, high-sensitive C - reactive protein (CRP) and aldosterone are significant

ISSN: 1471-2369