Table 3.

Logistic regression models evaluating associations between iron status and DTH-Candida^a (reference: iron replete)

	Model 1: crudea			Model 2: adjusted for identified confounding variablesb			Model 3: adjusted for all individual characteristicsc
	OR	95% CI	P-value	aOR	95% CI	P-value	aOR	95% CI	P-value
Outcome: DTH-Candida
Iron deficient erythropoesis	1.14	0.65, 2.00	0.656	1.33	0.74, 2.39	0.340	1.40	0.70, 2.76	0.347
Iron deficiency anemia	1.71	0.84, 3.48	0.139	2.34	1.08, 5.07	0.031	2.60	0.99, 6.85	0.053
Non-iron deficiency anemia	1.14	0.56, 2.33	0.716	1.27	0.61, 2.63	0.516	1.39	0.60, 3.20	0.444

^aN = 283.

^bControlling for age; N = 283.

^cControlling for age, sex, triceps skinfold thickness, elevated biomarkers of inflammation, early weaning, home construction materials, and household cattle ownership; N = 219.

aOR, adjusted odds ratio; CI, confidence interval; DTH-Candida, delayed-type hypersensitivity to Candida albicans; OR, odds ratio.