Diagnostic utility of zinc protoporphyrin to detect iron deficiency in Kenyan preschool children: a community-based survey

Table 3 Factors associated with ZPP-haem ratio measured in whole blood or erythrocytes, multiple linear regression analysis^a

	Whole blood ZPP-haem ratio (model 1)		Whole blood ZPP-haem ratio (model 2)		Erythrocyte ZPP-haem ratio
	Δ^b	(95% CI)	Δ^b	(95% CI)	Δ^b	(95% CI)
Plasma ferritin concentration, 5.11-fold change^c	−10.1%	(−14.3% to −5.6%)	−9.6%	(−14.0% to −4.9%)	−15.5%	(−21.1% to −9.5%)
Plasma sTfR concentration, 1.43-fold change^c	34.8%	(28.3% to 41.7%)	34.9%	(28.3% to 41.8%)	45.3%	(35.7% to 55.6%)
Plasma CRP concentration, 2.72-fold change^c	5.0%	(0.1% to 10.0%)	5.4%	(0.4% to 10.5%)	9.8%	(2.8% to 17.2%)
Haemoglobin concentration, change by 13.2 g/L^d	−13.9%	(−18.3% to −9.4%)	−14.9%	(−19.4% to −10.1%)	−14.5%	(−20.7% to −7.9%)
Plasmodium infection, any species^e	17.3%	(7.8% to 27.7%)	16.8%	(7.4% to 27.2%)	93.3%	(23.8% to 202.0%)
Plasma albumin concentration, change by 3.9 g/L^d	―	(eliminated)	3.0%	(−1.7% to 8.0%)	2.2%	(0.5% to 4.0%)

CRP C-reactive protein, sTfR soluble transferrin receptor, ZPP zinc protoporphyrin
^aZPP values were normalised by log-transformation; exponentiation of results yielded associations being expressed as relative differences. Plasma concentrations of α ₁-acid glycoprotein (log-transformed), vitamin B₁₂ (log-transformed), sex (binary) and age class (binary) were eliminated from all models through a manual stepwise backward elimination process with an removal criterion of p > 0.05; ^bDifference; ^cCorresponding to 1 geometric standard deviation; ^dCorresponding to 1 SD; ^eBased on HRP2- and pLDH-based dipstick test results

ISSN: 2052-1839