Table 3.

Hierarchical linear regression models exploring joint bleeding as explanatory factors for other oxidative stress biomarkers in all 122 subjects with Heme or Hb entered last after age, sex, and diagnostic group.

.	log10 sf-Heme as explanatory						log10 sf-Hb as explanatory
Dependent factors	aEffect, unstandardized (95%CI)	bStandardized effect	p-Value	Adjusted R2 for model	Adjusted R2 change when adding sf-Heme	Partial correlation	aEffect, unstandardized (95%CI)	bStandardized effect	p-Value	Adjusted R2 for model	Adjusted R2 change when adding sf-Hb	Partial correlation
sf-A1M	2.169 (0.634, 3.705)	0.267	0.006	0.063	0.057	0.257	0.420 (-0.250, 1.091)	0.118	0.217	0.009	0.005	0.114
Serum-A1M	1.934 (-0.272, 4.141)	0.150	0.085	0.227	0.013	0.163	0.263 (-0.674, 1.201)	0.047	0.579	0.206	-0.005	0.051
A1M ratio sf/serum	0.121 (0.036, 0.207)	0.264	0.006	0.083	0.056	0.257	0.038 (0.001, 0.075)	0.190	0.045	0.046	0.025	0.185
Log10 sf-Carbonyl	0.214 (0.015, 0.414)	0.206	0.036	0.035	0.030	0.198	0.082 (-0.002, 0.166)	0.183	0.056	0.026	0.023	0.176

^aEffect (regression coefficient): the estimate in average change in a dependent factor that corresponds to a 1-unit change in the explanatory factor log10-transformed sf-Heme or sf-Hb. ^bStandardized effect: the estimate in average change in a dependent factor expressed in standard deviations that corresponds to a 1 standard deviation change in the explanatory factor log10-transformed sf-Heme or sf-Hb.

Numbers in bold face indicate statistically significance at the 0.05 level.