Table 3.

Effects of PM_2.5 exposure on biomarkers of oxidative stress by supplementation groups.

	Soy oil		Fish oil
	β (SE)	p-Value	β (SE)	p-Value
Cu/Zn SOD
Intercept	0.75 (0.03)	< 0.001	0.60 (0.06)	< 0.001
Supplementationa	0.17 (0.09)	0.050	0.29 (0.14)	0.042
Indoor PM2.5	−0.06 (0.02)	< 0.001	−0.17 (0.05)	0.002
Indoor (PM2.5)2b	—		0.04 (0.02)	0.009
LPOc
Intercept	4.15 (0.23)	< 0.001	3.96 (0.21)	< 0.001
Supplementationa	0.49 (1.13)	0.669	−3.72 (1.04)	0.002
Indoor PM2.5	−0.02 (0.14)	0.904	−0.35 (0.19)	0.080
Indoor (PM2.5)2	—		0.16 (0.07)	0.024
GSHd
Intercept	1.21 (0.10)	< 0.001	1.44 (0.08)	< 0.001
Supplementationa	0.70 (0.22)	0.003	1.00 (0.19)	< 0.001
Indoor PM2.5	−0.03 (0.04)	0.406	−0.09 (0.04)	0.017

Data are regression coefficients; we adjusted models for date. For PM_2.5, the coefficients correspond to a change in 10 μg/m³. Regression models for Cu/Zn SOD include 20 subjects and 74 observations in the soy oil group and 24 subjects and 90 observations in the fish oil group; for LPO, 21 subjects and 51 observations in the soy oil group and 21 subjects and 55 observations in the fish oil group; for GSH, 20 subjects and 76 observations in the soy oil group and 23 subjects and 86 observations in the fish oil group.

^aReference category: presupplementation.

^bLinear relationship; indoor (PM_2.5)² does not apply.

^cSquare root transformed.

^dLog transformed.