TABLE 3.

Adjusted treatment effects for log₁₀-transformed serum concentrations of folate, glutathione, homocysteine, and vitamin B-12 in rural Chinese who had received placebo or garlic, vitamin and selenium, or amoxicillin and omeprazole treatments and percent difference from placebo-treated controls¹

Treatment	n	Serum analyte	Treatment effect (SEE)	P-value	Percent change (95% CI)
Garlic extracts	Treatment, 190 Placebo, 191	Folate	0.042 (0.015)	<0.006	10.2 (2.9 to 18.1)
		Glutathione	−0.007 (0.017)	0.67	−1.6 (−8.7 to 6.1)
		Homocysteine	0.015 (0.018)	0.40	3.4 (−4.6 to 12.2)
		Vitamin B-12	−0.016 (0.025)	0.51	−3.7 (−13.9 to 7.7)
Vitamins C and E and selenium	Treatment, 190 Placebo, 191	Folate	0.010 (0.016)	0.50	2.4 (−4.5 to 9.9)
		Glutathione	0.055 (0.016)	<0.001	13.4 (5.3 to 22.2)
		Homocysteine	−0.028 (0.018)	0.10	−6.4 (−13.5 to 1.3)
		Vitamin B-12	−0.009 (0.025)	0.71	−2.1 (−12.5 to 9.5)
Amoxicillin and omeprazole	Treatment, 131 Placebo, 125	Folate	−0.001 (0.020)	0.95	−0.3 (−9.0 to 9.2)
		Glutathione	0.003 (0.021)	0.88	0.0 (−8.2 to 10.5)
		Homocysteine	−0.013 (0.020)	0.51	3.0 (−11.6 to 6.4)
		Vitamin B-12	−0.047 (0.030)	0.12	−10.2 (−21.7 to 2.9)

¹Robust regression [MM option of SAS Procedure Robustreg (32)] was used to estimate treatment effects adjusted for age, gender, and smoking (see “Methods”).

The percent change, which is derived from the treatment effect as described in Methods, is 100× [(active treatment value/placebo treatment value) −1].