Table 5. Multiple logistic regression analysis of dyslipemia in Japanese alcoholic men according to combinations of ADH1B and ALDH2 genotypes after adjustment for other independent variables.

ADH1B genotype & ALDH2 genotype	TG level ≥150 mg/dl, vs. <150 mg/dl			TG level ≥250 mg/dl, vs. <150 mg/dl			HDL-C level >80 mg/dl, vs. ≤80 mg
	OR	95% CI	p	OR	95% CI	p	OR	95% CI	p
*1/*1 & *1/*2 (n = 90)	1	(referent)		1	(referent)		1	(referent)
*1/*1 & *1/*1 (n = 400)	0.95	(0.50–1.82)	0.89	3.13	(0.40–24.40)	0.28	0.87	(0.52–1.46)	0.60
*2 allele & *1/*2 (n = 188)	1.44	(0.72–2.89)	0.30	5.09	(0.62–42.06)	0.13	0.72	(0.40–1.27)	0.25
*2 allele & *1/*1 (n = 1128)	2.29	(1.25–4.18)	0.007	9.63	(1.29–72.09)	0.027	0.26	(0.16–0.43)	< .0001
P for homogeneity a			< .0001			0.0001			< .0001
P for interaction b			0.19			0.65			0.009
ADH1B genotype & ALDH2 genotype	HDL-C level ≥100 mg/dl, vs. ≤80 mg/dl			HDL-C level <40 mg/dl, vs. ≥40 mg/dl			LDL-C level ≥140 mg/dl, vs. <140 mg/dl
	OR	95% CI	p	OR	95% CI	p	OR	95% CI	p
*1/*1 & *1/*2	1	(referent)		1	(referent)		1	(referent)
*1/*1 & *1/*1	0.74	(0.38–1.46)	0.38	0.88	(0.43–1.79)	0.72	1.88	(0.97–3.66)	0.061
*2 allele & *1/*2	0.36	(0.16–0.82)	0.015	0.90	(0.41–1.97)	0.78	1.15	(0.55–2.43)	0.71
*2 allele & *1/*1	0.11	(0.05–0.22)	< .0001	1.50	(0.78–2.91)	0.23	1.01	(0.53–1.93)	0.97
P for homogeneity a			< .0001			0.007			0.001
P for interaction b			0.061			0.14			0.063

OR, odds ratio; CI, confidence interval.

ORs were adjusted for age, usual ethanol consumption, alcoholic beverage most frequently consumed, current smoking, BMI, presence of diabetes, hypertension, and liver cirrhosis.

^a P for overall no-association between the 4 combinatorial groups of ADH1B and ALDH2 genotypes.

^b P for interaction of ADH1B by ALDH2 genotypes; a significant p value means that the effect of ADH1B on lipid values is modified by ALDH2, or vice versa