Table 1.

Distribution (Number) of Participants by Ancestry in a Genome-Wide Meta-Analysis of Gene-Alcohol Interaction and Lipid Levels, 2017^a

Analysis Stage	Ancestry Group					Meta-Analysis
	European	African	Asian	Hispanic	Brazilian
1	89,893	20,989	12,450	3,994	0	127,326
2b	136,986	4,475	108,431	13,714	3,652	267,258
Totalc	226,879	25,464	120,881	17,708	3,652	394,584

Abbreviation: df, degrees of freedom.

^a For each lipid trait, association analyses were performed accounting for 2 alcohol consumption status variables: “current drinker” and “regular drinker.” For each ancestry group, study-specific results were combined to perform the 1-df test for an interaction effect and the 2-df joint test of the genetic main effect and interaction with drinking exposure. Persons from 5 ancestry groups were included: European, African, Asian, Hispanic, and Brazilian.

^b Variants selected for follow-up at P ≤ 1 × 10⁻⁶ using a 1-df interaction test and a joint 2-df interaction test.

^c Variants found to be significant at P ≤ 5 × 10⁻⁸ using a joint 2-df interaction test or a 1-df interaction test.