If a significant genetic correlation exists between smoking and alcohol dependence, smoking in one twin should correlate more highly in monozygotic than in dizygotic pairs, with alcohol dependence experienced by the co-twin. The two-way interaction between whether twins are monozygotic or dizygotic (i.e., twin-pair zygosity) and whether one twin reports a history of smoking provides a direct statistical test for this genetic correlation (Heath et al. 1998).
