The role of genotype in determining the effects of cigarette smoking on pulmonary function

The effect of cigarette smoking on pulmonary function is highly variable. Some heavy smokers retain normal pulmonary function and others experience profound pulmonary function loss. The role of genotype in this process is unknown. We tested for gene by environment interaction (GxE) in smoking-associated loss of forced expiratory volume in one second (FEV1) using repeated pulmonary function and smoke-exposure measures from 352 twin pairs enrolled in the NHLBI Male Veteran Twin Study. Inferences using global [Jinks and Fulker, 1970] and dichotomous [Ramakrishnan et al., 1992] procedures were supplemented with a new model for repeatedly observed twinships. The model facilitates testing of standard heritability hypotheses and families of hypotheses regarding the dependence of quantitative twin-twin phenotypic similarity on continuously varying twin-twin exposure concordance. The global and dichotomous procedures were suggestive of GxE in smoke-associated loss of FEV1 (p < 0.01, p = 0.08, respectively). With the new model, overall twin-twin correlation of FEV1 for concordant-smoking MZ and DZ twin pairs was estimated at 0.71 and 0.34, respectively. For twins with little or no difference in cigarette use, the intra-pair correlations of FEV(1) did not differ according to cigarette exposure over a wide range of exposures (0 - > or = 200 pack years). Even great twin-twin discordance in cigarette smoking (> or =10 pack years) had little effect of correlations. We conclude that a constant factor, such as genotype, appears to be interposed between the environmental toxin (cigarette smoke) and phenotype (FEV1).