Gene-Environment Interactions in Predicting Tobacco Use.

Robust standard errors in parentheses clustered at the state level.
***p<0.01,
**p<0.05,
*p<0.1.
Sample weights used.
Notes: Results for regression analyses testing GXE interaction effects on tobacco use reports. This table presents the final results where the main and interaction effects are entered simultaneously. All results use linear probability models (LPM), which is an ordinary least squares (OLS) regression predicting a binary variable outcome (Smoke  = 0/1). The main result is found in Column 2, where the statistical interaction between individual G/G genotype and state level tax rates is negative and statistically significant (P = <0.01). The regression coefficient for Log(Tax) suggests that individuals with the C/C or C/G genotypes are not responsive to higher rates of tobacco taxation. The regression coefficient for the G/G genotype suggests that individuals with this genotype are less likely to report current tobacco use than individuals with the C/G or C/C genotypes. The regression coefficient for the Interaction (G/G X Log (Tax)) suggests that a 10% tax increase reduces the likelihood of reported tobacco use for those with the G/G genotype by 0.73 percentage points more than those with C/G or C/C genotypes. Column 3 includes addition demographic in the analysis to test the robustness of the coefficient on Interaction. See Statistical Analysis section for further details.