Capitalising on Tax Variation to Estimate the Effect of Alcohol Consumption on Edentulism.

Objective: To estimate alcohol consumption's effect on edentulism using state alcohol taxes as an instrumental variable (IV).

Methods: Analysis of 514,357 U.S. Behavioural Risk Factor Surveillance System participants (2003-2006, 2008, 2010, 2012) linked to state alcohol taxes. We used IV regression modelling to assess the relationship between alcohol consumption and edentulism, plus potential mediators (body mass index, dental visits) and a positive control (coronary heart disease). Robustness to imperfect exogeneity was evaluated through sensitivity analyses and falsification testing using IV analysis on individuals under the age of 16 years.

Results: A 1.1-drink increment per day was associated with a 12% higher risk of edentulism (95% confidence interval: 9%-16%). Alcohol consumption was positively associated with body mass index, dental visits and coronary heart disease. No significant effect on edentulism was observed in the negative control population (individuals aged < 16 years).

Conclusions: The findings of this quasi-experimental study suggest that alcohol consumption increases the risk of edentulism.