Alcohol Consumption and Cigarette Smoking among Young Adults: An Instrumental Variable Analysis Using Alcohol Flushing

Abstract

Association between drinking and smoking has remained controversial since the association between two studies were influenced by various confounding. Thus, our study aimed to explore the causal effect of alcohol consumption and cigarette smoking using alcohol flushing as an instrument variable, which is free from confounders. We analyzed cross-sectional survey data from 2500 Korean young adults (1600 men and 900 women). Alcohol flushing was strongly associated with log transformed alcohol consumption (F = 272). In men, alcohol non-flushers were 1.41 times (95% CI 1.28–1.55) more likely to smoke 100 cigarettes in their lifetime in logistic regression analysis. Alcohol non-flushers were also 1.3 times (95% CI 1.21–1.40) more likely to become daily smokers and 1.39 times (95% CI 1.27–1.51) more likely to be current smokers than alcohol flushers. However, in an IV analysis, no causal relationships between alcohol consumption and smoking status were found. Alcohol consumption, on the other hand, was causally associated with lowering nicotine dependence and former smoking in men. Alcohol consumption determined by alcohol flushing status does not appear to be causally linked to the smoking behavior of young adults. The relationship between alcohol consumption and nicotine dependence and smoking cessation needs further study.