Using Latent Class Analysis to Identify Cigarette Smoking Patterns in US National Data

Cigarette smoking is the number one preventable cause of death in the US. Internationally, tobacco use is expected to kill 10 million people annually through 2030. For many years, the prototypical cigarette smoker was conceptualized as a daily smoker, consuming 20 or more cigarettes a day. However, interest has been increasing in so-called light and intermittent smoking. One recent paper reported that over 70% of Latino smokers in California were "low-frequency" smokers. Such large prevalences of light and intermittent smokers have important implications for both addiction theory and prevention/intervention approaches. However, there are no currently accepted definitions of light and intermittent smoking. This work explores adult cigarette smoking patterns in national data with latent class analysis. Measurement and model selection issues will be emphasized.