Assessing the Relationship Between Mental Distress and Tobacco Use in Post-Katrina and Rita Louisiana

Abstract

ABSTRACTBackground: The relationship between mental health status and smoking is complicated and often confounded by bi-directionality, yet most research on this relationship assumes exogeneity. Objectives: The goal of this article is to implement an instrumental variable approach to (1) test the exogeneity assumption and (2) report on the association between mental health status and smoking post-disaster. Methods: This analysis utilizes the 2006 and 2007 Louisiana Behavioral Risk Factor Surveillance Survey to examine the link between mental distress and smoking in areas affected by Hurricanes Katrina and Rita. Residence in a hurricane-affected parish (county) was used as an instrumental variable for mental distress. Results: Just over 22% of the sample resided in a hurricane-affected parish. Residents of hurricane-affected parishes were significantly more likely to report occasional and frequent mental distress. Residence in a hurricane-affected parish was not significantly associated with smoking status. With residence established as a salient instrumental variable for mental distress, the exogeneity assumption was tested and confirmed in this sample. A dose-response relationship existed between mental distress and smoking, with smoking prevalence increasing directly (and non-linearly) with mental distress. Conclusions: In this sample, the relationship between mental distress and smoking status was exogenous and followed a dose-response relationship, suggesting that the disasters did not result in an uptake of smoking initiation, but that the higher amounts of mental distress may lead to increased use among smokers. The findings suggest that tobacco control programs should devise unique strategies to address mentally distressed populations.