be relevant. For example, we have recently shown that state tobacco control policies may influence suicide risk (Grucza et al., 2014). Tobacco control policies also likely influence the prevalence of marijuana use (Chaloupka et al., 1999; Farrelly et al., 2001; Williams et al., 2004), and may influence alcohol use which could be an important determinant of suicide risk (Kaplan et al., 2014; Krauss et al., 2014; Young-Wolff et al., 2013a, 2013b). We also included measures of state political orientation, per-capita mental health spending, and health insurance coverage, all of which may be associated with suicide risk (Kposowa, 2013; Tondo et al., 2006; Yoon and Bruckner, 2009) and are plausibly related to state policy environments. If these factors changed concurrently with adoption of medical marijuana policy, lack of explicit control for them could lead to biased estimates of the association between medical marijuana policy and suicide. Finally, we incorporated more recent data into our analyses, reflecting newly adopted state medical marijuana policies.
