Specifically, the links between any anxiety disorders and COPD and PTSD and COPD are no longer significant after adjusting for former smoking. As both anxiety disorders and COPD are strongly related to cigarette smoking, these results are not surprising. But these results do call into question whether, thoroughly and to what degree, the link can be attributed to anxiety over having COPD and instead suggests that common risk factors for both COPD and anxiety disorders may explain this association. Finally, the relationship between GAD and COPD appears to be attributable to confounding by nicotine dependence. Previous studies have suggested that cigarette smoking may lead to GAD (Johnson et al., 2000), and strong associations have been found between nicotine dependence and GAD (Grant et al., 2004).

Third, we found that the relationship between mood disorders and COPD appears to be explained by confounding with nicotine dependence. Specifically, the association between mood disorders and COPD is no longer statistically significant after adjusting for nicotine dependence. This is consistent with our hypothesis that the link between COPD and mood disorders may be explained by exposure to a common risk factor for both, rather than a causal link between the two. As these data are cross sectional, we are only able to test for possible confounding but not for mediation. A recent study found that depressive symptoms remained associated with COPD after adjusting for smoking (Ng, Niti, Fones, Yap, & Tan, 2009).

Our findings are not necessarily inconsistent with this in that we also found that the link between major depression and COPD remained after adjusting for smoking, but only that it was no longer significant after adjusting for nicotine dependence. To our knowledge, this is the first study that has examined the role of nicotine dependence in this link. Limitations of this study should be considered when interpreting these results. First, we relied upon self-reports of COPD, which leave our results vulnerable to self-report bias. However, cases of COPD would more likely be underreported rather than overreported as many people with less severe COPD do not realize that they have the disease (Rennard & Vestbo, 2006). In addition, underreporting of COPD would unlikely to be influenced by whether or not a person has a mental disorder.

This type of potential nondifferential misclassification of disease would lead to an underestimate of the strength of the relationships between anxiety and mood disorders and COPD, and thus, our findings are likely to be conservative. Also, the question about COPD or emphysema was somewhat vague, so it is possible that those with rare lung disease such as primary pulmonary hypertension or sarcoidosis are included here, though this is likely GSK-3 to be a very small number since these diseases are rare.