Tobacco use is the leading preventable cause of illness, disability, and premature death in the United States (1). Smoking is responsible for 1 in 5 deaths, resulting in approximately 443,000 avoidable deaths per year (2). These rates are disproportionately higher among veterans because both active-duty personnel and veterans are more likely to have ever smoked or to currently smoke than the adult civilian nonveteran population (3,4). Thus, Department of Veteran Affairs (VA) provides smoking cessation interventions and programs to its health system users (5,6). However, most veterans who smoke and use VA health services report they do not receive tobacco cessation treatment (7). Identifying veteran characteristics related to tobacco use can clarify who is most likely to benefit from smoking prevention and cessation interventions and may enhance VA efforts to reduce smoking and smoking-related illnesses.

A nicotine dependence diagnosis is given to people who use tobacco regularly and have become chronically dependent on nicotine. Epidemiological studies have found a 13% point prevalence (8) and 24% lifetime prevalence (9) for nicotine dependence in the general US adult population. Two recent studies found that 26% to 27% of veterans smoke (10,11). No published research study could be found on the prevalence of diagnosed nicotine dependence among VA service users; thus, the extent to which VA clinicians are assessing and documenting nicotine use and dependence is unclear.

Factors related to smoking cessation have been widely studied; preventing nicotine dependence and identifying predictors of it, less so. Tobacco use is more prevalent and intense among psychiatric populations than the overall population. Up to 41% of adults with mental illness smoke (12-14). These adults may be particularly susceptible to nicotine addiction because tobacco positively influences mood (15). Many people who abuse other substances also smoke, and an especially strong correlation has been found between smoking and alcohol abuse (13). However, this association has not been fully investigated in large studies of veterans.

Research is inadequate on nicotine dependence in 2 groups of particular interest to the VA health system: homeless veterans and veterans who served in Iraq and Afghanistan in Operation Iraqi Freedom and Operation Enduring Freedom (OIF/OEF). Homelessness among veterans has been a national problem for more than 2 decades (16-19), and recently interest has been renewed in ending veterans' homelessness and providing all necessary health care interventions to this population (20). As the United States continues to wage war in the Middle East, health care providers have been especially concerned about OIF/OEF veterans who served in Iraq and Afghanistan, who are at risk of developing various physical and mental health problems postdeployment and after military discharge (21-24). Some studies suggest higher rates of smoking among these veterans (6). Because both of these groups are priorities for VA health services, identifying factors related to nicotine dependence in these 2 groups may help target prevention efforts and curb development of smoking-related illnesses.

The objective of this study was to examine all recent users of VA health services, a group readily available for smoking prevention and cessation interventions, to identify the prevalence of nicotine dependence diagnoses and determine the risk factors associated with receiving such a diagnosis. A secondary objective was to examine risk factors for nicotine dependence among homeless veterans and OEF/OIF veterans.

We found that 15% of all veterans who used VA health services in FY 2009 received a diagnosis of nicotine dependence. Because we analyzed administrative data, we likely underestimated how many veterans actually have nicotine dependence; recent estimates indicate that 26% to 27% of veterans smoke (10,11). Although no previous study to our knowledge has examined the prevalence of nicotine dependence in the population of veterans using health services, our finding suggests nicotine dependence may be underdiagnosed and not adequately documented in VA administrative records. Because smoking is a leading cause of many chronic diseases and deaths (1), it may be beneficial for VA clinicians to better document nicotine use. The benefit of better documentation assumes that assessment and diagnosis lead to increased likelihood of successful intervention; various smoking cessation interventions are effective for veterans (5,26,27).

In identifying major risk factors, veterans who had mental health or substance use disorders were at significantly higher risk of receiving a nicotine dependence diagnosis than veterans who did not have such diagnoses. Among VA mental health service users, one-fourth had a nicotine dependence diagnosis. This result is consistent with previous findings of increased rates of nicotine use among adults with mental illness or substance use disorders in the general population (12-14).

Having an alcohol use disorder was the strongest independent predictor of a nicotine dependence diagnosis, followed closely by a drug use disorder. Veterans who had an alcohol use disorder were more than 3 times as likely and veterans with a drug use disorder were almost 2 times as likely to receive a nicotine dependence diagnosis compared to veterans without such disorders and controlling for other influential factors. VA clinicians may need to pay particular attention to smoking behaviors among veterans with mental illness or substance use disorders, especially because nicotine dependence disproportionately reduces the quality and length of life of people with these disorders in the general population (28). Providing smoking prevention and cessation interventions with other substance abuse and mental health treatment for veterans may be useful; efforts to integrate nicotine cessation programs into VA mental health services have shown some success (5).

Homeless veterans were also at increased risk for nicotine dependence diagnosis (39%), independent of their increased risk for addictive disorders. This finding is consistent with recent studies, which have found that 69% to 73% of homeless people in the general population smoke (29,30). Interestingly, these studies also found that more than one-third of homeless smokers expressed a readiness to quit and more than half received advice to quit from their health care providers, but they were still less likely to quit compared to others in the general population. People with multiple episodes of homelessness were less likely to quit (29). Besides alcohol and drug use as factors associated with smoking in the homeless population, studies have also found out-of-home placement in childhood, victimization while homeless, and smoking intiation at an earlier age are significant factors (29,31). There has been little development of smoking prevention and cessation programs for homeless people, let alone homeless veterans, and more research is needed in this area.

OEF/OIF status was protective against nicotine dependence diagnosis, in contrast to previous studies, which relied on self-report (6). It is worth reiterating that we did not examine nicotine dependence, per se, but rather how often it was diagnosed, which may explain the difference in findings and suggests nicotine dependence is not adequately assessed among OEF/OIF veterans, who are likely seeking treatment for more pressing health issues. We found that 15% of OEF/OIF veterans who received VA health services in FY 2009 received a nicotine dependence diagnosis. Substance use disorders were still significant risk factors, but OEF/OIF status alone did not increase risk for a nicotine dependence diagnosis. Among both OEF/OIF and homeless veterans, we consistently found that veterans who were male, low-income, and living in a rural area were at higher risk of receiving a nicotine dependence diagnosis. Dementia was found to be a protective factor, which may be because of its effects on general life functioning and behaviors, including smoking.

This study has several limitations. Administrative records are not always complete or reliable. VA clinicians may have neglected to document nicotine dependence in the face of presenting primary diagnoses, which only illustrates the importance for VA clinicians to conduct comprehensive assessments of patients that include questions about smoking behaviors. We focused on identifying risk factors of a clinical diagnosis of nicotine dependence, which may be different from factors related to actual nicotine dependence. There may also be other correlates of nicotine dependence that we did not address in our analyses, such as certain medical conditions and unmeasured individual characteristics. Given our large sample size, analyses were sensitive to statistical significance, so we focused on odds ratios to identify major risk factors. Although we identified some correlates for nicotine dependence among veterans, we could not examine the causal pathways through which these factors increase risk because our data were cross-sectional. Future research and development of assessment, documentation, and interventions in this area are needed.

Our results suggest veterans are underdiagnosed for nicotine dependence and that better assessment and documentation methods are needed in the VA health system. Veterans who are homeless, have a mental illness, or have a substance use disorder may be particularly vulnerable to dependence on nicotine, and targeted outreach and intervention for these groups may be needed. This study may contribute to improved targeting of smokng prevention and cessation efforts in the VA health care system.