Tobacco addiction in the psychiatric population and in the general population 1

ABSTRACT Objective: To estimate the degree of tobacco addiction and identify independently associated factors by comparing the psychiatric population of secondary and tertiary care with the general population of the primary healthcare network. Method: This is a cross-sectional epidemiological study, conducted in a municipality of São Paulo, with 134 smokers of a Mental Health Outpatient Unit (MHOU), a Psychiatric Hospital (PH), and a Primary Healthcare Unit (PHU). Data were collected by means of individual interviews, recorded on a mobile device. Data were statistically processed using Stata/12 Results: Of the 134 participants, 54.5% were women. While 49.1% of the psychiatric population (MHOU/PH) had medium/high nicotine addiction, 58.3% of smokers of the general population had very low/low dependency. The Poisson regression model indicated a higher prevalence of smokers with high dependence among men (PR = 1.41), people aged 49 years or less (15 - 29 years, PR = 4.06, 30 - 39 PR = 2.96 years, 40 - 49 years PR = 1.84), with severe mental disorders (PR = 3.05), with anxiety disorders/other (PR = 3.98), and with high suicide risk (PR = 1.55). Conclusion: Nicotine dependence was greater in the psychiatric population than in the general population. The independent factors associated with severe dependence were sex, age group, diagnosis, and current risk of suicide. These results trigger reflection among nurses on the need to focus more attention on a neglected subject in mental health services.


Introduction
Mental disorders can cause intense anguish, reduce quality of life, self-esteem, and social and work performance, and affect the interpersonal relationships and family ties of sufferers (1)(2)(3) .
In addition to the emotional and social losses, psychiatric patients suffer from impaired health, a lower life expectancy, and a greater occurrence of somatic complications. People with severe and persistent mental disorders live an estimated 25 years less than the general population, and smoking is one of the main causes of reduced life expectancy (4)(5)(6) .
Although 22% of the US population suffers from a mental disorder, these patients consume half of the cigarettes produced in the country and represent 46% of tobacco-related deaths, which reveals the severity of the epidemic. Moreover, the highest levels of nicotine dependence and the lowest levels of smoking cessation are found among people with severe mental disorders (7)(8)(9)(10)(11)(12) .
High tobacco consumption has a significant impact on the financial life of people with mental disorders.
These smokers can spend up to 30% of their monthly income on cigarettes, whereas more than 4% is considered prejudicial. Difficulties in obtaining cigarettes can lead people with mental disorders to engage in demeaning practices such as stealing cigarettes and smoking cigarette butts, which reduce their self-esteem and self-respect (13)(14)(15) .
Furthermore, mental disorders can be more severe among smokers. Patients who smoke suffer more outbreaks (delusions and hallucinations), thoughts of suicide or suicide attempts, and psychiatric hospitalizations (7,11,(16)(17)(18) .  The variable importance of tobacco was obtained on a scale of 0 to 10, according to the importance the smokers' attributed to tobacco. In the descriptive analysis, the average importance of tobacco was 6.6.
This value was used to create two categories: below average and above average.  (19) 3) Fagerström Test for Nicotine Dependence (FTND): This test comprises six questions that investigate cigarette smoking patterns (first cigarette of the day, difficulty to refrain from smoking in non-smoking locations, most satisfying cigarette of the day, number of cigarettes, period smoking is more frequent, and whether subject smokes even when sick). Each response has a score, and the sum of these scores determines the level of tobacco dependence, as follows: very low (0 to 2 points); low (3 to 4 points); medium (5 points); high (6 to 7 points); and very high (8 to 10 points). The test was validated for use in Brazil (test retest 0.915 and Cronbach's alpha 0.642) and it is considered the "standard" test to assess nicotine dependence (20) .
Data were collected through individual interviews in a reserved room. They were conducted by a single interviewer. The answers of the participants were recorded on a mobile device (tablet) using the application TabacoQuest, especially designed for this project (21) .
The responses of the subjects, marked by the researcher in the application, were automatically transferred to Excel spreadsheet and subsequently transferred to Stata (version 12) for statistical processing.
We used descriptive statistics tools to characterize the participants (average, standard deviation, minimum, maximum, and absolute and relative frequency). The variable tobacco dependence was subjected to bivariate analysis by calculating the Prevalence Ratio (PR) and its respective confidence interval (CI 95%).
Multivariate analysis was performed using Poisson regression, with degree of tobacco addiction (FTND score: ≤ 5 and ≥ 6) as the outcome. The model was adjusted with time of tobacco use as the exposure control (offset variable).
It was possible to dichotomize the FTND scores since this cutoff point is recognized in scientific literature (9) .
We selected the independent variables with p< 0.20 in bivariate analysis and those considered relevant in scientific literature. The criterion of not exceeding the limit of 10 cases by variable was also observed (22) . The results were discussed according to scientific literature.

Results
Most of the 134 participants were single women who had finished elementary school. The average age of the participants was 46 years (15 to 78 years, SD = 14).
Approximately three-quarters of the participants were diagnosed with severe mental disorders (schizophrenia, schizoaffective disorder, mood or personality disorder), and a significant portion used first generation antipsychotics.
About a third of the subjects had a high risk of suicide. Most participants stated they were consuming alcohol, and a minority admitted using illegal substances (Table 1).
Seventy-one smokers (53%) considered that tobacco had above average importance.   In the model, the association of tobacco addiction with antipsychotics and with illicit substances was no longer detected, suggesting that this association occurred due to the interference of other variables ( Table 2).
The multiple regression model revealed that the prevalence of smokers with high dependence was 41% higher among men than among women.
The participants between 15 and 29 years old showed a prevalence of smokers with a degree of dependence that was 4.06 times greater than among the elderly participants. The prevalence ratio dropped as the aged increased, suggesting a dose-response effect ( Figure 3).    The high dependency on tobacco of this population is consistent with national and international findings (8)(9)(10)(11)23) .

Regardless of the variables included in the Poisson
In the general population, assisted in the primary care system, there was a predominance of smokers with very low/low tobacco addiction. A recent national survey indicated that 81% percent of smokers of the general Brazilian population showed low dependence on tobacco (24) .
The multivariate analysis revealed a greater dependence among men, younger age groups, people with severe mental disorders, anxiety disorders/other disorders, and people at high risk of suicide.
The higher prevalence of smokers with tobacco addiction among men is a reflection of the ideals promoted by the tobacco industry. In advertisements, smoking is associated, among other things, with the image of masculinity/virility. Although, years later, the tobacco industry started to include women in their target audience, smoking among women did not escalate due to social pressure (25)(26) .
In addition to the influence of the industry and social acceptance, the greater dependence of men in relation to women is attributed to the female hormones that protect women against more intense dependence on tobacco (27)(28) .
This fact is important because it indicates that strategies to treat tobacco dependence should be different for men and women. Men should respond better to medication that targets addiction while women should respond better to treatment oriented toward the behavioral aspects of smoking (27)(28) .
This insight is valuable for nurses since, as members of a healthcare team and a multidisciplinary setting, they must actively use their technical and scientific knowledge to plan the best strategies to help patients stop smoking.
Greater tobacco addiction among the young population, in comparison to the elderly, was maintained even after adjusting sex, psychiatric diagnosis, current use of antipsychotics, risk of suicide, alcohol use, illicit substance use, and importance attributed to tobacco.
This result causes concern since young people are less prone to seek psychiatric treatment. For antismoking interventions to reach this population, nurses and other health workers must create effective strategies for individuals who have not yet been included in the mental health network (29) .
The strong association between tobacco addiction patients acknowledged that relief of these symptoms was temporary, leading them to smoke more frequently (10,30) .
The association between intense tobacco addiction and the high risk of suicide was maintained in the multiple model.
The relationship between suicide and smoking was highlighted in cohort studies that identified tobacco use and high nicotine dependence as a risk factor for suicidal behavior, even after adjusting for psychiatric variables.
In terms of the dose-response, evidence shows that the greater the number of cigarettes consumed in a day, the higher the risk of suicide. There is also evidence that smoking cessation reduces the risk of suicide (31)(32)(33)(34)(35)(36)(37)(38)(39)(40) .
Some authors sought to explain the association between tobacco use and suicide with the use of neurotransmitters, and found that decreasing the activity of the enzyme monoaminaoxidase (MAO-A and MAO-B) can increase impulsive behavior, which is one of the predisposing factors for suicide attempts (41)(42)(43) .
More extensive research and social and therapeutic actions that address this issue are recommended to reduce the risk of suicide (32) .
A study conducted in the US detected a reduction in the risk of suicide after tobacco control actions, such as increasing tax on cigarettes and smoking-free locations.
It was estimated that an increase of one dollar in taxes on cigarettes could reduce the risk of suicide by 10.5%.
Considering the absolute number of suicides in 2012, this percentage corresponds to a reduction of 4,000 suicides per year in the US (44) .
The association between tobacco addiction and suicide risk is an important finding for nursing, as, despite the Brazilian law (12.546/2011) that prohibits smoking in collective environments, smoking is still found in many Brazilian mental health services. Considering nursing staff must closely observe patients with risk of suicide, especially during psychiatric hospitalization, allowing psychiatric patients to smoke goes against the ethical principle of non-malfeasance.
In this study, the association of tobacco addiction with the use of alcohol and illicit substances was not consistent and characterized the degree of dependence as dichotomous (≤ 5 points and ≥ 6 points), although these cutoff points are recognized in scientific literature (9) .
The results of this study can support the actions of nurses since they confirm a higher prevalence of smokers with intense tobacco dependence among the psychiatric population and identify, by means of a multiple model, the variables that are independently associated with addiction. The resulting knowledge and debate on smoking can help include smoking in the dialogue with patients and the multidisciplinary team.
One of the limitations of this study is the adopted method (cross-sectional study), which does not allow causal inference.

Conclusion
Tobacco addiction was higher in the psychiatric population at secondary and tertiary care levels than in the general population that uses the primary health network.
In the multiple model, the resulting independent factors associated with severe dependence were sex, age group, diagnosis, and current risk of suicide.
These results should encourage Brazilian nursing professionals to pay closer attention to a subject that is neglected in mental health services.