Prevalence and smokers' profile: comparisons between the psychiatric population and the general population*

Objectives: to identify the prevalence of smokers between the psychiatric population and the general population; to compare the personal, socio-demographic and clinical profile of smokers and non-smokers in the psychiatric population and the general population; to compare the reasons for smoking of these two population groups. Method: this is a cross-sectional descriptive-analytical epidemiological study with 378 patients from three services: Ambulatory Mental Health, Psychiatric Hospital, and Basic Health Unit. Interviews were conducted with three questionnaires. The Chi-square and Kruskal-Wallis tests were applied. Results: in the total of the 378 participants, 67% were women and 69% were over 40 years old. There was a higher prevalence of smokers among men, young people, illiterates, singles and with more than one government benefit. Smokers prevailed among schizophrenics, chronic patients, who used ≥ 3 psychotropic drugs and had a history of ≥ 4 psychiatric hospitalizations and/or suicide attempts. The main reason for smoking was the improvement of negative feelings. Conclusion: the prevalence of smokers is higher in the psychiatric population (especially among severely ill patients) and among men, young people, unmarried and with socioeconomic losses. The main reason for smoking is tension/relaxation relief. This study provides nurses and other professionals with knowledge capable of subsidizing the planning of smoking interventions in the Brazilian population.


Introduction
Currently, the world prevalence of smokers is 20.7%, while in 2007 it was 23.5%. This result shows an overall trend; however, it is observed that the reduction was more significant in countries with high per capita income (1) .
More localized studies show a divergence in the prevalence of smokers among different population groups, especially those considered vulnerable -poor, homosexual, with mental disorders, and users of alcohol and illicit substances (2)(3)(4) .
Smoking causes about seven million deaths a year, meaning one in 10 deaths by tobacco use. Despite the high mortality rate, 30 million lives may have been saved in the past ten years as a result of the World Health Organization and governments' efforts to control this epidemic (1) .
For more than a decade, the World Health Organization has proposed actions to control smoking, which include monitoring tobacco use, raising awareness about the harm for the person and passive smokers, encouraging advertisements to be banned tobacco use, tobacco cessation aid, and tobacco tax relief. About twothirds of the world's population is protected by these actions, as 121 countries adopt at least one of them (1) .
Although there has been big progress made in recent decades, the World Health Organization recognizes tobacco smoke as a lethal practice, advocating the urgent strengthening of control actions (1) .
Tobacco smoking among people with mental disorders has always been very frequent and encouraged even by health professionals. Currently, it is seen as a public health problem, since the prevalence of smokers is two or three times higher, compared to the general population. This fact leads to physical losses (high index of early mortality due to clinical comorbidities), mental losses (aggravated by psychiatric symptoms), social losses (social isolation) and financial losses (elimination of essential expenses to buy cigarettes) (3,(5)(6)(7) .
Bringing this discussion to the national level, the last Brazilian survey revealed a prevalence of smokers in the general population of 14.7%, while in 1989, it was 32.4%. Also, over a five-year period (2008 to 2013), the attempts to quit smoking increased from 41.3% to 47.2%, according to a survey of 39,425 Brazilians nationwide (8)(9) .
Brazil's commitment to tobacco control is indisputable, as it was one of the first countries to sign the "Framework Convention on Tobacco Control" (10) .
However, control actions are not only modifying the prevalence of smokers but also their distribution. This is in line with the World Health Organization, which argues that understanding the profile and trends of tobacco smoke contributes to the strengthening of tobacco control policies (1) .
From this perspective, it is necessary not only to question how many smokers there are, but who current smokers are and their reasons for smoking.
This study aimed to 1) Identify the prevalence of smokers between the psychiatric population and the general population; 2) Compare the personal, sociodemographic and clinical profile of smokers and nonsmokers of the psychiatric population and the general population; 3) Verify the reasons to smoke of these two population groups. A simple random probabilistic sample was calculated, estimating that the prevalence of smokers in the mental health ambulatory would be around 40% and 60% in the psychiatric hospital. The prevalence estimated was based on the experience of researchers in mental health services, as well as on the scientific literature (11)(12) . With a significance level (α) of 5% and beta (β) of 10%, the sample calculation indicated the need for 126 participants for each study place. Individuals were included in the survey according to the order of arrival at the service or date of hospitalization, during the collection period.

Method
The individual invited to participate should reside in the municipality and be at least 15 years old. Those who had difficulties or were unable to communicate due to vocal or hearing impairment, those who had a diagnosis of mental retardation and who declared problematic use of alcohol or illicit substances without psychiatric comorbidities were excluded.
The same inclusion and exclusion criteria were considered for the population of the Ambulatory of Mental Health, the Psychiatric Hospital, and the Basic Oliveira RM, Santos JLF, Furegato ARF.
Health Unit. Therefore, it was decided not to exclude people with mental disorders from the Basic Health Unit to maintain comparability with studies conducted by the World Health Organization and by other authors, who do not use the psychiatric diagnosis as an exclusion criterion to investigate smoking in the general population (1,(8)(9) .
In Figure 1, the process of study participants definition is illustrated:  respondents' answers were recorded on a mobile device (13) . = present in moderate degree, 4 = present in severe or extreme degree). The within-class reliability was 0.93 (14) .
The "Reasons for Modified Smoking Scale" was developed to evaluate the reasons people smoke tobacco (15) . Statistical analysis was performed in Stata/SE (version 12.1). Absolute and relative frequency (%) was calculated using the chi-square test, at the significance level (α) of 5%. The chi-square test was used to identify statistical evidence of an association between the variable "current tobacco smoke" and the other variables tested two to two.
Although the "Brief Scale of Psychiatric Evaluation" provides five possible classifications for symptoms -1) absent, 2) very mild or with dubious presence, 3) present in mild degree, 4) present in a moderate degree and 5) present in severe or extreme degree (14) , due to the sample size, for the statistical analysis of this study, three categories were chosen for The results were discussed based on the scientific literature on this topic.

Results
Of the total of 378 participants, 67% were women, 69% were over 40 years old and 56% had studied through elementary school. In the Basic Health Unit, 29% had a psychiatric diagnosis, recorded in their medical records. In the Mental Health Ambulatory and in the Psychiatric Hospital, this percentage was 100%.
The prevalence of smokers was different in the three places investigated (ambulatory=27%, hospital =60%, basic health unit=19%). Table 1 shows the personal and socio-demographic characteristics of smokers (n=134) and non-smokers (n=244).
Data showed that while approximately half of the men smoked tobacco, most of the women were nonsmokers. Non-smokers prevailed at all ages, but their highest frequency was noted in older people (≥ 60 years old). The prevalence of smokers was higher in young people (15 to 29 years old) and decreased as the aging process.
The highest prevalence of smokers was identified in the illiterate and in those who studied until elementary school. The non-smokers prevalence was identified in those with higher education.
While non-smokers were mostly married, separated/ divorced and widowed, nearly half singles were smokers.
Consistent with marital status, the highest prevalence of smokers were identified in those who lived without a partner.
There was a higher prevalence of smokers in those participants without occupation and in the retirees. The prevalence of non-smokers was higher among workers and housewives.
The prevalence of smokers was higher in those who stated that they did not have a current occupation, as it was higher in those who received more than one government benefit. Half of the smokers reported having abandoned some employment relationship when diagnosed with mental disorder.
Regarding the clinical profile, a higher prevalence of smokers was observed in people diagnosed with schizophrenia/schizoaffective disorder, followed by those with personality disorders. Non-smokers predominated in those without a psychiatric diagnosis, with mood disorders and with anxious disorders ( Table 2).
Oliveira RM, Santos JLF, Furegato ARF.  As observed in Table 2, the vast majority of those who had been diagnosed for less than one year were nonsmokers, among the 288 patients with mental disorders.
Smokers prevailed among those with longer diagnosis time.
However, there was no significant difference between the participants diagnosed between 12 years old and younger and those diagnosed for more than 12 years.
The highest prevalence of smokers was among those who used three or more psychotropic drugs and 1 st generation antipsychotics.
There was a difference in the history of psychiatric hospitalizations, according to the use of tobacco. Most who had never been hospitalized were non-smokers, while the majority of those who had had four or more hospitalizations were smokers. As the number of hospitalizations increased, the prevalence of smokers increased and non-smokers decreased.
While most participants who had never tried suicide did not smoke tobacco, most who had tried four or more times were smokers.
When assessing the presence and severity of psychiatric symptoms, during the three days before the interview, smokers had the most severe symptoms.
Three-quarters of the respondents classified in the total Oliveira RM, Santos JLF, Furegato ARF.
score of the "Brief Psychiatric Assessment Scale" as "major syndrome" smoked tobacco.
In Table 3, the psychiatric symptomatology related to thinking, sensory perception and behavior is compared between smokers and non-smokers.
In  *Variables: symptoms evaluated from the "Brief Psychiatric Assessment Scale" (14) ; † Evidence of statistical association (p< 0.05) higher (3.3) than smokers in the outpatient clinic (3.1) and the primary care unit (2.5).

Discussion
This study identified that smokers are predominantly male, young and single, and those with socioeconomic losses (illiterate or with few years of school, people with no employment relationship and receiving social benefits from the government).
This study is in line with the scientific literature regarding people with socioeconomic vulnerability, more likely to use tobacco (16)(17)(18)(19) . However, it is a vicious cycle in which social disadvantages make people more vulnerable to smoking, and becoming a smoker contributes to these disadvantages (smokers stop buying essential items such as food and medicine to buy cigarettes) (18) .
An American longitudinal study with 131 smokers and 120 non-smokers, looking for work, helps to understand this situation. Almost half of the smokers (45.8%) reported having been discriminated against in previous jobs by smoking tobacco and 8.4% admitted to having been dismissed for that reason. Although 29% Oliveira RM, Santos JLF, Furegato ARF.
acknowledged that being a smoker hindered to get a new job, tobacco purchases were listed as the highest financial priority, even exceeding food expenses (19) .
After a follow up of 12 months, those who did not smoke were more successful (55.6%) in re-entering the labor market than smokers (26.6%). If the 131 smokers stopped smoking, the percentage of reemployment would increase by 30%, regardless of unemployment time, age, school years, race/ethnicity and health conditions (19) .
Socioeconomic vulnerability helps to understand, in part, the lower prevalence of smokers in the mental health ambulatory compared to the psychiatric hospital.
Because psychiatric treatment is too expensive (as an example, each psychiatric visit is charged without return visits as in other specialties), it is common to find people with good economic conditions in mental health services.
Regarding the clinical profile, this study revealed a higher prevalence of smokers among the more severe psychiatric patients (diagnosis of schizophrenia or schizoaffective disorder), with intense symptoms, with a longer diagnosis, using three or more psychotropic drugs, especially antipsychotics of first generation, with a history of four or more psychiatric hospitalizations, as well as four or more suicide attempts.
The clinical profile of smokers was similar to the predominant characteristics of the participants in the psychiatric hospital, coincidentally, where there was the highest prevalence of smokers, in relation to the others.
As found in this study, the higher prevalence of smokers among schizophrenics, compared to those diagnosed with other mental disorders, is widely recognized in the scientific literature (3,(5)(6)20) .
The theory of self-medication exposes that tobacco would improve the cognitive symptoms of schizophrenia by increasing the release of dopamine and glutamate in the prefrontal cortex and by regulating the auditory sensory process so the schizophrenic can filter out those irrelevant stimuli from the environment that harm their cognitive functions (attention, concentration, memory).
Negative symptoms (anecdotal, affective blunting, psychomotor retardation, loss of initiative) would be ameliorated by the ability of tobacco to act on deficits in the brain reward system, commonly presented by schizophrenics, justifying the greatest cleavage among them (6,(21)(22) .
The data in this study lead to think of the verisimilitude of this theory, since almost two-thirds of those using only the first-generation antipsychotics were smokers, while the majority of those using only secondgeneration antipsychotics were non-smokers.
This finding is consistent with the scientific literature showing that individuals on first-generation antipsychotics are more likely to use tobacco than those on second-generation antipsychotics (23)(24)(25) .
Therefore, first-generation antipsychotics act only on positive symptoms (delusions, hallucinations, among others). The schizophrenic, using this type of psychoactive drug, would find in tobacco a way of temporarily reversing cognitive symptoms by inducing an increase in dopamine and glutamate in the prefrontal cortex (6,(26)(27)(28) . According to the self-medication theory, the between the smokers (18,22,29) .
In fact, a significant portion of the people who presented delusions and hallucinations, in the days before the interview, were smokers.
Regardless of whether the theory of selfmedication is true, smokers were those who presented more intense psychiatric symptoms (total score in the "Brief Psychiatric Evaluation Scale"). These results were consistent with other studies (22,(30)(31) .
The greater intensity of psychiatric symptoms among smokers is in agreement with some theories that although there may be an improvement of the negative, cognitive and anxiety symptoms at the onset of smoking, chronic use of tobacco can reverse this effect, increasing the symptoms (18,29,(32)(33) .
Complementing the theory of self-medication, studies showed that tobacco use interferes with the metabolism of psychoactive drugs, decreasing its concentration in plasma. Therefore, psychiatric patients would use tobacco more intensely as a way to alleviate side effects, especially in the case of first-generation antipsychotics (6,22,34) .
In the same line, the Brazilian study in the General Hospital Psychiatric Unit found that 50% of smokers with a diagnosis of schizophrenia justified the maintenance of smoking with the intention of alleviating the side effects of psychotropic drugs (35) .  (36) . In this study, suicide attempts were highlighted, since the prevalence of smokers increased according to the number of attempts.
Non-smokers followed the opposite direction of smokers.
The relationship between suicide and smoking was highlighted when cohort studies identified tobacco use and its high dependence as a risk factor for suicidal behavior, even after adjusting for psychiatric variables.
There is evidence of a dose-response effect as the higher the number of cigarettes smoked/day, the greater the risk of suicide. Moreover, there was evidence of a decrease in this risk when stopping smoking (37)(38)(39) .  (4,35,40) .
The main reason as the motivation of smokers who participated in this study to use tobacco was the reduction of tension/relaxation.
The use of tobacco as an attempt to alleviate anxiety is known. A Scottish study of 131 schizophrenics showed that 60% of smokers used tobacco to relax and 31% because they felt anxious or depressed (22) . Similarly, the Brazilian study with 270 psychiatric patients revealed that 79% of smokers believed in the anxiolytic function of tobacco (35) .
Despite the above, special care is needed for these results. While it is still possible to identify people who use tobacco to feel less anxious and safer in social interactions, this is a reality that is being modified, as tobacco smoke is moving from a glamorous act to conduct condemned by society (41) .
With the higher prevalence of smokers in the psychiatric population and the lower tolerance of society to smoking in collective settings, the trend is that those with mental disorders are even more discriminated and

Conclusion
The prevalence of smokers is higher in the psychiatric population, especially the hospitalized population.
When considering the psychiatric population and the general population, the study identified that smokers are predominantly male, young and single, and with socioeconomic losses (illiterate or with few years of school, people with no employment receive social benefits from the government).
Regarding the clinical profile, this study revealed a higher prevalence of smokers among the more severe psychiatric patients (diagnosis of schizophrenia or schizoaffective disorder), with intense symptoms, with a longer diagnosis, using three or more psychotropic drugs, especially antipsychotics of first generation, with a history of four or more psychiatric hospitalizations, as well as four or more suicide attempts.
The main reason alleged to justify maintaining cigarette smoke is to obtain tension relief and relaxation.
It is expected that this study will provide nurses and other health professionals with knowledge capable of subsidizing educational projects, as well as planning smoking interventions in the Brazilian psychiatric population.