- Email: [email protected]
The association of affective temperaments with smoking initiation and maintenance in adult primary care patients
Journal of Affective Disorders 172 (2015) 397–402
Contents lists available at ScienceDirect
Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad
Research report
The association of affective temperaments with smoking initiation and maintenance in adult primary care patients Ajandek Eory a,n, Sandor Rozsa b,c, Xenia Gonda d,e,f, Peter Dome d,f, Peter Torzsa a, Tatevik Simavorian i, Konstantinos N. Fountoulakis g, Maurizio Pompili h, Gianluca Serafini h, Knarig K. Akiskal i, Hagop S. Akiskal i, Zoltan Rihmer d,f, Laszlo Kalabay a a
Department of Family Medicine, Semmelweis University, 4 Kutvolgyi Street, Budapest H-1125, Hungary Center for Well-Being, Department of Psychiatry, Washington University School of Medicine, St. Louis, MO, United States c Department of Personality and Health Psychology, Eötvös Loránd University, Budapest, Hungary d Department of Clinical and Theoretical Mental Health, Kutvolgyi Clinical Center, Semmelweis University, Budapest, Hungary e Department of Pharmacodymanics, Semmelweis University, Budapest, Hungary f Laboratory for Suicide Research and Prevention, National Institute of Psychiatry and Addictions, Budapest, Hungary g 3rd Department of Psychiatry, Division of Neurosciences, School of Medicine, Aristotle University of Thessaloniki, Greece h Department of Neurosciences, Mental Health and Sensory Organs, Suicide Prevention Center, Sant'Andrea Hospital, Sapienza University of Rome, 00189 Rome, Italy i International Mood Center, University of California, San Diego, CA, USA b
art ic l e i nf o
a b s t r a c t
Article history: Received 23 July 2014 Received in revised form 16 October 2014 Accepted 18 October 2014 Available online 28 October 2014
Background: Smoking behaviour and its course is influenced by personality factors. Affective temperaments could allow a more specific framework of the role trait affectivity plays in this seriously harmful health-behaviour. The aim of our study was to investigate if such an association exists in an ageing population with a special emphasis on gender differences. Methods: 459 primary care patients completed the TEMPS-A, Beck Depression Inventory (BDI) and Hamilton Anxiety Rating Scale (HAM-A). Subjects were characterized according to their smoking behaviour as current, former or never smokers. Univariate analysis ANOVA and logistic regression were performed to analyse differences in the three smoking subgroups to predict smoking initiation and maintenance. Results: Current smokers were younger and less educated than former or never smokers. Males were more likely to try tobacco during their lifetime and were more successful in cessation. Depressive, cyclothymic and irritable temperament scores showed significant differences between the three smoking subgroups. Irritable temperament was a predictor of smoking initiation in females whereas depressive temperament predicted smoking maintenance in males with a small, opposite effect of HAM-A scores independent of age, education, lifetime depression and BDI scores. Whereas smoking initiation was exclusively predicted by a higher BDI score in males, smoking maintenance was predicted by younger age and lower education in females. Limitations: The cross-sectional nature of the study design may lead to selective survival bias and hinder drawing causal relationships. Conclusions: Affective temperaments contribute to smoking initiation and maintenance independently of age, education, and depression. The significant contribution of depressive temperament in males and irritable temperament in females may highlight the role of gender-discordant temperaments in vulnerable subgroups. & 2014 Elsevier B.V. All rights reserved.
Keywords: Affective temperament Smoking Sex Mood disorder
1. Introduction
n
Corresponding author. Tel.: þ 36 304003640; fax: þ36 12015500. E-mail address: [email protected] (A. Eory).
http://dx.doi.org/10.1016/j.jad.2014.10.036 0165-0327/& 2014 Elsevier B.V. All rights reserved.
Tobacco dependence is a chronically relapsing addictive disorder, the most important preventable cause of illness and premature death due to cardiovascular, pulmonary and cancer diseases, and is responsible for the death of approximately 6 million every year (WHO, 2013,
398
A. Eory et al. / Journal of Affective Disorders 172 (2015) 397–402
Clinical Practice Guideline for Treating Tobacco Use and Dependence: 2008 Update, 2008). Although the addictive features of nicotine are the most widely researched component of smoking behaviour, other chemical substances of tobacco fume, sensory cues, heritability, environmental factors (social bonds, education, accessibility of tobacco products), and personality patterns also play an important role in the initiation, maintenance or successful termination of tobacco use (Cosci et al., 2011, de Viron et al., 2013, Clinical Practice Guideline for Treating Tobacco Use and Dependence: 2008 Update, 2008; Kassel et al., 2003). Individual differences in the mood-modulating effect of nicotine can be approached from two different directions. Pre-existing trait markers (like personality traits, trait anxiety or depression) are responsible for the inter-individual differences and they are applicable to identify the fundamentals of variation at the population level and therefore can be used for exploratory purposes (Kassel et al., 2003). Mood states together with stressful situations may explain the individual response to smoking cessation (Perkins, 1999; Kassel et al., 2003; Shadel et al., 2004). Gender differences in smoking initiation and cessation are fairly well known, and initiatives to address gender differences are imperative (WHO, 2010; Zatonski et al., 2012), but recent research related to the influence of personality on tobacco use is scarce, even though the influence of personality on tobacco use is differentially indicative of distinct personality patterns in the two genders (Piñeiro et al., 2013; Nieva et al., 2011). The broad definition of personality refers to cognitions, affects and behaviours that determine the direction and pattern to the person's life. Affective temperaments with their roots in the ancient Greek humoral theory are trait-like temperaments, which are partially genetically-determined and which remain relatively stable throughout the lifespan and determine the emotional response by an individual to the environmental stimuli (Akiskal and Akiskal, 2007; Mischel et al., 2008; Rihmer et al., 2010) exploring the affect-related component of personality. They can be assessed on the depressive, hyperthymic, cyclothymic, irritable and anxious subscales of the Temperament Evaluation of Memphis, Pisa, Paris and San Diego Autoquestionnaire (TEMPS-A). This was developed thanks to an intensive clinical and genetic research, as well as broad clinical experience and collaboration (Akiskal and Akiskal, 2005a). Although these affective temperaments, especially when expressed at their trait level, contribute to the development of a healthy personality, they may serve as subaffective trait precursors or precursors of major affective disorders, constituting risk factors for specific clusters of symptoms and psychopathology (Iasevoli et al., 2013), course and prognosis of major mood disorders (Rihmer et al., 2010, Akiskal et al., 2005a) and treatment adherence (Fornaro et al., 2013). Due to its easy clinical applicability the affective temperamental model is utilized as an outcome measure in affective disorder research as well as in explorative research in the frontier of major affective disorders and chronic somatic conditions, including hypertension (Eory, 2014a) and acute cardiac complications in hypertension (Eory, 2014b), among others (Rihmer et al., 2010). The rationale for examining the relationship between tobacco use and affective temperaments are as follows: a) Research shows that neuroticism and negative affect are important contributors to smoking, and emotional dysregulation appears to be the possible variable of this association (Kassel et al., 2003; Malouff et al., 2006). The TEMPS-A questionnaire delineates those emotional components of personality, which relate to reactions to inner and outer adverse events, and therefore may add substantial information to more deeply investigate the question at hand. b) Affective temperaments having strong biological components may further extend existing psychological models of the relationship between addictions and personality (Akiskal and Akiskal, 2005b).
c) TEMPS-A may provide the background personality factors of increased tobacco use in major affective disorders (Rihmer et al., 2010) d) TEMPS-A is tested against other personality instruments (TCI, NEO-PI-R) (Rózsa et al., 2008; Akiskal et al., 2005b). e) Temperament traits remain stable from early adulthood over time and may be more characteristic predictors of chronic somatic conditions than major affective disorders (Akiskal and Akiskal, 2005a; Rafanelli et al., 2013; Rihmer et al., 2010; Eory, 2014a, 2014b). Therefore, research on the relationship of affective temperaments with harmful health-behaviour in non-psychiatric, adult sample could prove to be both useful and necessary. The aim of our study was therefore to explore the association between affective temperaments and smoking habits in a chronically ill outpatient population. Because the affective temperament profile differs among males and females we use gender difference in our computations.
2. Methods 2.1. Sample Data from 509 consecutively investigated primary care patients were collected to explore the association between affective temperaments and hypertension (Eory, 2014a) as well as acute cardiac complications (Eory, 2014b). Altogether six primary care practices were involved into the study (four practices in Budapest and two practices in the countryside) between February and May of 2011. All practices were single-handed, with an average of 1500 patients in the catchment area and a daily turnover of 20–40 patients in each. Inclusion criteria included patients older than 18 years. According to the primary aim of the research, patients with primary hypertension comprised the majority of the sample. Exclusion criteria were secondary hypertension, improper understanding of Hungarian language and legal incapacity. Other chronic medical conditions were not part of the exclusion criteria. The present sample is a subsample of the original population containing 459 regularly controlled outpatients with data on tobacco use. The frequency of chronic medical conditions across the sample was as follows: 76% of the sample suffered from primary hypertension. Hypertension related complications were cardiovascular complications (acute myocardial infarction, acute coronary syndrome, angina pectoris) (17%); peripheral artery disease (14%); cerebrovascular accidents (stroke or transient ischaemic attack) (11%) and kidney-related complications (10%). According to their medical history 29% of the sample suffered from diabetes mellitus, followed by musculoskeletal (25%), gastrointestinal (20%), anxiety (ICD-10 diagnosed or anxiolytic medication) (23%), depressive (ICD-10 diagnosed) (15%) and endocrine (10%) diseases. Subjects were classified as smokers if they had smoked more than 100 cigarettes in their lifetime and smoked on a daily basis; as quitters if they reported quitting smoking more than 30 days prior; or as never smokers if they stated that they had not smoked a cigarette throughout their lifetime (Smits et al., 2004; Ravara et al., 2011). Information on the number of cigarettes smoked per day, the length of abstinence in years, demographic factors (age, sex) and education (eight or less years, nine to twelve years, more than twelve years) was collected after receiving written informed consent by patients. The study has been approved by the Scientific and Research Ethics Committee of Scientific Health Council in charge of experimentation with human subjects and was carried out in accordance with the tenets of the Declaration of Helsinki.
A. Eory et al. / Journal of Affective Disorders 172 (2015) 397–402
399
(McDowell, 2006). The Cronbach alpha coefficient for the present study population was 0.89.
2.2. Measures 2.2.1. TEMPS-A The Temperament Evaluation of Memphis, Pisa, Paris and San Diego Autoquestionnaire (TEMPS-A) is a 110-item self-administered paper–pencil instrument, applying true/false statements (Akiskal et al., 2005a). It measures affective temperaments containing 21 items for depressive, cyclothymic, hyperthymic and irritable (20 for men on the latter) subscales, and 26 for anxious temperament. The self-report questionnaire was translated into more than 30 languages and validated in several countries, including Hungary in 2006 (Rózsa, 2006). The Hungarian version showed good reliability and internal consistency. It was externally validated across the Beck Depression Inventory (BDI), Profile of Mood States (POMS), the BarOn Emotional Quotient Inventory (BarOn EQ-i), Temperament and Character Inventory (TCI) and the NEO Personality Inventory-Revised (NEO-PI-R) (Rózsa et al., 2008). To establish internal consistency of TEMPS-A in this study population, Cronbach alpha coefficients for scores were measured. Alpha coefficients for each subscale are as follows: 0.77 for depressive temperament, 0.83 for cyclothymic temperament, 0.84 for hyperthymic temperament, 0.78 for irritable temperament and 0.92 for anxious temperament. All alpha coefficients ranged between 0.77 and 0.92. The magnitude of the coefficients alpha was satisfactory and good in all cases (Rózsa et al., 2008).
2.2.2. HAM-A/HARS The Hamilton Anxiety Rating Scale (HAM-A) is one of the most extensively used scales to measure the severity of anxiety symptoms with 14 items (each ranging 0–4) (Hamilton, 1959). Scores between 14 and 17 points indicate mild, 18–24 points indicate moderate and, scores greater than 25 signify severe anxiety. It has good reliability and validity, sensitivity to change over time and applicability for a wide range of patients (from adolescents to older adults) (McDowell, 2006).
2.2.3. BDI The Beck Depression Inventory (BDI) covers the emotional, behavioural and somatic symptoms of depression with 21 items on a four points intensity scale (0–3) (Beck et al., 1997). Scores between 10 and 18 indicate mild to moderate, 19–29 moderate to severe and, greater than 30 severe depression. This selfadministered questionnaire is extensively used in clinical research
2.3. Statistical analyses Chi-square test was applied for the assessment of group differences in categorical variables (gender, tobacco use, ICD-10 depression). Gender-tests were used for the comparison of continuous variables between groups. We applied one-way ANOVA to explore the differences in the TEMPS-A, BDI and HAMA scores and age between smoking groups in both gender. In cases of violation of the homogeneity of variance (depressive temperament in both genders and irritable and anxious temperaments as well as HAMA scores in females), we used the Brown–Forsythe F-ratio (Field, 2005). In cases where the subsets were not homogeneous (depressive and cyclothymic temperament scores in males and irritable temperament score in females) the Games–Howell post-hoc test was used (Field, 2005). In all other situations the Hochberg GT2 post-hoc test was applied due to unequal sample sizes (Field, 2005). To establish the relationship of affective temperaments with smoking behaviour we used a step forward likelihood ratio binary logistic regression with never vs ever smokers (smoking initiation) and current vs former smokers (smoking maintenance) as dependent variables; and considered affective temperaments, lifetime depression, age, education, BDI and HAMA as predictor variables.
3. Results 3.1. Baseline characteristics and affective temperament profile of the sample population The sample comprised 459 primary care patients (278 females). Never smokers constituted 55% of the total sample, and former smokers were represented in a slightly larger proportion (24.4%) than current smokers (20.6%). Current smokers were younger (54 714 years) than never (59 715 years) or former smokers (62 713 years) (F(2, 419) ¼7.7; p ¼ 0.001), and were less educated (χ2(4)¼21.615; po 0.0001). The average number of cigarettes smoked per day (CPD) was 15 (78). Among lifetime smokers, 112 (54%) quit successfully with a mean of 16 years ( 712 years) prior to data collection. Table 1 shows gender differences of demographic, psychosocial and temperamental characteristics in the three smoking groups.
Table 1 Distribution of demographic, psychosocial and temperament characteristics between smoking status by gender (n ¼459). Males (181) Smoking status Age Education ( r 8 yr/9–12 yr/ 412 yr) % ICD-10 depression (%)
Females (278)
Never (n¼ 82) 45% 58 (15) 34/33/33 5.4
Former (n¼ 63) 35% 62 (13) 17/37/46 11.3
Current (n ¼36 20%) 52 (14) 44/36/20n 20.6
Never (170) 61% 61 (14) 30/38/32 14.4
Former (49) 18% 61 (13) 8/49/43 13.0
Current (59) 21% 55 (14) 31/47/22n 25.9
6 (0–17)n 4 (0–19) 12 (0–20) 5 (0–17) 5 (0–26) 6 (0–34) 8 (0–39)
6 (2–13)n 5 (0–18) 12 (1–20) 5 (0–15) 6 (0–22) 7 (0–31) 9 (0–41)
8 (3–19)n 7 (0–19) 12 (0–21) 6 (0–13) 6 (0–26) 9 (0–38) 9 (0–38) 17 (9)
9 (0–21) 5 (0–25) 10 (0–21) 4 (0–12)nn 9 (0–25) 8 (7) 11 (9)
7 (2–16) 5 (0–16) 11 (2–19) 5 (1–13) 7 (0–17) 8 (6) 10 (7)
9 (3–20) 6 (0–18) 11 (0–17) 5 (1–17)nn 9 (0–26) 9 (9) 12 (11) 14 (7)
TEMPS-A Depressive Cyclothymic Hyperthymic Irritable Anxious BDI HAM-A CPD Length of abstinence (yr)
16 (13)
15 (12)
Values are expressed in mean (SD) unless otherwise indicated, psychometric scores are expressed in mean (range). TEMPS-A: Temperament Evaluation of Memphis, Pisa, Paris and San Diego Autoquestionnaire, BDI: Beck Depression Inventory, HAM-A: Hamilton Anxiety Rating Scale, CPD: number of cigarettes per day. n
po 0.05. p o0.01.
nn
400
A. Eory et al. / Journal of Affective Disorders 172 (2015) 397–402
Among males the proportion of ever smokers (vs. never smokers) was higher than among females (χ2(1)¼7.99; p¼ 0.005) with a higher rate of successful cessation (63.6% vs 45.4%) (χ2(1)¼6.94; p¼0.008). Although lifetime depression was four times more prevalent in current male smokers than in never smokers, the difference of ICD-10 diagnosed depression showed marginal significance between the three smoking groups in males (χ2(2)¼5.72; p¼0.057) and remained non-significant in females just like the BDI or HAMA scores in both sexes. Affective temperament scores showed gender-specific differences. Males scored higher on hyperthymic (U¼ 20,440, p ¼0.001) and irritable (U¼22,177, p ¼0.036) temperament scales, while females on the depressive (U¼ 17,914, p o0.0001) and the anxious (U¼ 17,484, p o0.0001). Depressive (F(2,424)¼6.23; p ¼0.002), cyclothymic (F(2, 424)¼ 7.06; p¼ 0.001) and irritable temperament (F(2,424) ¼10.23; p o0.0001) scores revealed significant differences among the three smoking subgroups. When it was further examined by gender, depressive (Brown–Forsythe F-ratio F (2,88.14) ¼ 5.07; p ¼0.008) and cyclothymic temperament scores (F(2,178) ¼3.4; p ¼0.035) proved to be significant in males while it was the irritable temperament score in females (Brown–Forsythe F-ratio: F(2,142.3) ¼7.35; p ¼0.001). Depressive temperament scores showed significant differences in males between current smokers and both never (p ¼0.023) and former smokers (p¼ 0.042) as well as cyclothymic temperament scores showed marginal significance between current smoker and never smoker males (p¼ 0.053). In females, irritable temperament showed significant difference between current and never smokers (p¼ 0.003).
3.2. Age, education and psychometric factors as predictors of smoking initiation To examine smoking initiation, the original smoking subgroups (never, former, current) were dichotomized as ever (current and former) versus never smokers. Binary logistic regression (step forward LR) was conducted to determine the role of age, education (r8 years, 9–12 years and o12 years), depressive, cyclothymic, hyperthymic, irritable and anxious temperament subscales, ICD-10 depression, BDI and HAMA scores in the initiation of smoking in both genders separately. 159 males were included in the analysis (with no missing data in any of the variables). The final model was statistically significant (χ2(1)¼4.6; p¼0.032, Cox and Schnell R2 ¼0.028). Table 2 shows that higher point scores on the BDI predicted smoking initiation independently of other predictors, increasing the odds of smoking initiation by 5%. In females (214 with no missing data on any of the variables), the model was also significant (χ2(2)¼21.01; po0.0001, Cox and Schnell R2 ¼ 0.094). Higher point scores on the irritable temperament subscale predicted smoking initiation independently of other predictors, increasing the odds of smoking initiation by 26%, whereas anxious
Table 2 Psychosocial predictors of smoking initiation in males and females (logistic regression, step forward LR). β
SE
Wald χ2 p
OR
95% CI
Males (n ¼159) BDI
0.046 0.022
4.15
0.042 1.047 1.002–1.094
Females (n¼214) Irritable temperament 0.234 0.056 17.529 Anxious temperament 0.066 0.027 5.736
0.000 1.264 1.133–1.411 0.017 0.936 0.887–0.988
temperament scores were inversely related to smoking initiation [β ¼ 0.066, SE¼ 0.027; Wald χ2(1)¼5.736; p¼0.017]. 3.3. Age, education and psychometric factors associated with smoking maintenance To explore the association between affective temperaments, ICD-10 depression, BDI, HAM-A scores, age, education, and smoking maintenance, we conducted binary logistic regression (step forward LR) with current smokers vs former smokers as a dependent variable in both genders. The results are presented in Table 3. In males, the regression analysis terminated after the third step. The final model was statistically significant (χ2(3)¼22.293; po 0.0001, Cox and Schnell R2 ¼0.226). Higher scores on the depressive temperament scale predicted smoking maintenance, increasing the odds of persistent smoking by 30.5%. On the contrary, age was inversely related to current smoker status, decreasing the odds by 7% just like higher HAM-A scores. The analysis was terminated after the second step in females, and no psychosocial predictors were included. Lower education (8 years or less) predicted smoking maintenance increasing the odds by 12 times while age had an inverse relationship, decreasing the odds of smoking maintenance by 4%.
4. Discussion We examined the relationship of affective temperaments and smoking as a harmful health-behaviour in an ageing population under regular control by their general practitioners. The gender distribution of affective temperament scores was similar to that of the three-decade younger Hungarian normative sample (Rózsa et al., 2008) with higher hyperthymic scores for males and higher depressive and anxious scores for females, further supporting the trait-like nature of affective temperaments (Akiskal and Akiskal, 2007). The fact that cyclothymic temperament did not reveal significant difference by gender can be explained by the negative correlation of this temperament with ageing (Rihmer et al., 2010). Additionally we found higher irritability scores in males, which are in accordance with male dominance on the irritable scale in international samples (Akiskal and Akiskal, 2005a; Gonda et al., 2011; Rihmer et al., 2010). 4.1. The contribution of affective temperaments to smoking and their relationships with other personality measures Our main finding is that depressive, cyclothymic and irritable temperament showed association with never, former or current smoking status in our ageing primary care sample. Similar results were found with the TEMPS-M in a college student population with irritable and cyclothymic temperament as predictors of selfreported nicotine abuse (Unseld et al., 2012). Depressive, cyclothymic and irritable temperaments as well as the anxious subscale showed a moderate or strong positive correlation with harm avoidance of the Temperament and Character Inventory (TCI), whereas a cyclothymic temperament also showed a moderate positive correlation with novelty seeking (Akiskal et al., 2005b), the latter of which was found to be related to smoking initiation in an adult sample (Gurpegui et al., 2007). The same affective temperaments are positively correlated with the neuroticism scale of the NEO-Personality Inventory-Revised (NEO-PI-R) (Rózsa et al., 2008). Neuroticism contributed to both former (Chapman et al., 2009) and current smoker (Terracciano and Costa, 2004) status in adulthood and was found to be a predictor of nicotine dependence after a 2.4 years follow-up of treatment in 250 patients (99 abstinent and 151 smoking patients) (Hooten et al., 2005).
A. Eory et al. / Journal of Affective Disorders 172 (2015) 397–402
401
Table 3 Demographic and psychosocial predictors of smoking cessation (logistic regression, step forward LR). β
SE
Wald χ2
p
0.266 0.068 0.075
0.093 0.021 0.039
8.208 10.388 3.819
0.004 0.001 0.051
1.305 0.935 0.928
2.515 0.038
0.864 0.018
8.474 4.404
0.004 0.036
12.36 0.963
OR
95% CI
Males (n¼87) Depressive temperament Age HAMA
1.09–1.57 0.90–0.97 0.86–1.00
Females (n¼91) o8 years of education Age
The correlation of TEMPS-A with other personality measures makes it possible to further refine the models postulating the role of personality in smoking behaviour through the medium of the affective temperamental components.
4.2. Gender differences in the predictive role of affective temperaments in smoking initiation and cessation Tobacco use is traditionally more prevalent in males than females (WHO, 2013). However, the increasing prevalence of smoking among females (WHO, 2010) and the less successful cessation attempts compared to males (Bjornson et al., 1995) make it plausible to explore the gender differences in underlying temperamental factors mediating the complex socio-cultural background. Our results show that higher depressive temperament scores are strongly associated with smoking maintenance and current smoker status in males. The available literature on personality measures and smoking shows that high scores on the depressive scale of the Millon Clinical Multiaxial Inventory-III make it less likely to quit smoking for males after 1 year of follow-up (Piñeiro et al., 2013). In females, we found that higher irritable temperament scores are strongly associated with smoking initiation and current smoker status. Although there is a lack of comparable studies in the field, it has been suggested that high impulsivitysensation seeking and general activity measured by the Alternative Five Factor Model has been related to high CPD in women (Nieva et al., 2011). Depressive temperament is predominant in females and encompasses empathy and self-denying as well as living in line with social norms (Akiskal and Akiskal, 2007). Dismantling persons with depressive temperament from such roles may result in clinical depression independent of gender (Akiskal and Akiskal, 2007). The positive correlation with harm avoidance and the negative correlation with novelty seeking (Akiskal and Akiskal, 2005b) may render this temperament protective against addictive behaviour; however it may render the introduction of new habits more difficult, hindering cessation. Irritable temperament is predominant in males (Rihmer et al., 2010) and characterized by explosive emotions to aversive events with negative affect (Akiskal and Akiskal, 2007; Kwapil et al., 2013; Akiskal and Mallya, 1987). This temperament is also related to heavy use of cocaine, other stimulants and alcohol in HIV infected patients (Moore et al., 2005), and is a predictor of selfreported nicotine dependence, alcohol abuse, and cannabis use along with cyclothymic temperament in a young adult population (Unseld et al., 2012). Moreover, irritable temperament was found to be a significant independent predictor of job stress related to interpersonal relationship in a Japanese population (Sakai et al., 2005). Our finding that higher scores on the irritable temperament subscale in women predicted smoking initiation and reached its highest values in the current smoker group is in line with earlier
2.27–67.22 0.93–0.998
reports of increased stress and negative affect experienced by smokers (Kassel et al., 2003). Although the relationship between affective temperaments and smoking habit is intricate and disentangling its many aspects exceeds the frame of this study, we propose the protective role of an affective temperament. These findings thereby support gender-specific behaviour on one hand, and may also carry vulnerability if “gender-atypical” behaviour is developed or if “gender-atypical” personality is predominant. Such considerations could be an explanation of the role of depressive temperament in smoking maintenance among males and irritable temperament in smoking initiation among females of our sample. Irritable temperament, traditionally characterized as “masculine”, may signal vulnerability to smoking initiation in females whereas depressive temperament with attributes traditionally labelled as “feminine” can serve to identify a subgroup of males more vulnerable to remain current smokers. 4.3. Limitations The present study should be interpreted in the light of the following limitations. First, our sample consists of chronically ill aging patients from primary care practices that is not representative of the general population, nonetheless carrying the extra advantage of greater impact of smoking on health. Second, all data on tobacco use was self-reported with no additional tests and nicotine dependence was not measured. Third, the cross-sectional design may make it difficult to draw within-subject conclusions. Finally, the effects of “selective survival bias” may further decrease the validity of the results. 4.4. Conclusion Our results illustrate the complex involvement of affective temperaments in smoking behaviour and course of smoking with differential gender-dependent effects. We submit that our results contribute not only to a better understanding of the risk factors related to smoking behaviour but also yield important contributions to a complex model of the involvement of personality in smoking, which could further assist in better screening and develop intervention targets. Further additional studies are required in order to investigate the nature of this complex relationship between personality, affective temperaments and smoking behaviour.
Role of funding source No funding was received for the work described in this paper.
Conflict of interest The authors declare no conflict of interest in relation to the present manuscript.
402
A. Eory et al. / Journal of Affective Disorders 172 (2015) 397–402
Acknowledgement Xenia Gonda and Peter Dome are recipients of the Janos Bolyai Research Fellowship of the Hungarian Academy of Sciences.
References Clinical Practice Guideline for Treating Tobacco Use and Dependence: 2008 Update. A U.S. Public Health Service Report. Am. J. Prev. Med. 35, 158–176. Akiskal, H.S., Akiskal, K.K., 2007. In search of Aristotle: temperament, human nature, melancholia, creativity and eminence. J. Affect. Disord. 100, 1–6. Akiskal, H.S., et al., 2005a. TEMPS-A: progress towards validation of a self-rated clinical version of the Temperament Evaluation of the Memphis, Pisa, Paris, and San Diego Autoquestionnaire. J. Affect. Disord. 85, 3–16. Akiskal, H.S., Akiskal, K.K., 2005a. Special issue. TEMPS: temperament evaluation of Memphis, Pisa, Paris and San Diego. J. Affect. Disord. 85, 1–242. Akiskal, H.S., Mallya, G., 1987. Criteria for the ‘soft’ bipolar spectrum: treatment implications. Psychopharmacol. Bull. 23, 68–73. Akiskal, H.S., et al., 2005b. TEMPS-A: validation of a short version of a self-rated instrument designed to measure variations in temperament. J. Affect. Disord. 85, 45–52. Akiskal, K.K., Akiskal, H.S., 2005b. The theoretical underpinnings of affective temperaments: implications for evolutionary foundations of bipolar disorder and human nature. J. Affect. Disord. 85, 231–239. Beck, A.T., et al., 1997. Screening for major depression disorders in medical inpatients with the Beck Depression Inventory for Primary Care. Behav. Res. Ther. 35, 785–791. Bjornson, W., et al., 1995. Gender differences in smoking cessation after 3 years in the lung health study. Am. J. Public Health 85, 223–230. Chapman, B., et al., 2009. Education and smoking: confounding or effect modification by phenotypic personality traits? Ann. Behav. Med. 38, 237–248. Cosci, F.P., F. Lazzarini, N., Carrozzi, L., 2011. Nicotine dependence and psychological distress: outcomes and clinical implications in smoking cessation. Psychol. Res. Behav. Manag. 4, 119–128. de Viron, S., et al., 2013. Environmental and genomic factors as well as interventions influencing smoking cessation: a systematic review of reviews and a proposed working model. Public Health Genomics 16, 159–173. Eory, A., et al., 2014a. Personality and cardiovascular risk: association between hypertension and affective temperaments – a cross-sectional observational study in primary care settings. Eur. J. Gen. Pract. PMID: 24456347. Eory, A., et al., 2014b. Affective temperaments contribute to cardiac complications in hypertension independently of depression. Psychother. Psychosom. 83, 187–189. Field, A., 2005. Discovering Statistics Using SPSS, first ed. SAGE Publications Ltd, London. Fornaro, M., et al., 2013. Treatment adherence towards prescribed medications in bipolar-II acute depressed patients: relationship with cyclothymic temperament and therapeutic sensation seeking in response towards subjective intolerance to pain. J. Affect. Disord. 151, 596–604. Gonda, X., et al., 2011. From putative genes to temperament and culture: cultural characteristics of the distribution of dominant affective temperaments in national studies. J. Affect. Disord. 131, 45–51. Gurpegui, M., et al., 2007. Personality traits associated with caffeine intake and smoking. Prog. Neuro-Psychopharmacol. Biol. Psychiatry 31, 997–1005. Hamilton, M., 1959. The assessment of anxiety states by rating. Br. J. Med. Psychol. 32, 50–55. Hooten, W., et al., 2005. Personality correlates related to tobacco abstinence following treatment. Int. J. Psychiatry Med. 35, 59–74.
Iasevoli, F., et al., 2013. Affective temperaments are associated with specific clusters of symptoms and psychopathology: a cross-sectional study on bipolar disorder inpatients in acute manic, mixed, or depressive relapse. J Affect Disord. 151, 540–550. Kassel, J.D., Stroud, L.R., Paronis, C.A., 2003. Smoking, stress, and negative affect: correlation, causation, and context across stages of smoking. Psychol. Bull. 129, 270–304. Kwapil, T.R., et al., 2013. Affective temperaments: unique constructs or dimensions of normal personality by another name? J. Affect. Disord. 151, 882–890. Malouff, J.M., Thorsteinsson, E.B., Schutte, N.S., 2006. The five-factor model of personality and smoking: a meta-analysis. J. Drug Educ. 36, 47–58. McDowell, I., 2006. Measuring Health. A Guide to Rating Scales and Questionnaires. Oxford University Press, Oxford. Mischel, W., et al., 2008. Introduction to Personality: Toward an Integrative Science of the Person, eighth ed John Wiley & Sons, http://eu.wiley.com/WileyCDA/ WileyTitle/productCd-EHEP000775.html. Moore, D.J., et al., 2005. Temperament and risky behaviors: a pathway to HIV? J. Affect. Disord. 85, 191–200. Nieva, G., et al., 2011. The alternative five-factor model of personality, nicotine dependence and relapse after treatment for smoking cessation. Addict. Behav. 36, 965–971. Perkins, K.A., 1999. Baseline-dependency of nicotine effects: a review. Behav. Pharmacol. 10, 597–615. Piñeiro, B., et al., 2013. Gender differences in personality patterns and smoking status after a smoking cessation treatment. BMC Public Health 13, 306. Ravara, et al., 2011. Smoking behaviour predicts tobacco control attitudes in a high smoking prevalence hospital: a cross-sectional study in a Portuguese teaching hospital prior to the national smoking ban. BMC Public Health 11, 720. Rafanelli, C., et al., 2013. Is depression the wrong treatment target for improving outcome in coronary artery disease? Psychother. Psychosom. 82, 285–291. Rihmer, Z., et al., 2010. Current research on affective temperaments. Curr. Opin. Psychiatry 23, 12–18. Rózsa, S., et al., 2008. A study of affective temperaments in Hungary: internal consistency and concurrent validity of the TEMPS-A against the TCI and NEOPI-R. Affective temperament, history of suicide attempt and family history of suicide in general practice patients. J. Affect. Disord. 106, 45–53. Rózsa, S., et al., 2006. Az affektív temperamentum: a TEMPS-A kérdőívvel szerzett hazai tapasztalatok. Psychiatr. Hung. 21, 147–160. Sakai, Y., et al., 2005. Temperament and job stress in Japanese company employees. Affective temperament, history of suicide attempt and family history of suicide in general practice patients. J. Affect. Disord. 85, 101–112. Shadel, W.G., et al., 2004. Investigating the big five personality factors and smoking: implications for assessment. J. Psychopathol. Behav. Assess. 26, 185–191. Smits, K., et al., 2004. Association of metabolic gene polymorphisms with tobacco consumption in healthy controls. Int. J. Cancer 110, 266–270. Terracciano, A., Costa, P.T., 2004. Smoking and the five-factor model of personality. Addiction 99, 472–481. Unseld, M., et al., 2012. The concept of temperament in psychoactive substance use among college students. Affective temperament, history of suicide attempt and family history of suicide in general practice patients. J. Affect. Disord. 141, 324–330. WHO. 2010. Gender, Women, and the Tobacco Epidemic. [online]. World Health Organization, Geneva. Available: 〈http://www.who.int/tobacco/publications/gender/ women_tob_epidemic/en/〉 (accessed 01.04 2014). WHO. 2013. WHO Report on the Global Tobacco Epidemic. 2013 [online]. World Health Organization, Geneva. Available: 〈http://www.who.int/tobacco/global_ report/2013/en/〉 (accessed 01.04 2014). Zatonski, W., et al., 2012. Tobacco smoking in countries of the European Union. Ann. Agric. Environ. Med. 19, 181–192.
Contents lists available at ScienceDirect
Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad
Research report
The association of affective temperaments with smoking initiation and maintenance in adult primary care patients Ajandek Eory a,n, Sandor Rozsa b,c, Xenia Gonda d,e,f, Peter Dome d,f, Peter Torzsa a, Tatevik Simavorian i, Konstantinos N. Fountoulakis g, Maurizio Pompili h, Gianluca Serafini h, Knarig K. Akiskal i, Hagop S. Akiskal i, Zoltan Rihmer d,f, Laszlo Kalabay a a
Department of Family Medicine, Semmelweis University, 4 Kutvolgyi Street, Budapest H-1125, Hungary Center for Well-Being, Department of Psychiatry, Washington University School of Medicine, St. Louis, MO, United States c Department of Personality and Health Psychology, Eötvös Loránd University, Budapest, Hungary d Department of Clinical and Theoretical Mental Health, Kutvolgyi Clinical Center, Semmelweis University, Budapest, Hungary e Department of Pharmacodymanics, Semmelweis University, Budapest, Hungary f Laboratory for Suicide Research and Prevention, National Institute of Psychiatry and Addictions, Budapest, Hungary g 3rd Department of Psychiatry, Division of Neurosciences, School of Medicine, Aristotle University of Thessaloniki, Greece h Department of Neurosciences, Mental Health and Sensory Organs, Suicide Prevention Center, Sant'Andrea Hospital, Sapienza University of Rome, 00189 Rome, Italy i International Mood Center, University of California, San Diego, CA, USA b
art ic l e i nf o
a b s t r a c t
Article history: Received 23 July 2014 Received in revised form 16 October 2014 Accepted 18 October 2014 Available online 28 October 2014
Background: Smoking behaviour and its course is influenced by personality factors. Affective temperaments could allow a more specific framework of the role trait affectivity plays in this seriously harmful health-behaviour. The aim of our study was to investigate if such an association exists in an ageing population with a special emphasis on gender differences. Methods: 459 primary care patients completed the TEMPS-A, Beck Depression Inventory (BDI) and Hamilton Anxiety Rating Scale (HAM-A). Subjects were characterized according to their smoking behaviour as current, former or never smokers. Univariate analysis ANOVA and logistic regression were performed to analyse differences in the three smoking subgroups to predict smoking initiation and maintenance. Results: Current smokers were younger and less educated than former or never smokers. Males were more likely to try tobacco during their lifetime and were more successful in cessation. Depressive, cyclothymic and irritable temperament scores showed significant differences between the three smoking subgroups. Irritable temperament was a predictor of smoking initiation in females whereas depressive temperament predicted smoking maintenance in males with a small, opposite effect of HAM-A scores independent of age, education, lifetime depression and BDI scores. Whereas smoking initiation was exclusively predicted by a higher BDI score in males, smoking maintenance was predicted by younger age and lower education in females. Limitations: The cross-sectional nature of the study design may lead to selective survival bias and hinder drawing causal relationships. Conclusions: Affective temperaments contribute to smoking initiation and maintenance independently of age, education, and depression. The significant contribution of depressive temperament in males and irritable temperament in females may highlight the role of gender-discordant temperaments in vulnerable subgroups. & 2014 Elsevier B.V. All rights reserved.
Keywords: Affective temperament Smoking Sex Mood disorder
1. Introduction
n
Corresponding author. Tel.: þ 36 304003640; fax: þ36 12015500. E-mail address: [email protected] (A. Eory).
http://dx.doi.org/10.1016/j.jad.2014.10.036 0165-0327/& 2014 Elsevier B.V. All rights reserved.
Tobacco dependence is a chronically relapsing addictive disorder, the most important preventable cause of illness and premature death due to cardiovascular, pulmonary and cancer diseases, and is responsible for the death of approximately 6 million every year (WHO, 2013,
398
A. Eory et al. / Journal of Affective Disorders 172 (2015) 397–402
Clinical Practice Guideline for Treating Tobacco Use and Dependence: 2008 Update, 2008). Although the addictive features of nicotine are the most widely researched component of smoking behaviour, other chemical substances of tobacco fume, sensory cues, heritability, environmental factors (social bonds, education, accessibility of tobacco products), and personality patterns also play an important role in the initiation, maintenance or successful termination of tobacco use (Cosci et al., 2011, de Viron et al., 2013, Clinical Practice Guideline for Treating Tobacco Use and Dependence: 2008 Update, 2008; Kassel et al., 2003). Individual differences in the mood-modulating effect of nicotine can be approached from two different directions. Pre-existing trait markers (like personality traits, trait anxiety or depression) are responsible for the inter-individual differences and they are applicable to identify the fundamentals of variation at the population level and therefore can be used for exploratory purposes (Kassel et al., 2003). Mood states together with stressful situations may explain the individual response to smoking cessation (Perkins, 1999; Kassel et al., 2003; Shadel et al., 2004). Gender differences in smoking initiation and cessation are fairly well known, and initiatives to address gender differences are imperative (WHO, 2010; Zatonski et al., 2012), but recent research related to the influence of personality on tobacco use is scarce, even though the influence of personality on tobacco use is differentially indicative of distinct personality patterns in the two genders (Piñeiro et al., 2013; Nieva et al., 2011). The broad definition of personality refers to cognitions, affects and behaviours that determine the direction and pattern to the person's life. Affective temperaments with their roots in the ancient Greek humoral theory are trait-like temperaments, which are partially genetically-determined and which remain relatively stable throughout the lifespan and determine the emotional response by an individual to the environmental stimuli (Akiskal and Akiskal, 2007; Mischel et al., 2008; Rihmer et al., 2010) exploring the affect-related component of personality. They can be assessed on the depressive, hyperthymic, cyclothymic, irritable and anxious subscales of the Temperament Evaluation of Memphis, Pisa, Paris and San Diego Autoquestionnaire (TEMPS-A). This was developed thanks to an intensive clinical and genetic research, as well as broad clinical experience and collaboration (Akiskal and Akiskal, 2005a). Although these affective temperaments, especially when expressed at their trait level, contribute to the development of a healthy personality, they may serve as subaffective trait precursors or precursors of major affective disorders, constituting risk factors for specific clusters of symptoms and psychopathology (Iasevoli et al., 2013), course and prognosis of major mood disorders (Rihmer et al., 2010, Akiskal et al., 2005a) and treatment adherence (Fornaro et al., 2013). Due to its easy clinical applicability the affective temperamental model is utilized as an outcome measure in affective disorder research as well as in explorative research in the frontier of major affective disorders and chronic somatic conditions, including hypertension (Eory, 2014a) and acute cardiac complications in hypertension (Eory, 2014b), among others (Rihmer et al., 2010). The rationale for examining the relationship between tobacco use and affective temperaments are as follows: a) Research shows that neuroticism and negative affect are important contributors to smoking, and emotional dysregulation appears to be the possible variable of this association (Kassel et al., 2003; Malouff et al., 2006). The TEMPS-A questionnaire delineates those emotional components of personality, which relate to reactions to inner and outer adverse events, and therefore may add substantial information to more deeply investigate the question at hand. b) Affective temperaments having strong biological components may further extend existing psychological models of the relationship between addictions and personality (Akiskal and Akiskal, 2005b).
c) TEMPS-A may provide the background personality factors of increased tobacco use in major affective disorders (Rihmer et al., 2010) d) TEMPS-A is tested against other personality instruments (TCI, NEO-PI-R) (Rózsa et al., 2008; Akiskal et al., 2005b). e) Temperament traits remain stable from early adulthood over time and may be more characteristic predictors of chronic somatic conditions than major affective disorders (Akiskal and Akiskal, 2005a; Rafanelli et al., 2013; Rihmer et al., 2010; Eory, 2014a, 2014b). Therefore, research on the relationship of affective temperaments with harmful health-behaviour in non-psychiatric, adult sample could prove to be both useful and necessary. The aim of our study was therefore to explore the association between affective temperaments and smoking habits in a chronically ill outpatient population. Because the affective temperament profile differs among males and females we use gender difference in our computations.
2. Methods 2.1. Sample Data from 509 consecutively investigated primary care patients were collected to explore the association between affective temperaments and hypertension (Eory, 2014a) as well as acute cardiac complications (Eory, 2014b). Altogether six primary care practices were involved into the study (four practices in Budapest and two practices in the countryside) between February and May of 2011. All practices were single-handed, with an average of 1500 patients in the catchment area and a daily turnover of 20–40 patients in each. Inclusion criteria included patients older than 18 years. According to the primary aim of the research, patients with primary hypertension comprised the majority of the sample. Exclusion criteria were secondary hypertension, improper understanding of Hungarian language and legal incapacity. Other chronic medical conditions were not part of the exclusion criteria. The present sample is a subsample of the original population containing 459 regularly controlled outpatients with data on tobacco use. The frequency of chronic medical conditions across the sample was as follows: 76% of the sample suffered from primary hypertension. Hypertension related complications were cardiovascular complications (acute myocardial infarction, acute coronary syndrome, angina pectoris) (17%); peripheral artery disease (14%); cerebrovascular accidents (stroke or transient ischaemic attack) (11%) and kidney-related complications (10%). According to their medical history 29% of the sample suffered from diabetes mellitus, followed by musculoskeletal (25%), gastrointestinal (20%), anxiety (ICD-10 diagnosed or anxiolytic medication) (23%), depressive (ICD-10 diagnosed) (15%) and endocrine (10%) diseases. Subjects were classified as smokers if they had smoked more than 100 cigarettes in their lifetime and smoked on a daily basis; as quitters if they reported quitting smoking more than 30 days prior; or as never smokers if they stated that they had not smoked a cigarette throughout their lifetime (Smits et al., 2004; Ravara et al., 2011). Information on the number of cigarettes smoked per day, the length of abstinence in years, demographic factors (age, sex) and education (eight or less years, nine to twelve years, more than twelve years) was collected after receiving written informed consent by patients. The study has been approved by the Scientific and Research Ethics Committee of Scientific Health Council in charge of experimentation with human subjects and was carried out in accordance with the tenets of the Declaration of Helsinki.
A. Eory et al. / Journal of Affective Disorders 172 (2015) 397–402
399
(McDowell, 2006). The Cronbach alpha coefficient for the present study population was 0.89.
2.2. Measures 2.2.1. TEMPS-A The Temperament Evaluation of Memphis, Pisa, Paris and San Diego Autoquestionnaire (TEMPS-A) is a 110-item self-administered paper–pencil instrument, applying true/false statements (Akiskal et al., 2005a). It measures affective temperaments containing 21 items for depressive, cyclothymic, hyperthymic and irritable (20 for men on the latter) subscales, and 26 for anxious temperament. The self-report questionnaire was translated into more than 30 languages and validated in several countries, including Hungary in 2006 (Rózsa, 2006). The Hungarian version showed good reliability and internal consistency. It was externally validated across the Beck Depression Inventory (BDI), Profile of Mood States (POMS), the BarOn Emotional Quotient Inventory (BarOn EQ-i), Temperament and Character Inventory (TCI) and the NEO Personality Inventory-Revised (NEO-PI-R) (Rózsa et al., 2008). To establish internal consistency of TEMPS-A in this study population, Cronbach alpha coefficients for scores were measured. Alpha coefficients for each subscale are as follows: 0.77 for depressive temperament, 0.83 for cyclothymic temperament, 0.84 for hyperthymic temperament, 0.78 for irritable temperament and 0.92 for anxious temperament. All alpha coefficients ranged between 0.77 and 0.92. The magnitude of the coefficients alpha was satisfactory and good in all cases (Rózsa et al., 2008).
2.2.2. HAM-A/HARS The Hamilton Anxiety Rating Scale (HAM-A) is one of the most extensively used scales to measure the severity of anxiety symptoms with 14 items (each ranging 0–4) (Hamilton, 1959). Scores between 14 and 17 points indicate mild, 18–24 points indicate moderate and, scores greater than 25 signify severe anxiety. It has good reliability and validity, sensitivity to change over time and applicability for a wide range of patients (from adolescents to older adults) (McDowell, 2006).
2.2.3. BDI The Beck Depression Inventory (BDI) covers the emotional, behavioural and somatic symptoms of depression with 21 items on a four points intensity scale (0–3) (Beck et al., 1997). Scores between 10 and 18 indicate mild to moderate, 19–29 moderate to severe and, greater than 30 severe depression. This selfadministered questionnaire is extensively used in clinical research
2.3. Statistical analyses Chi-square test was applied for the assessment of group differences in categorical variables (gender, tobacco use, ICD-10 depression). Gender-tests were used for the comparison of continuous variables between groups. We applied one-way ANOVA to explore the differences in the TEMPS-A, BDI and HAMA scores and age between smoking groups in both gender. In cases of violation of the homogeneity of variance (depressive temperament in both genders and irritable and anxious temperaments as well as HAMA scores in females), we used the Brown–Forsythe F-ratio (Field, 2005). In cases where the subsets were not homogeneous (depressive and cyclothymic temperament scores in males and irritable temperament score in females) the Games–Howell post-hoc test was used (Field, 2005). In all other situations the Hochberg GT2 post-hoc test was applied due to unequal sample sizes (Field, 2005). To establish the relationship of affective temperaments with smoking behaviour we used a step forward likelihood ratio binary logistic regression with never vs ever smokers (smoking initiation) and current vs former smokers (smoking maintenance) as dependent variables; and considered affective temperaments, lifetime depression, age, education, BDI and HAMA as predictor variables.
3. Results 3.1. Baseline characteristics and affective temperament profile of the sample population The sample comprised 459 primary care patients (278 females). Never smokers constituted 55% of the total sample, and former smokers were represented in a slightly larger proportion (24.4%) than current smokers (20.6%). Current smokers were younger (54 714 years) than never (59 715 years) or former smokers (62 713 years) (F(2, 419) ¼7.7; p ¼ 0.001), and were less educated (χ2(4)¼21.615; po 0.0001). The average number of cigarettes smoked per day (CPD) was 15 (78). Among lifetime smokers, 112 (54%) quit successfully with a mean of 16 years ( 712 years) prior to data collection. Table 1 shows gender differences of demographic, psychosocial and temperamental characteristics in the three smoking groups.
Table 1 Distribution of demographic, psychosocial and temperament characteristics between smoking status by gender (n ¼459). Males (181) Smoking status Age Education ( r 8 yr/9–12 yr/ 412 yr) % ICD-10 depression (%)
Females (278)
Never (n¼ 82) 45% 58 (15) 34/33/33 5.4
Former (n¼ 63) 35% 62 (13) 17/37/46 11.3
Current (n ¼36 20%) 52 (14) 44/36/20n 20.6
Never (170) 61% 61 (14) 30/38/32 14.4
Former (49) 18% 61 (13) 8/49/43 13.0
Current (59) 21% 55 (14) 31/47/22n 25.9
6 (0–17)n 4 (0–19) 12 (0–20) 5 (0–17) 5 (0–26) 6 (0–34) 8 (0–39)
6 (2–13)n 5 (0–18) 12 (1–20) 5 (0–15) 6 (0–22) 7 (0–31) 9 (0–41)
8 (3–19)n 7 (0–19) 12 (0–21) 6 (0–13) 6 (0–26) 9 (0–38) 9 (0–38) 17 (9)
9 (0–21) 5 (0–25) 10 (0–21) 4 (0–12)nn 9 (0–25) 8 (7) 11 (9)
7 (2–16) 5 (0–16) 11 (2–19) 5 (1–13) 7 (0–17) 8 (6) 10 (7)
9 (3–20) 6 (0–18) 11 (0–17) 5 (1–17)nn 9 (0–26) 9 (9) 12 (11) 14 (7)
TEMPS-A Depressive Cyclothymic Hyperthymic Irritable Anxious BDI HAM-A CPD Length of abstinence (yr)
16 (13)
15 (12)
Values are expressed in mean (SD) unless otherwise indicated, psychometric scores are expressed in mean (range). TEMPS-A: Temperament Evaluation of Memphis, Pisa, Paris and San Diego Autoquestionnaire, BDI: Beck Depression Inventory, HAM-A: Hamilton Anxiety Rating Scale, CPD: number of cigarettes per day. n
po 0.05. p o0.01.
nn
400
A. Eory et al. / Journal of Affective Disorders 172 (2015) 397–402
Among males the proportion of ever smokers (vs. never smokers) was higher than among females (χ2(1)¼7.99; p¼ 0.005) with a higher rate of successful cessation (63.6% vs 45.4%) (χ2(1)¼6.94; p¼0.008). Although lifetime depression was four times more prevalent in current male smokers than in never smokers, the difference of ICD-10 diagnosed depression showed marginal significance between the three smoking groups in males (χ2(2)¼5.72; p¼0.057) and remained non-significant in females just like the BDI or HAMA scores in both sexes. Affective temperament scores showed gender-specific differences. Males scored higher on hyperthymic (U¼ 20,440, p ¼0.001) and irritable (U¼22,177, p ¼0.036) temperament scales, while females on the depressive (U¼ 17,914, p o0.0001) and the anxious (U¼ 17,484, p o0.0001). Depressive (F(2,424)¼6.23; p ¼0.002), cyclothymic (F(2, 424)¼ 7.06; p¼ 0.001) and irritable temperament (F(2,424) ¼10.23; p o0.0001) scores revealed significant differences among the three smoking subgroups. When it was further examined by gender, depressive (Brown–Forsythe F-ratio F (2,88.14) ¼ 5.07; p ¼0.008) and cyclothymic temperament scores (F(2,178) ¼3.4; p ¼0.035) proved to be significant in males while it was the irritable temperament score in females (Brown–Forsythe F-ratio: F(2,142.3) ¼7.35; p ¼0.001). Depressive temperament scores showed significant differences in males between current smokers and both never (p ¼0.023) and former smokers (p¼ 0.042) as well as cyclothymic temperament scores showed marginal significance between current smoker and never smoker males (p¼ 0.053). In females, irritable temperament showed significant difference between current and never smokers (p¼ 0.003).
3.2. Age, education and psychometric factors as predictors of smoking initiation To examine smoking initiation, the original smoking subgroups (never, former, current) were dichotomized as ever (current and former) versus never smokers. Binary logistic regression (step forward LR) was conducted to determine the role of age, education (r8 years, 9–12 years and o12 years), depressive, cyclothymic, hyperthymic, irritable and anxious temperament subscales, ICD-10 depression, BDI and HAMA scores in the initiation of smoking in both genders separately. 159 males were included in the analysis (with no missing data in any of the variables). The final model was statistically significant (χ2(1)¼4.6; p¼0.032, Cox and Schnell R2 ¼0.028). Table 2 shows that higher point scores on the BDI predicted smoking initiation independently of other predictors, increasing the odds of smoking initiation by 5%. In females (214 with no missing data on any of the variables), the model was also significant (χ2(2)¼21.01; po0.0001, Cox and Schnell R2 ¼ 0.094). Higher point scores on the irritable temperament subscale predicted smoking initiation independently of other predictors, increasing the odds of smoking initiation by 26%, whereas anxious
Table 2 Psychosocial predictors of smoking initiation in males and females (logistic regression, step forward LR). β
SE
Wald χ2 p
OR
95% CI
Males (n ¼159) BDI
0.046 0.022
4.15
0.042 1.047 1.002–1.094
Females (n¼214) Irritable temperament 0.234 0.056 17.529 Anxious temperament 0.066 0.027 5.736
0.000 1.264 1.133–1.411 0.017 0.936 0.887–0.988
temperament scores were inversely related to smoking initiation [β ¼ 0.066, SE¼ 0.027; Wald χ2(1)¼5.736; p¼0.017]. 3.3. Age, education and psychometric factors associated with smoking maintenance To explore the association between affective temperaments, ICD-10 depression, BDI, HAM-A scores, age, education, and smoking maintenance, we conducted binary logistic regression (step forward LR) with current smokers vs former smokers as a dependent variable in both genders. The results are presented in Table 3. In males, the regression analysis terminated after the third step. The final model was statistically significant (χ2(3)¼22.293; po 0.0001, Cox and Schnell R2 ¼0.226). Higher scores on the depressive temperament scale predicted smoking maintenance, increasing the odds of persistent smoking by 30.5%. On the contrary, age was inversely related to current smoker status, decreasing the odds by 7% just like higher HAM-A scores. The analysis was terminated after the second step in females, and no psychosocial predictors were included. Lower education (8 years or less) predicted smoking maintenance increasing the odds by 12 times while age had an inverse relationship, decreasing the odds of smoking maintenance by 4%.
4. Discussion We examined the relationship of affective temperaments and smoking as a harmful health-behaviour in an ageing population under regular control by their general practitioners. The gender distribution of affective temperament scores was similar to that of the three-decade younger Hungarian normative sample (Rózsa et al., 2008) with higher hyperthymic scores for males and higher depressive and anxious scores for females, further supporting the trait-like nature of affective temperaments (Akiskal and Akiskal, 2007). The fact that cyclothymic temperament did not reveal significant difference by gender can be explained by the negative correlation of this temperament with ageing (Rihmer et al., 2010). Additionally we found higher irritability scores in males, which are in accordance with male dominance on the irritable scale in international samples (Akiskal and Akiskal, 2005a; Gonda et al., 2011; Rihmer et al., 2010). 4.1. The contribution of affective temperaments to smoking and their relationships with other personality measures Our main finding is that depressive, cyclothymic and irritable temperament showed association with never, former or current smoking status in our ageing primary care sample. Similar results were found with the TEMPS-M in a college student population with irritable and cyclothymic temperament as predictors of selfreported nicotine abuse (Unseld et al., 2012). Depressive, cyclothymic and irritable temperaments as well as the anxious subscale showed a moderate or strong positive correlation with harm avoidance of the Temperament and Character Inventory (TCI), whereas a cyclothymic temperament also showed a moderate positive correlation with novelty seeking (Akiskal et al., 2005b), the latter of which was found to be related to smoking initiation in an adult sample (Gurpegui et al., 2007). The same affective temperaments are positively correlated with the neuroticism scale of the NEO-Personality Inventory-Revised (NEO-PI-R) (Rózsa et al., 2008). Neuroticism contributed to both former (Chapman et al., 2009) and current smoker (Terracciano and Costa, 2004) status in adulthood and was found to be a predictor of nicotine dependence after a 2.4 years follow-up of treatment in 250 patients (99 abstinent and 151 smoking patients) (Hooten et al., 2005).
A. Eory et al. / Journal of Affective Disorders 172 (2015) 397–402
401
Table 3 Demographic and psychosocial predictors of smoking cessation (logistic regression, step forward LR). β
SE
Wald χ2
p
0.266 0.068 0.075
0.093 0.021 0.039
8.208 10.388 3.819
0.004 0.001 0.051
1.305 0.935 0.928
2.515 0.038
0.864 0.018
8.474 4.404
0.004 0.036
12.36 0.963
OR
95% CI
Males (n¼87) Depressive temperament Age HAMA
1.09–1.57 0.90–0.97 0.86–1.00
Females (n¼91) o8 years of education Age
The correlation of TEMPS-A with other personality measures makes it possible to further refine the models postulating the role of personality in smoking behaviour through the medium of the affective temperamental components.
4.2. Gender differences in the predictive role of affective temperaments in smoking initiation and cessation Tobacco use is traditionally more prevalent in males than females (WHO, 2013). However, the increasing prevalence of smoking among females (WHO, 2010) and the less successful cessation attempts compared to males (Bjornson et al., 1995) make it plausible to explore the gender differences in underlying temperamental factors mediating the complex socio-cultural background. Our results show that higher depressive temperament scores are strongly associated with smoking maintenance and current smoker status in males. The available literature on personality measures and smoking shows that high scores on the depressive scale of the Millon Clinical Multiaxial Inventory-III make it less likely to quit smoking for males after 1 year of follow-up (Piñeiro et al., 2013). In females, we found that higher irritable temperament scores are strongly associated with smoking initiation and current smoker status. Although there is a lack of comparable studies in the field, it has been suggested that high impulsivitysensation seeking and general activity measured by the Alternative Five Factor Model has been related to high CPD in women (Nieva et al., 2011). Depressive temperament is predominant in females and encompasses empathy and self-denying as well as living in line with social norms (Akiskal and Akiskal, 2007). Dismantling persons with depressive temperament from such roles may result in clinical depression independent of gender (Akiskal and Akiskal, 2007). The positive correlation with harm avoidance and the negative correlation with novelty seeking (Akiskal and Akiskal, 2005b) may render this temperament protective against addictive behaviour; however it may render the introduction of new habits more difficult, hindering cessation. Irritable temperament is predominant in males (Rihmer et al., 2010) and characterized by explosive emotions to aversive events with negative affect (Akiskal and Akiskal, 2007; Kwapil et al., 2013; Akiskal and Mallya, 1987). This temperament is also related to heavy use of cocaine, other stimulants and alcohol in HIV infected patients (Moore et al., 2005), and is a predictor of selfreported nicotine dependence, alcohol abuse, and cannabis use along with cyclothymic temperament in a young adult population (Unseld et al., 2012). Moreover, irritable temperament was found to be a significant independent predictor of job stress related to interpersonal relationship in a Japanese population (Sakai et al., 2005). Our finding that higher scores on the irritable temperament subscale in women predicted smoking initiation and reached its highest values in the current smoker group is in line with earlier
2.27–67.22 0.93–0.998
reports of increased stress and negative affect experienced by smokers (Kassel et al., 2003). Although the relationship between affective temperaments and smoking habit is intricate and disentangling its many aspects exceeds the frame of this study, we propose the protective role of an affective temperament. These findings thereby support gender-specific behaviour on one hand, and may also carry vulnerability if “gender-atypical” behaviour is developed or if “gender-atypical” personality is predominant. Such considerations could be an explanation of the role of depressive temperament in smoking maintenance among males and irritable temperament in smoking initiation among females of our sample. Irritable temperament, traditionally characterized as “masculine”, may signal vulnerability to smoking initiation in females whereas depressive temperament with attributes traditionally labelled as “feminine” can serve to identify a subgroup of males more vulnerable to remain current smokers. 4.3. Limitations The present study should be interpreted in the light of the following limitations. First, our sample consists of chronically ill aging patients from primary care practices that is not representative of the general population, nonetheless carrying the extra advantage of greater impact of smoking on health. Second, all data on tobacco use was self-reported with no additional tests and nicotine dependence was not measured. Third, the cross-sectional design may make it difficult to draw within-subject conclusions. Finally, the effects of “selective survival bias” may further decrease the validity of the results. 4.4. Conclusion Our results illustrate the complex involvement of affective temperaments in smoking behaviour and course of smoking with differential gender-dependent effects. We submit that our results contribute not only to a better understanding of the risk factors related to smoking behaviour but also yield important contributions to a complex model of the involvement of personality in smoking, which could further assist in better screening and develop intervention targets. Further additional studies are required in order to investigate the nature of this complex relationship between personality, affective temperaments and smoking behaviour.
Role of funding source No funding was received for the work described in this paper.
Conflict of interest The authors declare no conflict of interest in relation to the present manuscript.
402
A. Eory et al. / Journal of Affective Disorders 172 (2015) 397–402
Acknowledgement Xenia Gonda and Peter Dome are recipients of the Janos Bolyai Research Fellowship of the Hungarian Academy of Sciences.
References Clinical Practice Guideline for Treating Tobacco Use and Dependence: 2008 Update. A U.S. Public Health Service Report. Am. J. Prev. Med. 35, 158–176. Akiskal, H.S., Akiskal, K.K., 2007. In search of Aristotle: temperament, human nature, melancholia, creativity and eminence. J. Affect. Disord. 100, 1–6. Akiskal, H.S., et al., 2005a. TEMPS-A: progress towards validation of a self-rated clinical version of the Temperament Evaluation of the Memphis, Pisa, Paris, and San Diego Autoquestionnaire. J. Affect. Disord. 85, 3–16. Akiskal, H.S., Akiskal, K.K., 2005a. Special issue. TEMPS: temperament evaluation of Memphis, Pisa, Paris and San Diego. J. Affect. Disord. 85, 1–242. Akiskal, H.S., Mallya, G., 1987. Criteria for the ‘soft’ bipolar spectrum: treatment implications. Psychopharmacol. Bull. 23, 68–73. Akiskal, H.S., et al., 2005b. TEMPS-A: validation of a short version of a self-rated instrument designed to measure variations in temperament. J. Affect. Disord. 85, 45–52. Akiskal, K.K., Akiskal, H.S., 2005b. The theoretical underpinnings of affective temperaments: implications for evolutionary foundations of bipolar disorder and human nature. J. Affect. Disord. 85, 231–239. Beck, A.T., et al., 1997. Screening for major depression disorders in medical inpatients with the Beck Depression Inventory for Primary Care. Behav. Res. Ther. 35, 785–791. Bjornson, W., et al., 1995. Gender differences in smoking cessation after 3 years in the lung health study. Am. J. Public Health 85, 223–230. Chapman, B., et al., 2009. Education and smoking: confounding or effect modification by phenotypic personality traits? Ann. Behav. Med. 38, 237–248. Cosci, F.P., F. Lazzarini, N., Carrozzi, L., 2011. Nicotine dependence and psychological distress: outcomes and clinical implications in smoking cessation. Psychol. Res. Behav. Manag. 4, 119–128. de Viron, S., et al., 2013. Environmental and genomic factors as well as interventions influencing smoking cessation: a systematic review of reviews and a proposed working model. Public Health Genomics 16, 159–173. Eory, A., et al., 2014a. Personality and cardiovascular risk: association between hypertension and affective temperaments – a cross-sectional observational study in primary care settings. Eur. J. Gen. Pract. PMID: 24456347. Eory, A., et al., 2014b. Affective temperaments contribute to cardiac complications in hypertension independently of depression. Psychother. Psychosom. 83, 187–189. Field, A., 2005. Discovering Statistics Using SPSS, first ed. SAGE Publications Ltd, London. Fornaro, M., et al., 2013. Treatment adherence towards prescribed medications in bipolar-II acute depressed patients: relationship with cyclothymic temperament and therapeutic sensation seeking in response towards subjective intolerance to pain. J. Affect. Disord. 151, 596–604. Gonda, X., et al., 2011. From putative genes to temperament and culture: cultural characteristics of the distribution of dominant affective temperaments in national studies. J. Affect. Disord. 131, 45–51. Gurpegui, M., et al., 2007. Personality traits associated with caffeine intake and smoking. Prog. Neuro-Psychopharmacol. Biol. Psychiatry 31, 997–1005. Hamilton, M., 1959. The assessment of anxiety states by rating. Br. J. Med. Psychol. 32, 50–55. Hooten, W., et al., 2005. Personality correlates related to tobacco abstinence following treatment. Int. J. Psychiatry Med. 35, 59–74.
Iasevoli, F., et al., 2013. Affective temperaments are associated with specific clusters of symptoms and psychopathology: a cross-sectional study on bipolar disorder inpatients in acute manic, mixed, or depressive relapse. J Affect Disord. 151, 540–550. Kassel, J.D., Stroud, L.R., Paronis, C.A., 2003. Smoking, stress, and negative affect: correlation, causation, and context across stages of smoking. Psychol. Bull. 129, 270–304. Kwapil, T.R., et al., 2013. Affective temperaments: unique constructs or dimensions of normal personality by another name? J. Affect. Disord. 151, 882–890. Malouff, J.M., Thorsteinsson, E.B., Schutte, N.S., 2006. The five-factor model of personality and smoking: a meta-analysis. J. Drug Educ. 36, 47–58. McDowell, I., 2006. Measuring Health. A Guide to Rating Scales and Questionnaires. Oxford University Press, Oxford. Mischel, W., et al., 2008. Introduction to Personality: Toward an Integrative Science of the Person, eighth ed John Wiley & Sons, http://eu.wiley.com/WileyCDA/ WileyTitle/productCd-EHEP000775.html. Moore, D.J., et al., 2005. Temperament and risky behaviors: a pathway to HIV? J. Affect. Disord. 85, 191–200. Nieva, G., et al., 2011. The alternative five-factor model of personality, nicotine dependence and relapse after treatment for smoking cessation. Addict. Behav. 36, 965–971. Perkins, K.A., 1999. Baseline-dependency of nicotine effects: a review. Behav. Pharmacol. 10, 597–615. Piñeiro, B., et al., 2013. Gender differences in personality patterns and smoking status after a smoking cessation treatment. BMC Public Health 13, 306. Ravara, et al., 2011. Smoking behaviour predicts tobacco control attitudes in a high smoking prevalence hospital: a cross-sectional study in a Portuguese teaching hospital prior to the national smoking ban. BMC Public Health 11, 720. Rafanelli, C., et al., 2013. Is depression the wrong treatment target for improving outcome in coronary artery disease? Psychother. Psychosom. 82, 285–291. Rihmer, Z., et al., 2010. Current research on affective temperaments. Curr. Opin. Psychiatry 23, 12–18. Rózsa, S., et al., 2008. A study of affective temperaments in Hungary: internal consistency and concurrent validity of the TEMPS-A against the TCI and NEOPI-R. Affective temperament, history of suicide attempt and family history of suicide in general practice patients. J. Affect. Disord. 106, 45–53. Rózsa, S., et al., 2006. Az affektív temperamentum: a TEMPS-A kérdőívvel szerzett hazai tapasztalatok. Psychiatr. Hung. 21, 147–160. Sakai, Y., et al., 2005. Temperament and job stress in Japanese company employees. Affective temperament, history of suicide attempt and family history of suicide in general practice patients. J. Affect. Disord. 85, 101–112. Shadel, W.G., et al., 2004. Investigating the big five personality factors and smoking: implications for assessment. J. Psychopathol. Behav. Assess. 26, 185–191. Smits, K., et al., 2004. Association of metabolic gene polymorphisms with tobacco consumption in healthy controls. Int. J. Cancer 110, 266–270. Terracciano, A., Costa, P.T., 2004. Smoking and the five-factor model of personality. Addiction 99, 472–481. Unseld, M., et al., 2012. The concept of temperament in psychoactive substance use among college students. Affective temperament, history of suicide attempt and family history of suicide in general practice patients. J. Affect. Disord. 141, 324–330. WHO. 2010. Gender, Women, and the Tobacco Epidemic. [online]. World Health Organization, Geneva. Available: 〈http://www.who.int/tobacco/publications/gender/ women_tob_epidemic/en/〉 (accessed 01.04 2014). WHO. 2013. WHO Report on the Global Tobacco Epidemic. 2013 [online]. World Health Organization, Geneva. Available: 〈http://www.who.int/tobacco/global_ report/2013/en/〉 (accessed 01.04 2014). Zatonski, W., et al., 2012. Tobacco smoking in countries of the European Union. Ann. Agric. Environ. Med. 19, 181–192.