Available online at www.sciencedirect.com

y 30 (2008) 331–336

General Hospital Psychiatr

Association between anger and first-onset primaryspontaneous pneumothorax

Sang-Hyuk Lee, M.D., Ph.D.a, Ho Choi, M.D.b, Seoyoung Kim, M.D.c,Tae Kyou Choi, M.D., Ph.D.a, Sungsoo Lee, M.D.b,⁎, Borah Kim, M.D.a,

Shin Young Suh, M.D., Ph.D.a, Ki-Hwan Yook, M.D., Ph.D.a, Yong Woo Kim, M.D.aaDepartment of Psychiatry, Pochon CHA University College of Medicine, Seongnam-Si, Kyounggi-Do 463-712, Republic of Korea

bDepartment of Thoracic and Cardiovascular Surgery, Ajou University School of Medicine, Youngtong-Gu, Suwon-Si, Kyounggi-Do 443-721,Republic of Korea

cDepartment of Psychiatry, Sunchonhyang University College of Medicine, Cheonan-Si, Chungcheongnam-Do 336-745, Republic of Korea

Received 10 January 2008; accepted 25 February 2008

Abstract

Objective: Primary spontaneous pneumothorax (PSP) is a frequent and problematic disease, but its underlying causes and pathophysiologyremain unclear. This study examined whether anger, which is related to many psychosomatic diseases, is a psychosocial factor associatedwith first-onset PSP.Method:We administered the State–Trait Anger Expression Inventory, Stress Response Inventory, Coping Scale, Beck Depression Inventoryand Global Assessment of Recent Stress to 91 patients with first-onset PSP and to 77 patients with recent minor trauma as controls.Results: The scores on anger-in, anger-out, state anger and trait anger were significantly higher in the PSP group than in the control group.Logistic regression analysis revealed that low body mass index and trait anger could be associated with PSP.Conclusion: We hypothesize that anger could play a role in the pathophysiology of PSP.© 2008 Elsevier Inc. All rights reserved.

Keywords: Anger; Primary spontaneous pneumothorax

1. Introduction

Primary spontaneous pneumothorax (PSP) remains asignificant health problem. Spontaneous pneumothoraxoccurs without preceding trauma or obvious precipitatingcause in patients without clinically apparent lung disease,with estimated incidences of 7.4–18 male patients and 1.2–6female patients (age-adjusted incidence) per 100,000population per year [1,2]. Since PSP is defined as thepresence of air in the pleural cavity, chest radiography can beused to detect displacement of the visceral pleural line fromthe chest wall, which is the key to a definitive diagnosis [3].Cigarette smoking and low body mass index (BMI) arethought to be responsible for the induction of PSP [4].

The pathophysiology of pneumothorax remains unknown,but several assumptions have been made. First, emphysema-

⁎ Corresponding author. Tel.: +82 31 219 5213; fax: +82 31 219 5215.E-mail address: chestlee@ajou.ac.kr (S. Lee).

0163-8343/$ – see front matter © 2008 Elsevier Inc. All rights reserved.doi:10.1016/j.genhosppsych.2008.02.008

like changes (blebs or bullae) as found in imaging studiesmay play a role in the development of PSP, with PSP possiblybeing the result of bullae rupture [5]. However, histopatho-logical analysis of surgically resected subpleural blebs orbullae has not always demonstrated the defects responsiblefor the air leakage in the visceral pleura or resected bullae[6,7]. Second, the ectomorphic physique often noted in PSPpatients has been present since childhood and is drivenprimarily by the patient being taller than average, which isparticularly prominent in early teens (is as a low BMI). Therapid increase in the length of the thorax may affectintrathoracic pressure at the lung apex and drive subpleuralcyst formation (emphysema-like change formation) [8].Third, smoking has been strongly implicated in thepathogenesis of PSP and is associated with at least a ninefoldincrease in the risk of developing a first-onset PSP [9]. On theother hand, it is not clear whether cessation of cigarettesmoking reduces the likelihood of occurrence or recurrenceof PSP [10]. Such debates mean that the pathophysiology of

332 S.-H. Lee et al. / General Hospital Psychiatry 30 (2008) 331–336

PSP remains unclear. Therefore, these assumptions cannotexplain the development of PSP over the wide spectrum ofpatient characteristics.

According to Spielberger [11], anger is an emotional statethat consists of feelings of variable intensity, from mildirritation or annoyance to intense fury and rage. Anger isrelated to several psychosomatic diseases, including cere-brovascular disease, headache, pain disorders and, espe-cially, coronary heart disease [12–14]. The mechanism ofthis association is not clear, but it is assumed that emotionalstress stimulates heightened sympathetic activation, vagalwithdrawal, pressor responses and catecholamine release,and increases circulating interleukin-6 and other inflamma-tory markers, together with platelet activation and prothrom-botic responses [15,16]. Negative psychological states suchas hopelessness, pessimism, rumination and anxiety are alsoassociated with coronary artery disease [17], and emotionaltriggers are associated with the development of acutecoronary syndrome [18].

An association between anger and PSP has beensuggested [19]. However, that study involved a fairly smallsample (N=34) and had methodology limitations in that itcompared anger level between PSP patients and normalcontrols. Therefore, the association between anger and PSPhas not been closely examined.

The present study examined whether anger or other stress-related factors contribute to the development of first-onsetPSP in a larger number of patients with PSP.

2. Methods

One hundred seven patients with first-onset PSP at theDepartment of Chest Surgery, Ajou Medical Center, and83 patients with minor trauma at the Departments of ChestSurgery and Orthopedic Surgery, Ajou Medical Center andBundang CHA Hospital, were included in this study.We recruited both groups of patients from the two Koreangeneral hospitals from August 2005 to February 2007.The patients with minor trauma were age matched andgender matched with the PSP patients. The patients includedin the present study were aged ≥18 years and had beendiagnosed with PSP based on the presence of air in thepleural cavity on chest radiographs obtained by chestsurgery specialists (H.C. and S.L.). The exclusion criteriawere as follows: (a) presence of iatrogenic or traumaticpneumothorax; (b) failure to complete N80% of thequestionnaires; (c) history of pneumothorax; (d) brainlesion and/or seizure; (e) psychiatric illness or history, orpsychiatric family history; and (f) major acute comorbiditiesor major medical problems such as diabetes and hyperten-sion. Minor trauma referred to patients who did not needmajor orthopedic surgery and included those who hadsimply sprained their wrists and ankles, or had contusion onthe chest wall without a history of PSP. The research outlinewas explained to potential participants, and only patients

and their families who signed an informed consent werechosen to participate in this study. Overall, 91 patients withfirst-onset PSP and 77 patients with minor trauma (sprain,45 patients; chest contusion, 32 patients) were analyzedafter excluding those who met the exclusion criteria.Appropriate Ethics Committee approval was obtained forthe research. Patients in both groups took nonsteroidal anti-inflammatory drugs orally or via injection, and all patientsunderwent psychiatric interviews, including a mental statusexamination applying Diagnostic and Statistical Manual ofMental Disorders, Fourth Edition criteria [20]. Allexaminations were administered by subject-blinded board-certified psychiatrists (S.L. and T.C.); after the interview,the patients were asked to fill in questionnaires around3 days after relief of acute symptoms. Psychotropic drugs,including antidepressants or benzodiazepines, were notallowed in either group.

2.1. Psychometric measures

2.1.1. State–Trait Anger Expression Inventory (STAXI)The questionnaire included the Korean version [21] of the

STAXI [11], which is a 44-item self-rating instrumentdesigned to assess the levels of state anger, trait anger, anger-out, anger-in and anger control. The items on the STAXIwere rated on a 5-point scale ranging from 1 (not at all oralmost never) to 4 (very much so or almost always). Theinternal consistency of each of the five subscales wassignificantly high (Cronbach's αN.70). The test–retestreliabilities of the eight subscale scores and the total scorewere fairly high, ranging from 0.60 and 0.82. Constructvalidity was examined using factor analysis, and the resultsshowed that the Korean version of the STAXI had the samestructure as the original English scale. Criteria validity wasalso confirmed. The means±standard deviations (S.D.) of thesubscales were as follows: state anger, 12.5±4.1; trait anger,17.8±4.3; anger-in, 8.0±2.1; anger-out, 5.9±2.1; angercontrol, 15.9±5.1.

2.1.2. Global Assessment of Recent Stress (GARS)Perceived stress was quantified by total scores on the

GARS scale (Items 1–7) [22]. The GARS is a self-ratinginstrument used to assess the severity of recent ‘stressors’ inseven areas and one overall area during the week precedingits administration.

2.1.3. Coping ScaleThe Coping Scale developed by Folkman et al. [23] was

administered to assess the coping strategies employed instressful situations. Fifty items are scored on a 5-point scale.This Coping Scale consists of eight subscales: confrontation,distancing, self-control, seeking social support, acceptingresponsibility, escape–avoidance, planned problem solvingand positive reappraisal.

2.1.4. Stress Response Inventory (SRI)The SRI is a self-rated scale developed by Koh et al. [24]

that assesses the ‘stress responses’ of subjects experienced

Table 1Sociodemographic characteristics of patients with first-onset PSP andpatients with minor trauma

Patientswith PSP(n=91)

Patients withminor trauma(n=77)

t or χ2 df P

Age (years)[mean±S.D.]

21.1±5.0 20.7±1.1 0.9 113 .54

Gender [n (%)]Male 82 (90) 73 (96) .38 a

Female 9 (10) 3 (4)Education (years)[mean±S.D.]

12.3±2.2 12.7±1.3 −1.2 146 .19

Married [n (%)] 3 (4) 2 (3) .24 a

BMI (kg/m2)[mean±S.D.]

19.1±1.8 21.2±3.0 −5.3 162 .00 ⁎

Smoking amount(pack-years)[mean±S.D.]

0.76±2.6 0.59±1.9 0.5 157 .49

Smokers[n (%)] 21(26) 16(20) 0.5 1 .47Alcohol drinking(ml/month)[mean±S.D.]

20.3±50.7 20.3±20.7 0.03 143 .97

t=Student's t test; χ2=chi-square test.a Fisher's Exact Test.⁎ Pb.01.

Table 2Comparison of anger level and depression severity between patients withfirst-onset PSP and patients with minor trauma

Patients withPSP (n=91)[mean±S.D.]

Patients withminor trauma(n=77)[mean±S.D.]

t df P

State anger 20.7±6.4 15.0±7.8 5.1 166 .00 ⁎⁎

Trait anger 25.9±6.6 18.2±7.9 6.8 166 .00 ⁎⁎

333S.-H. Lee et al. / General Hospital Psychiatry 30 (2008) 331–336

over a 1-week period. The scale consists of seven subscales:tension, aggression, somatization, anger, depression, fatigueand frustration, with each item scored on a 5-point scale.

2.1.5. Beck Depression Inventory (BDI)The BDI is a self-rated scale developed by Beck et al. [25]

to assess the severity of depression. Twenty-one items arerated on a 3-point scale, with the total score obtained as thesummed score for all items. Lee et al. [26] assessed thevalidity and reliability of the version of the scale used here.

2.2. Data analysis

Independent t test and chi-square test were used tocompare sociodemographic findings, anger level and stress-related scales between the PSP group and the control group.Bivariate associations of sociodemographic variables (e.g.,age, smoking and alcohol consumption) and BMI with angerlevels were examined in patients with PSP using Pearsoncorrelation coefficient. Multivariate logistic regressionanalysis (SPSSWIN, version 11; Chicago, IL, USA) wasused to examine the relationship between anger levels andPSP and to calculate the odd ratios of associated factorsof PSP.

Anger levelAnger-in 12.5±6.3 8.4±5.4 5.8 165 .00 ⁎⁎

Anger-out 7.9±5.7 5.5±4.9 2.3 166 .02 ⁎

Anger control 18.9±4.8 17.2±4.5 1.1 166 .21DepressionBDI 8.1±7.0 4.6±4.5 4.7 166 .00 ⁎⁎

t=Student's t test.⁎ Pb.05.⁎⁎ Pb.01.

3. Results

3.1. Comparison of sociodemographic characteristicsbetween patients with PSP and the control group

Age, gender, smoking amount, percentage of smokersand alcohol consumption did not differ significantly between

the PSP group and the control group (Table 1). However,BMI was significantly lower in the PSP group than in thecontrol group.

3.2. Comparison of anger levels and depression severitybetween first-onset patients with PSP and the control group

Comparison of anger levels and depression severitybetween patients with PSP and the control group usingStudent's t test revealed that the scores on state anger, traitanger anger-in and anger-out andBDI scoreswere significantlyhigher in the PSP group than in the control group (Table 2).

3.3. Comparison of stress perceptions between patients withfirst-onset PSP and the control group

Student's t test was used to compare the scores of eachscale of stress perception between the two groups. In terms ofstressors, the GARS 1 (work/job/study), GARS 2 (inter-personal relationship) and GARS 8 (general assessment)scores were significantly higher in the PSP group than in thecontrol group (Table 3). In terms of stress coping, scores onthe confrontation, distancing, self-control, escape–avoid-ance, planned problem solving and positive reappraisalsubscales were significantly higher in the PSP group than inthe control group. In terms of stress response, scores on theanger, somatization, depression, tension, fatigue and frustra-tion subscales were significantly higher in the PSP groupthan in the control group.

3.4. Correlation between sociodemographic characteristicsand psychosocial factors in first-onset PSP

Simple correlation analysis revealed that BMI wasassociated with anger-in (r=−.25, Pb.01), state anger (r=−.22, Pb.01), trait anger (r=−.29, Pb.01), score on the SRIsomatization subscale (r=−.31, Pb.01), confrontation (r=−.18, Pb.01) and positive reappraisal (r=−.18, Pb.05), andthat age was associated with anger-out (r=−.14, Pb.05) andscore on the SRI aggression subscale (r=−.21, Pb.01).

able 4ignificant level of variables, including regression equation, on predictingrst-onset PSP

β Standarderror

Wald df P Oddsratio

95%confidencelimits

Table 3Comparison of recent stress perceptions between patients with first-onsetPSP and patients with minor trauma

Patients withPSP (n=91)[mean±S.D.]

Patients withminor trauma(n=77)[mean±S.D.]

t df P

StressorGARS (work/job/study) 4.6±2.2 3.2±2.0 4.5 166 .00 ⁎

GARS (interpersonalrelationship)

3.0±2.1 2.0±1.9 3.3 166 .00 ⁎

GARS (generalassessment)

3.5±2.5 1.9±1.8 4.8 166 .00 ⁎

GARS total 18.2±10.1 15.1±10.4 1.9 166 .05Coping strategyConfrontation 8.3±2.9 6.9±3.1 3.6 166 .00 ⁎

Distancing 7.4±3.6 5.4±3.2 3.7 166 .00 ⁎

Self-control 6.6±3.0 5.4±3.5 2.5 166 .01 ⁎

Escape–avoidance 9.0±3.7 7.4±4.5 2.8 166 .01 ⁎

Planned problem solving 7.2±3.4 5.7±3.5 3.2 166 .01 ⁎

Positive appraisal 9.8±4.0 7.7±4.5 3.5 166 .00 ⁎

Stress responseAggression 1.5±2.9 0.9±2.0 2.6 161 .09Somatization 2.7±2.3 1.3±1.8 4.3 165 .00 ⁎

Anger 4.2±4.4 1.9±2.9 4.8 157 .00 ⁎

Depression 5.9±3.7 3.6±4.9 3.3 163 .00 ⁎

Tension 4.2±3.5 2.3±3.1 3.7 165 .00 ⁎

Fatigue 4.2±4.4 3.7±3.5 4.3 164 .00 ⁎

Frustration 7.2±5.1 3.9±4.1 5.1 165 .00 ⁎

t=Student's t test.⁎ Pb.01.

334 S.-H. Lee et al. / General Hospital Psychiatry 30 (2008) 331–336

3.5. Identification of the psychosocial associated factors offirst-onset PSP using logistic regression analysis

Multivariate logistic regression analysis, with both groupsas dependent variables and with those that differedsignificantly between the PSP group and the control groupon independent t test as covariates, showed that BMI andtrait anger were significantly associated with PSP (Table 4;that is, the covariates were state anger, trait anger, anger-in,anger-out, GARS 1 (work/job/study), GARS 2 (interpersonalrelationship), GARS 8 (general assessment), and theconfrontation, distancing, self-control, escape–avoidance,planned problem solving, positive reappraisal, anger,somatization, depression, tension, fatigue and frustrationsubscales of the SRI. There was no collinearity among thecovariates (variance inflation factor of all covariates b10),and there were no interaction terms that should beconsidered. Although there was a significant correlationbetween BMI and trait anger, logistic regression analysisrevealed that these two variables are independently asso-ciated with PSP.

MI 0.51 0.14 12.9 1 .00 ⁎ 1.6 1.26–2.22rait anger −0.14 0.6 0.77 1 .03 ⁎⁎ 0.86 0.75–0.98onstant −7.6 2.9 6.4 1 .00 0.00

2 log likelihood=158.9; Cox and Snell R2=.33; Nagelkerke R2=.44;lassification accuracy=73.2.⁎ Pb.05.⁎⁎ Pb.01.

4. Discussion

The purpose of this study was to determine whether thelevels of anger and other stress-related factors are higher inpatients with first-onset PSP than in patients with minor

trauma (control group). The findings revealed that the levelsof anger-in, anger-out, state anger and trait anger were higherin patients with first-onset PSP than in patients with minortrauma. Logistic regression analysis showed that trait angercould be a predictive value for the development of PSP.These results are consistent with a previous study [19] thatsuggested for the first time that anger is associated with PSP,raising the possibility that negative emotions such as angerare associated with PSP.

In this study, BMI was lower in the PSP group than in thecontrol group, and logistic regression analysis also revealedthat BMI is a risk factor for PSP in Korea, consistent withprevious reports. However, we found that the amount ofsmoking did not differ significantly between the PSP groupand the control group, which was inconsistent with previousstudies [4–9]. The finding that there was no significantdifference in the amount of smoking between the PSPpatients and controls might be attributable to the character-istics of the controls and the PSP patients. Several inferencescan be made from this finding. First, the control group is aminor trauma group. Some young patients with minortrauma may have a tendency to engage in certain riskybehaviors such as fighting, reckless driving and substanceuse. Although young people with major psychiatric disorderswere excluded at screening, the characteristics of some of thepatients in the control group might have contributed to thehigh smoking rate in this group. Second, in Korea, mucheffort is being exerted to encourage young people to stopsmoking and to prohibit smoking in high schools, which maylead to lower smoking rates in ‘normal’ young people.Therefore, there is a possibility that PSP patients who werenot associated with smoking were enrolled in this study. Thefindings that the mean amount of smoking in the PSP groupwas only 0.76 pack-years and that the percentage of smokersin PSP was only 26% (77% in the previous study [4]) supportthis assumption. Studies with larger samples are thereforeneeded to clarify this inconsistency.

Furthermore, since anger or depression may be stronglyassociated with smoking, there is a possibility that anger,rather than smoking, might show a false-positive associationwith PSP. However, neither depression nor any of the angersubscales used in our study was significantly correlated with

TSfi

BTC

−c

Fig. 1. Hypothesis on the development of PSP.

335S.-H. Lee et al. / General Hospital Psychiatry 30 (2008) 331–336

the amount of smoking (not shown in Results). In addition,anger-in (r=−.22, P=.04) and anger-out (r=−.24, P=.03)only showed significant negative associations with smokingin the PSP group. Therefore, the possibility that anger, ratherthan smoking, is associated with PSP in this study isless likely.

From the results presented here, we can make severalassumptions and hypotheses. First, sustained anger mightplay a role in the pathophysiology of PSP, since ourlogistic regression analysis revealed that trait anger couldbe associated with PSP. Anger often leads to negativesocial interactions and more opportunities to experienceanger. An interpersonal relationship can deteriorate if theseemotions persist, finally leading to biological changes inthe body: heightened sympathetic activation, vagal with-drawal, pressor response, catecholamine release, increasesin circulating interleukin-6 and other inflammatory mar-kers, together with long-term platelet activation andprothrombotic responses [15,16]. A study [27] found thatpulmonary vasculopathy was associated with spontaneouspneumothorax in young subjects: pulmonary artery intimalfibrosis in 90% of cases, pulmonary vein intimal fibrosisin 80% of cases, and some fibrosis and chronicinflammation in 90% of cases. Pathology textbooks[28,29] have also reported these findings. Therefore, wehypothesize that biological changes induced by angermake pulmonary vessels more vulnerable to PSP. Inaddition, even though the evidence is relatively weak,cytokines and other inflammatory markers [30] or thechronic elevation of intrathoracic pressure induced bysustained anger in PSP patients might also affect thedevelopment of PSP.

Second, acute anger may play a role in the developmentof PSP. Our patients with first-onset PSP were young, had alow BMI and showed high levels of anger, depression andperceived stress. Especially, scores on GARS 1 (work/job/study), which assess ‘recent stress’ (within the previousweek), were higher than those of other stressors (GARS 1 vs.GARS total: t=5.6 vs. t=2.9). Even though logistic regressionanalysis did not reveal an association between state angerand first-onset PSP, these findings support the above

assumption. We summarized the hypothesis regarding thepathophysiological factors of PSP in Fig. 1.

In the present study, stress responses such as tension,somatization, fatigue and frustration were higher in first-onset PSP patients than in the control group, whereas nodifferences in personality, depression and anxiety were foundin a previous study [19]. This discrepancy might be due todifferences in sample size, characteristics of control subjectsand scales between the two studies.

This study was subjected to the following limitations.First, the sample was small, and we only recruited patientswho visited hospitals located in Kyounggi province inKorea. Therefore, the results might represent a manifestationpeculiar to a local area in Korea, although unlikely. A largemulticenter trial would be able to confirm this. Second, weevaluated the characteristics of first-onset patients with PSPcross-sectionally without employing intensive interviewsfocusing on anger. In addition, ethical considerationsprevented us from experimentally provoking anger in thepatients. A prospective study is also needed to examinewhether anger-prone people will develop PSP. Moreover,since we evaluated patients with first-onset PSP after thedevelopment of PSP, it is also possible that the results merelyreflect a stress response associated with the development ofPSP. However, reports on patients with recent stress and highlevels of trait anger also support the association betweenanger and the development of PSP.

In conclusion, first-onset PSP patients show significantlyhigher levels of anger and perceived stress. Logisticregression revealed that low BMI and trait anger could beassociated with first-onset PSP. However, a prospectivestudy with a larger sample is needed to verify these results.

References

[1] Melton III LJ, Hepper NNG, Offord KP. Incidence of spontaneouspneumothorax in Olmsted County, Minnesota: 1950 to 1974. Am RevRespir Dis 1979;120:1379–82.

[2] Bense L, Eklund G, Wiman LG. Smoking and the increased risk ofcontracting spontaneous pneumothorax. Chest 1987;92:1009–12.

[3] Sahn SA, Heffner JE. Spontaneous pneumothorax. N Engl J Med2000;342:868–74.

336 S.-H. Lee et al. / General Hospital Psychiatry 30 (2008) 331–336

[4] Ayed AK, Bazerbashi S, Ben-Nakhi M, Ben-Nakhi M, ChandrasekranC, Sukumar M, et al. Risk factors of spontaneous pneumothorax inKuwait. Med Princ Pract 2006;15:338–42.

[5] Baumann MH. Do blebs cause primary spontaneous pneumothorax?Pro: blebs do cause primary spontaneous pneumothorax. J Bronchol2002;9:313–8.

[6] Radomsky J, Becker HP, Hartel W. Pleural porosity in idiopathicspontaneous pneumothorax. Pneumologie 1989;43:250–3.

[7] Ohata M, Suzuki H. Pathogenesis of spontaneous pneumothorax. Withspecial reference to the ultrastructure of emphysematous bullae. Chest1980;77:771–6.

[8] Fujino S, Inoue S, Tezuka N, Hanaoka J, Sawai S, Ichinose M, et al.Physical development of surgically treated patients with primaryspontaneous pneumothorax. Chest 1999;116:899–902.

[9] Bense L, Eklund G, Wiman LG. Smoking and increased risk ofcontracting spontaneous pneumothorax. Chest 1987;2:1009–12.

[10] Sadikot RT, Greene T, Meadows K, Arnold AG. Recurrence of primaryspontaneous pneumothorax. Thorax 1997;52:805–9.

[11] Spielberger CD. State–Trait Anger Expression Inventory. Odessa(Fla): Psychological Assessment Resources; 1991.

[12] Stewart JC, Janicki DL, Muldoon MF, Sutton-Tyrrell K, Kamarck TW.Negative emotions and 3-year progression of subclinical atherosclero-sis. Arch Gen Psychiatry 2007;64:225–33.

[13] Nicholson RA, Houle TT, Rhudy JL, Norton PJ. Psychological riskfactors in headache. Headache 2007;47:413–26.

[14] Manchikanti L, Pampati V, Beyer C, Damron K, Barnhill RC.Evaluation of psychological status in chronic low back pain:comparison with general population. Pain Physician 2002;5:149–55.

[15] Steptoe A, Brydon L. Psychosocial factors and coronary heart disease:the role of psychoneuroimmunological processes. In: Ader R, editor.Psychoneuroimmunology. San Diego: Elsevier; 2006.

[16] von Kanel R, Mills PJ, Fainman C, Dimsdale JE. Effects ofpsychological stress and psychiatric disorders on blood coagulationand fibrinolysis: a biobehavioral pathway to coronary artery disease?Psychosom Med 2001;63:531–44.

[17] Kubzansky LD, Davidson KW, Rozanski A. The clinical impactof negative psychological states: expanding the spectrum of risk

for coronary artery disease. Psychosom Med 2005;67(Suppl 1):S10–4.

[18] Bhattacharyya MR, Steptoe A. Emotional triggers of acute coronarysyndromes: strength of evidence, biological processes, and clinicalimplications. Prog Cardiovasc Dis 2007;49(5):353–65.

[19] Martin Martin M, Cuesta Serrahima L, Rami Porta R, SolerInsa P, Mateu Navarro M. Study of the personality of patientswith spontaneous pneumothorax. Arch Bronconeumol 2001;37:424–8.

[20] American Psychiatric Association. Diagnostic and statistical manual ofmental disorders. 4th ed. Washington (DC): American PsychiatricAssociation; 1994.

[21] Lee JH, Han DW, Chun KK. Korean adaptation of the State–TraitAnger Expression Inventory. Korean J Health Psychol 1997;2:60–78.

[22] Linn MW. A Global Assessment of Recent Stress (GARS) scale. Int JPsychiatry Med 1985;15:47–59.

[23] Folkman S, Lazarus RS, Gruen RJ, DeLongis A. Appraisal, coping,health status, and psychological symptoms. J Pers Soc Psychol 1986;50:571–9.

[24] Koh KB, Park JK, Kim CH, Cho S. Development of the StressResponse Inventory and its application in clinical practice. PsychosomMed 2001;63:668–78.

[25] Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventoryfor measuring depression. Arch Gen Psychiatry 1961;4:561–71.

[26] Lee MK, Lee YH, Park SH, Son CH, Chung YJ, Hong SK, et al.A standardization study of Beck Depression Inventory (I): Koreanversion (K-BDI): reliability and factor analysis. Korean J Psycho-pharmacol 1995;4:77–95.

[27] Cyr PV, Vincic L, Kay JM. Pulmonary vasculopathy in idiopathicspontaneous pneumothorax in young subjects. Arch Pathol Lab Med2000;124:717–20.

[28] Dunnill MS. Pulmonary pathology. Edinburgh, UK: ChurchillLivingstone; 1982. p. 462–3.

[29] Churg A. Diseases of the pleura. In: ThurlbeckWM, Churg AM, editors.Pathology of the lung. New York (NY): Thieme; 1995. p. 1079–80.

[30] Zalcman SS, Siegel A. The neurobiology of aggression and rage: roleof cytokines. Brain Behav Immun 2006;20(6):507–14.