Behavioral Medicine
ISSN: 0896-4289 (Print) 1940-4026 (Online) Journal homepage: http://www.tandfonline.com/loi/vbmd20
Burnout and Risk Factors for Cardiovascular Diseases Samuel Melamed PhD , Talma Kushnir PhD & Arie Shirom PhD To cite this article: Samuel Melamed PhD , Talma Kushnir PhD & Arie Shirom PhD (1992) Burnout and Risk Factors for Cardiovascular Diseases, Behavioral Medicine, 18:2, 53-60, DOI: 10.1080/08964289.1992.9935172 To link to this article: https://doi.org/10.1080/08964289.1992.9935172
Published online: 09 Jul 2010.
Submit your article to this journal
Article views: 152
View related articles
Citing articles: 179 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=vbmd20
Burnout and Risk Factors for Cardiovascular Diseases Samuel Melamed, PhD; Talma Kushnir, PhD; and Arie Shirom, PhD
The burnout syndrome denotes a constellation of physical fatigue, emotional exhaustion, and cognitive weariness resulting from chronic stress. Although it overlaps considerably with chronic fatigue as defined in internal medicine, its links with physical illness have not been systematically investigated. Thk exploratory study, conducted among 104 male workers free from cardiovascular disease (CVD), tested the association between burnout and two of its common concomitants-tension and lktlessnessand cardiovascular risk factors. After ruling out five possible confounders (age, relative weight, smoking, alcohol use, and sports activity), the authors found that scores on burnout plus tension (tense-burnout) were associated with somatic complaints, cholesterol, glucose, triglycerides, uric acid, and, marginally, with ECG abnormalities. Workers scoring high on tense-burnout also had a sign$cantly higher low densiry lipoprorein (LDL) level. Conversely, scores on burnout plus listlessness were sign$ icantly associated with glucose and negatively with diastolic blood pressure. The findings warrant further study of burnout as a predictor of cardiovascular morbidity and mortality. INDEX TERMS: burnout, chronic stress, CVD risk factors, fatigue, emotional exhaustion, tension, lktlessness
The “burnout syndrome,” a set of symptoms commonly associated with stress in a variety of occupations,’” has attracted increasing interest during the past decade. Although no single conceptualization of burnout is accepted as standard, Shirom: in his extensive review, identifies the unique content of burnout as the chronic depletion of an individual’s energetic resources. This does not overlap with any other related concepts, such as stress and affective dysfunction (eg, depression or anxiety), and is conceptually distinct from a temporary state of fatigue,
which passes after a resting period. Shirom has conceptually defined burnout as a constellation of physical fatigue, emotional exhaustion, and cognitive weariness. Burnout has consistently been found to be related to work performance, job satisfaction, withdrawal behavior, quality of life, and psychological well-being.%’ Moreover, burnout, as conceptualized in psychology, appears to overlap, to some extent, with the syndrome of chronic fatigue, as defined in internal Chronic fatigue syndrome has been shown to be the chief complaint accounting for 4% to 9% of all office visits to internists and family physicians in the United States and Canada.9 Nevertheless, to the best of our knowledge, no one has conducted a systematic investigation of the links between burnout and physical illnesses, including cardiovascular disease. The present study is a step toward filling this void.
Dr Melamed is head of the Behavioral Medicine Unit at the Occupational Health and Rehabilitation Institute, Raanana, Israel, where Dr Kushnir is a senior researcher. Dr Shirom is a professor at the Department of Labour Studies, Tel Aviv University, Ramat A viv, Israel.
53
BURNOUT AND RISK FACTORS FOR CVD
Possible Links Between Burnout and Risk Factors of Cardiovascular Disease A possible link between burnout and ill health, particularly CVD, has been suggested by the findings of a series of studies conducted by Appels and his associates.'"*" These studies were prompted by the clinical observation that more than half of those who suffer a myocardial infarction (MI), or who die suddenly because of one, had previously complained of two categories of symptoms: (1) discomfort in the chest, and (2) feelings of fatigue and general malaise. Appels termed the symptoms in the second category the vital exhaustion and depression (VED) syndrome and devised the Maastricht Questionnaire (MQ) to gauge it. In prospective these scientists found VED to be predictive of future MI, even when blood pressure (BP), smoking, cholesterol, age, and the use of antihypertensive drugs were controlled for. Furthermore, VED did not correlate with these risk factors.'" These findings led Appels and Mulder" to conclude that a state of exhaustion is associated with increased risk of future MI, independent of the classic risk factors. Although some conceptual similarities between VED and burnout can be found, in practice only a small number of items in the MQ are similar to those constituting burnout scales. In addition, the MQ measures such other symptoms as irritability, somatic complaints, withdrawal behavior, hypochondriac feelings, and loss of libido. Thus, it is not possible to infer from Appels's data whether or not burnout will predict cardiovascular morbidity. As the first step in a systematic investigation into the predictive value of burnout for CVD, we designed the present study to test the association between burnout and cardiovascular risk factors. Several researchers have attempted to conceptualize burnout as a process resulting from chronic exposure to stress. Cherni~s'~ and his associates constructed a developmental model that has received partial support from empirical evidence.' According to this model, the initial stage of burnout is that of physical fatigue and emotional exhaustion caused by actively coping with the perceived stress. During this stage, individuals report burnout symptoms coupled with anxiety and tension. These feelings result from unsuccessful attempts to cope with workrelated stress that is too demanding. With the progression of unsuccessful coping attempts, according to this hypothesis, defensive coping emerges,'3(pp17-'8' resulting in several attitudinal and behavioral changes, notably withdrawal, defensiveness, and emotional detachment. This developmental model led us to view burnout as essentially a two-phase process. In the early stages, individuals
54
may employ active and direct coping strategies to enhance and protect their resources, and they will thus report both burnout and tension. In the more advanced stages, however, when burnout appears to be coupled with depression, listlessness, and apathy, indirect and inactive coping behavior prevails. Study Hypotheses We hypothesized, first, that burnout would be positively and strongly associated with both tension and listlessness. As noted, past theory and research suggest that tension and listlessness may be linked to different stages of b ~ r n o u t Because .~ there are no theoretical grounds to support an association between tension and listlessness, we expected them to be either uncorrelated or only mildly correlated, with the relationship between them explained by their respective&nks to burnout. Second, we hypothesized that the combinations of burnout plus tension and of burnout plus listlessness would be related differentially to somatic complaints and CVD risk factors. Given the exploratory nature of this study and the absence of any prior data, we did not develop more specific hypotheses. METHOD Subjects The sample consisted of 114 full-time male employees of a telecommunications firm in Israel, representing 37% of the f m ' s total workforce. The study was part of a company-sponsored medical check-up plan offered on a voluntary basis to all employees as part of their fringe benefits package. Because female employees constituted a small minority of the sample, we included only male employees in the study. Exclusions Of the initial sample, we excluded 10 subjects: 4 had incomplete data, and 6 reported CVD morbidity on a medical questionnaire. The criteria for CV morbidity were medical history of previous MI; valvular disease; diagnosed hypertension; reported angina pectoris symptoms (according to Rose et all4); and/or regular medication usage, such as antihypertensive drugs and anticoagulents. Thus, the final sample consisted of 104 healthy individuals. Age ranged from 24 to 68 years (M = 39, SD = 9.3); job tenure ranged from 1 to 26 years (M = 6.8, SD = 6.5); 74010 had a graduate-level education; 56% were white-collar employees, mostly in managerial positions; and 44% were employed in blue-collar jobs.
Behavioral Medicine
MELAMED ET AL
Procedure The medical examination was a routine general health check-up, but physiological risk factors in CVD were especially emphasized; it took place on the company premises and on company time, and was performed in two stages. First, a nurse collected medical data, including tests and measurements, then an experienced interviewer administered a questionnaire to obtain background information. A physician performed the general health check-up and administered the psychological questionnaire, including burnout and somatic complaint scales, at the end of the examination as an integral part of the medical procedure.
Measures
Medical data Blood pressure. BP was measured after a 5-minute lying-down rest, with three consecutive measurements taken at 1-minute intervals. The first was taken while the subjects were lying down and the other two while they were sitting up. The average of the second and third measurements was used. Height and weight. Height was measured accurate to 0.5 cm; weight determined with 0.5 kg accuracy. Quetelet index (relative weight). The formula used to calculate this index was weight in kg/(height in m)’ = Quetelet. ECG. This was recorded using a standard 12-channel ECG recorder. A cardiologist analyzed the ECGs, and they were coded according to the Minnesota Code 1982. For the present study, we used only the classification of ECGs into normal and abnormal. Blood tests. We used vacutainers to take blood samples after an 8- to 10-hour fast. Within 2 hours, the serum was separated, refrigerated, and sent to the laboratory. Analysis, performed either on the same or the following day using the Abbott VP Autoanalyzer, tested levels of glucose, total cholesterol, triglycerides, and uric acid, all measured by enzymatic methods, and HDL cholesterol, which was measured after precipitation with magnesium phosphotungstate (sigma). Background information Data on demographic, life-style, job characteristics, and medical background were collected in a structured interview that used a computerized questionnaire. To exclude the possibility that the difference in the outcome variables among subjects high and low on burnout might stem from other factors, we tested such possible con-
Vol 18, Summer 1992
founders as age, relative weight, and habits (smoking, drinking, physical activity).
Psychological measures Three scales measuring burnout, tension, and listlessness were constructed, based on the results of factor analyses applied to several sets of data obtained from independent samples of teachers, white-collar employees, and military officer^.'^ (These results are available upon request from the third author; the complete list of items appears in the Appendix.) Burnout was assessed by 8 items (numbers 1 4 , 6, 8-10 on our scale) measuring physical fatigue and emotional exhaustion. Responses for each item were scored on a 7-point scale ranging from 1 (almost never) to 7 (almost always). The total score was averaged by dividing by the number of items; mean score on this scale was 1.29 (SD = 0.48); the reliability coefficient (Cronbach’s alpha) was .84. Tension was measured by 4 items (numbers 13-16). The response scale was the same as for burnout; mean score was 1.45 (SD = 1.11); Cronbach’s alpha, .90. Lktlessnas was measured by 4 items (numbers 5, 7, 11, 12). The mean score was 2.48 (SD = 0.69), and Cronbach’s alpha was .73. Somatic complaints were assessed using 12 items, each with four points, based on the index reported by Caplan,’6 which measures the frequency of symptoms experienced during the month prior to participation (eg, dizziness, shortness of breath, clammy hands, headaches). Total scores on this scale ranged from 1.08 to 1.92 with a mean of 1.20 (SD = 0.15); Cronbach’s alpha was .87. RESULTS Correlations Between Burnout and Its Concomitants The intercorrelations computed among the above three variables indicate that, as expected, both tension (.72) and listlessness (.62) correlated with burnout. Tension and listlessness were also intercorrelated (r = .38). However, after partialling out their relationship to burnout, the correlation between tension and listlessness was - .lo. This finding supports the first hypothesis as formulated.
Predictor Combinations and Some Possible Confounders The lack of association between tension and Listlessness, after partialling out burnout, enabled us to test the separate impact on CV risk factors of the combinations of burnout plus tension and burnout plus listlessness, as suggested by our second hypothesis. For this purpose, we created two new scales, tense-burnout and listless-burnout, by combining scores on the respective scales.
55
BURNOUT AND RISK FACTORS FOR CVD
Before examining the relationship between the predictor variables and risk factors, we examined the possible effects of some confounding variables, namely, age, relative weight, smoking, alcohol consumption, and physical activities. Each of the predictor combinations was divided into low, medium, and high levels on the basis of the score distribution. The association between these levels and the possible confounders was tested by applying either chi-square tests for nominally scaled data or one-way ANOVAs for continuous measures (see Table 1). The results shown in Table 1 indicate that, of the five confounders examined, only participation in sport activities was significantly related to tense-bumout. Subjects reporting medium and high levels of tense-bumout participated less than low-level individuals. None of the possible confounders was related to listless-burnout except, but only marginally, relative weight (with no consistent trend). Thus, in subsequent analyses, sports and relative weight were taken as confounders (as was age, because of its strong association with CVD risk factors). Predictor Combinations and CV Risk Factors We examined the relationship of tense-bumout and listless-bumout to CV risk factors through a series of multiple regressions controlling for age, relative weight, and sports (see Table 2). As the data in Table 2 indicate, tense-bumout was positively related to five outcome variables: serum cholesterol, glucose, uric acid, and somatic complaints; the re-
lationship for triglycerides was only marginally significant. Listless-burnout showed less predictive power; it was positively related to glucose levels and to somatic complaints but negatively to diastolic blood pressure. These results supported the first hypothesis-that bumout is related to CVD risk factors. The pattern of associations between tense-bumout and listless-burnoutwith the outcome variables was somewhat different. Whereas both predictor combinations were related positively to glucose and somatic complaints, tenseburnout was also positively related to serum cholesterol, uric acid, and triglyceride levels, and listless-burnout was negatively related to diastolic blood pressure. These differences partially support the second hypothesis regarding differential association to risk factors. To test the association between tense-bumout and listless-bumout and ECG abnormalities, we canied out two logistic regression analyses. Age, relative weight, and sport were taken as confounders. The results indicated a marginally significant association between tense-bumout and ECG abnormalities (p = .43, xz = 3.31, p < .069) in the expected direction. No significant association was revealed for listless-burnout (p = .049, x2 = 1 . 3 1 , ~< .25). The next step in this exploratory study was to examine the possibility of increased risk (in terms of the number of elevated risk factors) of CVD for persons scoring high on burnout. For this purpose, subjects who were high (upper tertile) on tense-bumout and high on listless-bumout were compared with all other subjects in terms of the
TABLE 1 Associations Between Burnout Levels and Possible Confounding Variables
Tense-bumout
Listless-bumout
Variable
Low
Medium
High
X2
Low
Medium
High
X2
Smoking (Yo) Sport (Yo nonparticipation) Alcohol consumption (Oro)
15.63 43.75 31.25
20.00 76.67 36.67
31.82 70.45 15.91
2.99 8.59+ 4.50
22.50 52.50 30.00
24.32 75.68 29.73
24.14 65.52 17.24
4.52 1.73
Mean age (years) Mean Quetelet index
40.3 25.9
39.2 26.7
38.6 25.2
0.32 1.29
39.7 25.6
40.4 26.9
37.4 24.7
0.91 2.93t
.@I
Note: Quetelet index = weight (kg)/height (m)’. *p < .01; t.05 < p < .lo. I
56
Behavioral Medicine
MELAMED ET AL
TABLE 2 Multiple Regression Analysis of the Effects of Burnout on Cardiovascular Risk Factors Tense-burnout Standardized regression coefficient Systolic BP Burnout Age Quetelet index sport Diastolic BP Burnout Age Quetelet index sport Cholesterol Burnout Age Quetelet index sport LDL Burnout Age Quetelet index sport HDL Burnout Age Quetelet index sport Triglycerides Burnout Age Quetelet index sport Glucose Burnout Age Quetelet index sport Uric acid Burnout Age Quetelet index sport Somatic complaints Burnout Age Quetelet index sport
- .04 .34 .18 - .02
R2 .I5
.65 .001
- .07 .33 .18 - .02
.05
.80
.05
.82
.m
.09
- .26
.005
.25 .22 - .ll
.01 .02 .21
.23 .22 - .10
.01 .02 .25
.33
.o006
.14
.14 .41
- .07
.11
.15 .03
.04 .32 .12
.63 .001
.20
- .09
.28
.01 .36 .04 - .14
.88 .0003 .67 .13
- .01
.86 .86 .02 .39
.13
.I5 .14 .37 .05 - .12
.12 .0002 .57 .18
.02
.02
.04 .02 - .22 .09
.67 .82 .03 .36
.01 - .23
.08
.10
.12 .18 .02 .37 .03
.06
.09
.83 .0002 .76
.01 .36 .01
.33 .88 .0004 .93 .19
.15 .32 .30 .08 .06
.37 .31 .08
.001 .002 .37 * 54
.03
.OOol .001
.38 .77
.04
.05
.19 - .08 .24 - .02
.05
.22 - .17
.02 .08 .37 .39
.43 .02 .81
.16 - .08 .23 - .04
.ll .43 .02
24 - .17
.01
.66
.08 - .09 - .08
R .16
.16
.m
Adjusted
.38 .001
- .16
Note: Quetelet index = weight (kg)/helght (m)’.
Vol18, Summer 1992
Adjusted
P
List less-burnou t Standardized regression coefficient P
.09 - .09
- .10
.09
.34 .27
BURNOUT AND RISK FACTORS FOR CVD
jects in the upper one third of the tense-bumout scores. Compared with the others, these subjects showed significantly higher levels of cholesterol, LDL, triglycerides, and uric acid. The finding that a number of risk factors turned out to be significantly elevated among high scorers on tense-bumout suggested an overall threshold effect and higher risk. Evidence from epidemiologic studies indicates that the presence of a number of CVD risk factors increases the probability of developing ischemic heart disease (IHD)." We observed a different pattern of associations with risk factors for listless-bumout. Only three of the outcome variables studied were significantly associated with listless-bumout. Moreover, these results became even less pronounced when we compared subjects in the upper thirtiles with the rest of the sample. In addition, listlessburnout was significantly, but negatively, associated with diastolic blood pressure in all analyses we performed. It is admittedly difficult to explain this latter finding. We might surmise that the factors mediating between bumout and biochemical risk factors are different from those mediating blood pressure. At any rate, similar puzzling results concerning SBP were reported by Appels and Mulder" and Appels" in a large sample of healthy individuals. Negative associations between stress and blood pressure have also been observed in other studies.'%*'Another noteworthy finding is that the study groups marginally differed on ECG at rest, in contrast to an observation by Appels.'* The different patterns of association of tense-bumout and listless-bumout with the outcome variables is consis-
outcome variables studied. One-way analyses of covariance (ANCOVAs), with age and relative weight as covariates, were used. Sport was not controlled for because it was not found to be a significant confounder in any of the multiple regression analyses presented in Table 2. For the results of the ANCOVA analyses, see Table 3. For tense-bumout, the ANCOVAs yielded more extensive results than did the regression analyses. Workers scoring high on tense-bumout showed significantly elevated levels of serum cholesterol, LDL, triglycerides, and uric acid. For listless-burnout, the results of the ANCOVA analyses yielded similar, but less pronounced, results when compared with those obtained in the regression analyses. Only somatic complaints and diastolic blood pressure turned out to be significant, with the latter again showing a negative relationship to listless-bumout. Higher scores on listless-bumout, thus, do not signify higher risk of CVD. Hence, these results further reinforce our previous findings that tense-bumout and listlessburnout combinations have different predictive patterns. DISCUSSION Our results supported the main hypothesis of this study, namely, that burnout is associated with cardiovascular risk factors (after controlling for the effects of age, relative weight, and sport participation). This association was found to be strongest for tense-bumout. Tense-bumout was related to reduced well-being (increased somatic complaints) as well as to elevated levels of cholesterol, uric acid, and glucose, and marginally higher levels of triglycerides. This trend was more pronounced among sub-
TABLE 3 Adjusted Mean Scores of Outcome Variables Among the Predictor Combinations and Results of ANCOVA Tests of Significance
Low Outcome variables Systolic blood pressure Diastolic blood pressure Cholesterol LDL HDL Triglycerides Glucose
58
(n
=
68)
124.7 80.3 1%.7 128.2 42.9 127.4 %.3 4.9 1.2
Tense-bumout High
(n
=
36)
124.0 78.2 224.3 145.2 44.2 174.9 102.7 5.4 1.2
Low P .748 .a2 .MI2
.030 .554 7@ I. .177 .045 .150
(n
=
68)
125.6 81.0 206.8 136.3 43.1 136.5 95.5 5.0 1.2
Listless-bumout High
( n = 36)
P
122.3 76.8 205.2 129.8 43.8 157.8 104.1 5.2 1.3
.127 .012 .865 .415 .753 244 .073 .521 .051
Behavioral Medicine
MELAMED ET AL
tent with our second hypothesis, but not with the developmental model suggested by Cherniss13and Shirom4 on which it was based. Tense-burnout was hypothesized to represent the early stages of burnout characterized by tension and anxiety resulting from unsuccessful attempts to cope with overly demanding work stress. On the other hand, listless-burnout was hypothesized to signify more advanced stages of burnout, characterized by withdrawal, emotional detachment, and defensive behavior. Our findings, however, indicate that tense-burnout was positively and strongly associated with CVD risk factors but that listless-burnout was not. At least two possible explanations may account for this discrepancy with Cherniss's model. First, the increase in biochemical risk factors may be a transient phase, and a more advanced stage of burnout might be characterized by reduction in these risk factors and emergence of others not measured in our study. Second, tension and listlessness may not represent developmental stages of burnout, as suggested by ChernissI3 and S h i r ~ m .Rather, ~ they might be considered separate concomitants of burnout, with each characterizing different response patterns to job stress. Further examination of the data indicated that subjects scoring high on tense-burnout were different from those scoring high on listless-burnout, with only a 24% overlap. Only longitudinal studies would enable these suggestions to be tested directly. The findings here suggest that tense-burnout has negative implications for the cardiovascular system, whereas listless-burnout does not. It is likely that the relationship of tense-burnout to cardiovascular risk factors is a result of the tension concomitant of burnout. Thaye? highlighted the negative impact of tension on well-being in his discussion of tense arousal. Tension was also found to be related to serum cholesterol,u albeit in a sample of coronary-prone individuals. Because of the high multicolinearity between the predictors, very few subjects scored high on tension only. Therefore, the data did not allow for the separate testing of the impact of tension alone on the outcome variables. The same constraint applied to listlessness. In the present study, we took steps to assure that the associations between predictor combinations and CVD risk factors did not stem from external confounding factors. Our study was conducted among CVD-free individuals, and five possible confounders of the bumout-risk factors association-age, relative weight, participation in sports activities, smoking, and alcohol consumptionwere considered. Only sport was found to be related negatively to burnout. In the multiple regression analyses, this variable was included as a confounder, as were age
Vol 18, Summer 1992
and relative weight. Thus, we may conclude that tenseburnout in our sample contributed to CVD risk factors independently of possible confounders. It is interesting that, in the present study, tense-burnout was associated with biochemical risk factors for CVD, whereas studies conducted in the Netherlands by Appels and his associates'0*''did not find such associations. One possible reason for these discrepancies might be the difference in the predictor variables examined. Although burnout here and vital exhaustion in Appels's studies share some common elements in terms of the components constituting these predictors (namely physical fatigue and emotional exhaustion), Appels's studies did not include the element of tension. Listless-burnout, which is conceptually close to VED," however, does not appear to be associated with CVD risk factors. This is congruent with the Appels findings. We are aware that this study has several shortcomings that should be addressed in future research. First, it was conducted among employees in a very demanding and competitive high-tech firm. The possibility that the relationship between burnout and CVD risk factors is weaker in less demanding occupations must be considered. Second, sample size was small and, therefore, the results ought to be cross validated in larger samples. Third, although mental exhaustion has been identified as one of the core elements of b ~ r n o u tit, ~was not included in our burnout measure. Further work is needed to overcome these shortcomings, as well as to elucidate possible mechanisms linking burnout and increased CVD risk factors. Measures of stress hormones (catecholamines and corticosteroids) might be included as possible mediators. In conclusion, this exploratory study showed a generalized measure of burnout (composed of emotional exhaustion and physical fatigue), combined with tension, was positively associated with both subjective measures of health complaints and objective measures of physiological CVD risk factors in a sample of healthy workers. This measure of tense-burnout was unrelated to such other CV risk factors as systolic and diastolic blood pressure, smoking, and relative weight. Prospective studies now seem warranted to examine whether either tense-bumout or listless-burnout, especially for those with high scores, is predictive of CVD morbidity and mortality, thus playing a similar role to Appels's VED. Appendix Items in the burnout (BO), tension (TE), and listlessness (LIS) scales; those marked with an asterisk were reverse scored. 1. I feel tired (BO) *2. I feel refreshed. (BO)
59
BURNOUT AND RISK FACTORS FOR CVD
3. I feel physically exhausted. (BO) 4. 1 feel “fed-up.’’ (BO) * 5 . I feel full of vitality. (LIS)
6. My “batteries” are “dead.” (BO) 1 feel alert. (LIS) I feel burned out. (BO) I feel mentally fatigued. (BO) I feel no energy for going to work in the morning. (BO) I feel active. (LIS) I feel sleepy. (LIS) 13. I am tense. (TE) *14. I feel relaxed. (TE) 15. I feel restless. (TE) 16. I feel intense inner tension. (TE) *7. 8. 9. 10. 11. 12.
NOTE This study was supported by a grant from the Committee for Prevention and Research in Occupational Health, Ministry of Work and Social Welfare, Jerusalem, Israel. For further information, write Samuel Melamed, PhD, Occupational Health and Rehabilitation Institute at Loewenstein Hospital, PO Box 3, Raanana 43 100, Israel.
REFERENCES I . Farber BA. Introduction: A critical perspective on burnout. In: Farber B, ed. Stress and Burnout in Human Service Professions. New York: Pergamon; 1983. 2. MacNeille D. The relationship of occupational stress and burnout. In: Jones JW, ed. The Burnout Syndrome. Park Ridge: London House; 1983. 3. Paine WS. Job Stress and Burnout. Beverley Hills: Sage; 1982. 4. Shirom A. Burnout in work organization. In: Cooper CL, Robertson 1. eds. International Review of Industrial and Organizational Psychologv. New York: Wiley; 1989. 5 . Burke RJ, Greenglass ER. Psychological burnout among men and women in teaching: An examination of the Cherniss model. Human Rel. 1989;42:261-273. 6. Jackson SE, Maslach C. After-effects of job-related stress: Families as victims. J Occup Behav. 1982;3:63-77. 7. Jackson SE, Schwab RL, Schuler RS. Toward an understanding of the burnout phenomenon. J Appl Psychol. 1986;71:630-640. 8. Holmes GA, Kaplan JE, Gantz NM, et al. Chronic fatigue syndrome: A working case definition. Ann Intern Med. 1988;108: 387-389. 9. Mann, P, Lane TJ, Mathews DA. Somatization disorder in patients with chronic fatigue. Psychosom. 1989;388-395. 10. Appels A, Mulder MA. Imminent myocardial infarction: A psychological study. J Human Stress. Fall 1984;10:129-134. 1 I. Appels A, Mulder MA. Excess fatigue as a precursor of myocardial infarction. European Heart J. 1988;9:758-764. 12. Appels A. Tiredness before myocardial infarction. In: Lacky JH, Stungeon DA, eds. Proceedings of the Fvteenth European Conference on Psychosomatic Research. London: John Libbey; 1986. 13. Cherniss C. Stsff Burnout: Job Stress in the Human Services. Beverley Hills, CA: Sage; 1980. 14. Rose GA, Blackbum H, Gillum RF, Prineas RJ. Cardiovascular Survey Metho&. Geneva: WHO; 1982. 15. Ezrahi N. Burnout in military officers ranks: A construct validation. Tel Aviv, Israel; Tel Aviv University, 1985. Dissertation.
60
16. Caplan RD, Cobb S, French JRP Jr, Van Harrison R, Pinneau SR Jr. Job Demands and Worker Health: Main Effects and Occupational Differences. Washington, DC: Report to National Institute for Occupational Safety and Health, HEW Publication 75-16 (NIOSH); 1975. 17. Fraser GE. Preventive Cardiology. New York: Oxford University Press; 1986. 18. Appels A. The year before myocardial infarction. In: Dembroski TM, Schmidt T, Blumchen G, eds. Biobehavioral Bases of Coronary Heart Disease. New York: Karger; 1983:18-37. 19. Linden W, Pardhus DL, Dobson KS. Effects of response styles on the report of psychological and somatic distress. J Consult CIin Psychol. 1986;54:309-313. 20. Pieper C, LaCroix AZ, Karasek RA. The relation of psychosocial dimensions of work with coronary heart disease risk factors: A meta-analysis of five United States data bases. A m J Epidemiol. 1989;129:483-494. 21. Winkleby MA, Ragland DR, Syme SL. Self-reported stressors and hypertension: Evidence of an inverse relationship. A m J Epidemiol. 1988;127:124-134. 22. Thayer RE. The Biopsychology of Mood and Arousal. New York: Oxford University Press; 1989. 23. Trevisan M, Tsong Y, Stamler J, et al. Nervous tension and serum cholesterol: Findings from the Chicago Coronary Prevention Evaluation Program. J Human Stress. 1983;9:12-16. 24. Appels A. Vital exhaustion and depression as precursors of myocardial infarction. In: Spielberger IG, Defares PB, eds. Stress and Anxiety, vol 2. Washington: Hemisphere; 1988.
The UMI Article Clearinghouse offers articles from more than 11,OOO copyright-cleared periodicals in a wide range of subjects. You can place your orders electronically,as well as by phone, mail, and telefacsimile. For more information, please complete and mail this coupon to UMI Article Clearinghouse, 300 North Zeeb Road, Box 11. Ann Arbor, MI 48106 USA. Or call toll-free for an immediate response: 800-52 1-0600. From Alaska and Michigan call collect 3 1 3 - 7 6 1 - 4 7 0 0 . From Canada, call toll-Free 800-343-5299.
YES! I’d like to know more about UMI Article Clearinghouse Name ~
Company/lnstitution
Telephone-(
1
U M-I
A Bell & Howell Company 300 North Zeeb Road Ann Arbor, MI 48106 USA
Behavioral Medicine
ISSN: 0896-4289 (Print) 1940-4026 (Online) Journal homepage: http://www.tandfonline.com/loi/vbmd20
Burnout and Risk Factors for Cardiovascular Diseases Samuel Melamed PhD , Talma Kushnir PhD & Arie Shirom PhD To cite this article: Samuel Melamed PhD , Talma Kushnir PhD & Arie Shirom PhD (1992) Burnout and Risk Factors for Cardiovascular Diseases, Behavioral Medicine, 18:2, 53-60, DOI: 10.1080/08964289.1992.9935172 To link to this article: https://doi.org/10.1080/08964289.1992.9935172
Published online: 09 Jul 2010.
Submit your article to this journal
Article views: 152
View related articles
Citing articles: 179 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=vbmd20
Burnout and Risk Factors for Cardiovascular Diseases Samuel Melamed, PhD; Talma Kushnir, PhD; and Arie Shirom, PhD
The burnout syndrome denotes a constellation of physical fatigue, emotional exhaustion, and cognitive weariness resulting from chronic stress. Although it overlaps considerably with chronic fatigue as defined in internal medicine, its links with physical illness have not been systematically investigated. Thk exploratory study, conducted among 104 male workers free from cardiovascular disease (CVD), tested the association between burnout and two of its common concomitants-tension and lktlessnessand cardiovascular risk factors. After ruling out five possible confounders (age, relative weight, smoking, alcohol use, and sports activity), the authors found that scores on burnout plus tension (tense-burnout) were associated with somatic complaints, cholesterol, glucose, triglycerides, uric acid, and, marginally, with ECG abnormalities. Workers scoring high on tense-burnout also had a sign$cantly higher low densiry lipoprorein (LDL) level. Conversely, scores on burnout plus listlessness were sign$ icantly associated with glucose and negatively with diastolic blood pressure. The findings warrant further study of burnout as a predictor of cardiovascular morbidity and mortality. INDEX TERMS: burnout, chronic stress, CVD risk factors, fatigue, emotional exhaustion, tension, lktlessness
The “burnout syndrome,” a set of symptoms commonly associated with stress in a variety of occupations,’” has attracted increasing interest during the past decade. Although no single conceptualization of burnout is accepted as standard, Shirom: in his extensive review, identifies the unique content of burnout as the chronic depletion of an individual’s energetic resources. This does not overlap with any other related concepts, such as stress and affective dysfunction (eg, depression or anxiety), and is conceptually distinct from a temporary state of fatigue,
which passes after a resting period. Shirom has conceptually defined burnout as a constellation of physical fatigue, emotional exhaustion, and cognitive weariness. Burnout has consistently been found to be related to work performance, job satisfaction, withdrawal behavior, quality of life, and psychological well-being.%’ Moreover, burnout, as conceptualized in psychology, appears to overlap, to some extent, with the syndrome of chronic fatigue, as defined in internal Chronic fatigue syndrome has been shown to be the chief complaint accounting for 4% to 9% of all office visits to internists and family physicians in the United States and Canada.9 Nevertheless, to the best of our knowledge, no one has conducted a systematic investigation of the links between burnout and physical illnesses, including cardiovascular disease. The present study is a step toward filling this void.
Dr Melamed is head of the Behavioral Medicine Unit at the Occupational Health and Rehabilitation Institute, Raanana, Israel, where Dr Kushnir is a senior researcher. Dr Shirom is a professor at the Department of Labour Studies, Tel Aviv University, Ramat A viv, Israel.
53
BURNOUT AND RISK FACTORS FOR CVD
Possible Links Between Burnout and Risk Factors of Cardiovascular Disease A possible link between burnout and ill health, particularly CVD, has been suggested by the findings of a series of studies conducted by Appels and his associates.'"*" These studies were prompted by the clinical observation that more than half of those who suffer a myocardial infarction (MI), or who die suddenly because of one, had previously complained of two categories of symptoms: (1) discomfort in the chest, and (2) feelings of fatigue and general malaise. Appels termed the symptoms in the second category the vital exhaustion and depression (VED) syndrome and devised the Maastricht Questionnaire (MQ) to gauge it. In prospective these scientists found VED to be predictive of future MI, even when blood pressure (BP), smoking, cholesterol, age, and the use of antihypertensive drugs were controlled for. Furthermore, VED did not correlate with these risk factors.'" These findings led Appels and Mulder" to conclude that a state of exhaustion is associated with increased risk of future MI, independent of the classic risk factors. Although some conceptual similarities between VED and burnout can be found, in practice only a small number of items in the MQ are similar to those constituting burnout scales. In addition, the MQ measures such other symptoms as irritability, somatic complaints, withdrawal behavior, hypochondriac feelings, and loss of libido. Thus, it is not possible to infer from Appels's data whether or not burnout will predict cardiovascular morbidity. As the first step in a systematic investigation into the predictive value of burnout for CVD, we designed the present study to test the association between burnout and cardiovascular risk factors. Several researchers have attempted to conceptualize burnout as a process resulting from chronic exposure to stress. Cherni~s'~ and his associates constructed a developmental model that has received partial support from empirical evidence.' According to this model, the initial stage of burnout is that of physical fatigue and emotional exhaustion caused by actively coping with the perceived stress. During this stage, individuals report burnout symptoms coupled with anxiety and tension. These feelings result from unsuccessful attempts to cope with workrelated stress that is too demanding. With the progression of unsuccessful coping attempts, according to this hypothesis, defensive coping emerges,'3(pp17-'8' resulting in several attitudinal and behavioral changes, notably withdrawal, defensiveness, and emotional detachment. This developmental model led us to view burnout as essentially a two-phase process. In the early stages, individuals
54
may employ active and direct coping strategies to enhance and protect their resources, and they will thus report both burnout and tension. In the more advanced stages, however, when burnout appears to be coupled with depression, listlessness, and apathy, indirect and inactive coping behavior prevails. Study Hypotheses We hypothesized, first, that burnout would be positively and strongly associated with both tension and listlessness. As noted, past theory and research suggest that tension and listlessness may be linked to different stages of b ~ r n o u t Because .~ there are no theoretical grounds to support an association between tension and listlessness, we expected them to be either uncorrelated or only mildly correlated, with the relationship between them explained by their respective&nks to burnout. Second, we hypothesized that the combinations of burnout plus tension and of burnout plus listlessness would be related differentially to somatic complaints and CVD risk factors. Given the exploratory nature of this study and the absence of any prior data, we did not develop more specific hypotheses. METHOD Subjects The sample consisted of 114 full-time male employees of a telecommunications firm in Israel, representing 37% of the f m ' s total workforce. The study was part of a company-sponsored medical check-up plan offered on a voluntary basis to all employees as part of their fringe benefits package. Because female employees constituted a small minority of the sample, we included only male employees in the study. Exclusions Of the initial sample, we excluded 10 subjects: 4 had incomplete data, and 6 reported CVD morbidity on a medical questionnaire. The criteria for CV morbidity were medical history of previous MI; valvular disease; diagnosed hypertension; reported angina pectoris symptoms (according to Rose et all4); and/or regular medication usage, such as antihypertensive drugs and anticoagulents. Thus, the final sample consisted of 104 healthy individuals. Age ranged from 24 to 68 years (M = 39, SD = 9.3); job tenure ranged from 1 to 26 years (M = 6.8, SD = 6.5); 74010 had a graduate-level education; 56% were white-collar employees, mostly in managerial positions; and 44% were employed in blue-collar jobs.
Behavioral Medicine
MELAMED ET AL
Procedure The medical examination was a routine general health check-up, but physiological risk factors in CVD were especially emphasized; it took place on the company premises and on company time, and was performed in two stages. First, a nurse collected medical data, including tests and measurements, then an experienced interviewer administered a questionnaire to obtain background information. A physician performed the general health check-up and administered the psychological questionnaire, including burnout and somatic complaint scales, at the end of the examination as an integral part of the medical procedure.
Measures
Medical data Blood pressure. BP was measured after a 5-minute lying-down rest, with three consecutive measurements taken at 1-minute intervals. The first was taken while the subjects were lying down and the other two while they were sitting up. The average of the second and third measurements was used. Height and weight. Height was measured accurate to 0.5 cm; weight determined with 0.5 kg accuracy. Quetelet index (relative weight). The formula used to calculate this index was weight in kg/(height in m)’ = Quetelet. ECG. This was recorded using a standard 12-channel ECG recorder. A cardiologist analyzed the ECGs, and they were coded according to the Minnesota Code 1982. For the present study, we used only the classification of ECGs into normal and abnormal. Blood tests. We used vacutainers to take blood samples after an 8- to 10-hour fast. Within 2 hours, the serum was separated, refrigerated, and sent to the laboratory. Analysis, performed either on the same or the following day using the Abbott VP Autoanalyzer, tested levels of glucose, total cholesterol, triglycerides, and uric acid, all measured by enzymatic methods, and HDL cholesterol, which was measured after precipitation with magnesium phosphotungstate (sigma). Background information Data on demographic, life-style, job characteristics, and medical background were collected in a structured interview that used a computerized questionnaire. To exclude the possibility that the difference in the outcome variables among subjects high and low on burnout might stem from other factors, we tested such possible con-
Vol 18, Summer 1992
founders as age, relative weight, and habits (smoking, drinking, physical activity).
Psychological measures Three scales measuring burnout, tension, and listlessness were constructed, based on the results of factor analyses applied to several sets of data obtained from independent samples of teachers, white-collar employees, and military officer^.'^ (These results are available upon request from the third author; the complete list of items appears in the Appendix.) Burnout was assessed by 8 items (numbers 1 4 , 6, 8-10 on our scale) measuring physical fatigue and emotional exhaustion. Responses for each item were scored on a 7-point scale ranging from 1 (almost never) to 7 (almost always). The total score was averaged by dividing by the number of items; mean score on this scale was 1.29 (SD = 0.48); the reliability coefficient (Cronbach’s alpha) was .84. Tension was measured by 4 items (numbers 13-16). The response scale was the same as for burnout; mean score was 1.45 (SD = 1.11); Cronbach’s alpha, .90. Lktlessnas was measured by 4 items (numbers 5, 7, 11, 12). The mean score was 2.48 (SD = 0.69), and Cronbach’s alpha was .73. Somatic complaints were assessed using 12 items, each with four points, based on the index reported by Caplan,’6 which measures the frequency of symptoms experienced during the month prior to participation (eg, dizziness, shortness of breath, clammy hands, headaches). Total scores on this scale ranged from 1.08 to 1.92 with a mean of 1.20 (SD = 0.15); Cronbach’s alpha was .87. RESULTS Correlations Between Burnout and Its Concomitants The intercorrelations computed among the above three variables indicate that, as expected, both tension (.72) and listlessness (.62) correlated with burnout. Tension and listlessness were also intercorrelated (r = .38). However, after partialling out their relationship to burnout, the correlation between tension and listlessness was - .lo. This finding supports the first hypothesis as formulated.
Predictor Combinations and Some Possible Confounders The lack of association between tension and Listlessness, after partialling out burnout, enabled us to test the separate impact on CV risk factors of the combinations of burnout plus tension and burnout plus listlessness, as suggested by our second hypothesis. For this purpose, we created two new scales, tense-burnout and listless-burnout, by combining scores on the respective scales.
55
BURNOUT AND RISK FACTORS FOR CVD
Before examining the relationship between the predictor variables and risk factors, we examined the possible effects of some confounding variables, namely, age, relative weight, smoking, alcohol consumption, and physical activities. Each of the predictor combinations was divided into low, medium, and high levels on the basis of the score distribution. The association between these levels and the possible confounders was tested by applying either chi-square tests for nominally scaled data or one-way ANOVAs for continuous measures (see Table 1). The results shown in Table 1 indicate that, of the five confounders examined, only participation in sport activities was significantly related to tense-bumout. Subjects reporting medium and high levels of tense-bumout participated less than low-level individuals. None of the possible confounders was related to listless-burnout except, but only marginally, relative weight (with no consistent trend). Thus, in subsequent analyses, sports and relative weight were taken as confounders (as was age, because of its strong association with CVD risk factors). Predictor Combinations and CV Risk Factors We examined the relationship of tense-bumout and listless-bumout to CV risk factors through a series of multiple regressions controlling for age, relative weight, and sports (see Table 2). As the data in Table 2 indicate, tense-bumout was positively related to five outcome variables: serum cholesterol, glucose, uric acid, and somatic complaints; the re-
lationship for triglycerides was only marginally significant. Listless-burnout showed less predictive power; it was positively related to glucose levels and to somatic complaints but negatively to diastolic blood pressure. These results supported the first hypothesis-that bumout is related to CVD risk factors. The pattern of associations between tense-bumout and listless-burnoutwith the outcome variables was somewhat different. Whereas both predictor combinations were related positively to glucose and somatic complaints, tenseburnout was also positively related to serum cholesterol, uric acid, and triglyceride levels, and listless-burnout was negatively related to diastolic blood pressure. These differences partially support the second hypothesis regarding differential association to risk factors. To test the association between tense-bumout and listless-bumout and ECG abnormalities, we canied out two logistic regression analyses. Age, relative weight, and sport were taken as confounders. The results indicated a marginally significant association between tense-bumout and ECG abnormalities (p = .43, xz = 3.31, p < .069) in the expected direction. No significant association was revealed for listless-burnout (p = .049, x2 = 1 . 3 1 , ~< .25). The next step in this exploratory study was to examine the possibility of increased risk (in terms of the number of elevated risk factors) of CVD for persons scoring high on burnout. For this purpose, subjects who were high (upper tertile) on tense-bumout and high on listless-bumout were compared with all other subjects in terms of the
TABLE 1 Associations Between Burnout Levels and Possible Confounding Variables
Tense-bumout
Listless-bumout
Variable
Low
Medium
High
X2
Low
Medium
High
X2
Smoking (Yo) Sport (Yo nonparticipation) Alcohol consumption (Oro)
15.63 43.75 31.25
20.00 76.67 36.67
31.82 70.45 15.91
2.99 8.59+ 4.50
22.50 52.50 30.00
24.32 75.68 29.73
24.14 65.52 17.24
4.52 1.73
Mean age (years) Mean Quetelet index
40.3 25.9
39.2 26.7
38.6 25.2
0.32 1.29
39.7 25.6
40.4 26.9
37.4 24.7
0.91 2.93t
.@I
Note: Quetelet index = weight (kg)/height (m)’. *p < .01; t.05 < p < .lo. I
56
Behavioral Medicine
MELAMED ET AL
TABLE 2 Multiple Regression Analysis of the Effects of Burnout on Cardiovascular Risk Factors Tense-burnout Standardized regression coefficient Systolic BP Burnout Age Quetelet index sport Diastolic BP Burnout Age Quetelet index sport Cholesterol Burnout Age Quetelet index sport LDL Burnout Age Quetelet index sport HDL Burnout Age Quetelet index sport Triglycerides Burnout Age Quetelet index sport Glucose Burnout Age Quetelet index sport Uric acid Burnout Age Quetelet index sport Somatic complaints Burnout Age Quetelet index sport
- .04 .34 .18 - .02
R2 .I5
.65 .001
- .07 .33 .18 - .02
.05
.80
.05
.82
.m
.09
- .26
.005
.25 .22 - .ll
.01 .02 .21
.23 .22 - .10
.01 .02 .25
.33
.o006
.14
.14 .41
- .07
.11
.15 .03
.04 .32 .12
.63 .001
.20
- .09
.28
.01 .36 .04 - .14
.88 .0003 .67 .13
- .01
.86 .86 .02 .39
.13
.I5 .14 .37 .05 - .12
.12 .0002 .57 .18
.02
.02
.04 .02 - .22 .09
.67 .82 .03 .36
.01 - .23
.08
.10
.12 .18 .02 .37 .03
.06
.09
.83 .0002 .76
.01 .36 .01
.33 .88 .0004 .93 .19
.15 .32 .30 .08 .06
.37 .31 .08
.001 .002 .37 * 54
.03
.OOol .001
.38 .77
.04
.05
.19 - .08 .24 - .02
.05
.22 - .17
.02 .08 .37 .39
.43 .02 .81
.16 - .08 .23 - .04
.ll .43 .02
24 - .17
.01
.66
.08 - .09 - .08
R .16
.16
.m
Adjusted
.38 .001
- .16
Note: Quetelet index = weight (kg)/helght (m)’.
Vol18, Summer 1992
Adjusted
P
List less-burnou t Standardized regression coefficient P
.09 - .09
- .10
.09
.34 .27
BURNOUT AND RISK FACTORS FOR CVD
jects in the upper one third of the tense-bumout scores. Compared with the others, these subjects showed significantly higher levels of cholesterol, LDL, triglycerides, and uric acid. The finding that a number of risk factors turned out to be significantly elevated among high scorers on tense-bumout suggested an overall threshold effect and higher risk. Evidence from epidemiologic studies indicates that the presence of a number of CVD risk factors increases the probability of developing ischemic heart disease (IHD)." We observed a different pattern of associations with risk factors for listless-bumout. Only three of the outcome variables studied were significantly associated with listless-bumout. Moreover, these results became even less pronounced when we compared subjects in the upper thirtiles with the rest of the sample. In addition, listlessburnout was significantly, but negatively, associated with diastolic blood pressure in all analyses we performed. It is admittedly difficult to explain this latter finding. We might surmise that the factors mediating between bumout and biochemical risk factors are different from those mediating blood pressure. At any rate, similar puzzling results concerning SBP were reported by Appels and Mulder" and Appels" in a large sample of healthy individuals. Negative associations between stress and blood pressure have also been observed in other studies.'%*'Another noteworthy finding is that the study groups marginally differed on ECG at rest, in contrast to an observation by Appels.'* The different patterns of association of tense-bumout and listless-bumout with the outcome variables is consis-
outcome variables studied. One-way analyses of covariance (ANCOVAs), with age and relative weight as covariates, were used. Sport was not controlled for because it was not found to be a significant confounder in any of the multiple regression analyses presented in Table 2. For the results of the ANCOVA analyses, see Table 3. For tense-bumout, the ANCOVAs yielded more extensive results than did the regression analyses. Workers scoring high on tense-bumout showed significantly elevated levels of serum cholesterol, LDL, triglycerides, and uric acid. For listless-burnout, the results of the ANCOVA analyses yielded similar, but less pronounced, results when compared with those obtained in the regression analyses. Only somatic complaints and diastolic blood pressure turned out to be significant, with the latter again showing a negative relationship to listless-bumout. Higher scores on listless-bumout, thus, do not signify higher risk of CVD. Hence, these results further reinforce our previous findings that tense-bumout and listlessburnout combinations have different predictive patterns. DISCUSSION Our results supported the main hypothesis of this study, namely, that burnout is associated with cardiovascular risk factors (after controlling for the effects of age, relative weight, and sport participation). This association was found to be strongest for tense-bumout. Tense-bumout was related to reduced well-being (increased somatic complaints) as well as to elevated levels of cholesterol, uric acid, and glucose, and marginally higher levels of triglycerides. This trend was more pronounced among sub-
TABLE 3 Adjusted Mean Scores of Outcome Variables Among the Predictor Combinations and Results of ANCOVA Tests of Significance
Low Outcome variables Systolic blood pressure Diastolic blood pressure Cholesterol LDL HDL Triglycerides Glucose
58
(n
=
68)
124.7 80.3 1%.7 128.2 42.9 127.4 %.3 4.9 1.2
Tense-bumout High
(n
=
36)
124.0 78.2 224.3 145.2 44.2 174.9 102.7 5.4 1.2
Low P .748 .a2 .MI2
.030 .554 7@ I. .177 .045 .150
(n
=
68)
125.6 81.0 206.8 136.3 43.1 136.5 95.5 5.0 1.2
Listless-bumout High
( n = 36)
P
122.3 76.8 205.2 129.8 43.8 157.8 104.1 5.2 1.3
.127 .012 .865 .415 .753 244 .073 .521 .051
Behavioral Medicine
MELAMED ET AL
tent with our second hypothesis, but not with the developmental model suggested by Cherniss13and Shirom4 on which it was based. Tense-burnout was hypothesized to represent the early stages of burnout characterized by tension and anxiety resulting from unsuccessful attempts to cope with overly demanding work stress. On the other hand, listless-burnout was hypothesized to signify more advanced stages of burnout, characterized by withdrawal, emotional detachment, and defensive behavior. Our findings, however, indicate that tense-burnout was positively and strongly associated with CVD risk factors but that listless-burnout was not. At least two possible explanations may account for this discrepancy with Cherniss's model. First, the increase in biochemical risk factors may be a transient phase, and a more advanced stage of burnout might be characterized by reduction in these risk factors and emergence of others not measured in our study. Second, tension and listlessness may not represent developmental stages of burnout, as suggested by ChernissI3 and S h i r ~ m .Rather, ~ they might be considered separate concomitants of burnout, with each characterizing different response patterns to job stress. Further examination of the data indicated that subjects scoring high on tense-burnout were different from those scoring high on listless-burnout, with only a 24% overlap. Only longitudinal studies would enable these suggestions to be tested directly. The findings here suggest that tense-burnout has negative implications for the cardiovascular system, whereas listless-burnout does not. It is likely that the relationship of tense-burnout to cardiovascular risk factors is a result of the tension concomitant of burnout. Thaye? highlighted the negative impact of tension on well-being in his discussion of tense arousal. Tension was also found to be related to serum cholesterol,u albeit in a sample of coronary-prone individuals. Because of the high multicolinearity between the predictors, very few subjects scored high on tension only. Therefore, the data did not allow for the separate testing of the impact of tension alone on the outcome variables. The same constraint applied to listlessness. In the present study, we took steps to assure that the associations between predictor combinations and CVD risk factors did not stem from external confounding factors. Our study was conducted among CVD-free individuals, and five possible confounders of the bumout-risk factors association-age, relative weight, participation in sports activities, smoking, and alcohol consumptionwere considered. Only sport was found to be related negatively to burnout. In the multiple regression analyses, this variable was included as a confounder, as were age
Vol 18, Summer 1992
and relative weight. Thus, we may conclude that tenseburnout in our sample contributed to CVD risk factors independently of possible confounders. It is interesting that, in the present study, tense-burnout was associated with biochemical risk factors for CVD, whereas studies conducted in the Netherlands by Appels and his associates'0*''did not find such associations. One possible reason for these discrepancies might be the difference in the predictor variables examined. Although burnout here and vital exhaustion in Appels's studies share some common elements in terms of the components constituting these predictors (namely physical fatigue and emotional exhaustion), Appels's studies did not include the element of tension. Listless-burnout, which is conceptually close to VED," however, does not appear to be associated with CVD risk factors. This is congruent with the Appels findings. We are aware that this study has several shortcomings that should be addressed in future research. First, it was conducted among employees in a very demanding and competitive high-tech firm. The possibility that the relationship between burnout and CVD risk factors is weaker in less demanding occupations must be considered. Second, sample size was small and, therefore, the results ought to be cross validated in larger samples. Third, although mental exhaustion has been identified as one of the core elements of b ~ r n o u tit, ~was not included in our burnout measure. Further work is needed to overcome these shortcomings, as well as to elucidate possible mechanisms linking burnout and increased CVD risk factors. Measures of stress hormones (catecholamines and corticosteroids) might be included as possible mediators. In conclusion, this exploratory study showed a generalized measure of burnout (composed of emotional exhaustion and physical fatigue), combined with tension, was positively associated with both subjective measures of health complaints and objective measures of physiological CVD risk factors in a sample of healthy workers. This measure of tense-burnout was unrelated to such other CV risk factors as systolic and diastolic blood pressure, smoking, and relative weight. Prospective studies now seem warranted to examine whether either tense-bumout or listless-burnout, especially for those with high scores, is predictive of CVD morbidity and mortality, thus playing a similar role to Appels's VED. Appendix Items in the burnout (BO), tension (TE), and listlessness (LIS) scales; those marked with an asterisk were reverse scored. 1. I feel tired (BO) *2. I feel refreshed. (BO)
59
BURNOUT AND RISK FACTORS FOR CVD
3. I feel physically exhausted. (BO) 4. 1 feel “fed-up.’’ (BO) * 5 . I feel full of vitality. (LIS)
6. My “batteries” are “dead.” (BO) 1 feel alert. (LIS) I feel burned out. (BO) I feel mentally fatigued. (BO) I feel no energy for going to work in the morning. (BO) I feel active. (LIS) I feel sleepy. (LIS) 13. I am tense. (TE) *14. I feel relaxed. (TE) 15. I feel restless. (TE) 16. I feel intense inner tension. (TE) *7. 8. 9. 10. 11. 12.
NOTE This study was supported by a grant from the Committee for Prevention and Research in Occupational Health, Ministry of Work and Social Welfare, Jerusalem, Israel. For further information, write Samuel Melamed, PhD, Occupational Health and Rehabilitation Institute at Loewenstein Hospital, PO Box 3, Raanana 43 100, Israel.
REFERENCES I . Farber BA. Introduction: A critical perspective on burnout. In: Farber B, ed. Stress and Burnout in Human Service Professions. New York: Pergamon; 1983. 2. MacNeille D. The relationship of occupational stress and burnout. In: Jones JW, ed. The Burnout Syndrome. Park Ridge: London House; 1983. 3. Paine WS. Job Stress and Burnout. Beverley Hills: Sage; 1982. 4. Shirom A. Burnout in work organization. In: Cooper CL, Robertson 1. eds. International Review of Industrial and Organizational Psychologv. New York: Wiley; 1989. 5 . Burke RJ, Greenglass ER. Psychological burnout among men and women in teaching: An examination of the Cherniss model. Human Rel. 1989;42:261-273. 6. Jackson SE, Maslach C. After-effects of job-related stress: Families as victims. J Occup Behav. 1982;3:63-77. 7. Jackson SE, Schwab RL, Schuler RS. Toward an understanding of the burnout phenomenon. J Appl Psychol. 1986;71:630-640. 8. Holmes GA, Kaplan JE, Gantz NM, et al. Chronic fatigue syndrome: A working case definition. Ann Intern Med. 1988;108: 387-389. 9. Mann, P, Lane TJ, Mathews DA. Somatization disorder in patients with chronic fatigue. Psychosom. 1989;388-395. 10. Appels A, Mulder MA. Imminent myocardial infarction: A psychological study. J Human Stress. Fall 1984;10:129-134. 1 I. Appels A, Mulder MA. Excess fatigue as a precursor of myocardial infarction. European Heart J. 1988;9:758-764. 12. Appels A. Tiredness before myocardial infarction. In: Lacky JH, Stungeon DA, eds. Proceedings of the Fvteenth European Conference on Psychosomatic Research. London: John Libbey; 1986. 13. Cherniss C. Stsff Burnout: Job Stress in the Human Services. Beverley Hills, CA: Sage; 1980. 14. Rose GA, Blackbum H, Gillum RF, Prineas RJ. Cardiovascular Survey Metho&. Geneva: WHO; 1982. 15. Ezrahi N. Burnout in military officers ranks: A construct validation. Tel Aviv, Israel; Tel Aviv University, 1985. Dissertation.
60
16. Caplan RD, Cobb S, French JRP Jr, Van Harrison R, Pinneau SR Jr. Job Demands and Worker Health: Main Effects and Occupational Differences. Washington, DC: Report to National Institute for Occupational Safety and Health, HEW Publication 75-16 (NIOSH); 1975. 17. Fraser GE. Preventive Cardiology. New York: Oxford University Press; 1986. 18. Appels A. The year before myocardial infarction. In: Dembroski TM, Schmidt T, Blumchen G, eds. Biobehavioral Bases of Coronary Heart Disease. New York: Karger; 1983:18-37. 19. Linden W, Pardhus DL, Dobson KS. Effects of response styles on the report of psychological and somatic distress. J Consult CIin Psychol. 1986;54:309-313. 20. Pieper C, LaCroix AZ, Karasek RA. The relation of psychosocial dimensions of work with coronary heart disease risk factors: A meta-analysis of five United States data bases. A m J Epidemiol. 1989;129:483-494. 21. Winkleby MA, Ragland DR, Syme SL. Self-reported stressors and hypertension: Evidence of an inverse relationship. A m J Epidemiol. 1988;127:124-134. 22. Thayer RE. The Biopsychology of Mood and Arousal. New York: Oxford University Press; 1989. 23. Trevisan M, Tsong Y, Stamler J, et al. Nervous tension and serum cholesterol: Findings from the Chicago Coronary Prevention Evaluation Program. J Human Stress. 1983;9:12-16. 24. Appels A. Vital exhaustion and depression as precursors of myocardial infarction. In: Spielberger IG, Defares PB, eds. Stress and Anxiety, vol 2. Washington: Hemisphere; 1988.
The UMI Article Clearinghouse offers articles from more than 11,OOO copyright-cleared periodicals in a wide range of subjects. You can place your orders electronically,as well as by phone, mail, and telefacsimile. For more information, please complete and mail this coupon to UMI Article Clearinghouse, 300 North Zeeb Road, Box 11. Ann Arbor, MI 48106 USA. Or call toll-free for an immediate response: 800-52 1-0600. From Alaska and Michigan call collect 3 1 3 - 7 6 1 - 4 7 0 0 . From Canada, call toll-Free 800-343-5299.
YES! I’d like to know more about UMI Article Clearinghouse Name ~
Company/lnstitution
Telephone-(
1
U M-I
A Bell & Howell Company 300 North Zeeb Road Ann Arbor, MI 48106 USA
Behavioral Medicine
