Burnout Research 2 (2015) 1924

Contents lists available at ScienceDirect

Burnout Research
journal homepage: www.elsevier.com/locate/burn

Research Article

Life stress as potential risk factor for depression and burnout

T. Plieger a, , M. Melchers a , C. Montag b , R. Meermann c , M. Reuter a,d
a
Department of Psychology, Biological Psychology, University of Bonn, Germany
b
Department of Psychology, Molecular Psychology, University of Ulm, Germany
c
AHG Psychosomatic Hospital, Bad Pyrmont, Germany
d
Center for Economics & Neuroscience (CENs), Laboratory of Neurogenetics, University of Bonn, Germany

a r t i c l e i n f o a b s t r a c t

Article history: Objective: Depression and burnout are two psychopathological labels that have been subject to an exten-
Received 26 September 2014 sive discussion over the last decades. The crucial question is whether they can be seen as conceptually
Received in revised form 9 January 2015 equal or as two distinct syndromes. One argument for the distinction is that depression impacts on the
Accepted 6 March 2015
whole life of a suffering person whereas burnout is restricted to the job context. Depression has been
shown to be affected by life stress. The more stressful life events a person experiences, the more he or
Keywords:
she is susceptible for developing a depression. As there is the widespread but controversial opinion that
Life stress
burnout is a prodromal syndrome of depression, the present study examined whether the number of
Stressful life events
Depression
stressful life events is also associated with an increased risk for burnout.
Burnout Methods: N = 755 healthy participants and N = 397 depressed patients completed the Maslach Burnout
ICD-10 diagnosis Inventory (MBI), the Beck Depression Inventory (BDI II) and reported the extent of experienced life stress.
Results: A signicantly closer relation between depression and life stress than between burnout and life
stress was found in the healthy (z = 3.01, p = .003) as well as in the depressed sample (z = 3.41, p = .001).
This nding was supported in both samples by means of a path analytic approach where the associations
between life stress, burnout, and depression were controlled for possible mediator and moderator effects,
also considering the inuence of age.
Conclusion: By considering the inuence of life stress it could be demonstrated that depression and
burnout are not identical although they share substantial phenotypic variance (r = .46.61). Most impor-
tant, the trivariate associations are the same in a representative employee sample and in an inpatient
clinical sample suggesting the same underlying mechanisms covering the whole range from normal
behavior to psychopathology. However, only longitudinal data can show if burnout necessarily turns into
depression with the consequence that the burnout life stress association approaches the depression
life stress association over time.
2015 The Authors. Published by Elsevier GmbH. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction denition of this construct. Moreover, there is even no accordance

whether burnout can be seen as a distinct syndrome apart from
There has been an extensive discussion on the association depression. Burnout was initially thought to emerge only in
between the constructs of depression and burnout over the past socially oriented professions, which require working with or help-
twenty years (e.g., Leiter & Durup, 1994; Maslach, Schaufeli, & ing clients/patients (Freudenberger, 1975; Maslach, 1976). Finally,
Leiter, 2001; Toker, Shirom, Shapira, Berliner, & Melamed, 2005). the term burnout was broadened and extended to all kinds of
Particularly, the fact that burnout again will neither appear in the occupational groups (Pines & Aronson, 1988) and more and more
upcoming version of the International Classication of Diseases studies tried to identify this syndrome in specic occupations. The
(ICD-11) nor is recognized in the updated Diagnostic and Statistical rising number of publications concerning burnout underlined the
Manual of Mental Disorders (DSM V) has intensied this debate. problem of its inconsistent denition and differentiation from other
Although burnout was mentioned rst in the 1970s psychopathologic or psychosomatic syndromes. Furthermore, it
(Freudenberger, 1975; Maslach, 1976), there still is no undisputed became increasingly clear that the exact characteristics of burnout
differ between professions. Therefore, the most prominent ques-
tionnaire measuring the burnout construct, the Maslach Burnout
Corresponding author. Tel.: +49 228 73 4312; fax: +49 228 73 62331. Inventory (MBI), was modied in a way which allows application to
E-mail address: [email protected] (T. Plieger). all kinds of occupational groups (Maslach et al., 2001). As the MBI is

http://dx.doi.org/10.1016/j.burn.2015.03.001
2213-0586/ 2015 The Authors. Published by Elsevier GmbH. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/
by-nc-nd/4.0/).
20 T. Plieger et al. / Burnout Research 2 (2015) 1924

the most prominent and most administered questionnaire for the moderating factors increasing the risk for burnout in adverse
assessment of burnout, most preliminary denitions of burnout working conditions such as personality (Swider & Zimmermann,
use the three symptoms that are measured by the three subscales 2010) or sleep duration (Sderstrm, Jeding, Ekstedt, Perski, &
of the MBI (1. emotional exhaustion, 2. depersonalization or cyn- kerstedt, 2012). This t in causality might be responsible for the
icism, and 3. reduced personal accomplishment) to characterize above mentioned overlap of 20% between both constructs, because
burnout. Although there are many different conceptualizations of the origin of burnout can also serve as one out of many risk factors
burnout, there is one characteristic all denitions have in common: for depression whereas other risk factors (e.g., in private life) might
an exhaustion of the organism which is caused by work stress. lead to depression but are not related to burnout. Therefore, it can
Due to this blurry picture, there is much space for discussion be hypothesized that depression and burnout can be dissociated
whether burnout can be seen as a distinct construct or whether it by their relation to the number of stressful life events.
is just another label for the same group of psychopathologic dis- Life stress has repeatedly been associated with depression (e.g.,
eases. Because of a very similar symptomatology there has been a Mazure, 1998) and is mostly operationalized by self-report check-
long debate especially concerning the overlap and distinctive char- lists of specic traumatic life experiences such as divorce, loss of
acteristics of burnout and depression. Among others, symptoms closely related persons, serious diseases, or sexual abuse in child-
of depression include fatigue, social withdrawal, feelings of failure hood. The more stressful life events are experienced by a person,
or worthlessness and various somatic symptoms such as insom- the more likely is the development of a depression (e.g., Kendler,
nia or hypersomnia, gain or loss of weight, and decreased interest Karkowski, & Prescott, 1999). There is also evidence that this asso-
in sexual activities (DSM-IV-TR; American Psychiatric Association, ciation is moderated by individual factors (for an overview see
2000). While fatigue is quite similar to the emotional and physi- Hammen, 2005). However, if burnout can be traced back to job rel-
cal exhaustion component of burnout, the depersonalization facet evant factors only, there should be no association between burnout
has similarities with social withdrawal. Finally, reduced personal and the number of stressful life events in general, but, if at all,
accomplishment can be seen as closely related to depressive feel- only with events that deal with work stress (e.g., unemployment
ings of failure and learned helplessness (Abramson, Seligman, & or nancial issues). This result would support the distinctiveness
Teasdale, 1978). These results are in line with some newer publi- of both constructs and would provide evidence for the genesis of
cations reporting similar symptoms in burned-out and depressed burnout.
patients and therefore come to the conclusion that burnout can- Hence, if the job environment is implied to be of specic rele-
not be seen as a distinct entity (Bianchi, Boffy, Hingray, Truchot, vance for burnout while environmental inuences on depression
& Laurent, 2013; Bianchi, Schonfeld, & Laurent, 2014). Bianchi and are unspecic, we expect (a) burnout to be associated with depres-
colleagues furthermore suppose an underestimation of the asso- sion, but (b) no relation of burnout to stressful life events, whereas
ciation between depression and burnout because most work in (c) depression is related to the reported number of stressful life
the eld is limited to healthy samples. However, in a conrma- events.
tory factor analysis the items measuring burnout and depression To test our hypotheses, we followed a stepwise strategy: First,
were found to load on different factors. Furthermore, the three- the hypotheses regarding the relations of life stress, depression,
factor structure of the MBI could be replicated so that the authors and burnout were tested in a healthy sample. Second, results were
conclude that both syndromes are distinct (Leiter & Durup, 1994). validated in a clinical sample. Based on the assumption of a con-
This might also be due to the form in which the items are presented tinuous transition of traits between healthy and psychopathologic
(Toker & Biron, 2012): burnout is, as a part of its denition, closely persons, similar results were expected in both samples, the healthy
related to job-context which is why the items are rather specic controls and the depressed patients.
or situational (e.g., I doubt the signicance of my work or I feel
emotionally drained from my work) whereas depression is more
2. Methods
global and not restricted to dened situations. Therefore, the items
assessing depression are more global as well (e.g., I am too tired
2.1. Participants
or fatigued to do most of the things I used to do or I feel guilty all
of the time).
2.1.1. Controls
Although there is some evidence that burnout and depression
A total of N = 755 (301 males; 454 females) healthy participants
can be seen as two different psychopathologies, the relation still
took part in this study. Data was collected within an ongoing project
remains unclear (e.g., Ahola & Hakanen, 2007; Bakker et al., 2000;
investigating the genetic basis of burnout. Inclusion criteria for par-
Glass & McKnight, 1996; Maslach et al., 2001; Toker & Biron, 2012).
ticipation were an age of over 18 years and actual employment.
There is evidence that the association between burnout and depres-
For participation in the research project participants were com-
sion is particularly based on the emotional exhaustion component
pensated with 15D.
of the MBI (e.g., Ahola et al., 2005; Nyklicek & Pop, 2005; Peterson
To recruit participants, we conducted an internet research to
et al., 2008). Iacovides, Fountoulakis, Kaprinis, and Kaprinis (2003)
identify businesses and public institutions which were contacted
suggest that especially severe burnout can lead to depression, a
and asked whether they would allow advertising the project to
result which could be replicated by other authors (e.g., Ahola &
their employees. Thus, participants of this study worked in vari-
Hakanen, 2007). Further data suggesting burnout as a preceding
ous companies and professions (e.g., medical staff, administration
syndrome or a step in the development of a depression respec-
employees, policemen, management-consultants, or civil servants)
tively were presented by Hakanen and Schaufeli (2012). Despite the
so that results are not restricted to a single profession with its spe-
reported ndings, there is still no consensus of any causal direction
cic demands. When employers agreed to take part in the study,
of the association (Hakanen, Schaufeli, & Ahola, 2008). However,
employees were informed about the project and asked for their
most studies nd an overlap between these two syndromes of about
voluntary participation.
20% shared variance (Iacovides et al., 2003).
Depression affects the whole life in general and can, among
others, also be caused by job related factors (Ahola, Hakanen, 2.1.2. Depressed patients
Perhoniemi, & Mutanen, 2014). In contrast, burnout is thought The second sample consisted of N = 397 (266 females, 131
to be mainly driven by such problems or chronic stress at work males) inpatients from psychosomatic hospitals. Due to the fact
(Maslach et al., 2001), although there have been identied other that burnout is not an acknowledged diagnosis, patients suffering
 T. Plieger et al. / Burnout Research 2 (2015) 1924 21

from burnout symptoms were diagnosed by ICD-10 F32 or F33 with depression and/or on burnout. So rst of all, we conducted one-way
an additional job specic anamnesis of exhaustion. Also pure ANOVAs comparing the healthy sample with the clinical sample to
depressive patients without a job specic anamnesis were included make sure that our samples differed with respect to the dependent
to allow a differentiation between burnout and depression. Not all variables. As expected, there was a difference between patients
hospitals provided us the exact clinical diagnosis of their patients and the healthy group in both, depression (F(1, 1150) = 603.527,
so that we relied preliminary on the scores obtained by the p < .001) and MBI burnout (F(1, 1150) = 693.601, p < .001). Depres-
administered assessment tools BDI-2 and MBI. We collaborated sive (M = 22.77, SD = 11.11) as well as burnout symptomatology
with several psychosomatic hospitals spread across Germany, (M = 50.44, SD = 15.75) in the sample of depressed patients sig-
most of them belong to the AHG consortium. The AHG comprises nicantly exceeded the values in the control group (depression:
more than 40 hospitals and has an own research department that M = 8.84, SD = 7.92; burnout: M = 26.06, SD = 14.49) on both vari-
supported us in this scientic project. As in the healthy sample, ables.
depressed participants had to be 18 years or older and also worked
in various occupational areas. Therefore, there were various 3.2. Control of sex and age
backgrounds/etiologies explaining their disease. However, these
participants all underwent an inpatient (cognitive behavioral) Next, it was tested whether age and sex had an impact on
psychotherapy during participation and, therefore, were not on depression, burnout, or the number of stressful life events. With
the job but on sick leave in the course of this study. Of course, par- respect to sex, there was a signicant effect on the MBI sub-
ticipation was voluntary and participants gave written informed scale cynicism in both samples (controls: F(1, 753) = 5.613, p = .018;
consent. Furthermore, patients also agreed that the respective patients: F(1, 394) = 7.722, p = .006) with higher scores in the male
psychotherapists sent us their clinical diagnoses (only ICD-10 keys). (controls: M = 8.57, SD = 6.94; patients: M = 17.58, SD = 7.25) than
The study was approved by the local ethics committee at the in the female (controls: M = 7.42, SD = 6.26; patients: M = 15.32,
University of Bonn. SD = 7.81) subsamples. Furthermore, there was a signicant differ-
ence between males (M = 20.36, SD = 10.97) and females (M = 23.96,
SD = 11.00) on BDI depression in the clinical sample (F(1, 395) = 9.549,
2.2. Questionnaire measures
p = .002). Age was only signicantly associated with the number of
stressful life events (r = .203, p < .001) in the healthy sample, but not
All used self-report measures and the corresponding reliabili-
in the clinical sample. However, age was weakly, but signicantly
ties observed in the present study are outlined below. To assess
associated with depression (r = .125, p = .012) and with burnout
burnout, participants were asked to complete the Maslach Burnout
(r = .113, p = .025) in the clinical subgroup. Hence, for analyses
Inventory General Survey (Maslach, Jackson, & Leiter, 1996) that
including the aficted variables, age and sex were considered as
consists of 16 items belonging to the three subscales Emotional
covariate (age) or second xed factor (sex) respectively.
Exhaustion (controls: = .88, patients: = .87), Cynicism (controls:
= .83, patients: = .81), and Professional Efcacy (controls: = .82,
3.3. Burnout and depression
patients: = .81). In order to facilitate comparison of the three sub-
scales, the Professional Efcacy scale was inverted so that higher
Because cynicism was signicantly related to sex, correlations
scores indicate a higher proportion of burnout strain as in the Emo-
between burnout subscales and BDI 2 depression score were com-
tional Exhaustion and Cynicism scales.
puted separately for both sexes and correlation coefcients were
Becks Depression Inventory II (Beck et al., 1996; BDI 2) as one of
compared by means of z-statistics (for procedure see Montag &
the most common clinical inventories was used to measure depres-
Reuter, 2008). As there were no signicant differences in the
sion. The BDI 2 consists of 21 items which cover affective, cognitive,
heights of correlation coefcients between males and females, only
and somatic symptoms of depression asking specically for partic-
coefcients of the total samples are reported (see Table 1). Correla-
ipants feelings during the last two weeks. Items are answered on
tions between the total MBI score and BDI 2 depression score were
a 4 point Guttman scale. Cronbachs alpha of the BDI 2 depression
r = .613 (p < .001) in the healthy sample and r = .456 (p < .001) in the
score was = .91 for the healthy sample and = .92 for the clinical
depressed sample.
sample.
Finally, a list of 30 stressful life events was administered to the
3.4. Stressful life events, burnout, and depression
participants. Each item consisted of a stressful experience that was
answered dichotomously with yes or no depending on whether
When correcting for age, number of stressful life events was
the answering person had experienced the event. For example, we
correlated with BDI 2 depression scores and MBI emotional exhaus-
asked for the following stressful life events: the death of a beloved
tion scores in both samples. The association with MBI Cynicism did
pet, a stressful and painful split-up from a partner, a serious acci-
not reach signicance in any of the subsamples (Table 2). In the
dent, the death of closely related persons, sexual abuse, or assault.
control sample, the total MBI score was weakly, but signicantly
To avoid that an experienced stressful life event was not covered by
correlated with the number of experienced stressful life events
the items, an additional item was Were there any other traumatic
whereas there was a signicant, but small association between SLE
situations or experiences that have not been mentioned here.
and MBI Efcacy in the clinical group. Correlation coefcients were
compared by means of z-statistics. In the control group, the asso-
3. Results ciation between BDI 2 and stressful life events did not differ from
the association between MBI emotional exhaustion and number of
3.1. Differences in subsample characteristics regarding burnout stressful life events (z = 1.60, p = .111), but there was a signicant
and depression difference regarding the association between number of stress-
ful life events and BDI 2 or total MBI score respectively (z = 3.01,
As the samples were independent and not divided post hoc by p = .003) indicating a closer relation between depression and num-
means of psychometrical criteria, it has to be mentioned that there ber of stressful life events than between burnout and number of
were participants in the control sample being highly depressed stressful life events. Comparison of correlations in the clinical sam-
and/or scored very high on burnout and that there were patients ple revealed similar results with signicantly different coefcients
that nearly had recovered and therefore had relatively low scores on between SLE-BDI 2 and SLE-MBI total score (z = 3.41, p = .001). The
22 T. Plieger et al. / Burnout Research 2 (2015) 1924

Table 1
Correlations between MBI subscales emotional exhaustion (EE), cynicism (CY) and efcacy (EF) and BDI 2 depression score. (Controls: top line, black
colored; patients: bottom line, gray colored.).

relationships between life stress depression and life stress MBI

emotional exhaustion also reached signicance in the clinical sam-
ple (z = 1.99, p = .047).
However, when associations between life stress and depres-
sion or life stress and burnout were additionally corrected for
one another, the association between depression and life stress
remained signicant whereas MBI emotional exhaustion was no
longer related to life stress in both samples. Contrarily, MBI efcacy
and MBI total score were associated with number of stressful life
events when correcting for age and depression (Table 2). Compari-
son of these partial correlation coefcients by means of z-statistics
revealed a signicant difference in life stress depression associa-
tion and life stress efcacy association (controls: z = 7.12, p < .001; Fig. 1. The inuence of AGE and stressful life events (SLE) on burnout (MBI) and
patients: z = 6.67, p < .001) as well as a signicant difference in life depression (BDI): Integrated results of the path analyses calculated for both samples
stress depression and life stress total burnout score association independently. Saturated models with perfect t. (Coefcients in black: controls;
(controls: z = 6.26, p < .001; patients: z = 5.53, p < .001). coefcients in gray: patients; nonsignicant coefcients in brackets.)

3.5. The relation between life stress, depression, and burnout in perfect t. Second, we conducted a multiple group analysis, esti-
the healthy vs. in the clinical sample mating all coefcients in the control sample and setting the paths
in the patients sample as invariant. The t of the multiple group
In the very last step, we investigated whether there is a dif- analysis was excellent (Chi2 = 10.88, df = 6, p = .092; RMSEA = 0.038;
ference in the associations reported above between the healthy CFI = 0.99). Third, we stepwise set all paths invariant one after
and the clinical sample. Therefore, we conducted a series of path another (but only one at a time) in the patient sample and cal-
analyses by means of the LISREL software package (Science Soft- culated Chi2 and RMSEA values for the resulting models. There was
ware International, Inc.) (Fig. 1). At rst, we tted the path model no signicant Chi2 -difference between the healthy and the clinical
portrayed in Fig. 1 in each subsample separately to estimate the sample in any of the paths dened in our models (Table 3) indicating
path-coefcients. Both models were saturated and therefore had a the same underlying relations between age, life stress, burnout, and

Table 2
Correlations between stressful life events (SLE) and burnout/depression when correcting for age, depression and age or burnout and age respectively.
 T. Plieger et al. / Burnout Research 2 (2015) 1924 23

Table 3 to depression than to burnout. The emotional exhaustion dimen-

All dened paths are tested one by one for signicant differences between the
sion was related to the number of stressful life events, whereas the
healthy and the depressed sample.
association between life stress and cynicism missed signicance.
Model Chi2 df p RMSEA p(Chi2 ) Reduced efcacy was weakly related to life stress only in the sample
All free 0 0 1.000 0.000 of depressed patients. An association between life stress opera-
MBI BDI 0.21 1 0.65036 0.000 0.64 tionalized as childhood trauma and fatigue, which is quite similar
SLE BDI 1.92 1 0.16566 0.040 0.17 to the emotional exhaustion component of burnout, has recently
SLE MBI 1.95 1 0.16257 0.041 0.16
been found by Borsini, Hepgul, Mondelli, Chalder, and Pariante
AGE BDI 2.23 1 0.13535 0.046 0.14
AGE MBI 2.15 1 0.14303 0.045 0.14 (2013). Similarly, Carlier, Lamberts, and Gersons (1997) found a
AGE SLE 2.46 1 0.11672 0.051 0.12 signicantly increased level of emotional exhaustion in policemen
that had experienced traumatic events and suffered from PTSD. An
explanation for the association between life stress and emotional
depression independent of group status (patients/controls). Finally, exhaustion is that it is the burnout dimension with the closest
the path models showed for both groups that the association relation to depression (Ahola et al., 2005; Nyklicek & Pop, 2005;
between SLE and depression is still signicant when controlling Peterson et al., 2008) while depression is indubitably related to life
for age and burnout. stress. This explanation is furthermore supported by our own data,
because after correcting the association between MBI emotional
4. Discussion exhaustion and critical life events for depression, the correlation
between both variables is no longer signicant in both samples.
Because of the sustained debate regarding the distinctiveness When correlations of number of stressful life events were cor-
of burnout and depression and due to the lack of a conclusive rected for the association between depression and burnout, there
denition of this construct, the aim of the present study was only remained a substantial association with depression, whereas
to test whether life stress accounts for the relationship between the association between life stress and burnout was albeit signif-
depression and burnout. As there is the widespread assump- icant close to zero indicating that burnout is mainly related to life
tion that depression arises from and affects many areas of life, stress by its close relation to depression and not by itself. Hence,
whereas burnout is not only but mainly restricted to job stress- when potentially mediating effects are not considered in the rela-
ors (Hakanen & Schaufeli, 2012), we assessed experienced life tion between either depression or burnout and other constructs,
stress in a large sample of employed adults and in a sample of the similarity between both syndromes might be overestimated.
depressed inpatients who were not employed during participa- To further investigate this assumption, we also set up path analy-
tion. We hypothesized that both constructs are related, but that ses for both samples. Again, the association between stressful life
the number of stressful life events is closer related to depression events and depression was signicant in both samples whereas
than to burnout. the path from life stress to burnout was much smaller in the
In our data, we found a moderate correlation between burnout healthy group and did not even reach signicance in the patients.
and depression, which is quite similar to that reported by Toker Moreover, we used the path models to compare the respective
et al. (2005) who found correlation coefcients of about r = .50 associations between the observed variables in our samples. There
between burnout and depression for men and for women. With were no differences in any of the dened paths so that we suggest
an amount of about 36% of shared variance in the healthy control the same underlying associations between life stress, depression,
sample, we found a closer relation than other authors suggested and burnout in both samples. These associations therefore seem
in the past (e.g., Iacovides et al., 2003). However, the association to be valid for the whole continuum, from the normal/healthy
was still moderate which means that important shares of variance range to psychopathology. We also considered that because most
were not equal for both constructs. Furthermore, correlations of patients were diagnosed according to the denition of depression
the three MBI subscales drew the typical picture of closest relation (as burnout is no valid ICD-10 diagnosis) it could be possible that
between depression and emotional exhaustion and lowest associ- both samples would not differ concerning their burnout scores.
ation between depression and reduced efcacy. These ndings are When comparing our samples on both variables, we found not only
in line with shared variance estimates by Schaufeli and Enzmann a signicant difference in depression with patients scoring signi-
(1998), who found an overlap with depression of 1238% (emo- cantly higher but also in burnout symptoms in the same direction,
tional exhaustion), 229% (depersonalization), and 320% (reduced i.e. the clinical sample having a higher mean value on MBI as com-
efcacy) respectively. In the clinical sample, there was an amount pared to the healthy sample. Taken together, the results in the
of shared variance of about 20%, which meets the shared variance clinical sample can be seen as an independent replication for the
estimates of the literature very well. Worthy of note, Bianchi et al. found relations between life stress, depression, and burnout found
(2014) also reported similar correlations between depression and in the healthy sample. Therefore, results indicate that these asso-
burnout for their whole sample but report a much closer rela- ciations can also be applied to participants in a psychopathological
tion when only taking highly burned-out participants into account. context.
They argue that the fact that most studies only include healthy Nevertheless, assessment of stressful life events has been sub-
participants might lead to an underestimation of the correlation ject of discussions for a long time as it is mostly operationalized
between burnout and depression as these results cannot be gen- by self-report scales asking for experiences of specic events. This
eralized to a clinical setting. However, our data suggest the same procedure neglects subjective ratings of what is considered as
underlying mechanisms in the healthy as well as in the inpatient stressful: One might rate the sudden death of a goldsh as trau-
depressed sample questioning this assumption. Moreover, the cor- matic or stressful, whereas others might not classify their heart
relation between both constructs was smaller in the depressed attack as serious and, therefore, not as stressful (Monroe & Reid,
sample than in the healthy sample (r = .456 vs. r = .613). 2008; Monroe & Roberts, 1990). Therefore, the problem arises that
We found signicant associations between the number of stress- participants might report a parental divorce or the death of a pet
ful life events and depression in both samples, whereas burnout even though it was not traumatic if they are just asked, whether
was only related to life stress in the healthy group. When compar- they have experienced a particular life event or not. Because of
ing the correlation coefcients, there was a signicant difference in the individual differences in perceived life stress when experienc-
the strength of association indicating that life stress is closer related ing particular events, the assessment of life stress is not valid, as
24 T. Plieger et al. / Burnout Research 2 (2015) 1924

subjective burden in sense of life stress is not measured, but the Ahola, K., Hakanen, J., Perhoniemi, R., & Mutanen, P. (2014). Relationship between
objective experience of potentially burdening events. Hence, the burnout and depressive symptoms: A study using the person-centred approach.
Burnout Research.
validity of life stress assessment might be questioned. This should Ahola, K., Honkonen, T., Isomets, E., Kalimo, R., Nykyri, E., Aromaa, A., et al.
be considered in future research. A further limitation is that our (2005). The Relationship between job-related burnout and depressive disorders
study is cross-sectional. In future, one could think of controlling Results from the Finnish Health 2000 Study. Journal of Affective Disorders, 88,
5562.
for the elapsed time of the particular SLEs. Alternatively, a longi- American Psychiatric Association. (2000). Diagnostic and statistical manual of men-
tudinal study could be conducted with the rst measurement at tal disorders, fourth edition, text revision: DSM-IV-TR. Washington, DC: American
the earliest possible time to capture the predictive power of SLEs Psychiatric Publishing, Inc.
Bakker, A. B., Schaufeli, W. B., Demerouti, E., Janssen, P. M. P., Van der Hulst, R.,
over the life span. Results show that life stress is related to both
& Brouwer, J. (2000). Using equity theory to examine the difference between
burnout and depression. This effect may be caused by the tempo- burnout and depression. Anxiety, Stress & Coping: An International Journal, 13,
ral overlap between burnout and depression. This could also be 247268.
Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Beck Depression Inventory manual (2nd
claried by a longitudinal study investigating the relation between
ed.). San Antonio, TX: Psychological Corporation.
all three variables dependent on time. When burnout is seen as a Bianchi, R., Boffy, C., Hingray, C., Truchot, D., & Laurent, E. (2013). Comparative
prodromal syndrome of depression (e.g., Iacovides et al., 2003), it symptomatology of burnout and depression. Journal of Health Psychology, 18(6),
is possible that parts of the burnout life stress association basi- 782787.
Bianchi, R., Schonfeld, I. S., & Laurent, E. (2014). Is burnout a depressive disorder?
cally represent the depression life stress association at an earlier A reexamination with special focus on atypical depression. International Journal
stage. This would in turn question the existence or distinctiveness of Stress Management, 21(4), 307.
of burnout itself. Longitudinal data could shed more light on this Borsini, A., Hepgul, N., Mondelli, V., Chalder, T., & Pariante, C. M. (2013). Childhood
stressors in the development of fatigue syndromes: A review of the past 20 years
issue by observing the development of the associations reported of research. Psychological Medicine, 115.
here during a potential shift from burnout to depression over time. Carlier, I. V., Lamberts, R. D., & Gersons, B. P. (1997). Risk factors for posttraumatic
But if so, the explanation is still missing why the depression life stress symptomatology in police ofcers: A prospective analysis. The Journal of
Nervous and Mental Disease, 185(8), 498506.
stress association gets closer over time or with growing severity of Freudenberger, H. J. (1975). The staff burn-out in alternative institutions. Psychother-
depression. This would suggest differential and unique aspects of apy: Theory, Research & Practice, 12, 7382.
burnout independent of whether it is a preceding syndrome of Glass, D. C., & McKnight, J. D. (1996). Perceived control, depressive symptomatology
and professional burnout: A review of the evidence. Psychology and Health, 11(1),
depression or not.
2348.
This study tried to provide further evidence regarding the rela- Hakanen, J. J., & Schaufeli, W. B. (2012). Do Burnout and work engagement predict
tion between depression and the burnout syndrome. Therefore, the depressive symptoms and life satisfaction? A three-wave seven-year prospec-
tive study. Journal of Affective Disorders, 141(2), 415424.
relation of stressful life events to both depression and burnout was
Hakanen, J. J., Schaufeli, W. B., & Ahola, K. (2008). The Job Demands-Resources model:
measured. First of all, our data suggest an overlap of both syn- A three-year cross-lagged study of burnout, depression, commitment, and work
dromes that is in line with previous ndings (Schaufeli & Enzmann, engagement. Work & Stress, 22(3), 224241.
1998). With respect to life stress, we found an association between Hammen, C. (2005). Stress and depression. Annual Review of Clinical Psychology, 1,
293319.
the number of experienced stressful life events and burnout that Iacovides, A., Fountoulakis, F. N., Kaprinis, S., & Kaprinis, G. (2003). The relationship
was as hypothesized weaker than the relation between depres- between job stress, burnout and clinical depression. Journal of Affective Disorders,
sion and life stress. Hence, our data support the assumption that 75, 209221.
Kendler, K. S., Karkowski, L. M., & Prescott, C. A. (1999). Causal relationship between
depression aficts all areas of life whereas burnout has a rather spe- stressful life events and the onset of major depression. American Journal of Psy-
cic scope even though it has parallels to depression. This seems to chiatry, 156(6), 837841.
hold for the healthy population as well as for depressed patients. Leiter, M. P., & Durup, J. (1994). The discriminant validity of burnout and depression:
A conrmatory factor analytic study. Anxiety, Stress & Coping: An International
Journal, 7(4), 357373.
Conict of interest Maslach, C. (1976). Burned out. Human Behavior, 5, 1622.
Maslach, C., Jackson, S. E., & Leiter, M. P. (1996). Maslach Burnout Inventory manual
(3rd ed.). Palo Alto, CA: Consulting Psychologists Press.
The authors declare that there are no conicts of interest. Maslach, C., Schaufeli, W. B., & Leiter, M. P. (2001). Job burnout. Annual Review of
The Daimler Benz Foundation was neither involved in collection, Psychology, 52(1), 397422.
analysis, and interpretation of data, nor in the writing of the report Mazure, C. M. (1998). Life stressors as risk factors in depression. Clinical Psychology:
Science and Practice, 5(3), 291313.
or in the decision to submit the article.
Monroe, S. M., & Reid, M. W. (2008). Geneenvironment interactions in depression
research genetic polymorphisms and life-stress polyprocedures. Psychological
Acknowledgements Science, 19(10), 947956.
Monroe, S. M., & Roberts, J. E. (1990). Conceptualizing and measuring life stress:
Problems, principles, procedures, progress. Stress Medicine, 6(3), 209216.
We thank the Daimler Benz Foundation (Ladenburg, Germany) Montag, C., & Reuter, M. (2008). Does speed in completing an online questionnaire
who nanced this study by a research grant. have an inuence on its reliability? Cyber Psychology & Behavior, 11(6), 719721.
Nyklicek, I., & Pop, V. J. (2005). Past and familial depression predict current symptoms
Furthermore, we want to thank the colleagues who helped col-
of professional burnout. Journal of Affective Disorders, 88, 6368.
lecting the data for the clinical sample. In particular, we thank Peterson, U., Demerouti, E., Bergstrm, G., Samuelsson, M., Asberg, M., & Nygren,
(in alphabetical order) Dr. Jrgen Borgart (AHG hospital Bad Pyr- A. (2008). Burnout and physical and mental health among Swedish healthcare
mont, Germany), Dr. Mario Gartmann (AHG hospital Bad Drkheim, workers. Journal of Advanced Nursing, 62(1), 8495.
Pines, A., & Aronson, E. (1988). Career burnout. New York: The Free Press.
Germany), Dr. Michael Rolffs (AHG hospital Daun, Germany), and Schaufeli, W. B., & Enzmann, D. (1998). The burnout companion to study and practice.
Dr. Albrecht Schumacher (AHG hospital Waren, Germany) for London: Taylor & Francis.
organizing the data collection in the respective psychosomatic hos- Sderstrm, M., Jeding, K., Ekstedt, M., Perski, A., & kerstedt, T. (2012). Insufcient
sleep predicts clinical burnout. Journal of Occupational Health Psychology, 17(2),
pitals. 175.
The position of CM is funded by a Heisenberg-grant by the Ger- Swider, B. W., & Zimmerman, R. D. (2010). Born to burnout: A meta-analytic path
man Research Foundation (MO 2363/3-1). model of personality, job burnout, and work outcomes. Journal of Vocational
Behavior, 76(3), 487506.
Toker, S., & Biron, N. (2012). Job burnout and depression: Unraveling their tempo-
References ral relationship and considering the role of physical activity. Journal of Applied
Psychology, 97(3), 699710.
Abramson, L. Y., Seligman, M. E., & Teasdale, J. D. (1978). Learned helplessness in Toker, S., Shirom, A., Shapira, I., Berliner, S., & Melamed, S. (2005). The associ-
humans: Critique and reformulation. Journal of Abnormal Psychology, 87(1), 49. ation between burnout, depression, anxiety, and inammation biomarkers:
Ahola, K., & Hakanen, J. (2007). Job strain, burnout and depressive symptoms: A C-reactive protein and brinogen in men and women. Journal of Occupational
prospective study among dentists. Journal of Affective Disorders, 104, 103110. Health Psychology, 10(4), 344362.