Journal of Psychosomatic Research 168 (2023) 111217

Contents lists available at ScienceDirect

Journal of Psychosomatic Research

journal homepage: www.elsevier.com/locate/jpsychores

Somatic symptoms in burnout in a general adult population

Patrik Hammarström, Simon Rosendahl, Michael Gruber, Steven Nordin *
Department of Psychology, Umeå University, Sweden

A R T I C L E I N F O A B S T R A C T

Keywords: Objective: Burnout results in individual suffering and high societal costs, and is associated with somatic symp­
Anxiety toms, anxiety and depression, calling for prevention by early identification. The objectives were to (i) determine
Chronic stress prevalence of specific somatic symptoms in burnout, (ii) identify the symptoms, referred to as somatic symptoms
Depression
of burnout (SS-B), that are particularly associated with burnout, (iii) determine their extent of association with
Exhaustion syndrome
Mental health
burnout, and (iv) determine how well a certain number of SS-B differentiates participants with and without
Public health burnout in a general Swedish adult population.
Methods: Cross-sectional, population-based data were used from 687 participants high on burnout, and 2544
referents based on the Shirom-Melamed Burnout Questionnaire. The Patient Health Questionnaire 15-item So­
matic Symptom Severity Scale was used to assess common somatization symptoms, and the Hospital Anxiety and
Depression Scale to assess anxiety and depression.
Results: Feeling tired/having low energy, back pain, joint/limb pain, trouble sleeping, headaches, stomach pain,
nausea/gas/indigestion, and constipation/loose bowels/diarrhea were most prevalent in burnout (57.2–95.0%).
These symptoms, except for joint/limb pain, and dizziness, were also identified as the SS-B, with odds ratios of
2.34–12.74 and 1.95–9.11 when adjusted for background variables, and for anxiety and depression, respectively.
Corresponding odds ratios for each additional number of SS-B were 1.69 and 1.52, respectively. The highest
balanced accuracy (71.6%) for predicting burnout was found for ≥4 SS-B.
Conclusion: Fatigue, pain and gastrointestinal symptoms are particularly common in burnout. Further studies
may show whether clinicians should consider screening for burnout when patients present with SS-B without
pathophysiological explanations.

1. Introduction
Burnout is strongly related to the combination of long-term stress
and limited recovery. A recent study shows that 17.5% of a general adult
Swedish population meet a criterion for burnout, and that it is particu­
larly prevalent in young women [1]. It is a persistent condition as
medical guidelines recommend 6–12 months of sick leave [2], and high
level of burnout may persist after 18 months, despite rehabilitation [3].
Burnout was early defined by Maslach and Jackson [4] as emotional
exhaustion, depersonalization, and low sense of personal accomplish­
ment, with focus on the working population. A later definition of
burnout, and the one presently used, includes the dimensions fatigue,
cognitive weariness, tension, and listlessness [5]. It is often considered
an occupational phenomenon, for example, according to the 11th
Revision of the International Classification of Diseases (ICD-11). How­
ever, this view of burnout being a specific occupational phenomenon has
been questioned [6]. From the perspective of public health, it is
important to have an inclusive approach, and thus not focus only on the
working population, which would exclude afflicted adolescents, the
unemployed and old persons. In a population-based study about 11% of
those aged 60–69 years, and 16% of these aged 70–79 years were found
to meet a commonly used criterion for burnout [1].
The persistent symptoms often include cognitive difficulties that
affect speed and quality of work [7–10]. Apart from suffering of the
afflicted individual, burnout has a major economic cost for society in
terms of productivity loss and healthcare [11]. Apart from the afore­
mentioned core dimensions of burnout, other dimensions have been
suggested [12]. Burnout is the crucial component of stress-related
exhaustion disorder (SED) [3]; F43.8A in the Swedish version of the
ICD-10. The Swedish National Board of Health Welfare [13] defines SED
as distinct mental and somatic impact after long-term stress. In Sweden,
1–7% of patients in psychiatric care, 11–16% in primary care, and
17–29% in occupational care are diagnosed with SED [2]. Multimodal
interventions for treatment of SED have demonstrated improved health

* Corresponding author at: Department of Psychology, Umeå University, SE-90187 Umeå, Sweden.
E-mail address: steven.nordin@umu.se (S. Nordin).

https://doi.org/10.1016/j.jpsychores.2023.111217
Received 4 November 2022; Received in revised form 5 March 2023; Accepted 5 March 2023
Available online 8 March 2023
0022-3999/© 2023 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
P. Hammarström et al.
outcome, and may include components such as cognitive behavioral
interventions, interventions for improving sleep, relaxation exercises,
and aerobic and cognitive training [14,15]. The treatment can be
considered as safe to administer and reasonable in cost.
Anxiety and depression show high comorbidity with burnout
[16–20]. Maladaptive aspects of perfectionism have been shown to be a
factor that may partly explain this comorbidity [17], but also overlap in
diagnostic criteria and symptoms [21].
As many as 132 different symptoms have been reported to be asso­
ciated with burnout [22]. The symptoms include, for example, neck/
back pain, tingling/numbness, fatigue, headache, gastrointestinal
problems, respiratory problems, sleep difficulties, and pain sensitivity
[23–26]. They may be referred to changes in the endocrine, immune,
nervous and digestive systems, and to its contribution to symptoms of
anxiety and depression [27–30]. This may be caused partly by high
release of cortisol from the hypothalamic-pituitary-adrenal (HPA) axis,
but chronic stress resulting in burnout may instead cause HPA down­
regulation and hypocortisolism. In a similar vein, sleep disturbance,
which is common in burnout [31] may not only potentiate the stress
reactivity of the HPA axis, but if becoming chronic also contribute to
hypocortisolism [32].
By asking the patient about symptoms that are particularly associ­
ated with burnout, indications of such symptoms can motivate the
clinician to screen for burnout to prevent clinical burnout, SED and
related serious health problems. The present objective was therefore to
identify symptoms that are particularly associated with burnout. In
contrast to most previous studies [24–26,31,33], this was conducted in a
general population, including a reference group, and by controlling for
anxiety and depression symptoms.
Whereas past research on burnout has focused predominantly on
work and organizational factors (e.g., workload, time pressure, and role
conflict) [34] the present study focused on intrapsychic factors, and had
four objectives. Firstly, we determined the prevalence of specific
symptoms, among a set of common psychosomatic symptoms, in in­
dividuals high and low in burnout in a general Swedish adult popula­
tion. Secondly, we identified a subset of these symptoms, referred to as
somatic symptoms of burnout (SS-B), and particularly associated with
burnout. Thirdly, we determined extent to which (i) each specific SS-B
and (ii) each additional number of SS-B are associated with burnout
relative to a reference group. Fourthly, we determined how accurately a
certain number of SS-B correctly classifies the participants as having and
not having burnout, respectively.
A Swedish study has shown that nausea/gas/indigestion, headaches,
dizziness, constipation/loose bowels/diarrhea, feeling heart pound/
race, and back pain were the most prevalent symptoms in patients with
SED, and that they were reported by more than half of that sample [35].
Based on these results from patients with SED, but instead with focus on
less severe cases, the present study tested the hypothesis of these
symptoms also being among those most strongly associated with
burnout in a general population. It was further hypothesized that the
associations would be weaker when controlled for symptoms of
depression and anxiety.
2. Methods
2.1. Study population
Data were obtained from the population-based Västerbotten Envi­
ronmental Health Study, which investigates mental and somatic health
in a general Swedish population [36]. The county of Västerbotten in
Northern Sweden has a population of approximately 260,000 in­
habitants (about 195,000 aged 18–79 years), and an age and sex dis­
tribution that is very similar to the general Swedish population [37]. A
random sample of 8520 individuals aged 18 to 79 years, drawn from the
population register of Västerbotten, was invited to the study in 2010, of
which 3406 (40.0%) agreed to participate. The sample was stratified for
2
Journal of Psychosomatic Research 168 (2023) 111217
age and sex according to the age strata 18–29, 30–39, 40–49, 50–59,
60–69 and 70–79 years.
2.2. Questions and questionnaire instruments
Single questions were used for a background description of the
participants regarding demographics, physical exercise, smoking, self-
rated health, and self-report of having been given diagnoses by a
physician of hypertension, SED, depression, and various types of anxiety
syndrome and functional somatic syndrome. The Swedish version [38]
of the 10-item Perceived Stress Scale [39] was used to assess stress level.
The Swedish version [40] of the Shirom Melamed Burnout Ques­
tionnaire (SMBQ) [5] was used to assess burnout. This particular
burnout instrument was chosen as it, in contrast to other instruments
[41], has no particular emphasis on occupational burnout, and thus is
appropriate for a general population. It has 22 items that are categorized
into the subscales Exhaustion (e.g. “I feel tired”), Listlessness (e.g. “I feel
alert”), Tension (e.g. “I feel tense”) and Weariness (e.g. “I can't think
clearly”). Each item is responded to on a scale ranging from 1 (“Almost
never”) to 7 (“Almost always”). The global scale was used that is the
average rating across all items. The SMBQ has good factor structure,
reliability and construct and convergent and validity [40,42]. A cutoff
score of 3.75 on the global SMBQ scale was used for the criterion of
burnout [40]. This was chosen to include persons at an early stage of
burnout, in contrast to that of, for example, 4.47 that refers to clinical
burnout [43]. The categorization of participants enabled assessment of
the risk for each SS-B in burnout expressed as an odds ratio, and
assessment of sensitivity, specificity and balanced accuracy for meeting
a certain cutoff of number of SS-B.
The Swedish version [44] of the Patient Health Questionnaire 15-
Item Somatic Symptom Severity Scale (PHQ-15) [45] was used to
assess prevalence of specific symptoms or symptom clusters. These ac­
count for >90% of the somatic complaints reported by non-clinical re­
spondents, and includes 14 of the most common somatization symptoms
in DSM-IV. The symptom menstrual problems refers to women only. The
items are responded to on a scale ranging from 0 (“Not bothered at all”),
to 1 (“Bothered a little”) and to 2 (“Bothered a lot”). A symptom was
considered present if rated as being bothered by it a little or a lot. The
PHQ-15 has good reliability and validity [44–46].
The Swedish version [47] of the Hospital Anxiety and Depression
Scale (HADS) [48] was applied to assess symptoms of anxiety and
depression. It consists of seven items for anxiety (HADS-A; e.g.,
“Worrying thoughts go through my mind”) and seven for depression
(HADS-D; e.g., “I have lost interest in my appearance”). Each item is
responded to on a scale ranging from 0 to 3 (high score referring to high
level of anxiety or depression symptoms). The HADS has good validity
and reliability [49,50].
2.3. Procedure
The questionnaire was sent to the participants with the instruction to
return it via mail with prepaid postage, and non-responders received up
to two reminders. All participants responded to the questionnaire during
the period March–April 2010, before onset of the pollen season in
Västerbotten. The study was carried out in accordance with the Helsinki
Declaration and approved by the Umeå Regional Ethics Board (Dnr 09-
171 M) and the Swedish Ethical Review Authority (Dnr 2022–05265-
02). All participants gave their informed consent to participate.
2.4. Statistical analysis
Fully conditional Markov chain Monte Carlo method with 10
maximum iterations was applied to estimate missing values on the
SMBQ (0.97%), PHQ-15 (1.31%) and HADS (0.79%). This multiple
imputation created five imputed datasets, and the estimate values were
obtained by pooling the five datasets. The burnout and reference groups
P. Hammarström et al. Journal of Psychosomatic Research 168 (2023) 111217

were compared on demographics and other background variables, total Table 1

PHQ-15 score, number of PHQ-15 symptoms, and prevalence rate of Characteristics of the burnout and reference groups and the total sample, and
specific PHQ-15 symptoms with independent sample t-test and chi- results from group comparisons with independent t-test and chi-square analysis.
square analysis. Burnout Referent Total
Identification of the SS-B was based on Spearman correlation co­ sample
efficients between somatic symptom severity (rating 0–2) for each PHQ- (n = 687) (n = 2544) (n = 3231)
15 symptom and global SMBQ score. Logistic regression analyses
Age, years, mean (SD) 47.4 (16.8) 51.6 (16.6) 50.6 (16.7)
determined extent to which each specific SS-B and each additional ***
number of SS-B were associated with increased risk of burnout. This Women, % (n) 66.2 (455) 53.1 55.9 (1806)
provided odds ratios (ORs) and confidence intervals, with the referent *** (1351)
group as referents. Separate analyses were conducted when (i) not Married/living with partner, % (n) 68.4 (470) 75.2 73.8 (2383)
*** (1913)
adjusting, (ii) adjusting for the background variables age, sex, married/ Having children, % (n) 37.8 (260) 30.4 (773) 32.0 (1033)
living with partner, having children, physical exercise ≥2 times per ***
week, and (iii) adjusting for these background variables and HADS-A University education, % (n) 42.4 (291)ns 42.3 42.3 (1366)
and HADS-D scores, the latter representing symptoms of anxiety and (1075)
ns
Smoking regularly, % (n) 9.2 (62) 8.4 (214) 8.5 (276)
depression.
Physical exercise >2 times/week, % 54.6 (375) 69.1 66.0 (2132)
Based on number of PHQ-15 symptoms, classification accuracy into (n) *** (1757)
the groups of burnout and referents was calculated. The accuracy Self-rated health, % (n)
included sensitivity (percentage of burnout cases correctly identified as Excellent/very good 17.6 (121) 46.3 40.2 (1300)
cases based on meeting a cutoff for having a certain number of SS-B), *** (1179)
Good 32.9 (226) 34.4 (875) 34.1 (1101)
specificity (percentage of referents that were correctly identified as Fairly good/poor 48.6 (334) 18.3 (465) 24.7 (799)
referents based on not meeting the cutoff) and balanced accuracy PSS-10, mean (SD) 20.1 (5.71) 12.0 (5.10) 13.7 (6.19)
(overall measure of accuracy due to unequal numbers of participants in ***
the burnout and reference groups). The guidelines 0.90–1 = excellent, Diagnosis, % (n)
Hypertension 23.1 (159)ns 24.2(616) 24.0 (775)
0.80–0.90 = good, 0.70–0.80 = fair, 0.60–0.70 = poor, and 0.50–0.60 =
Stress-related exhaustion disorder 11.9 (82)*** 2.2 (57) 4.3 (139)
fail [51] were used to interpret accuracy. The α-level was set at 0.05, and Generalized anxiety disorder 3.6 (25)*** 0.2 (6) 1.0 (31)
the Statistical Package for the Social Sciences (IBM SPSS Statistics for Panic disorder 4.8 (33)*** 0.6 (16) 1.5 (49)
Windows, Version 26.0. Armonk, NY) was used for the data analyses. Posttraumatic stress disorder 2.5 (17)*** 0.4 (9) 0.8 (26)
Depression 17.0 (117) 2.0 (50) 5.2 (167)
***
3. Results Fibromyalgia 5.2 (36)*** 1.4 (36) 2.2 (72)
Irritable bowel syndrome 4.5 (31)*** 1.8 (47) 2.4 (78)
3.1. Study sample Chronic fatigue syndrome 1.9 (13)*** 0.4 (9) 0.7 (22)
SMBQ subscales, mean (SD)
Exhaustion 4.6 (0.88)*** 2.4 (0.82) 2.8 (1.23)
Among the 3406 participants, 175 were excluded due to having
Listlessness 5.0 (0.93)*** 3.0 (1.10) 3.4 (1.33)
responded to less than half of the items on the SMBQ, PHQ-15 and/or Tension 4.4 (1.05)*** 2.5 (0.93) 2.9 (1.23)
HADS, resulting in a study sample of 3231 participants. Of these, 687 Weariness 4.4 (1.12)*** 2.1 (0.87) 2.6 (1.34)
met the criterion for burnout (global SMBQ score ≥ 3.75), and the HADS-A, mean (SD) 8.3 (3.98)*** 3.4 (2.83) 4.4 (3.70)
remaining 2544 participants constituted referents. The two samples are HADS-D, mean (SD) 6.2 (3.56)*** 2.2 (2.15) 3.1 (3.00)

described in Table 1. The burnout sample was significantly younger than
the referents, and consisted to a larger extent of participants who were
women, not married/living with a partner, having children, and not
exercising ≥2 times/week. The burnout sample also rated their health as
poorer, scored higher on the PSS-10, was more likely to have reported
being given a diagnosis of SED, generalized anxiety disorder, panic
disorder, posttraumatic stress disorder, depression, fibromyalgia, irri­
table bowel syndrome and chronic fatigue syndrome, and scored higher
on all SMBQ subscales and on the HADS-A and HADS-D. The samples did
not differ in prevalence of university education, smoking or
hypertension.
3.2. Prevalence of somatic symptoms
Table 2 shows mean global PHQ-15 score and number of PHQ-15
symptoms, and prevalence rates of the specific PHQ-15 symptoms.
The burnout sample had, compared to the referents, significantly higher
global PHQ-15 score (t = 25.33, p < 0.001), larger number of symptoms
(t = 25.56, p < 0.001), and higher prevalence rate on all PHQ-15
symptoms (χ2 > 16.48, p < 0.001).
3.3. Identification of SS-B
Correlation coefficients between PHQ-15 symptom ratings and
global SMBQ score are presented in Table 3, with coefficients ranging
from 0.143 to 0.614 (p < 0.001). Correlation coefficients of 0.20 and
0.35 were considered as adequate and good, respectively [52]. The two
cutoffs would have identified as many as twelve SS-B, and as few as two
PSS-10 = 10-item Perceived Stress Scale.
SMBQ = Shirom Melamed Burnout Questionnaire.
HADS-A = Anxiety subscale of the Hospital Anxiety and Anxiety Scale.
HADS-D = Depression subscale of the Hospital Anxiety and Depression Scale.
***p < 0.001, nsnon-significant.
SS-B, respectively. A cutoff of 0.30 was therefore chosen, which iden­
tified the following eight SS-B: feeling tired/having low energy, trouble
sleeping, stomach pain, nausea/gas/indigestion, back pain, headaches,
constipation/loose bowels/diarrhea, and dizziness.
3.4. Odds ratios for specific SS-B and number of SS-B
Table 4 shows the association between burnout and each SS-B in
terms of ORs. For all SS-B, the OR differed significantly from unity (p <
0.001) when unadjusted (2.69–14.70), adjusted for background vari­
ables (2.34–12.74), and adjusted for background variables and HADS-A
and HADS-D scores (1.95–9.11). The OR (95% CI) for each additional
number of SS-B was 1.71 (1.62–1.80) when unadjusted, 1.69
(1.60–1.78) when adjusted for background variables, and 1.52
(1.44–1.62) when adjusted for background variables and HADS-A and
HADS-D scores. All three ORs were found to differ significantly from
unity (p < 0.001).
3.5. Sensitivity, specificity and balanced accuracy
Measures of sensitivity, specificity and balanced accuracy for
meeting a certain cutoff of number of SS-B are given in Table 5. The

3
P. Hammarström et al. Journal of Psychosomatic Research 168 (2023) 111217

Table 2 Table 4
PHQ-15 score and number of PHQ-15 symptoms, and prevalence rates of specific Odds ratios [95% CI] for each Somatic Symptom of Burnout (SS-B). All odds
PHQ-15 symptoms in the burnout and reference groups, and the total sample. ratios were significant at p < 0.001.
Independent t-test and chi-square analysis showed significant groups differences SS-B Unadjusted Adjusteda Adjustedb
on all group comparisons (p < 0.001).
Feeling tired/having 14.70 12.74 9.11
Burnout Referent Total low energy [10.28–21.04] [8.88–18.28] [6.19–14.42]
sample Trouble sleeping 4.47 [3.72–5.37] 4.42 2.92
Global PHQ-15 score, mean (SD) 9.72 4.90 (3.60) 5.93 (4.29) [3.64–5.35] [2.36–3.62]
(4.48) Stomach pain 3.56 [2.97–4.26] 3.30 2.26
Number of PHQ-15 symptoms, mean 7.17 4.11 (2.68) 4.76 (3.96) [2.78–3.99] [1.81–2.81]
(SD) (2.70) Constipation/loose 3.09 [2.58–3.69] 2.95 2.33
Stomach pain, % (n) 64.5 (436) 32.7 (833) 39.3 (1269) bowels/diarrhea [2.45–3.55] [1.89–2.89]
Back pain, % (n) 78.2 (537) 57.2 61.6 (1991) Dizziness 3.02 [2.51–3.63] 3.03 2.27
(1454) [2.49–3.69] [1.81–2.85]
Joint/limb pain, % (n) 74.7 (513) 58.6 62.1 (2005) Nausea/gas/ 2.86 [2.40–3.42] 2.70 2.03
(1492) indigestion [2.25–3.25] [1.64–2.50]
Menstrual problems, % (n)a 36.4 (166) 26.4 (357) 29.0 (523) Headaches 2.72 [2.28–3.25] 2.34 1.95
Headaches, % (n) 65.4 (449) 41.0 46.1 (1491) [1.93–2.83] [1.56–2.48]
(1042) Back pain 2.69 [2.20–3.28] 2.67 2.40
Chest pain, % (n) 24.3 (167) 8.9 (227) 12.2 (394) [2.18–3.29] [1.89–3.04]
Dizziness, % (n) 39.3 (270) 17.6 (448) 22.2 (717) a
Age, sex, married/living with partner, having children, and physical exercise
Fainting, % (n) 9.0 (62) 2.7 (69) 4.1 (131)
≥2 times per week.
Feeling the heart pound/race, % (n) 35.7 (245) 15.1 (383) 19.4 (628) b
Shortness of breath, % (n) 26.8 (184) 12.5 (317) 15.5 (501) Age, sex, married/living with partner, having children, physical exercise ≥2
Sexual pain/problems, % (n) 14.8 (102) 6.4 (164) 8.2 (266) times per week, and score on the Anxiety and Depression subscales of the Hos­
Constipation/loose bowels/diarrhea, 57.2 (393) 30.1 (767) 35.9 (1160) pital Anxiety and Depression Scale.
% (n)
Nausea/gas/indigestion, % (n) 63.5 (436) 37.7 (959) 43.2 (1395)
Feeling tired/having low energy, % (n) 95.0 (652) 55.7 64.0 (2067) Table 5
(1416) Sensitivity, specificity, and balanced accuracy as functions of number of Somatic
Trouble sleeping, % (n) 70.2 (482) 34.4 (876) 42.0 (1358)
Symptoms of Burnout (SS-B).
PHQ-15 = Patient Health Questionnaire 15-Item Somatic Symptom Severity Number of SS-Ba Sensitivity (%) Specificity (%) Balanced
Scale. accuracy (%)
a
Women only.
≥1 99.8 10.8 55.3
≥2 97.6 26.2 61.9
≥3 91.8 43.2 67.5
Table 3 ≥4 82.5 60.6 71.6
Spearman correlation coefficients between rating on each PHQ-15 ≥5 66.9 74.8 70.9
symptom and global SMBQ score. All coefficients were significant ≥6 51.6 85.7 68.7
at p < 0.001. ≥7 29.4 93.9 61.7
8 12.0 98.3 55.2
Stomach pain 0.342
a
Back pain 0.305 Feeling tired/having low energy, trouble sleeping, stomach pain, nausea/
Joint/limb pain 0.238 gas/indigestion, back pain, headaches, constipation/loose bowels/diarrhea, and
Menstrual problemsa 0.165 dizziness.
Headaches 0.304
Chest pain 0.235
Dizziness 0.301 data from a general adult population that includes a reference group,
Fainting 0.143 and by control for anxiety and depression symptoms.
Feeling the heart pound/race 0.265 In addition to the burnout group having higher global PHQ-15 score
Shortness of breath 0.223
and a larger number of symptoms than the referents, the results showed
Sexual pain/problems 0.171
Constipation/loose bowels/diarrhea 0.304 that the burnout group had higher prevalence rate on all PHQ-15
Nausea/gas/indigestion 0.331 symptoms. The most prevalent symptoms in the burnout group were
Feeling tired/having low energy 0.614 feeling tired/having low energy, back pain, joint/limb pain, trouble
Trouble sleeping 0.389
sleeping, headaches, stomach pain, nausea/gas/indigestion, and con­
PHQ-15 = Patient Health Questionnaire 15-Item Somatic Symptom stipation/loose bowels/diarrhea, all of which were reported by more
Severity Scale. than half of the group. These symptoms overlap to a large extent with
SMBQ = Shirom Melamed Burnout Questionnaire. those previously shown to be most commonly reported by patients with
a
Women only. SED [35]. In this latter study, feeling tired/having low energy and
trouble sleeping were not addressed since they are criteria for SED [13].
highest balanced accuracy (71.6%), was found for ≥4 SS-B, with a Hence, the results suggest that burnout and SED are very similar
sensitivity of 82.5% and a specificity of 60.6%. regarding somatic symptoms.
With one exception, the data provided support for the hypothesis of
4. Discussion nausea/gas/indigestion, headaches, dizziness, constipation/loose
bowels/diarrhea, feeling the heart pound/race, and back pain being
The objectives of the current study were to determine prevalence of those that strongest correlated with degree of burnout, and thus were
specific somatic symptoms, of relevance for somatization, in burnout, identified as the SS-B. The exception was feeling the heart pound/race,
identify SS-B among these symptoms, determine extent to which each although its correlation coefficient (0.265) with burnout was close to the
specific SS-B and each additional number of SS-B are associated with cutoff of 0.30 for SS-B. The other identified SS-B were feeling tired/
burnout in relation to a reference group, and determine accuracy of having low energy, trouble sleeping, and stomach pain.
burnout discriminability based on the number of SS-B. The study adds to When adjusted for background variables, there was an approximate
the existing literature by addressing identification of burnout based on 2- to 13-fold increased risk of burnout if having a certain SS-B. The

4
P. Hammarström et al.
highest risk was found for feeling tired/having low energy and trouble
sleeping. This is not surprising considering that sleep disturbance and
fatigue are cardinal symptoms in burnout [31]. Sleep disturbance and
stress enhance each other, and both conditions may result in both high
release of cortisol and, in the long run, hypocortisolism, with conse­
quences for the other identified SS-B. Hence, stress and sleep distur­
bance may contribute to changes in the digestive (nausea/gas/
indigestion, constipation/loose bowels/diarrhea and stomach pain) and
nervous systems (dizziness and feeling the heart pound/race) [27,28] as
well as lead to increased pain sensitivity (headaches, back pain, and
stomach pain) [53,54]. When controlled for background variables, the
analyses showed that the risk of burnout increased by a factor of 1.69 for
each additional SS-B. Giving examples, this implies that there is an 8.2
times increased risk for burnout if having four SS-B, and a 66.5 times
increased risk if having all eight SS-B.
Regarding the discriminability, thus the ability to correctly classify
of participants as having burnout or not based on a cutoff of number of
SS-B, the highest balanced accuracy (71.6%; good according to [51])
was found for ≥4 symptoms, providing a sensitivity of 82.5% and a
specificity of 60.6%. However, this cutoff is appropriate assuming that it
is as important to correctly identify cases without burnout as it is to
correctly identify cases with burnout. If the latter condition is more
important, which is likely to be the case for precaution in healthcare, a
cutoff of ≥3 symptoms may be more appropriate. This cutoff showed a
sensitivity of 91.8%, but on the expense of the specificity (43.2%) and
the balanced accuracy (67.5%; poor according to [51]).
The present data also support the hypothesis of the associations be­
tween symptoms and burnout becoming weaker when controlled for
anxiety and depression symptoms. Thus, the OR for all SS-B were smaller
(1.95–9.11) when controlled for anxiety and depression than when not
controlled for these conditions (2.34–12.74). Importantly, though, the
associations remained significant after this control, as would be ex­
pected due to the three conditions being different and robust constructs
[55]. In a similar vein, the OR for each additional SS-B decreased from
1.69 to 1.52. Using a similar approach as in the present study, Carlehed
et al. [56] reported feeling tired/having low energy, trouble sleeping,
stomach pain and nausea/gas/indigestion being strongly associated
with both anxiety and depression, and headaches also with anxiety, and
constipation/loose bowels/diarrhea also with depression. However,
although becoming weaker, the associations between symptoms and
burnout remained even after controlling for anxiety and depression
symptoms. This implies that although there is some overlap between
burnout and anxiety/depression, burnout is, nevertheless, an indepen­
dent construct separate from anxiety and depression.
The categorization of the study sample into burnout and reference
groups is somewhat validated by the fact that the burnout group was
more likely than the referents to have been given diagnoses of SED,
anxiety syndrome (generalized anxiety disorder, panic disorder and
posttraumatic stress disorder) and depression, and also rated their
health as poorer. In accordance with this, results from a recent study
showed that high levels of anxiety, depression, insomnia, somatic
symptoms and poor self-rated health were found to be risk factors for
development of burnout [57]. Thus, persistent mental burden of these
kinds may result in excessive and prolonged stress and the inability to
regain lost resources, which characterizes burnout [34]. Work-related
and/or non-work-related factors may be underlying this mental
burden. Furthermore, apart from scoring higher than the referents on
the PSS-10, they scored higher on all four SMBQ subscales, thus on
Exhaustion, Listlessness, Tension and Weariness, which suggests high
level of burnout in several respects. As would be expected, the burnout
group was also more likely to have been given diagnoses of fibromyal­
gia, irritable bowel syndrome and chronic fatigue syndrome, which are
known to be associated with burnout [58–60].
Regarding clinical implications, the results evoke the question as to
whether primary care clinicians should, in addition to fatigue and sleep
problems, take digestive problems, dizziness, heart pounding and pain
5
Journal of Psychosomatic Research 168 (2023) 111217
as indications of need of screening for long-term stress for prevention of
clinical burnout or SED. However, this should be accompanied by
screening for depression, which otherwise may cause the risk of
perpetuating the problem of major depression being missed and attrib­
uted to burnout. Further research on the relation between burnout and
somatic symptoms, preferably with a longitudinal design to study cau­
sality, is needed for definite recommendations to be made.
Although not addressed in the present study with the PHQ-15,
cognitive and respiratory problems and tingling/numbness may be
additional symptoms to be attentive to [7,23,26]. In these cases, and if
the burnout appears to be strongly related to the work place, the patient
may be referred to the employer's occupational health care for further
evaluation and intervention. The latter often aims at restoring a healthy
balance between effort and rest, recovery from chronic stress, and
improving coping skills [61]. Furthermore, the comorbidity in somatic
symptoms between burnout and anxiety/depression implies that the
clinician should not only focus on a primary diagnosis, but also recog­
nize the need of multimodal treatment [35].
The present study had certain strengths, such as having a large scale,
population-based sample, with an age and sex distribution that is very
similar to that of the general Swedish adult population [37]. However,
the study also had limitations, such as a relatively low response rate
(40.0%). Although effects of low response rate have been shown to vary
very little between rates of 30–70% [62], it is possible that a selection
bias has resulted in individuals with severe burnout, due to their con­
dition, being less likely to participate. If so, the variance in the data
would be smaller than otherwise, resulting in underestimation of the
ORs.
5. Conclusion
The results suggest that (i) prevalence of all PHQ-15 symptoms are
higher than normal in burnout, (ii) feeling tired/having low energy,
trouble sleeping, stomach pain, nausea/gas/indigestion, back pain,
headaches, constipation/loose bowels/diarrhea, and dizziness are
particularly strong indications of burnout, (iii) the risk of burnout in­
creases with number of these symptoms, and (iv) having four or more of
these symptoms best discriminates between being and not being at risk
for burnout. Further research may show whether primary care clinicians
should consider screening for burnout when patients present with SS-B
without a pathophysiological explanation.
Funding
This work was supported by AFA Insurance (190082).
Declaration of Competing Interest
The authors have no competing interest to report.
Data availability statement
The data that support the findings of this study are available from the
corresponding author upon reasonable request.
Acknowledgement
We are thankful to Dr. Eva Palmquist for help with the database. This
work is an extension of an undergraduate thesis by Patrik Hammarström
and Simon Rosendahl.
References
[1] P. Höglund, C. Hakelind, S. Nordin, Severity and prevalence of various types of
mental ill-health in a general adult population: age and sex differences, BMC
Psychiat. 20 (2020) 1–11, https://doi.org/10.1186/s12888-020-02557-5.
P. Hammarström et al.
[2] Swedish National Audit Office, Försäkringsmedicinskt beslutsstöd: ett stöd för
Försäkringskassan vid psykisk ohälsa? Insurance medicine decision support: a
support for the Swedish Social Insurance Agency in cases of mental illness? Report
2018:22, Stockholm 2018, 2018.
[3] K. Glise, G. Ahlborg Jr., I.H. Jonsdottir, Course of mental symptoms in patients
with stress-related exhaustion: does sex or age make a difference? BMC Psychiat.
12 (2012) 18. http://www.biomedcentral.com/1471-244X/12/18.
[4] C. Maslach, S.E. Jackson, The measurement of experienced burnout, J. Organ.
Behav. 2 (1981) 99–113, https://doi.org/10.1002/job.4030020205.
[5] S. Melamed, T. Kushnir, A. Shirom, Burnout and risk factors for cardiovascular
diseases, Behav. Med. 18 (1992) 53–60, https://doi.org/10.1080/
08964289.1992.9935172.
[6] R. Bianchi, R. Brisson, Burnout and depression: causal attributions and construct
overlap, J. Health Psychol. 24 (2019) 1574–1580, https://doi.org/10.1177/
1359105317740415.
[7] L. Öhman, S. Nordin, J. Bergdahl, L. Slunga Birgander, A. Stigsdotter-Neely,
Cognitive function in outpatients with perceived chronic stress, Scand. J. Work
Environ. Health 33 (2007) 223–232, https://doi.org/10.5271/sjweh.1131.
[8] K. Österberg, B. Karlson, B. Malmberg, Å.M. Hansen, A follow-up of cognitive
performance and diurnal salivary cortisol changes in former burnout patients,
Stress 15 (2012) 589–600, https://doi.org/10.3109/10253890.2011.648972.
[9] B.G. Oosterholt, J.H. Maes, D. Van der Linden, M.J. Verbraak, M.A. Kompier,
Getting better, but not well: a 1.5 year follow-up of cognitive performance and
cortisol levels in clinical and non-clinical burnout, Biol. Psychol. 117 (2016)
89–99, https://doi.org/10.1016/j.biopsycho.2016.02.009.
[10] A.S. Dahlman, K. Blennow, H. Zetterberg, K. Glise, I.H. Jonsdottir, Growth factors
and neurotrophins in patients with stress-related exhaustion disorder,
Psychoneuroendocrinol. 109 (2019), 104415, https://doi.org/10.1016/j.
psyneuen.2019.104415.
[11] J. Hassard, K.R.H. Teoh, G. Visockaite, P. Dewe, T. Cox, The cost of work-related
stress to society: a systematic review, J. Occup. Health Psychol. 23 (2018) 1–17,
https://doi.org/10.1037/ocp0000069.
[12] N. Almén, B. Jansson, The reliability and factorial validity of different versions of
the Shirom-Melamed burnout measure/questionnaire and normative data for a
general Swedish sample, Int. J. Stress. Manag. 28 (2021) 314, https://doi.org/
10.1037/str0000235.
[13] Swedish National Board of Health and Welfare, Utmattningssyndrom:
Stressrelaterad Psykisk ohälsa. (Exhaustion syndrome: Stress-related mental poor
health). Stockholm, Bjurner & Bruno, 2003.
[14] J. Wallensten, M. Åsberg, M. Wiklander, A. Nager, Role of rehabilitation in chronic
stress-induced exhaustion disorder: a narrative review, J. Rehabil. Med. 51 (2019)
331–342, https://doi.org/10.2340/16501977-2545.
[15] J.C. van de Leur, M. Buhrman, F. Åhs, A. Rozental, G.B. Jansen, Standardized
multimodal intervention for stress-induced exhaustion disorder: an open trial in a
clinical setting, BMC Psychiat. 20 (2020) 1–14, https://doi.org/10.1186/s12888-
020-02907-3.
[16] K. Ahola, T. Honkonen, E. Isometsä, R. Kalimo, E. Nykyri, A. Aromaa, J. Lönnqvist,
The relationship between job-related burnout and depressive disorders: results
from the Finnish Health 2000 Study, J. Affect. Disord. 88 (2005) 55–62.
[17] N.W. Van Yperen, M. Verbraak, E. Spoor, Perfectionism and clinical disorders
among employees, Personal. Individ. Differ. 50 (2011) 1126–1130, https://doi.
org/10.1016/j.paid.2011.01.040.
[18] Y. Ding, J. Qu, X. Yu, S. Wang, The mediating effects of burnout on the relationship
between anxiety symptoms and occupational stress among community healthcare
workers in China: a cross-sectional study, PLoS One 9 (2014), e107130, https://
doi.org/10.6084/m9.figshare.1136113.
[19] J. Zhou, Y. Yang, X. Qiu, X. Yang, H. Pan, B. Ban, W. Wang, Relationship between
anxiety and burnout among Chinese physicians: a moderated mediation model,
PLoS One 11 (2016), e0157013, https://doi.org/10.1371/journal.pone.0157013.
[20] A. Adamsson, S. Bernhardsson, Symptoms that may be stress-related and lead to
exhaustion disorder: a retrospective medical chart review in Swedish primary care,
BMC Fam. Pract. 19 (2018) 172, https://doi.org/10.1186/s12875-018-0858-7.
[21] R. Bianchi, E. Laurent, I.S. Schonfeld, J. Verkuilen, C. Berna, Interpretation bias
toward ambiguous information in burnout and depression, Personal. Individ.
Differ. 135 (2018) 216–221, https://doi.org/10.1016/j.paid.2018.07.028.
[22] W. Schaufeli, D. Enzmann, The Burnout Companion to Study and Practice: A
Critical Analysis, CRC press, London, 1998, https://doi.org/10.1201/
9781003062745.
[23] R.C. Gorter, M.A. Eijkman, J. Hoogstraten, Burnout and health among Dutch
dentists, Eur. J. Oral Sci. 108 (2000) 261–267, https://doi.org/10.1034/j.1600-
0722.2000.108004261.x.
[24] D.C.L. Mohren, G.M.H. Swaen, I.J. Kant, L.G.P.M. van Amelsvoort, P.J.A. Borm, J.
M.D. Galama, Common infections and the role of burnout in a Dutch working
population, J. Psychosom. Res. 55 (2003) 201–208, https://doi.org/10.1016/
S0022-3999(02)00517-2.
[25] U. Peterson, E. Demerouti, G. Bergström, M. Samuelsson, M. Åsberg, Å. Nygren,
Burnout and physical and mental health among Swedish healthcare workers,
J. Adv. Nurs. 62 (2008) 84–95, https://doi.org/10.1111/j.1365-2648.2007.04580.
x.
[26] D.A.J. Salvagioni, F.N. Melanda, A.E. Mesas, A.D. González, F.L. Gabani, S.M. de
Andrade, Physical, psychological and occupational consequences of job burnout: a
systematic review of prospective studies, PLoS One 12 (2017), e0185781, https://
doi.org/10.1371/journal.pone.0185781.
[27] S. Cohen, D. Janicki-Deverts, G.E. Miller, Psychological stress and disease, J. Am.
Med. Assoc. 298 (2007) 1685–1687, https://doi.org/10.1001/jama.298.14.1685.
6
Journal of Psychosomatic Research 168 (2023) 111217
[28] M.U. Nater, S. Fischer, U. Ehlert, Stress as a pathophysiological factor in functional
somatic syndromes, Curr. Psychiatr. Rev. 7 (2011) 152–169, https://doi.org/
10.2174/157340011796391184.
[29] I.H. Jonsdottir, A. Sjörs Dahlman, Endocrine and immunological aspects of
burnout: a narrative review, Eur. J. Endocrinol. 180 (2019) R147–R158, https://
doi.org/10.1530/EJE-18-0741.
[30] P. Koutsimani, A. Montgomery, K. Georganta, The relationship between burnout,
depression, and anxiety: a systematic review and meta-analysis, Front. Psychol. 10
(2019) 284, https://doi.org/10.3389/fpsyg.2019.00284.
[31] M. Söderström, K. Jeding, M. Ekstedt, A. Perski, T. Åkerstedt, Insufficient sleep
predicts clinical burnout, J. Occup. Health Psychol. 17 (2012) 175–183, https://
doi.org/10.1037/a0027518.
[32] J.H. van Dalfsen, C.R. Markus, The influence of sleep on human hypothalamic-
pituitary-adrenal (HPA) axis reactivity: a systematic review, Sleep Med. Rev. 39
(2018) 187–194, https://doi.org/10.1016/j.smrv.2017.10.002.
[33] P.M.C. Mommersteeg, C.J. Heijnen, A. Kavelaars, L.J.P. van Doornen, Immune and
endocrine function in burnout syndrome, Psychosom. Med. 68 (2006) 879–886,
https://doi.org/10.1097/01.psy.0000239247.47581.0c.
[34] C. Aumayr-Pintar, C. Cerf, A. Parent-Thirion, Eurofound (2018), Burnout in the
Workplace: A Review of Data and Policy Responses in the EU, Publications Office
of the European Union, Luxembourg, 2018, p. 2018.
[35] K. Glise, G. Ahlborg, I.H. Jonsdottir, Prevalence and course of somatic symptoms in
patients with stress-related exhaustion: does sex or age matter, BMC Psychiat. 14
(2014) 118.
[36] E. Palmquist, A.-S. Claeson, G. Neely, B. Stenberg, S. Nordin, Overlap in prevalence
between various types of environmental intolerance, Int. J. Hyg. Environ. Health
217 (2014) 427–434.
[37] Statistics Sweden, Tables on the population in Sweden 2009, in: Population by Sex,
Age, Marital Status by County Dec. 31, 2009 According to the Administrative
Subdivisions of Jan. 1, 2010, 2010.
[38] M. Nordin, S. Nordin, Psychometric evaluation and normative data of the Swedish
version of the 10-item perceived stress scale, Scand. J. Psychol. 54 (2013)
502–507, https://doi.org/10.1111/sjop.12071.
[39] S. Cohen, G. Williamson, (1988). Perceived stress in a probability sample of the
United States, in: S. Spacapan, S. Oskamp (Eds.), The Social Psychology of Health:
Claremont Symposium on Applied Social Psychology, Sage, Newbury Park, CA,
1988, pp. 31–67.
[40] G. Grossi, A. Perski, B. Evengård, V. Blomkvist, K. Orth-Gomér, Physiological
correlates of burnout among women, J. Psychosom. Res. 55 (2003) 309–316,
https://doi.org/10.1016/S0022-3999(02)00633-5.
[41] Y. Shoman, S.C. Marca, R. Bianchi, L. Godderis, H.F. Van der Molen, I.G. Canu,
Psychometric properties of burnout measures: a systematic review, Epidemiol.
Psychiatr. Sci. 30 (2021), e8, https://doi.org/10.1017/S2045796020001134.
[42] A. Sundström, A. Söderholm, M. Nordin, S. Nordin, Construct validation and
normative data for different versions of the Shirom-Melamed burnout
questionnaire/measure in a Swedish population sample, Stress. Health (2022),
https://doi.org/10.1002/smi.3200.
[43] Å. Lundgren-Nilsson, I.H. Jonsdottir, J. Pallant, G. Ahlborg, Internal construct
validity of the Shirom-Melamed burnout questionnaire (SMBQ), BMC Public
Health 12 (2012) 1. http://www.biomedcentral.com/1471-2458/12/1.
[44] S. Nordin, E. Palmquist, M. Nordin, Psychometric evaluation and normative data
for a Swedish version of the patient health questionnaire 15-item somatic symptom
severity scale, Scand. J. Psychol. 54 (2013) 112–117, https://doi.org/10.1111/
sjop.12029.
[45] K. Kroenke, R.L. Spitzer, J.B.W. Williams, The PHQ-15: validity of a new measure
for evaluating the severity of somatic symptoms, Psychosom. Med. 64 (2002)
258–266, https://doi.org/10.1111/sjop.12029.
[46] H. van Ravesteijn, K. Wittkampf, P. Lucassen, E. van de Lisdonk, H. van den
Hoogen, H. van Weert, J. Huijser, A. Schene, C. van Weel, A. Speckens, Detecting
somatoform disorders in primary care with the PHQ-15, Ann. Fam. Med. 7 (2009)
232–238, https://doi.org/10.1370/afm.985.
[47] J. Lisspers, A. Nygren, E. Söderman, Hospital anxiety and depression scale (HAD):
some psychometric data for a Swedish sample, Acta Psychiatr. Scand. 96 (1997)
281–286, https://doi.org/10.1111/j.1600-0447.1997.tb10164.x.
[48] A.S. Zigmond, R.P. Snaith, The hospital anxiety and depression scale, Acta
Psychiatr. Scand. 67 (1983) 361–370, https://doi.org/10.1111/j.1600-0447.1983.
tb09716.x.
[49] I. Bjelland, A.A. Dahl, T.T. Haug, D. Neckelmann, The validity of the hospital
anxiety and depression scale: an updated literature review, J. Psychosom. Res. 52
(2002) 69–77, https://doi.org/10.1016/S0022-3999(01)00296-3.
[50] M. Hansson, J. Chotai, A. Nordstöm, O. Bodlund, Comparison of two self-rating
scales to detect depression: HADS and PHQ-9, Brit. J. Gen. Pract. 59 (2009)
e283–e288, https://doi.org/10.3399/bjgp09X454070.
[51] European Federation of Psychology Associations, EFPA Review Model for the
Description and Evaluation of Psychological and Educational Tests (Version 4.2.6),
European Federation of Psychology Associations, Brussels, 2013.
[52] P.H. Finan, B.R. Goodin, M.T. Smith, The association of sleep and pain: an update
and a path forward, J. Pain 14 (2013) 1539–1552, https://doi.org/10.1016/j.
jpain.2013.08.007.
[53] C.G. Abdallah, P. Geha, Chronic pain and chronic stress: two sides of the same
coin? Chronic Stress 1 (2017) 1–10, https://doi.org/10.1177/2470547017704763.
[54] R.H. El Khouli, K.J. Macura, P.B. Barker, M.R. Habba, M.A. Jacobs, D.A. Bluemke,
The relationship of temporal resolution to diagnostic performance for dynamic
contrast enhanced (DCE) MRI of the breast, Magnet. Res. Imag. 30 (2009)
999–1004, https://doi.org/10.1002/jmri.21947.
P. Hammarström et al.
[55] P. Koutsimani, A. Montgomery, K. Georganta, The relationship between burnout,
depression, and anxiety: a systematic review and meta-analysis, Front. Psychol. 10
(2019) 284, https://doi.org/10.3389/fpsyg.2019.00284.
[56] G. Carlehed, J. Katz, S. Nordin, Somatic symptoms of depression and anxiety: a
population based study, Mental Health Prevent. 6 (2017) 57–62, https://doi.org/
10.1016/j.mhp.2017.03.005.
[57] P. Höglund, C. Hakelind, M. Nordin, S. Nordin, Risk factors for insomnia and
burnout: a longitudinal population-based cohort study, Stress. Health (2023),
https://doi.org/10.1002/smi.3218.
[58] M. Bernik, T. Sampaio, L. Gandarela, Fibromyalgia comorbid with anxiety
disorders and depression: combined medical and psychological treatment, Curr.
Pain Headache Rep. 17 (2013) 1–9, https://doi.org/10.1007/s11916-013-0358-3.
Journal of Psychosomatic Research 168 (2023) 111217
[59] C. Hausteiner-Wiehle, P. Henningsen, Irritable bowel syndrome: relations with
functional, mental, and somatoform disorders, World J. Gastroenterol. 20 (2014)
6024–6030, https://doi.org/10.3748/wjg.v20.i20.6024.
[60] N. Coetzee, D. Maree, B. Smit, The relationship between chronic fatigue syndrome,
burnout, job satisfaction, social support and age among academics at a tertiary
institution, Int. J. Occup. Med. Environ. Health 32 (2019) 75–85, https://doi.org/
10.13075/ijomeh.1896.01274.
[61] A. Van Dam, A clinical perspective on burnout: diagnosis, classification, and
treatment of clinical burnout, Eur. J. Work Organizat. Psychol. 30 (2021) 732–741,
https://doi.org/10.1080/1359432X.2021.1948400.
[62] S. Galea, M., Tracy participation rates in epidemiologic studies, Ann. Epidemiol. 17
(2007) 643–653, https://doi.org/10.1016/j.annepidem.2007.03.013.