Journal of Psychosomatic Research 173 (2023) 111467

Contents lists available at ScienceDirect

Journal of Psychosomatic Research

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

Fatigue and somatization in shift-workers: Effects of depression and sleep

Jin Won Seo, Jooyoung Lee, Sehyun Jeon, Yunjee Hwang, Jichul Kim, Somi Lee, Seog Ju Kim *
Department of Psychiatry, Sungkyunkwan University College of Medicine, Samsung Medical Center, Seoul, Republic of Korea

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

Keywords: Objective: We investigated the differences in fatigue and somatization between shift and non-shift workers and
Shift workers explored the effects of sleep and depression on fatigue and somatization in shift workers.
Fatigue Methods: In total, 4543 shift workers and 2089 non-shift workers completed self-reported questionnaires. The
Somatization
Center for Epidemiologic Studies Depression Scale (CES–D), Pittsburgh Sleep Quality Index (PSQI), Fatigue
Depression
Sleep
Severity Scale (FSS), and Somatization subscale of the Symptom Checklist 90-item version (SCL-SOM) were used
to measure depression, sleep quality, fatigue, and somatization, respectively. Fatigue and somatization were
compared between shift and non-shift workers after controlling for different sets of covariates.
Results: Compared to non-shift workers, shift workers reported higher FSS (mean difference: 2.19 ± 0.30, p <
0.01) and SCL-SOM (mean difference: 1.77 ± 0.21, p < 0.01) scores after controlling for age, gender, presence of
medical illness, occupational category, monthly income, length of service, and weekly working hours. The
between-group difference in FSS score was no longer significant after additionally controlling for CES-D (p =
0.15) or PSQI (p = 0.18). The between-group difference in SCL-SOM score showed only non-significant trends
after additionally controlling for CES-D (p = 0.09) or PSQI (p = 0.07). The group difference in SCL-SOM scores
disappeared after controlling for both CES-D and PSQI scores (p = 0.99).
Conclusions: Shift workers had higher fatigue and somatization levels than non-shift workers and the group
difference was associated with disturbed sleep and depressed mood in shift workers.

1. Introduction depressive symptoms [16]. These previous studies did not consider the
close interaction of sleep, mood, and medical illness with fatigue in shift
Shift work is ubiquitous worldwide. Approximately 10–30% of em­ workers. Few studies have explored the potentially complicated in­
ployees work in shifts or at night [1]. Shift workers often perform teractions among sleep, depression, medical illness, and fatigue in shift
stressful jobs. The physical and mental health of shift workers have workers.
received significant attention because of their impact on work efficiency In addition to fatigue, somatization, defined as subjective distress
and safety [2,3]. due to body dysfunction, also produces physical complaints that reduce
Fatigue worsens the quality of life of shift workers and increases the the quality of life in shift workers [17]. Somatization and fatigue sub­
risk of accidents [4]. Several previous studies have reported a rela­ stantially overlap with each other [18] and share several risk factors,
tionship between shift work and fatigue [5–7]. Shift workers have including depression and disturbed sleep [19,20]. Despite the similar­
several potential risk factors for fatigue. Shift workers have higher risks ities, previous studies suggested clear differences in the conceptual
of sleep disturbances [8], depressed mood [9], and medical problems definition and prevalence of somatization and fatigue. Fatigue is
[10]. Fatigue is also associated with depression, medical illness, and generally defined as mental or physical tiredness, weakness, or
disturbed sleep in the general population [11–14]. However, it remains exhaustion [38], while somatization is generally defined as unexplained
unclear whether increased sleep disturbances and/or depression in shift medical symptoms without organic disease [39]. According to a pro­
workers would lead to increased fatigue. A previous study reported that spective study, it is suggested that there might be a distinct and
disturbed sleep aggravated fatigue in shift workers [15]. Another study consistent population who only has fatigue without psychosomatic
reported that fatigue in shift-working nurses was independent of symptoms [38]. Whereas, in the longitudinal study following up

* Corresponding author at: Department of Psychiatry, Sungkyunkwan University School of Medicine, Samsung Medical Center, 50 Ilwon-dong, Kangnam-gu,
#06351, Seoul 135-710, Korea.
E-mail address: [email protected] (S.J. Kim).

https://doi.org/10.1016/j.jpsychores.2023.111467
Received 20 May 2022; Received in revised form 12 August 2023; Accepted 15 August 2023
Available online 17 August 2023
0022-3999/© 2023 Elsevier Inc. All rights reserved.
J.W. Seo et al. Journal of Psychosomatic Research 173 (2023) 111467

somatization patients, somatization was rather suggested as an incon­ Apparently inappropriate answers to the questions were automatically
sistent condition [40]. Compared to fatigue in shift workers, somatiza­ rejected by survey program. The online survey company tried to
tion is less well studied. Although a few studies have reported an improve authenticity of the data by eliminating outlying data or
association between shift work and somatization [17,22,23], they had examining data by panels.
small sample sizes and/or included participants with only certain oc­ Information regarding demographic characteristics, presence of
cupations (e.g., nurses and police officers). Furthermore, these studies medical illness, occupation, monthly income, length of service, and
did not account for risk factors of somatization, such as sleep, depres­ weekly working hours was gathered using an online survey. Regarding
sion, medical illness, and interactions thereof. medical illness, the participants were requested to answer the question,
The current study explored the fatigue and somatization in shift and ‘Do you have any medical or physical illness?’ and “If you have any
non-shift workers, and the effects of risk factors (such as sleep and medical or physical illness, please write the name of medical illness.”
mood) on the fatigue and somatization levels in shift workers. We hy­ The additional information was acquired, including occupation,
pothesized that fatigue and somatization would be greater in shift monthly income (from 1 [no income] to 6 [> 4,500,000 Korean won]),
workers than non-shift workers, and that the higher fatigue and soma­ length of service, and weekly working hours. Occupation was classified
tization levels in shift workers would be associated with the presence of into 3 categories (A [Managers, Professionals, and Clerks], B [Service
depressive symptoms and sleep disturbances. and Sales workers], and C [Skilled Agricultural, Forestry and Fishery
Workers, Craft and Related Trades Workers and elementary workers])
2. Methods based on the Korean Standard Classification of Occupations (KSCO)
[43]. The self-completed questionnaires conducted to evaluate depres­
2.1. Participants sion, sleep quality, fatigue, and somatization were also administered
online.
The initial recruitment process employed both online and off-line The Korean version of a short form of the Center for Epidemiologic
methods, advertising on a bulletin board within the Samsung Medical Studies Depression Scale (CES–D) was used to assess the severity of
Center. Then, the recruited participants referred others through depressive symptoms [25]. The short-form of CES-D is composed of 11
Kakaotalk, the biggest social networking service in South Korea. Par­ items and has a similar construct to the original CES-D [26]. A higher
ticipants initially recruited had various professions, including health­ CES-D score indicates more severe depression [25]. The Korean version
care workers, police officers, drivers, firefighters, and factory workers. of the short-form of CES-D has good reliability and validity. Cronbach's α
However, among participants initially recruited, the proportion of coefficient for internal consistency was 0.838 [27].
young female shift workers were larger [448 men and 806 women; 961 The Korean version of the Pittsburgh Sleep Quality Index (PSQI) was
shift workers and 293 non-shift workers; 32.6 ± 7.9 years old). There­ used to assess sleep quality [28]. The PSQI includes 19 questions and
fore, an online survey company (Macromill Embrain Co., Ltd., Seoul, rates sleep problems in seven domains (i.e., sleep quality, sleep latency,
South Korea) recruited 5400 additional participants (2693 men and sleep duration, habitual sleep efficiency, sleep disturbances, use of sleep
2707 women; 3600 shift workers and 1800 non-shift workers; 38.3 ± medications, and daytime dysfunction) [28]. The summed scores of the
9.9 years old), with more men, middle-aged individuals, and non-shift seven domains provide an index of sleep quality [28]. The Korean
workers. The online survey company possessed its own research panel version of the PSQI has good reliability and validity. Cronbach's α co­
consisting of more than one million individuals. This panel was carefully efficient for internal consistency was 0.84. The test–retest correlation
balanced in terms of gender, age, and geographical distribution across coefficient was 0.65 [29].
South Korea. These individuals either voluntarily joined the panel or The Korean version of the Fatigue Severity Scale (FSS) was used to
were recruited through referrals from existing panel members. Their evaluate fatigue [13]. The FSS includes 9 items, which explore the motor
profiles were regularly updated on an annual basis, removing in­ aspects of fatigue, daytime impairment by fatigue, and vulnerability to
dividuals who were not contacted. Participants aged at least 18 years fatigue. A higher FSS score indicates more severe fatigue [13]. The
and working full or part time were included, whereas those who were Korean version of the FSS has good reliability and test-retest variability.
unable to complete the online surveys were excluded. Cronbach's a coefficient for internal consistency was 0.929, and the
Among the 6665 individuals identified, 11 were excluded because of Pearson correlation coefficient for test-retest was 0.916 [30].
difficulty in classifying their work schedule. In addition, 22 participants The Korean version of the somatization subscale of the Symptom
were excluded because they failed to complete the questionnaires. Checklist 90-item version (SCL-SOM) was used to evaluate somatization
Overall, 6632 participants were included in the study. There were no symptoms [31]. The SCL-SOM consists of 12 questions related to sub­
significant differences in age, gender, or proportion of shift workers jective distress of bodily dysfunction [31]. A higher SCL-SOM score in­
between the 33 excluded individuals and 6632 included participants. dicates more severe somatization [32]. The Korean version of the SCL-
According to the classification of previous studies [24,41,42], the SOM has acceptable reliability and validity. Cronbach's a coefficient
current study defined the shift workers with wide definition of shift for internal consistency was 0.72 [44].
workers. Workers whose schedule was rotating or outside of regular
working hours were defined as ‘shift worker’ group. The shift workers in 2.3. Statistical analyses
the current study includes participants whose schedule was evening or
night, those whose schedule was rotating, and those whose schedule is Initially, group differences between shift and non-shift workers were
casual or flexible. examined using the independent t-test, or ANCOVA. Pearson's correla­
The study procedures were conducted in accordance with the ethical tion was used to examine the association between continuous variables.
standards of the relevant national and institutional committees on Partial correlation was conducted to explore the associations between
human experimentation and the 1964 Declaration of Helsinki (2013 CES–D, PSQI, FSS, and SCL-90, after controlling for age, gender, med­
revision). All participants submitted written informed consent. The ical illness, occupational category, monthly income, length of service,
study protocol involving human subjects was approved by the Institu­ and weekly working hours. An analysis of covariates (ANCOVA) was
tional Review Board of Samsung Medical Center (protocol code: repeated using four different sets of covariates in a between-group
2019–04-095). comparison of the FSS and SCL-SOM. The first ANCOVA was conduct­
ed with age, gender, presence of medical illness, occupational category,
2.2. Questionnaires monthly income, length of service, and weekly working hours included
as covariates, whereas the second and third ANCOVAs added CES-D or
The questionnaires were self-completed by the participants online. PSQI to the covariates included in the first ANCOVA, respectively. The

2
J.W. Seo et al. Journal of Psychosomatic Research 173 (2023) 111467

final ANCOVA added CES-D and PSQI simultaneously to the covariates A) FSS
included in the first ANCOVA. 34

3. Results 33.5

3.1. Demographic and psychiatric characteristics 33

Participants who worked in rotating shifts or at night were classified

32.5
as shift workers (2135 men and 2408 women, 37.0 ± 9.8 years old).

FSS
Participants who worked regularly during the day were classified as
non-shift workers (998 men and 1091 women, 37.8 ± 9.7 years old). 32

There was no significant difference in gender composition, presence

of medical illness, and monthly income between shift workers and non- 31.5
shift workers. Shift workers were younger than non-shift workers (t =
2.99, p < 0.01). The proportion of occupational category of shift workers 31
and non-shift workers was different from each other (chi = 391.10, p <
0.01). The length of services of non-shift workers was longer than shift 30.5
workers (t = 9.72, p < 0.01). Shift workers worked more hours per week Total sample Shift workers Non-shift workers
than non-shift workers (t = − 8.42, p < 0.01). Shift workers reported
higher CES-D (t = − 10.34, p < 0.01) and PSQI (t = − 13.22, p < 0.01)
than non-shift workers (Table 1). B) SCL-SOM
8.5

3.2. Comparison of fatigue between shift and non-shift workers

8
Shift workers had a higher FSS score (t = − 7.42, p < 0.01) than non-
shift workers (Table 1, Fig. 1), which remained significant despite
controlling for age, gender, presence of medical illness, occupational
7.5
category, monthly income, length of service, and weekly working hours
SCL-SOM

(F = 48.46, p < 0.01). However, between-group difference of FSS scores

7
Table 1
Depression, sleep, somatization, and fatigue of shift workers and non-shift
workers.
6.5
Total Shift Non-shift
(N = workers workers
6632) (N = (N = 2089)
4543) 6
Total sample Shift workers Non-shift workers
mean ± mean ± mean ± SD t P
SD SD FSS = Fatigue Severity Scale, SCL-SOM = Symptom Checklist-90 somatization scale

Age (age) 37.3 ± 37.0 ± 37.8 ± 9.7 2.99 <0.01

Fig. 1. Comparison of fatigue and somatization between shift and non-
9.8 9.8
shift workers.
Gender (Women) 3499 2408 1091 chi = 0.56
(52.8%) (53.0%) (52.2%) 0.35
Medical illness 1555 1091 464 chi = 0.11 was no longer significant after additionally controlling for CES-D (F =
(23.4%) (24.0%) (22.2%) 2.59 2.12, p = 0.15), PSQI (F = 1.79, p = 0.18), or both CES-D and PSQI (F <
Occupational chi =
0.01, p = 0.95) (Table 2).
<0.01
category 391.10
A 4492 2728 1764
(67.7%) (60.0%) (85.5%)
B 1189 1002 187 (9.1%)
3.3. Comparison of somatization between shift and non-shift workers
(17.9%) (22.1%)
C 842 730 112 (5.4%) Shift workers reported higher SCL-SOM (t = − 8.92, p < 0.01) than
(12.7%) (16.1%) non-shift workers (Table 1, Fig. 1). Adjusted for age, gender, presence of
Monthly income 4.22 ± 4.22 ± 4.23 ± 0.54 0.59
medical illness, occupational category, monthly income, length of ser­
1.00 1.00 0.99
Length of service 76.47 ± 70.29 ± 89.98 ± 9.72 <0.01 vice, and weekly working hours, SCL-SOM was higher in shift workers
(months) 77.06 68.80 91.15 than non-shift workers (F = 67.07, p < 0.01). The higher SCL-SOM of
Weekly working 31.29 ± 32.68 ± 28.29 ± − 8.42 <0.01 shift workers showed only non-significant trends after additionally
hours 19.83 20.34 18.33 controlling for CES-D (F = 2.86, p = 0.09) or PSQI (F = 3.13, p = 0.07).
CES-D 8.23 ± 8.74 ± 30 7.10 ± − 10.34
After both CES-D and PSQI were additionally controlled, between-group
<0.01
6.20 5.84
PSQI 7.05 ± 7.42 ± 6.25 ± − 13.22 <0.01 difference of SCL-SOM disappeared (F < 0.01, p = 0.99) (Table 2).
3.52 3.59 3.23
FSS 34.09 ± 34.79 ± 32.59 ± − 7.42 <0.01
11.21 11.08 11.35 3.4. Predictors of fatigue and somatization
SCL-SOM 8.63 ± 9.19 ± 7.42 ± − 8.92 <0.01
7.90 8.14 7.20 Women exhibited higher FSS score than men (t = − 10.74, p < 0.01).
Abbreviation: CES-D = A short form of the Center for Epidemiological Studies FSS had significant negative correlations with age (r = − 0.11, p < 0.01),
Depression, PSQI = Pittsburgh sleep quality index, FSS = Fatigue Severity Scale, monthly income (r = − 0.07, p < 0.01), and length of service (r = − 0.10,
SCL-SOM = symptom checklist-90 somatization scale. p < 0.01). FSS had significant positive correlations with CES-D (r = 0.54,

3
J.W. Seo et al. Journal of Psychosomatic Research 173 (2023) 111467

Table 2
Comparison of depression, sleep, fatigue, and somatization between shift workers and non-shift workers after controlling covariates.
After controlling age, gender, After controlling age, gender, medical After controlling age, gender, medical After controlling age, gender, medical
medical illness, and job-related illness, job-related factors, and CES-D illness, job-related factors, and PSQI illness, job-related factors, CES–D, and
factors* PSQI

F P F P F P F P

CES-D 66.67 <0.01 N/A 11.17 <0.01 N/A

PSQI 110.60 <0.01 54.72 <0.01 N/A N/A
FSS 31.49 <0.01 2.12 0.15 1.79 0.18 <0.01 0.95
SCL-SOM 42.33 <0.01 2.86 0.09 3.13 0.07 <0.01 0.99

Abbreviation: CES-D = A short form of the Center for Epidemiological Studies Depression, PSQI = Pittsburgh sleep quality index, SCL-SOM = Symptom Checklist-90
somatization scale, FSS = Fatigue Severity Scale.
*
Job-related factors included occupational category, monthly income, length of service, and weekly working hours.

p < 0.01) and PSQI (r = 0.42, p < 0.01) after controlling for age, gender, Negative correlation between SCL-SOM and weekly working hours was
presence of medical illness, occupational category, monthly income, only observed in shift workers (r = − 0.05, p < 0.01), not in non-shift
length of service, and weekly working hours. Participants with medical workers (r = − 0.01, p = 0.72) (Table 4).
illness exhibited higher FSS score than those without (F = 31.10, p <
0.01), after controlling for age, gender, occupational category, monthly 4. Discussion
income, length of service, weekly working hours, CES–D, and PSQI.
Similar results were obtained when the analysis was repeated for both In the present study, shift workers showed increased fatigue, soma­
shift and non-shift workers (Table 3). tization, depression, and sleep disturbances, in line with previous
Female exhibited higher SCL-SOM score than male (t = − 6.64, p < studies [4,17,34]. The current study is the first to explore the effects of
0.01). SCL-SOM had negative correlations with age (r = − 0.05, p < depression and sleep on the increased fatigue and somatization in shift
0.01), monthly income (r = − 0.07, p < 0.01), length of service (r = workers. Increased fatigue and somatization in shift workers were
− 0.05, p < 0.01), and weekly working hours (r = − 0.03, p = 0.04). SCL- associated with increased depression and sleep disturbances.
SOM had positive correlation with CES-D (r = 0.64, p < 0.01) and PSQI In line with our hypothesis, shift workers had a higher level of fatigue
(r = 0.47, p < 0.01) after controlling for age, gender, presence of than non-shift workers. However, the higher fatigue in shift workers was
medical illness, occupational category, monthly income, length of ser­ associated with potential mediators. The current study suggested
vice, and weekly working hours. Participants with medical illness depression or sleep disturbances might be accountable for higher fatigue
exhibited higher SCL-SOM score than those without (F = 49.90, p < in shift workers. The current study supports the previous study sug­
0.01), after controlling for age, gender, occupational category, monthly gesting the association between poor sleep and fatigue [16] or between
income, length of service, weekly working hours, CES–D, and PSQI. In depression and fatigue [45] in shift workers. However, previous studies
the shift worker group, SCL-SOM was significantly different between did not examine whether the difference in fatigue between shift and non-
three occupational categories (A = 8.83, B = 9.53, C = 10.01, F = 7.30, shift workers would be associated with the difference in sleep and mood
p < 0.01). SCL-SOM of C occupational category was higher than A between shift and non-shift workers. The current study suggested the
occupational category (mean difference = 1.18 ± 0.34, p < 0.01). Such possibility that depression or poor sleep of shift workers could explain
difference between occupational categories was not observed in non- the difference of fatigue between two groups.
shift workers (A = 7.42, B = 7.58, C = 7.36, F = 0.05, p = 0.96). The current study found a higher somatization level in shift workers
than non-shift workers. The increased somatization of shift workers

Table 3
Predictors of fatigue in shift workers and non-shift workers. Table 4
Predictors of Somatization in shift workers and non-shift workers.
Total sample Shift workers Non-shift workers
Total sample Shift workers Non-shift workers
r p- r p- r p-
value value value r p- r p- r p-
value value value
Age − 0.11 <0.01 − 0.10 <0.01 − 0.10 <0.01
Gender T= <0.01 T= <0.01 T= <0.01 Age − 0.05 <0.01 − 0.04 0.01 − 0.05 0.03
-10.74 -9.81 -4.71 Gender T= <0.01 T= <0.01 T= 0.01
Medical illness* F= <0.01 F= <0.01 F= <0.01 -6.64 -6.17 -2.58
31.10 15.82 16.95 Medical illness* F= <0.01 F= <0.01 F= <0.01
Occupational F= 0.83 F= 0.21 F= 0.31 49.90 40.09 9.09
category 0.19 1.58 1.19 Occupational F= <0.01 F= <0.01 F= 0.96
Monthly income − 0.07 <0.01 − 0.05 <0.01 − 0.10 <0.01 category 15.03 7.30 0.05
Length of service − 0.10 <0.01 − 0.05 <0.0 − 0.15 <0.01 Monthly income − 0.07 <0.01 − 0.07 <0.01 − 0.06 0.01
[1] Length of service − 0.05 <0.01 − 0.03 0.05 − 0.05 0.02
Weekly working 0.02 0.19 0.01 0.71 0.01 0.67 Weekly working − 0.03 0.04 − 0.05 <0.01 − 0.01 0.72
hours hours
CES-D** 0.54 <0.01 0.54 <0.01 0.53 <0.01 CES-D** 0.64 <0.01 0.65 <0.01 0.61 <0.01
PSQI** 0.42 <0.01 0.42 <0.01 0.40 <0.01 PSQI** 0.47 <0.01 0.51 <0.01 0.44 <0.01

Abbreviation: FSS = Fatigue Severity Scale, CES-D = A short form of the Center Abbreviation: FSS = Fatigue Severity Scale, CES-D = A short form of the Center
for Epidemiological Studies Depression, PSQI = Pittsburgh sleep quality index. for Epidemiological Studies Depression, PSQI = Pittsburgh sleep quality index.
* *
The analysis was conducted adjusting for age, gender, presence of medical The analysis was conducted adjusting for age, gender, presence of medical
illness, occupational category, monthly income, length of service, weekly illness, occupational category, monthly income, length of service, weekly
working hours, CES–D, and PSQI. working hours, CES–D, and PSQI.
** **
The analysis was conducted adjusting for age, gender, presence of medical The analysis was conducted adjusting for age, gender, presence of medical
illness, occupational category, monthly income, length of service, and weekly illness, occupational category, monthly income, length of service, and weekly
working hours. working hours.

4
J.W. Seo et al. Journal of Psychosomatic Research 173 (2023) 111467

disappeared after controlling for depression or disturbed sleep. This interest, financial or otherwise(s).
result supports previous studies, suggesting shift workers have more
somatization [17,22,23]. However, those studies didn't consider the Funding
potential risk factors such as depression and sleep, and what makes shift
workers' somatization higher than non-shift workers. The current study This research was supported by the National Research Foundation of
revealed that somatization of shift workers could be explained by their Korea (NRF) grant funded by the Korean government (MSIT) (No.
depression and poor sleep. 2020R1F1A1049200, 2022R1A2C2008417), the Bio & Medical Tech­
The mechanism underlying the interdependent effects of mood and nology Development Program of the National Research Foundation
sleep on physical discomfort is unclear, but it may involve alterations to (NRF) funded by the Korean government (MSIT) (No.
the immune system. Sleep disturbance induced by circadian misalign­ 2020M3E5D9080561), and a grant of the Korea Health Technology R&D
ment in shift workers might disrupt the immune system [35], resulting Project through the Korea Health Industry Development Institute
in mood changes and physical discomfort [36]. Another study suggested (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea
that genetic factors are a possible mediator, and a common genetic (No. HR21C0885).
factor might account for depression and fatigue [37].
The current study found that difference of fatigue and somatization
between shift workers and non-shift workers disappeared when Declaration of Competing Interest
depression and sleep were controlled. In addition, correlation analysis
showed that both fatigue and somatization shared potential associated The authors declare no conflict of interest.
factors, such as age, length of service, depression, and poor sleep. These
result may supports previous studies suggesting that the mechanisms of Acknowledgments
fatigue and somatization could be overlapped [18–20]. The previous
studies suggested that inflammation might be common mechanism for This research was supported by the National Research Foundation of
both fatigue and somatization [18,46], although it remains unclear Korea (NRF) grant funded by the Korean government (MSIT) (No.
which inflammation process would be shared for fatigue and somati­ 2020R1F1A1049200, 2022R1A2C2008417), the Bio & Medical Tech­
zation of shift workers. nology Development Program of the National Research Foundation
Despite the lack of evidence regarding a causal relationship, the (NRF) funded by the Korean government (MSIT) (No.
assumption that fatigue and somatization are associated with depression 2020M3E5D9080561), and a grant of the Korea Health Technology R&D
and sleep disturbances in shift workers may have some clinical impli­ Project through the Korea Health Industry Development Institute
cations. Early interventions to treat sleep and mood problems in shift (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea
workers would help prevent or reduce somatization and fatigue. In (No. HR21C0885).
addition, when shift workers complain of fatigue and somatization, their
sleep and mood problems should be carefully explored and managed. References
The presence of medical illness was also associated with fatigue and
somatization, even after depression and poor sleep were controlled. [1] Humans IWGotIoCHt, IARC Monographs on the Identification of Carcinogenic
Although it was statistically insignificant, shift workers had medical Hazards to Humans. Night Shift Work, International Agency for Research on
Cancer © International Agency for Research on Cancer, Lyon (FR), 2020. For more
illness more commonly than non-shift workers. Medical illness of shift information contact [email protected].; 2020.
workers could contribute to more fatigue and somatization of them. [2] S.Z. Aemmi, E. Mohammadi, H. Heidarian-Miri, M. Fereidooni-Moghadam,
However, the medical illness could not explain the associations between H. Boostani, K. Zarea, The effectiveness of bright light exposure in shift-worker
nurses: a systematic review and meta-analysis, Sleep Sci 13 (2020) 145–151.
depression/sleep and fatigue/somatization, as these association
[3] T. Alterman, S.E. Luckhaupt, J.M. Dahlhamer, B.W. Ward, G.M. Calvert, Prevalence
remained significant after controlling for the presence of medical illness. rates of work organization characteristics among workers in the U.S.: data from the
The present study had some limitations. First, this was a cross- 2010 National Health Interview Survey, Am. J. Ind. Med. (56) (2013) 647–659.
[4] K. Richter, J. Acker, S. Adam, G. Niklewski, Prevention of fatigue and insomnia in
sectional study, which made it difficult to determine whether or not
shift workers-a review of non-pharmacological measures, EPMA J. 7 (2016) 16.
the relationships between the factors are causal. Therefore, future [5] M. Härmä, K. Karhula, S. Puttonen, A. Ropponen, A. Koskinen, A. Ojajärvi,
studies with a longitudinal design are needed. Second, the data collected M. Kivimäki, Shift work with and without night work as a risk factor for fatigue and
by online self-reported questionnaires may include inappropriate data changes in sleep length: a cohort study with linkage to records on daily working
hours, J. Sleep Res. 28 (2019), e12658.
although online survey program and company checked the authenticity [6] J. Shen, L.C. Botly, S.A. Chung, A.L. Gibbs, S. Sabanadzovic, C.M. Shapiro, Fatigue
of the data. Third, this study only included Korean participants. How­ and shift work, J. Sleep Res. 15 (2006) 1–5.
ever, the study participants worked in various fields of work. It is unclear [7] S.C. Yuan, M.C. Chou, C.J. Chen, Y.J. Lin, M.C. Chen, H.H. Liu, H.W. Kuo,
Influences of shift work on fatigue among nurses, J. Nurs. Manag. 19 (2011)
whether our findings may be applicable to shift workers in other cultures 339–345.
and societies. Fourth, only presence or absence of medical illness was [8] T. Akerstedt, K.P. Wright Jr., Sleep loss and fatigue in shift work and shift work
analyzed as a binary variable. Although a variety of medical illness could disorder, Sleep Med. Clin. 4 (2009) 257–271.
[9] L. Torquati, G.I. Mielke, W.J. Brown, N.W. Burton, T.L. Kolbe-Alexander, Shift
be taken into account, the severity and the number of medical illness work and poor mental health: a meta-analysis of longitudinal studies, Am. J. Public
could not be considered in the current study. Finally, the current study Health 109 (2019) e13–e20.
explored only two core psychiatric features (i.e., depression and sleep [10] K.P. Wright Jr., R.K. Bogan, J.K. Wyatt, Shift work and the assessment and
management of shift work disorder (SWD), Sleep Med. Rev. 17 (2013) 41–54.
disturbances) of shift workers. Future studies should focus on additional
[11] Association AP, Depressive Disorders: DSM-5®, Selections: American Psychiatric
individual features (e.g., personality, chronotype, and genetic poly­ Pub, 2015.
morphisms) to identify other mediating factors. [12] I. Bozoky, E.J. Corwin, Fatigue as a predictor of postpartum depression, J. Obstet.
Gynecol. Neonatal. Nurs. 31 (2002) 436–443.
In conclusion, fatigue and somatization were greater in shift workers
[13] L.B. Krupp, N.G. LaRocca, J. Muir-Nash, A.D. Steinberg, The fatigue severity scale.
compared to non-shift workers, but not after controlling for sleep quality Application to patients with multiple sclerosis and systemic lupus erythematosus,
or depression. Our findings suggest that the higher fatigue and somati­ Arch. Neurol. 46 (1989) 1121–1123.
zation levels in shift workers are associated with depression and [14] K.A. Rimes, R. Goodman, M. Hotopf, S. Wessely, H. Meltzer, T. Chalder, Incidence,
prognosis, and risk factors for fatigue and chronic fatigue syndrome in adolescents:
disturbed sleep. a prospective community study, Pediatrics 119 (2007) e603-e9.
[15] M. Lavidor, A. Weller, H. Babkoff, How sleep is related to fatigue, Br. J. Health
Disclosure statement Psychol. 8 (2003) 95–105.
[16] N.M. Øyane, S. Pallesen, B.E. Moen, T. Akerstedt, B. Bjorvatn, Associations
between night work and anxiety, depression, insomnia, sleepiness and fatigue in a
To the best of our knowledge, the named authors have no conflict of sample of Norwegian nurses, PLoS One 8 (2013), e70228.

5
J.W. Seo et al. Journal of Psychosomatic Research 173 (2023) 111467

[17] Y. Selvi, P.G. Özdemir, O. Özdemir, A. Aydın, L. Beşiroğlu, Influence of night shift [32] X. Chen, P. Li, F. Wang, G. Ji, L. Miao, S. You, Psychological results of 438 patients
work on psychologic state and quality of life in health workers, Dusunen Adam J. with persisting gastroesophageal reflux disease symptoms by symptom checklist
Psychiat. Neurol. Sci. 23 (2010) 238. 90-revised questionnaire, Euroasian J. Hepato-gastroenterol. 7 (2017) 117.
[18] M. Maes, Inflammatory and oxidative and nitrosative stress pathways [34] E.C. Corfield, N.G. Martin, D.R. Nyholt, Co-occurrence and symptomatology of
underpinning chronic fatigue, somatization and psychosomatic symptoms, Curr. fatigue and depression, Compr. Psychiatry 71 (2016) 1–10.
Opin. Psychiatry 22 (2009) 75–83. [35] R.M. Benca, J. Quintans, Sleep and host defenses: a review, Sleep 20 (1997)
[19] J.Y. Jun, S.J. Kim, Y.-J. Lee, S.-J. Cho, Effect of major depressive disorder and 1027–1037.
insomnia on somatization, Sleep Med. Psychophysiol. 19 (2012) 84–88. [36] C.-H. Lee, F. Giuliani, The role of inflammation in depression and fatigue, Front.
[20] W. Katon, A. Kleinman, G. Rosen, Depression and somatization: a review, Part I. Immunol. (2019) 1696.
Am J Med 72 (1982) 127–135. [37] I. Hickie, K. Kirk, N. Martin, Unique genetic and environmental determinants of
[22] M.-K. Shin, H.L. Kang, Effects of emotional labor and occupational stress on prolonged fatigue: a twin study, Psychol. Med. 29 (1999) 259–268.
somatization in nurses, J. Korean Acad. Nurs. Adm. 17 (2011) 158–167. [38] S.B. Harvey, S. Wessely, D. Kuh, M. Hotopf, The relationship between fatigue and
[23] I. Yun, S.-G. Kim, S. Jung, S. Borhanian, A study on police stressors, coping psychiatric disorders: evidence for the concept of neurasthenia, J. Psychosom. Res.
strategies, and somatization symptoms among south Korean frontline police 66 (2009) 445–454.
officers, Policing: An Int. J. Pol. Strat. Manag. 36 (2013) 787–802. [39] G. Van der Leeuw, M. Gerrits, B. Terluin, M. Numans, C. van der Feltz-Cornelis,
[24] A. Knutsson, Methodological aspects of shift-work research, Chronobiol. Int. 21 H. van der Horst, B. Penninx, H. van Marwijk, The association between
(2004) 1037–1047. somatization and disability in primary care patients, J. Psychosom. Res. 79 (2015)
[25] L.S. Radloff, The CES-D scale: a self-report depression scale for research in the 117–122.
general population, Appl. Psychol. Meas. 1 (1977) 385–401. [40] G.E. Simon, O. Gureje, Stability of somatization disorder and somatization
[26] Z.D. Gellis, Assessment of a brief CES-D measure for depression in homebound symptoms among primary care patients, Arch. Gen. Psychiatry 56 (1999) 90–95.
medically ill older adults, J. Gerontol. Soc. Work. 53 (2010) 289–303. [41] A.S. Rivera, M. Akanbi, L.C. O’Dwyer, M. McHugh, Shift work and long work hours
[27] E.-Y. Park, J.-H. Kim, Unidimensionality and measurement invariance of the 11- and their association with chronic health conditions: a systematic review of
item Korean CES-D scale among examining parents of individuals with cerebral systematic reviews with meta-analyses, PLoS One 15 (2020), e0231037.
palsy, J. Child Fam. Stud. 29 (2020) 895–903. [42] S. Ijaz, J. Verbeek, A. Seidler, M.-L. Lindbohm, A. Ojajärvi, N. Orsini, G. Costa,
[28] D.J. Buysse, C.F. Reynolds 3rd, T.H. Monk, S.R. Berman, D.J. Kupfer, The K. Neuvonen, Night-shift work and breast cancer—a systematic review and meta-
Pittsburgh sleep quality index: a new instrument for psychiatric practice and analysis, Scand. J. Work Environ. Health (2013) 431–447.
research, Psychiatry Res. 28 (1989) 193–213. [43] Korean Standard Classification of Occupations, Available from, http://kssc.kostat.
[29] S.I. Sohn, D.H. Kim, M.Y. Lee, Y.W. Cho, The reliability and validity of the Korean go.kr/.
version of the Pittsburgh sleep quality index, Sleep Breath. 16 (2012) 803–812. [44] J.H. Kim, S.Y. Lee, C.H. Kim, S.L. Lim, J.N. Shin, W.H. Chung, B.H. Yu, S.H. Hong,
[30] K.-I. Chung, C.-H. Song, Clinical usefulness of fatigue severity scale for patients Reliability and validity of a Korean adaptation of the tinnitus handicap inventory,
with fatigue, and anxiety or depression, Korean J. Psychosom. Med. 9 (2001) Kor. J. Otorhinolaryngol. Head Neck Surg. 45 (2002) 328–334.
164–173. [45] W.A.A. Khan, R. Conduit, G.A. Kennedy, M.L. Jackson, The relationship between
[31] L.R. Derogatis, SCL-90-R: administration, scoring & procedures manual-II for the shift-work, sleep, and mental health among paramedics in Australia, Sleep Health 6
(revised) version and other instruments of the psychopathology rating scale series, (2020) 330–337.
Clin. Psychom. Res. (1992) 1–16. [46] G. Anderson, M. Berk, M. Maes, Biological phenotypes underpin the physio-somatic
symptoms of somatization, depression, and chronic fatigue syndrome, Acta
Psychiatr. Scand. 129 (2014) 83–97.