RESEARCH ARTICLE

The impact of occupational psychological

hazards and metabolic syndrome on the 8-
year risk of cardiovascular diseases—A
longitudinal study
Wei-Liang Chen1, Chung-Ching Wang1, Sheng-Ta Chiang1, Ying-Chuan Wang1, Yu-
Shan Sun1, Wei-Te Wu2*, Saou-Hsing Liou1,2,3*
a1111111111 1 Division of occupational medicine, Division of fanily medicine, Department of Family and Community
a1111111111 Medicine, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan, Republic of China,
a1111111111 2 National Institute of Environmental Health Science, National Health Research Institutes, Miaoli, Taiwan,
a1111111111 Republic of China, 3 Department of Public Health, National Defense Medical Center, Taipei, Taiwan,
Republic of China
a1111111111
* ader.una@gmail.com (WTU); shliou@nhri.org.tw (SHL)

OPEN ACCESS Abstract

Citation: Chen W-L, Wang C-C, Chiang S-T, Wang There was little information concerning the combined effect of occupational psychosocial
Y-C, Sun Y-S, Wu W-T, et al. (2018) The impact of
hazards such as long working hours, high job stress, and high fatigue on the risk of cardio-
occupational psychological hazards and metabolic
syndrome on the 8-year risk of cardiovascular vascular and cerebrovascular diseases (CVD). The aim of this study was to investigate the
diseases—A longitudinal study. PLoS ONE 13(8): interaction among occupational psychosocial hazards and the impact of metabolic syn-
e0202977. https://doi.org/10.1371/journal. drome (MetS) on the risk of CVD among bus drivers. The Taiwan Bus Driver Cohort Study
pone.0202977
involving 1014 professional drivers was established in 2005 and comprehensively studied.
Editor: Antonio Palazón-Bru, Universidad Miguel The interactions among occupational psychosocial hazards and the impact of MetS on the
Hernandez de Elche, SPAIN
risk of CVD were measured. A working pattern questionnaire, job stress questionnaires, the
Received: January 24, 2018 Swedish occupational fatigue inventory, the stress satisfaction offset score, biochemical
Accepted: August 13, 2018 measurements, and physical examinations were used to assess psychosocial hazards and
Published: August 27, 2018 the presence of metabolic syndrome. There were 707 eligible bus drivers with a mean age
of 43.5years old. During the 8-years of follow-up, 77 drivers were diagnosed with CVD. Long
Copyright: © 2018 Chen et al. This is an open
access article distributed under the terms of the working hours, high job stress, and high fatigue were associated with an increased risk of
Creative Commons Attribution License, which cardiovascular disease incidence in the multivariate analysis. There were synergistic effects
permits unrestricted use, distribution, and among long working hours, high job stress, and high fatigue only in drivers with MetS. A
reproduction in any medium, provided the original
author and source are credited.
combination of long working hours, high job stress, and high fatigue increased the risk of
developing CVD in bus drivers with MetS.
Data Availability Statement: The data set is
owned by the Institutional Review Board (IRB) of
Tri-Service General Hospital (TSGH). TSGH IRB
only approved the data analysis in our study and
did not approve data sharing. Therefore, we do not
have permission to share the data set. Interested Introduction
researchers can submit data access requests to the
The prevention of cardiovascular and cerebrovascular diseases (CVD) was still a major public
Tri-Service General Hospital IRB using the
following email address: tsghirb@ndmctsgh.edu.
health challenge worldwide despite modern medical advances. Psychosocial factors at work
tw. The authors did not have special access and health behaviors played a pivotal role in the pathogenesis and progression of CVD.
privileges. Recently, research efforts had been increasingly concerned about the effect of occupational

PLOS ONE | https://doi.org/10.1371/journal.pone.0202977 August 27, 2018 1 / 12

 Association between occupational psychological hazards and metabolic syndrome

Funding: Funded by National Health Research hazards on CVD. In a meta-analysis of 603,838 individuals, long working hours (55 hours
Institutes of Taiwan (98-EO-PP01, 99-EO-PP01, per week) were associated with a 13% increase in coronary heart disease risk and a 33%
00-EO-PP01, EO-101-PP-01, EO-102-PP-01 and
increase in stroke risk [1]. In the Nurses’ Health Study, rotating night shift work was associated
EO-103-PP-01) to Dr. Saou-Hsing Liou National
Institute of Occupational Safety and Health, Japan with a 4% increased risk of ischemic stroke for every 5 years[2]. Previously, 13 European epide-
(IOSH96-M102 and IOSH97-M102) to Dr. Saou- miologic cohort studies had reported that the association between job strain and coronary
Hsing Liou. heart disease was significant for men and women, including those younger and older than 50
Competing interests: The authors have declared years of age, and at all levels of socioeconomic status[3]. Based on abovementioned evidence,
that no competing interests exist. overwork was a multifactorial, work-related psychosocial factor that interacted with environ-
mental and lifestyle factors to cause CVD or karoshi. In the transportation industry, concur-
rent individual risk factors and work environmental risk factors among professional drivers
contributed to cumulative CVD impact, such as age, metabolic syndrome, smoking, sedentary
work, unconventional and irregular hours, shift work, long working hours, a tight delivery
schedule, and job strain. Among approximately 120,000 Japanese workers from a national rep-
resentative insurance organization, several business categories, including transportation, con-
struction, and mining had the highest prevalence for metabolic syndrome (MetS) at 25.7%,
21.0%, and 20.5%, respectively[4]. Among them, metabolic Syndrome (MetS) is a cluster of
metabolic abnormalities and is considered as a multi-morbid condition.
Metabolic syndrome (MetS), characterized by central obesity, glucose intolerance, athero-
genic dyslipidemia, and fluctuations in blood pressure, was strongly associated with increased
risk of coronary heart disease, cerebrovascular diseases, diabetes, and mortality[5, 6]. Patients
with MetS were associated with a 2-fold increase in cardiovascular outcomes and a 1.5-fold
increase in all-cause mortality compared with those without MetS[7]. Moreover, previous Nor-
wegian HUNT Study mentioned that the association between metabolic syndrome and the
risk of death from cardiovascular disease was stronger among persons with sedentary work
and physically heavy work compared with persons with a lot of walking/lifting at work[8].
These findings implied that occupational physical activity or job categories may moderate the
detrimental effects of MetS on CVD. Despite the tight link between MetS and CVD risk, little
empirical evidence had been found that the psychosocial factors in workplace play which role
in the development of MetS and CVD.
Importantly, light could be shed on psychosocial risks in workplace which was still largely
uninvestigated. Therefore, studies on whether or not psychosocial factors contributes to or
exacerbates CVD with/without MetS, could support a novel target for cardiovascular risk
reduction in bus drivers. Thus, we performed a perspective cohort study to investigate the
interaction of occupational psychosocial hazards with the risk of CVD among drivers with
MetS.

Methods
Ethics statement
Our study protocol was approved by the institutional review board (IRB) of the National
Health Research Institutes and Tri-Service General Hospital in accordance with the revised
Helsinki Declaration. All subjects completed consent forms and agreed to participate in the
survey after a detailed explanation of the protocol of the study.

Study population
A total of 1,650 professional drivers from a largest transportation company were recruited in
the Taiwan Bus Driver Cohort Study (TBDCS). After completion of informed consents, all
participants received the basic and working related questionnaire and biochemistry tests. This
cohort was linked to the Driving Hours Dataset during the period from 2005 to 2007[9]. At

PLOS ONE | https://doi.org/10.1371/journal.pone.0202977 August 27, 2018 2 / 12

 Association between occupational psychological hazards and metabolic syndrome

baseline, all 1,650 eligible drivers were followed up at the time of the evaluation and received
comprehensive examinations. Our exclusion criteria included previous CVD history (n = 85),
driving days < 100 days (n = 613), missing information on demographics (n = 23), and bio-
chemical examinations (n = 222). To ensure some homogeneity of social background, all sub-
jects were selected from the same bus company. Finally, 707 eligible drivers were followed up
from 2005 to 2012.

Definition of metabolic syndrome

The diagnosis of metabolic syndrome (MetS) was defined as the presence of three or more of
the following components according to the Taiwan Health Promotion Administration of the
Ministry of Health and Welfare in 2007[10]. Among the MetS criteria, the first component
was waist circumference, which was replaced by BMI >27 kg/m2 in our study. Based on the
receiver operating characteristic curve analysis from the Taiwan National Nutrition and
Health Survey, a BMI cutoff of 27 kg/m2 was used to predict abdominal visceral obesity and
waist circumference of 90 cm for men and 80 cm for women[11], [12]. The second component
of MetS was blood pressure, which was determined to be abnormal by a systolic BP130 mm
Hg, diastolic BP85 mm Hg, or self-reported hypertension. The third component was serum
triglycerides (TGs), which was classified abnormal as serum TGs150 mg/dL (1.7 mmol/L).
The fourth component was serum fasting glucose based on serum fasting glucose 100 mg/dL
or self-reported diabetes mellitus. The last component represented decreased serum high-den-
sity lipoprotein cholesterol (HDL) according to HDL <40 mg/dL (1.03 mmol/L) for men and
<50 mg/dL (1.3 mmol/L) for women.

Job stress assessment

We used the Chinese version of the Job Content Questionnaire (JCQ) to assess the psychoso-
cial job characteristics of all participants[13]. The scores of the validated JCQ were based on
the Demand-Control-Support model established by Karasek and Theorell[14]. Among the 22
items, the JCQ had a minimum set of questions for the assessment of the four factors, includ-
ing job control, psychological demands, supervisor support, and co-worker support. A psycho-
logical demand score greater than 30 was classified as high job demand, and a decision latitude
score that was greater 60 was categorized as high job control. The definition of high strain
included high job demand and low job control, whereas the category of low strain was com-
posed of low job strain (low job demand and high job control), an active job (high job demand
and high job control), and a passive job (low job demand and low job control).

Workplace stress evaluation tools

The questionnaires of Stress Satisfaction Offset Score (SSOS), a four-item scale created by
Shain, is commonly used to evaluate the influence of workplace stressors in the workers. The
four domains of SSOS include Demand Control model and the Effort Reward Imbalance
model of workplace stress[15, 16]. There are control, effort, reward, and demand in the ques-
tionnaires. In terms of scoring scale in the SSOS questionnaires, total score ranges from -2
through to +2. A negative SSOS indicates that stress outweighs satisfaction, whereas a positive
SSOS indicates that satisfaction outweighs stress[17].

Dimensional assessment of fatigue questionnaire

The Swedish Occupational Fatigue Inventory (SOFI) is commonly used to assess the five
domains of work-related perceived fatigue, including lack of energy, physical exertion, physical

PLOS ONE | https://doi.org/10.1371/journal.pone.0202977 August 27, 2018 3 / 12

 Association between occupational psychological hazards and metabolic syndrome

discomfort, lack of motivation, and sleepiness[18]. The participants scores their perceived
intensity of fatigue based on 25 items.

Sociodemographic variables
All eligible subjects were interviewed using a structured questionnaire, including questions
regarding demographics (age, marital status, and education status), work conditions (work
hours, years of driving experience, frequency of driving more than 12 hours a day, and schedule
of rotating shifts), lifestyle habits (smoking, betel nut chewing, and drinking), and past medical
histories. Long working hours were analyzed as a categorical variable (driving hours more than
12 hour/day and less than or equal to 12 hour/day). Body weight and height were evaluated
using a digital scale, and BMI level was calculated as the subject’s weight in kilograms divided
by the square of their height in meters (kg/m2). We used a digital automatic sphygmomanome-
ter to measure systolic and diastolic blood pressure in a sitting position after 10 to 15 minutes of
rest. The blood testing was collected after 8 to 10 hours of overnight fasting to determine serum
levels of high-density lipoprotein cholesterol, triglyceride and fasting blood glucose.

Outcome measures
Each eligible subject was traced from the date of the baseline evaluation to the development of
CVD or the end of the follow-up. The primary endpoint in our study was incident CVD,
including ischemic heart disease, congestive heart failure, dissecting aneurysm of the aorta,
serious cardiac arrhythmia, and cerebrovascular disease (ICD9: 390–459 and ICD10: I00-I99)
based on the recognition criteria for overwork-related CVD in Taiwan [19, 20].

Statistical analysis
All statistical analyses were examined using SPSS (Version 18.0 for Windows, SPSS, Inc., Chi-
cago, IL, USA). Two-sided P-values < 0.05 were used to indicate significant differences. Con-
cerning the association analysis, Cox proportional hazards regression was used to investigate
the hazard ratios (HRs) for cardiovascular and cerebrovascular disease incidence. The poten-
tial confounders adjusted in the HRs analysis included age, BMI, education, drinking, smok-
ing, and exercise. Five models were used to clarify the correlations among metabolic
syndrome, work-related factors, and CVD incidences. Dichotomized occupational psychoso-
cial hazards in theses analyses were long working hour (driving hour>12 hour/day vs 12
hour/day), high job stress (SSOS score < +2 vs +2), and high fatigue (SOFI 3.5 vs < 3.5). The
five models analyzed include: Model 1: a single occupational psychosocial hazard was investi-
gated for CVD incidences; Model 2: long working hours and high job stress were investigated
for CVD incidence; Model 3: long working hours and high fatigue were investigated for CVD
incidence; Model 4: high job stress and high fatigue were investigated for CVD incidences.
Model 5: three variables, including long working hours, high job stress, and high fatigue, were
investigated for CVD incidence. Finally, sensitivity analyses were performed to determine the
robustness of the findings using different exclusion criteria. In the sensitivity analyses (S1 and
S2 Tables) of 778 drivers, the exclusion definition of previous CVD history did not include
hypertension (n = 71).

Results
Baseline characteristics
Demographic characteristics and biochemical indices were presented in Table 1. At study
baseline, there were 327 drivers with MetS and 380 drivers without MetS (Non-MetS). The

PLOS ONE | https://doi.org/10.1371/journal.pone.0202977 August 27, 2018 4 / 12

 Association between occupational psychological hazards and metabolic syndrome

Table 1. Demographic characteristics and biochemical indices among bus drivers and stratified by metabolic syndrome (n = 707).
Variables All drivers (N = 707) MetS (N = 327) Non-MetS (N = 380) p-value
Continuous variables
Age (years), mean (SD) 43.5 (6.9) 43.8 (6.5) 43.2 (7.2) 0.285
BMI (kg/m2), mean (SD) 26.1 (4.0) 28.4 (3.8) 24.1 (3.0) <0.001
Systolic blood pressure (mmHg), mean (SD) 126.9 (13.1) 132.2 (12.2) 122.4 (12.1) <0.001
Diastolic blood pressure (mmHg), mean (SD) 80.9 (10.3) 85.1 (10.0) 77.3 (9.1) <0.001
Total cholesterol (mg/dl), mean (SD) 203.5 (38.0) 210.1 (38.0) 197.8 (37.1) <0.001
Triglyceride (mg/dl), mean (SD) 209.9 (181.3) 287.1 (226.0) 143.6 (87.9) <0.001
HDL cholesterol (mg/dl), mean (SD) 37.7 (9.3) 33.6 (7.0) 41.2 (9.7) <0.001
Fasting blood glucose (mg/dl), mean (SD) 102.7 (41.5) 115.2 (53.2) 91.9 (22.7) <0.001
Categorical variables
CVD after follow-up 0.012
Yes 77(10.9) 46(14.1) 31(8.2)
No 630(90.1) 281(85.9) 349(91.8)
Marital status, n (%) 0.899
Unmarried 123 (17.4) 59 (18.0) 64 (16.8)
Married 506 (71.2) 233 (71.3) 273 (71.8)
Others 78 (11.4) 35 (10.7) 43 (11.4)
Education, n (%) 0.563
 Junior high school 206 (29.1) 93 (28.4) 113 (29.7)
Senior high and vocational school 451 (63.8) 214 (65.4) 237 (62.4)
University and College 50 (7.1) 20 (6.2) 30 (7.9)
Cigarette smoking, n (%) 0.033
Never smokers 241 (34.1) 99 (30.3) 142 (37.4)
Ex-smokers 47 (6.6) 17 (5.2) 30 (7.9)
Current smokers 414 (59.3) 207 (64.5) 207 (54.7)
Missing n = 5
Alcohol use, n (%) 0.489
No 557 (82.4) 253 (77.4) 304 (80.0)
Yes 146 (17.6) 71 (22.6) 75 (20.0)
Missing n = 4
Betel nut chewing 0.110
No 583 (82.5) 262 (80.1) 321 (84.5)
Yes 123 (17.5) 65 (19.9) 58 (15.5)
Missing n = 1
Moderate exercise, n (%) 0.159
No 516 (72.9) 248 (75.8) 268 (70.5)
Yes 181 (27.1) 76 (24.2) 105 (29.5)
Missing n = 10
Working hour, n (%) 0.134
 12 hr 424 (59.9) 186 (56.9) 238 (62.6)
> 12 hr 278 (40.1) 138 (43.1) 140 (37.4)
Missing n = 5
SSOS, n (%) 0.966
+ 2 ~ +1 161 (22.8) 66 (20.2) 95 (25.0)
-2 ~ 0 544 (77.2) 261 (79.8) 283 (75.0)
Missing n = 2
SOFI, n (%) 0.610
(Continued)

PLOS ONE | https://doi.org/10.1371/journal.pone.0202977 August 27, 2018 5 / 12

 Association between occupational psychological hazards and metabolic syndrome

Table 1. (Continued)

Variables All drivers (N = 707) MetS (N = 327) Non-MetS (N = 380) p-value

< 3.0 316 (44.7) 143 (43.7) 173 (45.5)
3.0 390 (55.3) 184 (56.3) 206 (54.2)
Missing n = 1
Shift work, n (%) 0.083
No 127 (17.9) 50 (15.3) 77 (20.3)
Yes 579 (82.1) 277 (84.7) 302 (79.5)
Missing n = 1

SD, standard deviation; MetS, metabolic syndrome; BMI, body mass index; HDL, high-density lipoprotein; SSOS, Stress Satisfaction Offset Score; SOFI, Swedish
Occupational Fatigue Inventory

https://doi.org/10.1371/journal.pone.0202977.t001

mean age of participants with MetS and Non-MetS was 43.8 and 43.2 years old, respectively.
With comparison to participants with non-MetS, participants with MetS were older and had
higher mean BMI, systolic blood pressure, diastolic blood pressure, total cholesterol, triglycer-
ide, and fasting blood glucose.

Metabolic syndrome, psychosocial hazards, and CVD

During the 8-years of follow-up, there were 77 drivers (10.9%) developed CVD (Table 1). The
crude hazard ratio among the individuals with MetS was 1.77 (P value = 0.014) compared to
those without MetS (Table 2). For long working hours, working hour >12 hour/ day was sig-
nificantly associated with an increased risk of CVD incidence in the multivariate analysis
(HR = 1.76, 95% CI = 1.11–2.79). In terms of high job stress, risk for CVD incidence was not
significantly increased among the drivers with high job stress (multivariate analysis HR = 1.20,
95% CI = 0.67–2.14) when compared with those with low job stress (scores of SSOS +2~+1).
We observed a borderline significant trend association between high fatigue (elevated scores of
SOFI) and the risk of CVD.

Table 2. Unadjusted and adjusted hazard ratios of work patterns and psychosocial hazards on CVD in bus drivers (n = 707).
Univariate Analysis Multivariate Analysis a.
b
Independent variables HR 95% CI p-value HR 95% CI p-value
1 Metabolic syndrome (Ref.< No)
Yes 1.77 1.12 2.79 0.014 1.38 0.80 2.36 0.247
2 Long working hours (Ref. Working hour 12 hour/ day)
Working hour >12 hour/ day 1.93 1.23 3.02 0.004 1.76 1.11 2.79 0.016
3 Shift work (Ref. No shift work)
Shift work 0.74 0.43 1.28 0.281 0.68 0.39 1.18 0.172
4 Job Content Questionnaire (Ref. Low strain)
High strain 1.17 0.71 1.94 0.538 1.08 0.65 1.80 0.772
5 Stress Satisfaction Offset Score (Ref. = +2~+1)
-2 ~ 0 1.38 0.78 2.42 0.270 1.20 0.67 2.14 0.539
6 Swedish Occupational Fatigue Inventory (Ref. <3)
>=3 1.51 0.95 2.39 0.080 1.58 0.98 2.55 0.060

CVD, cardiovascular and cerebrovascular diseases; HR, hazard ratios; CI, confidence interval; BMI, body mass index.
a.
Adjusted for age, BMI, education, drinking, smoking, and exercise.
b.
Each independent variable (1–6) was solely included in the models

https://doi.org/10.1371/journal.pone.0202977.t002

PLOS ONE | https://doi.org/10.1371/journal.pone.0202977 August 27, 2018 6 / 12

 Association between occupational psychological hazards and metabolic syndrome

Table 3. Hazard ratios f of interactions among psychosocial hazards or on CVD classified by psychosocial hazards in participants (n = 707).
Dependent variable Univariate Analysis Multivariate Analysis
Model 1 HR 95%CI p-value HR 95%CI p-value
Long working hours 1.76 1.11 2.81 0.017 1.65 1.03 2.66 0.037
High job stress 1.03 0.56 1.87 0.936 0.94 0.51 1.74 0.852
High fatigue 1.31 0.81 2.12 0.280 1.43 0.87 2.36 0.160
MetS 1.69 1.07 2.67 0.025 1.33 0.77 2.29 0.308
Model 2
Ref: Working hours < = 12 hrs and SSOS = +2
Either long working hours or high job stress only 1.80 0.83 3.89 0.135 1.64 0.76 3.57 0.210
Both long working hours and high job stress 2.48 1.14 5.37 0.022 2.17 0.99 4.79 0.054
MetS 1.69 1.07 2.67 0.025 1.33 0.77 2.30 0.302
Model 3
Ref: Working hours < = 12 hr & SOFI<3
Either long working hours or high fatigue only 1.27 0.71 2.28 0.419 1.35 0.75 2.43 0.321
Both long working hours and high fatigue 2.29 1.28 4.09 0.005 2.31 1.26 4.23 0.007
MetS 1.69 1.07 2.67 0.025 1.32 0.77 2.26 0.321
Model 4
Ref: SSOS = +2 & SOFI<3
Either high job stress or high fatigue only 2.10 0.92 4.77 0.078 1.90 0.83 4.35 0.130
Both high job stress and high fatigue 2.15 0.96 4.83 0.064 2.07 0.90 4.73 0.086
MetS 1.70 1.08 2.69 0.022 1.32 0.76 2.27 0.321
Model 5
Ref: Working hours < = 12 hr & SSOS = +2 & SOFI<3
One of three variables (long working hours, high job stress, or high fatigue) 2.76 0.95 8.02 0.062 2.43 0.83 7.08 0.105
Two of three variables (long working hours, high job stress, or high fatigue) 2.66 0.92 7.67 0.070 2.57 0.89 7.45 0.082
All of three variables (long working hours, high job stress, or high fatigue) 4.10 1.42 11.80 0.009 3.71 1.26 10.91 0.017
MetS 1.69 1.07 2.67 0.025 1.31 0.76 2.26 0.331

Long working hours = driving hour>12 hour/ day; High job stress = SSOS_scores < +1; High fatigue = SOFI > = 3.
CVD, cardiovascular and cerebrovascular diseases; HR, hazard ratios; CI, confidence interval; MetS, metabolic syndrome; SSOS, Stress Satisfaction Offset Score; SOFI,
Swedish Occupational Fatigue Inventory.
a.
Adjusted for age, BMI, education, drinking, smoking, and exercise.

https://doi.org/10.1371/journal.pone.0202977.t003

Relationship among metabolic syndrome, occupational psychosocial

hazards, and CVD
In the univariate analysis (model 1 of Table 3), only long working hours were significantly asso-
ciated with increased risk of CVD among participants with MetS (multivariate analysis
HR = 1.65, 95% CI = 1.03–2.66). The univariate and multivariate analyses were performed to
determine the association between long working hours and high job stress (model 2 of Table 3);
the adjusted HR ratio of participants with both long working hours and high job stress com-
pared with those with neither long working hours nor high job stress was 2.17 (95% CI = 0.99–
4.79, P value = 0.054). Compared to the group with neither long working hours nor high fatigue
(model 3 of Table 3), the adujsted HR for CVD was 2.31 (95% CI = 1.26–4.23, P value = 0.007)
among those with both long working hours and high fatigue. The adjusted HR ratio of partici-
pants with both high job stress and high fatigue compared to those with neither high job stress
nor high fatigue was 2.07 (95% CI = 0.90–4.73, P value = 0.086). In the final model, the univari-
ate and multivariate analysis HRs of MetS participants with three occupational psychosocial
hazards (model 5 of Table 3) were 4.10 (95% CI = 1.42–11.08, P value = 0.009) and 3.71 (95%

PLOS ONE | https://doi.org/10.1371/journal.pone.0202977 August 27, 2018 7 / 12

 Association between occupational psychological hazards and metabolic syndrome

Table 4. Combinations of individual psychosocial hazards and metabolic syndrome for predicting CVD (n = 707).
Univariate Analysis Multivariate Analysis a.
Combinations LWH HJS HF MetS HR 95%CI p-value HR 95%CI p-value
1 - - - - Reference Reference
2 + - - - 2.95 0.49 17.66 0.237 2.38 0.39 14.66 0.349
3 - + - 1.38 0.33 5.78 0.662 1.22 0.29 5.20 0.788
4 - - + - 1.74 0.29 10.49 0.544 1.99 0.33 12.15 0.458
5 + + - 0.61 0.06 5.83 0.664 0.66 0.07 6.42 0.721
6 - + + - 2.27 0.60 8.63 0.227 2.31 0.60 8.91 0.223
7 + + - 6.75 0.70 65.09 0.099 6.94 0.69 69.77 0.100
8 + + + - 2.58 0.69 9.62 0.158 2.46 0.64 9.43 0.190
9 + - - + 2.53 0.26 24.37 0.422 2.19 0.23 21.38 0.499
10 - + - + 3.45 0.95 12.60 0.061 2.62 0.69 9.95 0.158
11 - - + + 3.56 0.59 21.33 0.165 2.97 0.48 18.28 0.241
12 + + - + 4.46 1.15 17.29 0.031 3.24 0.80 13.14 0.100
13 - + + + 1.48 0.35 6.20 0.595 1.30 0.30 5.60 0.723
14 + - + + 12.01 2.39 60.43 0.003 10.91 2.11 56.30 0.004
15 + + + + 4.76 1.39 16.34 0.013 3.72 1.02 13.56 0.046

LWH, long working hours; HJS, high job stress; HF, high fatigue; MetS, metabolic syndrome
a.
Adjusted for age, BMI, education, drinking, smoking, and exercise.

https://doi.org/10.1371/journal.pone.0202977.t004

CI = 1.26–10.91, P value = 0.017), respectively, which were higher than the HRs of participants
with two psychosocial hazards. Our results still remained robust to sensitivity analyses using the
different exclusion definitions (S1 Table).
In the Table 4, we showed the different combinations of individual psychosocial hazards
and metabolic syndrome for predicting CVD. Compared with those without any psychosocial
hazards, there were no significant associations with CVD among non-MetS participants with
1–3 occupational psychosocial hazards. Notably, multivariate analysis HRs of MetS partici-
pants with two (long working hours+ high fatigue) and three occupational psychosocial haz-
ards were 10.91 (95% CI = 1.02–56.30, P value = 0.004) and 3.72 (95% CI = 1.02–13.56, P
value = 0.046), respectively. In the sensitivity analyses (S2 Table), our results were still robust
to using the different exclusion definitions.

Discussion
To the best of our knowledge, our study was the first investigation to examine the combined effect
of metabolic syndrome and occupational psychosocial hazards on the risk of CVD. Thus, we
sought to describe an approach to evaluate the status of overwork using by multiple clustered
traits. In this 8-year longitudinal cohort study, we found that the risk of CVD was highest among
the MetS participants who had long working hours, high job stress, and high fatigue. However, no
evidence of associations between occupational psychosocial hazards and CVD was observed in
the subjects without MetS. There was a linear increase in the HR among MetS participants with
increased numbers of occupational psychosocial hazards. Among three occupational psychosocial
hazards, long working hours were most associated with an elevated risk of CVD.
For detrimental effect of long working hours, our study finding was concordant with the
findings of several previous studies. The results from several prospective observational studies
suggested an approximately 40% increase in the risk of CHD among employees with long
working hours[21]. In a systematic review of epidemiological evidence, long working hours

PLOS ONE | https://doi.org/10.1371/journal.pone.0202977 August 27, 2018 8 / 12

 Association between occupational psychological hazards and metabolic syndrome

had significant adverse effects on many health outcomes, including depressive state, anxiety,
sleep condition, and coronary heart disease[22]. Moreover, employees who work long hours
have an elevated risk of stroke than those who work standard hours[1]. There was one plausi-
ble explanation for this. Sleep deprivation and fragmentation caused by long working hours
increased the risk of type 2 diabetes through β-cell dysfunction and insulin resistance, indi-
rectly leading to accelerate insult to vascular endothelial cells and an increased risk of CVD
[23].
High job stress was linked with a broader spectrum of diseases via an unhealthy lifestyle
and obesity, including type 2 diabetes, coronary heart disease, and stroke[24, 25]. Meta-analy-
sis of individual participant data from over 47,000 participants showed that the association
between job strain and elevated cardiovascular risks was attributable to the higher prevalence
of diabetes, smoking, and physical inactivity among those reporting job strains[26]. In the
cross-sectional data of 46,573 Finnish public sector employees, an association between higher
work stress and lower leisure time physical activity was observed[27]. High stress attenuated
the willingness or the ability of employees to engage in regular exercise and other physical
activity[28]. Another possible pathophysiological mechanism underlying the association
between job strain and CVD was attributed to an imbalance of the autonomic nervous system
and elevated catecholamines. Emerging evidence had shown that high strain jobs were
reported to be significantly associated with decreased vagal tone indicators[29, 30]. It was sug-
gested that reduced activity of the parasympathetic nervous system increased the risk of coro-
nary heart disease[31].
In terms of high fatigue, MetS participants with high fatigue had a higher risk of CVD com-
pared with those without three occupational psychosocial hazards. Consistent with our find-
ings, Appels et al observed that undue fatigue was a frequent premonitory symptom of cardiac
events. They delineated a triad of symptoms as vital exhaustion (VE) or exhaustion, including
excessive fatigue, feelings of demoralization, and increased irritability[32]. In a meta-analysis
of 17 studies concerning VE/exhaustion, a significant association between VE/exhaustion and
CVD events remained after adjusting for a number of clinical and psychosocial factors[33].
The psychosocial consequences of fatigue were becoming increasingly addressed because high
levels of fatigue, as measured by the Short Form 36 vitality domain, were associated with
increased mortality in the general population[34]. Collectively, it was possible that fatigue or
VE/exhaustion was a potential underlying psychophysiological mediator of CVD.
Even through this body of research had the undeniable merit of offering valuable insights
into the interaction between MetS and psychosocial hazards, it also some limitations. The first
limitation concerned the assessments used in the current study. Although there were many
work-related psychosocial hazards that had a substantial impact on CVD outcomes, we tried
to adopted several valuable questionnaires to evaluate these hazards. Second, the generalizabil-
ity of the results to other populations with different occupations may be limited since all partic-
ipants in the present study were drivers. Third, the interaction between MetS and changes in
psychosocial hazards over time was not analyzed, because the pertinent variables were mea-
sured only once at enrollment into the study. Last, all measures were only taken once at base-
line but not repeated measures during follow-up period. Moreover, the interpretation of our
observations was limited because of absence of long-term changes in psychosocial hazards and
cardiometabolic risks.

Conclusion
Our findings clearly supported the notion that a combination of long working hours, high job
stress, and high fatigue in MetS participants increased the risk of developing CVD. These

PLOS ONE | https://doi.org/10.1371/journal.pone.0202977 August 27, 2018 9 / 12

 Association between occupational psychological hazards and metabolic syndrome

results underscored the importance of recognizing these occupational psychosocial hazards in

the drivers with MetS, especially combined with long working hours, high job stress, and high
fatigue. It was tempting to speculate that since metabolic syndrome insulted cardiovascular
healthy and ensuing psychosocial hazards could contribute to the deterioration of vascular
function in individuals predisposed for CVD. Psychosocial interventions in the workplace
may have beneficial effects on CVD morbidity and mortality. Further research might extend
the investigation of psychosocial hazards to different occupations to examine the impact of
MetS on CVD or mortality.

Supporting information
S1 Table. Hazard ratios f of interactions among psychosocial hazards or on CVD classified
by psychosocial hazards in participants.
(DOCX)
S2 Table. Combinations of individual psychosocial hazards and metabolic syndrome for
predicting CVD.
(DOCX)

Author Contributions
Conceptualization: Wei-Liang Chen, Wei-Te Wu, Saou-Hsing Liou.
Data curation: Wei-Liang Chen, Chung-Ching Wang, Sheng-Ta Chiang, Ying-Chuan Wang,
Yu-Shan Sun, Wei-Te Wu, Saou-Hsing Liou.
Formal analysis: Wei-Liang Chen, Chung-Ching Wang, Sheng-Ta Chiang, Ying-Chuan
Wang, Yu-Shan Sun, Wei-Te Wu, Saou-Hsing Liou.
Funding acquisition: Saou-Hsing Liou.
Investigation: Wei-Liang Chen, Chung-Ching Wang, Sheng-Ta Chiang, Ying-Chuan Wang,
Yu-Shan Sun, Wei-Te Wu, Saou-Hsing Liou.
Methodology: Wei-Liang Chen, Chung-Ching Wang, Sheng-Ta Chiang, Ying-Chuan Wang,
Yu-Shan Sun, Wei-Te Wu, Saou-Hsing Liou.
Project administration: Wei-Liang Chen, Chung-Ching Wang, Sheng-Ta Chiang, Ying-
Chuan Wang, Yu-Shan Sun, Wei-Te Wu, Saou-Hsing Liou.
Resources: Wei-Liang Chen.
Supervision: Wei-Te Wu, Saou-Hsing Liou.
Validation: Wei-Liang Chen, Wei-Te Wu, Saou-Hsing Liou.
Visualization: Wei-Liang Chen, Wei-Te Wu, Saou-Hsing Liou.
Writing – original draft: Wei-Liang Chen.
Writing – review & editing: Wei-Liang Chen, Wei-Te Wu, Saou-Hsing Liou.

References
1. Kivimaki M, Jokela M, Nyberg ST, Singh-Manoux A, Fransson EI, Alfredsson L, et al. Long working
hours and risk of coronary heart disease and stroke: a systematic review and meta-analysis of pub-
lished and unpublished data for 603,838 individuals. Lancet (London, England). 2015; 386
(10005):1739–46. Epub 2015/08/25. https://doi.org/10.1016/s0140-6736(15)60295-1 PMID: 26298822.

PLOS ONE | https://doi.org/10.1371/journal.pone.0202977 August 27, 2018 10 / 12

 Association between occupational psychological hazards and metabolic syndrome

2. Brown DL, Feskanich D, Sanchez BN, Rexrode KM, Schernhammer ES, Lisabeth LD. Rotating night
shift work and the risk of ischemic stroke. American journal of epidemiology. 2009; 169(11):1370–7.
Epub 2009/04/10. https://doi.org/10.1093/aje/kwp056 PMID: 19357324; PubMed Central PMCID:
PMCPmc2727250.
3. Kivimaki M, Nyberg ST, Batty GD, Fransson EI, Heikkila K, Alfredsson L, et al. Job strain as a risk factor
for coronary heart disease: a collaborative meta-analysis of individual participant data. Lancet (London,
England). 2012; 380(9852):1491–7. Epub 2012/09/18. https://doi.org/10.1016/s0140-6736(12)60994-5
PMID: 22981903; PubMed Central PMCID: PMCPmc3486012.
4. Hidaka T, Hayakawa T, Kakamu T, Kumagai T, Hiruta Y, Hata J, et al. Prevalence of Metabolic Syn-
drome and Its Components among Japanese Workers by Clustered Business Category. PloS one.
2016; 11(4):e0153368. Epub 2016/04/16. https://doi.org/10.1371/journal.pone.0153368 PMID:
27082961; PubMed Central PMCID: PMCPmc4833411.
5. Malik S, Wong ND, Franklin SS, Kamath TV, L’Italien GJ, Pio JR, et al. Impact of the metabolic syn-
drome on mortality from coronary heart disease, cardiovascular disease, and all causes in United
States adults. Circulation. 2004; 110(10):1245–50. Epub 2004/08/25. https://doi.org/10.1161/01.CIR.
0000140677.20606.0E PMID: 15326067.
6. Liaw FY, Kao TW, Wu LW, Wang CC, Yang HF, Peng TC, et al. Components of Metabolic Syndrome
and the Risk of Disability among the Elderly Population. Scientific reports. 2016; 6:22750. Epub 2016/
03/08. https://doi.org/10.1038/srep22750 PMID: 26948125; PubMed Central PMCID:
PMCPmc4780009.
7. Mottillo S, Filion KB, Genest J, Joseph L, Pilote L, Poirier P, et al. The metabolic syndrome and cardio-
vascular risk a systematic review and meta-analysis. Journal of the American College of Cardiology.
2010; 56(14):1113–32. Epub 2010/09/25. https://doi.org/10.1016/j.jacc.2010.05.034 PMID: 20863953.
8. Moe B, Mork PJ, Holtermann A, Nilsen TI. Occupational physical activity, metabolic syndrome and risk
of death from all causes and cardiovascular disease in the HUNT 2 cohort study. Occupational and envi-
ronmental medicine. 2013; 70(2):86–90. Epub 2012/10/02. https://doi.org/10.1136/oemed-2012-
100734 PMID: 23022656.
9. Wu WT, Tsai SS, Liao HY, Lin YJ, Lin MH, Wu TN, et al. Usefulness of overnight pulse oximeter as the
sleep assessment tool to assess the 6-year risk of road traffic collision: evidence from the Taiwan Bus
Driver Cohort Study. International journal of epidemiology. 2016. Epub 2016/08/02. https://doi.org/10.
1093/ije/dyw141 PMID: 27477030.
10. The Definition of the Metabolic Syndrome in Adults: Health Promotion Administration, Ministry of Health
and Welfare; [01 May 2017]. Available from: https://www.hpa.gov.tw/Pages/Detail.aspx?nodeid=
1127&pid=1887.
11. Lin YC, Yen LL, Chen SY, Kao MD, Tzeng MS, Huang PC, et al. Prevalence of overweight and obesity
and its associated factors: findings from National Nutrition and Health Survey in Taiwan, 1993–1996.
Preventive medicine. 2003; 37(3):233–41. Epub 2003/08/14. PMID: 12914829.
12. Pan WH, Flegal KM, Chang HY, Yeh WT, Yeh CJ, Lee WC. Body mass index and obesity-related meta-
bolic disorders in Taiwanese and US whites and blacks: implications for definitions of overweight and
obesity for Asians. The American journal of clinical nutrition. 2004; 79(1):31–9. Epub 2003/12/20.
https://doi.org/10.1093/ajcn/79.1.31 PMID: 14684394.
13. Chien TW, Lai WP, Wang HY, Hsu SY, Castillo RV, Guo HR, et al. Applying the revised Chinese Job
Content Questionnaire to assess psychosocial work conditions among Taiwan’s hospital workers. BMC
public health. 2011; 11:478. Epub 2011/06/21. https://doi.org/10.1186/1471-2458-11-478 PMID:
21682912; PubMed Central PMCID: PMCPmc3146858.
14. Karasek R, Baker D, Marxer F, Ahlbom A, Theorell T. Job decision latitude, job demands, and cardio-
vascular disease: a prospective study of Swedish men. American journal of public health. 1981; 71
(7):694–705. Epub 1981/07/01. PMID: 7246835; PubMed Central PMCID: PMCPmc1619770.
15. Karasek R TT. Healthy Work: Stress, productivity, and the reconstruction of working life.: Basic Books:
New York.; 1990.
16. Siegrist J, Starke D, Chandola T, Godin I, Marmot M, Niedhammer I, et al. The measurement of effort-
reward imbalance at work: European comparisons. Social science & medicine (1982). 2004; 58
(8):1483–99. Epub 2004/02/05. https://doi.org/10.1016/s0277-9536(03)00351-4 PMID: 14759692.
17. M S. Stress and Satisfaction. OHS Canada. 1999; 15:10.
18. Ahsberg E G, F, Kjellberg A. Perceived quality of fatigue during different occupational tasks: Develop-
ment of a questionnaire. International Journal of Industrial Ergonomics. 1997; 20(2):15.
19. Lin RT, Lin CK, Christiani DC, Kawachi I, Cheng Y, Verguet S, et al. The impact of the introduction of
new recognition criteria for overwork-related cardiovascular and cerebrovascular diseases: a cross-
country comparison. 2017; 7(1):167. https://doi.org/10.1038/s41598-017-00198-5 PMID: 28279019.

PLOS ONE | https://doi.org/10.1371/journal.pone.0202977 August 27, 2018 11 / 12

 Association between occupational psychological hazards and metabolic syndrome

20. Park J, Kim Y, Cheng Y, Horie S. A comparison of the recognition of overwork-related cardiovascular
disease in Japan, Korea, and Taiwan. Industrial health. 2012; 50(1):17–23. Epub 2011/12/08. PMID:
22146142.
21. Virtanen M, Heikkila K, Jokela M, Ferrie JE, Batty GD, Vahtera J, et al. Long working hours and coro-
nary heart disease: a systematic review and meta-analysis. American journal of epidemiology. 2012;
176(7):586–96. Epub 2012/09/07. https://doi.org/10.1093/aje/kws139 PMID: 22952309; PubMed Cen-
tral PMCID: PMCPmc3458589.
22. Bannai A, Tamakoshi A. The association between long working hours and health: a systematic review
of epidemiological evidence. Scandinavian journal of work, environment & health. 2014; 40(1):5–18.
Epub 2013/10/09. https://doi.org/10.5271/sjweh.3388 PMID: 24100465.
23. Cappuccio FP, D’Elia L, Strazzullo P, Miller MA. Quantity and quality of sleep and incidence of type 2
diabetes: a systematic review and meta-analysis. Diabetes care. 2010; 33(2):414–20. Epub 2009/11/
17. https://doi.org/10.2337/dc09-1124 PMID: 19910503; PubMed Central PMCID: PMCPmc2809295.
24. Kivimaki M, Singh-Manoux A, Nyberg S, Jokela M, Virtanen M. Job strain and risk of obesity: systematic
review and meta-analysis of cohort studies. International journal of obesity (2005). 2015; 39(11):1597–
600. Epub 2015/06/05. https://doi.org/10.1038/ijo.2015.103 PMID: 26041697; PubMed Central PMCID:
PMCPmc4579559.
25. Heraclides AM, Chandola T, Witte DR, Brunner EJ. Work stress, obesity and the risk of type 2 diabetes:
gender-specific bidirectional effect in the Whitehall II study. Obesity (Silver Spring, Md). 2012; 20
(2):428–33. Epub 2011/05/20. https://doi.org/10.1038/oby.2011.95 PMID: 21593804.
26. Nyberg ST, Fransson EI, Heikkila K, Alfredsson L, Casini A, Clays E, et al. Job strain and cardiovascular
disease risk factors: meta-analysis of individual-participant data from 47,000 men and women. PloS
one. 2013; 8(6):e67323. Epub 2013/07/11. https://doi.org/10.1371/journal.pone.0067323 PMID:
23840664; PubMed Central PMCID: PMCPmc3688665.
27. Kouvonen A, Kivimaki M, Elovainio M, Virtanen M, Linna A, Vahtera J. Job strain and leisure-time physi-
cal activity in female and male public sector employees. Preventive medicine. 2005; 41(2):532–9. Epub
2005/05/27. https://doi.org/10.1016/j.ypmed.2005.01.004 PMID: 15917049.
28. Hellerstedt WL, Jeffery RW. The association of job strain and health behaviours in men and women.
International journal of epidemiology. 1997; 26(3):575–83. Epub 1997/06/01. PMID: 9222783.
29. Jarczok MN, Jarczok M, Mauss D, Koenig J, Li J, Herr RM, et al. Autonomic nervous system activity
and workplace stressors—a systematic review. Neuroscience and biobehavioral reviews. 2013; 37
(8):1810–23. Epub 2013/07/31. https://doi.org/10.1016/j.neubiorev.2013.07.004 PMID: 23891906.
30. Collins S, Karasek R. Reduced vagal cardiac control variance in exhausted and high strain job subjects.
International journal of occupational medicine and environmental health. 2010; 23(3):267–78. Epub
2010/10/12. https://doi.org/10.2478/v10001-010-0023-6 PMID: 20934956.
31. Hayano J, Sakakibara Y, Yamada M, Ohte N, Fujinami T, Yokoyama K, et al. Decreased magnitude of
heart rate spectral components in coronary artery disease. Its relation to angiographic severity. Circula-
tion. 1990; 81(4):1217–24. Epub 1990/04/01. PMID: 2317904.
32. Appels A HP, Mulder P. 15e24. A questionnaire to assess premonitory symptoms of myocardial infarc-
tion. Int J Cardiol. 1987; 17:10.
33. Cohen R, Bavishi C, Haider S, Thankachen J, Rozanski A. Meta-Analysis of Relation of Vital Exhaus-
tion to Cardiovascular Disease Events. The American journal of cardiology. 2017; 119(8):1211–6. Epub
2017/02/22. https://doi.org/10.1016/j.amjcard.2017.01.009 PMID: 28215416.
34. Basu N, Yang X, Luben RN, Whibley D, Macfarlane GJ, Wareham NJ, et al. Fatigue is associated with
excess mortality in the general population: results from the EPIC-Norfolk study. BMC medicine. 2016;
14(1):122. Epub 2016/08/21. https://doi.org/10.1186/s12916-016-0662-y PMID: 27543008; PubMed
Central PMCID: PMCPmc4992307.

PLOS ONE | https://doi.org/10.1371/journal.pone.0202977 August 27, 2018 12 / 12