Work xx (20xx) x–xx 1

DOI:10.3233/WOR-192968
IOS Press

1 Job strain and psychological distress

2 in office workers: The role of coping

f
roo
3 Deokhoon Juna,f,∗ , Shaun O’Learya,b , Steven M McPhailc,d and Venerina Johnstona,e
4
a School of Health and Rehabilitation Sciences, University of Queensland,
5 Brisbane, Australia
6
b Department of Physiotherapy, Royal Brisbane and Women’s Hospital, Metro North Health,

7 Brisbane, Australia

rP
8
c Centre for Functioning and Health Research, Metro South Health, Brisbane, Australia

9
d School of Public Health and Institute of Health and Biomedical Innovation, Queensland University

10 of Technology, Brisbane, Australia

11
e RECOVER Injury Research Centre, Faculty of Health and Behavioural Sciences, The University

tho
12 of Queensland, Brisbane, Australia
15
f Department of Physical Therapy, Kyoungsung University, Busan, South Korea

13

14 Received 24 October 2018

Accepted 2 May 2019
Au
16 Abstract.
17 BACKGROUND: Work-related psychosocial factors such as job strain are thought to contribute to elevated psychological
d

18 stress in office workers. One factor that may impact the relationship between job strain and psychological distress is the
19 individual’s coping resources.
OBJECTIVE: The purpose of this study is to examine the interaction effect of coping resources on the relationship between
cte

20

21 job strain and psychological distress in office workers.

22 METHODS: 220 office workers in Australia and Korea completed the Job Content Questionnaire (to evaluate job strain and
23 social support at work), the Depression, Anxiety, and Stress Scale, (DASS-21, a measure of psychological distress), and the
24 Coping with Job Stress Scale to assess control and escape coping. Hierarchical regression analyses were executed to examine
the interaction and moderating effect of coping resources.
rre

25

26 RESULTS: Job strain had a direct positive relationship with all three domains of psychological distress. The relationship
27 between job strain and depression was positively moderated by escape coping, but negatively moderated by social support.
28 Use of higher levels of escape coping predicted higher levels of depression and anxiety symptoms when higher levels of job
29 strain were perceived.
co

30 CONCLUSIONS: Findings suggest there may be a direct relationship between job strain and psychological distress in office
31 workers. This relationship, however, may be moderated by the office workers coping resources (coping strategies and social
32 support). It is suggested that the evaluation of coping might be a key consideration in the elements of the assessment of
33 psychological distress in office workers.
Un

34 Keywords: Interaction effect, coping strategy, social support, work stress

1. Introduction 35

∗ Address for correspondence: Deokhoon Jun, Division of

Physiotherapy, School of Health and Rehabilitation Sciences, The
Psychological distress can impact the health of 36

University of Queensland, Saint Lucia, QLD 4072, Australia. Tel.: office workers and is thought to be a major deter- 37

+61 431231348, E-mail: deokhoon.jun@uq.net.au. minant of some conditions such as back and neck 38

1051-9815/19/$35.00 © 2019 – IOS Press and the authors. All rights reserved
 2 D. Jun et al. / Job strain and psychological distress in office workers

39 pain [1]. Epidemiological evidence supports a link distress with greater distress reported when work- 91

40 between work-related psychosocial factors and psy- ers used greater coping strategies and less palliative 92

41 chological distress. Systematic reviews [2, 3] suggest coping. Lazarus (2000) suggested that analyzing 93

42 that low job control combined with high work the consequences of problem-focused and emotion- 94

43 demands in the work environment [4] may result in focused strategies separately, leads to a limited 95

44 psychological distress in the work environment. One understanding of the role of coping [18]. Beyond 96

45 factor potentially affecting the relationship between the psychological investigation, for instance in mus- 97

46 work-related psychosocial factors and psychological culoskeletal disorder studies, coping strategies were 98

f
47 distress is an individual’s coping resources such as not oriented to the perspective of work-related stres- 99

roo
48 coping strategies and social support [5–7]. sors. This was highlighted by Cote and colleagues 100

49 Coping is defined as the cognitive and behav- in their review [19] concerning the role of coping 101

50 ioral efforts to manage stressful situations that tax or behaviors in the development of work-related neck 102

51 exceed the resources of the individual [8]. A variety of pain. Specifically, coping with stressful situations in 103

52 coping strategies have been proposed in the evidence the workplace may differ to stressful situations out- 104

rP
53 literature and in general they can be described in two side the work environment and may require specific 105

54 dimensions; control coping (direct action, problem- assessment instruments. To date, no study has investi- 106

55 focused, active) referred to as targeting the causes of gated the relationship between work-oriented coping 107

56 stress in practical ways and escape coping (palliative, strategies and psychological distress. 108

tho
57 emotion-focused, passive, avoidance) referred to as The purpose of this study was to investigate the 109

58 trying to reduce or ignoring the negative emotional role of coping on the relationship between job strain 110

59 responses caused by stress [9]. (a work-place psychosocial factor) and psychological 111

60 There is evidence that an individual’s coping strate- distress in office workers. Specifically, we proposed 112

61 gies can influence their level of psychological distress two hypotheses: 1) that job strain would be associated 113
Au
62 [10]. For example, control coping strategies have with psychological distress in a population of office 114

63 been shown to reduce psychological distress [11, 12], workers and 2) that coping resources would moderate 115

64 while escape coping strategies have been shown to the relationship between job strain and psychological 116

65 increase psychological distress [6, 13]. Support from distress. 117

66 colleagues and supervisors is hypothesized to buffer The findings of this study will inform future 118

the negative impact of high job demands and low studies investigating the relationship between work-
d

67 119

68 job control with evidence that low social support is place psychosocial factors, psychological distress, 120
cte

69 associated with higher risk of psychological distress and health-related disorders such as musculoskeletal 121

70 in workers [5, 14]. However, the literature exploring conditions. 122

71 the interaction effect and moderating role of coping

72 resources (coping strategies and social support) on
73 psychological distress is inconsistent. Some studies 2. Methods 123
rre

74 have shown that an excessive risk of psychological

75 distress was reported when low job control, high job 2.1. Design and sample 124

76 demand and low job support coexist [5, 15]. Lian

77 and his colleagues also reported that increased coping A cross-sectional exploratory study of 220 office 125
co

78 resources (e.g. self-care, social support, recreation workers was conducted in two cities, Brisbane, Aus- 126

79 and rational coping) had a buffering effect on the tralia, and Daegu, South Korea. Participants were 127

80 relationship between work-related psychosocial fac- recruited from multiple organizations employing 128

81 tors and psychological distress [15]. On the contrary, office workers in both regions through advertise- 129
Un

82 other studies have only provided weak [7, 16] or no ments, social media, word of mouth, and email 130

83 [9, 17] interaction effect between coping resources contact. All participants provided written consent to 131

84 and psychological distress in the work environment. participate and ethical approval for the study was 132

85 Uncertainty as to the role of coping resources with granted by the institute Human Research Ethics Com- 133

86 regards to the modulation of psychological distress mittee in accordance with the Declaration of Helsinki. 134

87 in the workplace probably reflects methodological There were 571 office workers who volunteered to 135

88 issues associated with clarifying any relationship. participate from six organizations in two regions. The 136

89 One study in Spain [9] only reported the interaction majority of participants were university educational 137

90 effect between coping strategies and psychological personnel or faculty members from the University of 138
 D. Jun et al. / Job strain and psychological distress in office workers 3

139 Queensland (n = 428), and Daegu University (n = 81). freedom, and influence on work and policy) was 189

140 To be eligible, participants had to be an office worker measured by nine items. Psychological demand and 190

141 aged 18 to 60 years and working longer than 30 decision latitude were combined to create the linear 191

142 hours per week, performing computer, document, and scale of job strain score to equally weight the magni- 192

143 phoning intensive work for more than 20 hours per tude of psychological demand and decision latitude 193

144 week. Participants were excluded if they reported (subtraction of decision latitude from psychological 194

145 musculoskeletal pain of the neck and back region in job demand) [22]. Thus, a higher job strain score 195

146 the past twelve months, or took sick or health care reflected higher psychological demands and lower 196

f
147 leave due to musculoskeletal neck and back pain. The levels of decision latitude. Cronbach’s ␣ coefficients 197

roo
148 definition of a musculoskeletal disorder to neck and for job demand and decision latitude were accept- 198

149 back region was derived from a conceptual model able in this study sample reporting 0.70 and 0.85 for 199

150 classified by the Neck Pain Task Force, as follows: Brisbane, and 0.70 and 0.78 for Daegu, respectively. 200

151 “Interfering neck or back pain which prompts the A Korean version of the JCQ was utilized for data 201

152 participant to consider action” [20]. If there was any collection in the Korean participants [23]. 202

rP
153 uncertainty regarding an individual’s eligibility based Coping Resources: Two different strategies with 203

154 on the online survey an interview with the investigator the potential to influence the degree of psychologi- 204

155 was arranged to clarify eligibility. Non-eligible par- cal distress experienced were selected and measured; 205

156 ticipants did not complete the main questionnaires. coping strategies (escape and control coping) to work 206

tho
157 In Brisbane, 441 potential participants were stress, and social support from supervisor and co- 207

158 screened via email or by interview with 160 partici- workers. 208

159 pants meeting the study inclusion criteria. Exclusion Coping strategies to stressful workplace situa- 209

160 from participation occurred due to performing non- tions were measured using the Coping with Job 210

161 office work (n = 114), the presence of pain (n = 81), Stress Scale [24]. This instrument is a compact, 211
Au
162 non-response to follow up communication (n = 72), comprehensive, and reliable inventory widely used 212

163 working part-time (n = 7), upcoming leave (n = 5), to assess behavioural strategies preferred by people 213

164 and other miscellaneous reasons (n = 2). In Daegu, when faced with stressful conditions at work. 214

165 130 potential participants underwent initial screening The five scales were developed for work stress 215

166 with 60 satisfying the study inclusion criteria. Rea- by integrating three different conceptual coping 216

sons for exclusion were the presence of pain (n = 42), frameworks; control (direct actions and proactive
d

167 217

168 performing non-office work (n = 15), no response cognitive evaluations to stressful situation in work- 218
cte

169 (n = 8), and working part-time (n = 5). In total, 220 place), escape (avoidance from stressful situation), 219

170 office workers consented to participate in the study and symptom management (manage job stress in 220

171 and completed all required questionnaires. Partici- different aspect such as alcohol consumption). This 221

172 pants were recruited over a 20 month period between instrument demonstrated good internal consistency 222

173 May 2014 and December 2015. and reliability for each scale (coefficient alpha: 223
rre

help-seeking = 0.61, avoidance/resignation = 0.74, 224

174 2.2. Questionnaires positive-thinking = 0.76, direct action = 0.69, and 225

alcohol use = 0.83). With acceptable validity and 226

175 Data were collected via an online survey and reliability, this was the only scale appropriate for 227
co

176 included measures of job strain, coping, psycho- assessing the strategies of coping with workplace 228

177 logical distress, as well as potentially confounding stress [25]. Following the general distinction of 229

178 individual and work-related measures. coping strategies, the five domains were divided 230

179 Job Strain: Job strain was assessed with the Job into two contrary domains; control coping (positive 231
Un

180 Content Questionnaire (JCQ) which incorporates thinking, direct action, and help seeking) and escape 232

181 measures of psychological demand and decision lati- coping (avoidance/resignation and alcohol use) [24]. 233

182 tude [21]. Psychological demands (work fast, work The internal consistency for escape coping and 234

183 hard, over load, time constraints, and conflicting control coping was 0.61 and 0.82 for Brisbane and 235

184 demands) were measured by five items. Decision lat- 0.70 and 0.87 for Deagu, respectively. 236

185 itude which consists of skill discretion (occasions for Social support was evaluated with eight items in 237

186 new learning, repetitive work, work creativity, high the JCQ, four items each relating to support from 238

187 level skills, task variation, and skill development) coworkers (coworkers’ component, coworker inter- 239

188 and decision authority (decision allowance, decision est in me, friendly coworkers, and helpful coworkers), 240
 4 D. Jun et al. / Job strain and psychological distress in office workers

241 and support from supervisors (concerned about me, sented as numbers and percentages. To standardize 289

242 pays attention to me, helpful supervisor, and good different units of measurement in exploratory vari- 290

243 organizer), respectively. Both scales were combined ables (job strain and coping resources), z-score 291

244 to calculate job support score. Cronbach’s ␣ coeffi- transformation was performed for all four exploratory 292

245 cients for social support in this study sample was 0.83 variables. Potential differences in each variable 293

246 for Brisbane and 74 for Daegu, respectively. between office workers from Brisbane and Daegu 294

247 Psychological Distress: Psychological distress were examined using the Student t-test (continuous 295

248 was evaluated with the short version of the Depres- variables) and Fisher exact test (categorical vari- 296

f
249 sion, Anxiety and Stress Scale consisting of twenty ables). The correlation between job strain, coping 297

roo
250 one questions (DASS-21) [26]. This instrument resources, and psychological distress were evaluated 298

251 is a widely used quantitative screening tool of using Spearman’s r. 299

252 psychological distress for both clinical and non- Mixed effects linear regression analyses with two 300

253 clinical purposes. The three scales each contain seven cities as a random effects variable were executed to 301

254 items assessing symptoms of depression (displeasure, detect interaction effects of coping resources on each 302

rP
255 hopelessness, devaluation of life, self-depreciation, of the psychological distress symptoms (depression, 303

256 lack of interest or involvement, anhedonia, and anxiety, and stress). For the highest prediction accu- 304

257 inertia), anxiety (autonomic arousal, skeletal muscu- racy with a parsimonious set of independent variables 305

258 lature effects, situational anxiety, subjective anxious without overfitting each model least absolute shrink- 306

tho
259 affect), and stress (relaxation, nervous arousal, eas- age and selection operator (Lasso) variable selection 307

260 ily upset, irritability, and impatience). Good internal method was applied [28]. The initial models were 308

261 consistency (Cronbach’s ␣) in this study sample for formed by the variable selections from the Lasso 309

262 Depression, Anxiety, and Stress scales have been method. Next, potential interaction terms between 310

263 attained at 0.82, 0.76, and 0.82 for Brisbane, and 0.85, variables were entered into the initial models. Ver- 311
Au
264 0.80, 84 for Daegu, respectively. ification that the study data met the assumptions of 312

the regression analyses was performed for all models 313

265 2.3. Covariates (linearity, normality of residuals, homoscedasticity, 314

and no multicollinearity). 315

266 Several individual and work-related variables were Interaction terms (two-way or three-way combi- 316

recorded as potential covariates and considered in nations of job strain with control coping, negative
d

267 317

268 all adjusted analytic models. Individual variables coping, and social support) were entered into the 318
cte

269 included; gender, age, body mass index, alcohol con- regression models to examine Hypothesis 2. The 319

270 sumption, geographic location (Brisbane, Daegu), newly created models with interaction terms were 320

271 self-rated perceived health (five scales - poor to excel- compared with initial models (without interaction 321

272 lent), job insecurity (JCQ) and physical activity levels term) using the Akaike’s information criterion (AIC) 322

273 as measured with the International Physical Activ- [29]. Given a set of variables, AIC estimates the good- 323
rre

274 ity Questionnaire (IPAQ – short version) [27]. The ness of fit of each model on the basis of the maximized 324

275 IPAQ measures four domains of physical activity log-likelihood and includes a penalty for each addi- 325

276 (work, transport, domestic, and leisure time) dur- tional parameter that discourages overfitting. When 326

277 ing the previous seven days to estimate physical interaction effect exists, it was measured by plotting 327
co

278 activity levels (high, moderate, or low level of activ- the change of marginal effect of variables on the out- 328

279 ity). Work-related factors included hours of computer come of psychological distress symptoms (Figs. 1, 2) 329

280 work per week in work, duration of break per day and [30]. Demographic factors considered to be important 330

281 self-satisfaction with the physical work environment covariates (age, gender, and body mass index (BMI)) 331
Un

282 (lighting, acoustics, temperature, air quality, room were included in all models. 332

283 space) with each item rated from 1 to 5 (very poor

284 to excellent).
3. Results 333

285 2.4. Statistics

A total of 220 office workers (98 male and 122 334

286 Analyses were performed with Stata version 13. female) met the inclusion criteria. Descriptive details 335

287 Continuous variables were presented as mean ± SD, regarding participants and variables assessed are 336

288 and dichotomous or categorical variables were pre- found in Table 1. Male and female workers did 337
 D. Jun et al. / Job strain and psychological distress in office workers 5

338 not significantly differ in variables and confounders anxiety symptoms, less satisfaction with the office 343

339 included in the analysis, except for BMI scale environment, and lower physical activity, compared 344

340 (25.2 kg2 /m ± 3.19 for men and 23.05 kg2 /m ± 4.66 to the office workers in Brisbane (Table 1). Cor- 345

341 for women). Office workers in Daegu reported sig- relations between the study variables are shown in 346

342 nificantly higher job strain, greater depression and Table 2. Briefly, psychological distress symptoms 347

were positively associated with job strain and escape 348

coping strategy, but negatively associated with social 349

support from supervisors and colleague. For each of 350

f
the mixed effect models, there was no significant 351

roo
rP
Fig. 1. Linear prediction of depression symptom on job strain in

tho
different condition of escape coping and social support.
Note: The X-axis ranges from –3 to 3 by z-score transformed scale
(originally range from –74 to –2). The incremental levels of escape
coping by ± SD z-score from mean were created to predict anxiety Fig. 2. Linear prediction of anxiety symptom on job strain in
symptom. different level of escape coping.
Au
Table 1
Characteristics of participants (n = 220) from Brisbane and Daegu cities
Variables Mean or SD or % Mean or SD or % Mean or SD or %
cases in cases in cases in
Brisbane Daegu Total
(n = 160) (n = 60) (n = 220)
d

Age (years) 37.4 ±10.14 37.4 ±9.9 37.4 ±10.1

Female (n) 98 61.3% 24 40.0% 122 55.5%
BMI (kg2 /m)∗ ±4.3 ±3.6 ±4.2
cte

25.2 23.1 24.0

Non-smoker (n)∗ 152 95% 49 81.8% 201 91.4%
Drinking (scale)∗ 2.4 ±1.4 2.08 ±0.8 2.4 ±1.3
Job positions (n)∗
Manager 12 7.5% 4 6.7% 16 7.3%
Professional 81 50.6% 10 16.7% 91 41.4%
rre

Administrative 67 41.9% 46 76.7% 113 51.4%

Years working in current organization∗ 4.4 ±5.7 9.9 ±9.2 5.9 ±7.3
Computer hours per day (hr) 7.3 ±1.1 7.4 ±1.2 7.3 ±1.1
Hours of work before break (n)
<2hour work 93 58.1 % 40 66.7% 133 60.5%
>2hour work 67 41.9 % 20 33.3% 87 39.6%
co

Satisfaction to workplace environment (scale)∗ 3.2 ±0.5 2.5 ±0.5 3.0 ±0.6
Physical activity level (n)∗
Low activity 18 11.3% 20 33.3% 38 17.3%
Moderate activity 77 48.1% 26 42.3% 103 46.8%
Un

High activity 65 40.6% 14 23.3% 79 35.9%

Job insecurity 5.51 ±1.94 5.37 ±2.07 5.47 ±1.98
Job strain (score)∗ –39.39 ±12.54 –33.4 ±6.51 –33.76 ±11.52
Control coping (score)∗ 63.35 ±8.54 59.75 ±9.44 62.37 ±8.92
Escape coping (score) 19.74 ±4.03 19.12 ±4.49 19.57 ±4.16
Social support (score)∗ 25.22 ±3.30 23.95 ±2.38 24.87 ±3.12
Depression (score)∗ 2.31 ±2.62 3.22 ±2.71 2.56 ±2.67
Anxiety (score)∗ 1.62 ±2.32 2.63 ±2.54 1.90 ±2.42
Stress (score) 4.5 ±3.46 5.23 ±3.03 4.7 ±3.36
∗:p < 0.05 for 2 × 2 Fisher’s exact test (categorical variables) or t-test (continuous variables) of the comparisons between
Brisbane and Daegu regions.
 6 D. Jun et al. / Job strain and psychological distress in office workers

Table 2
Correlations of the study variables
1 2 3 4 5 6 7
1 Depression 1
2 Anxiety 0.52∗∗ 1
3 Stress 0.62∗∗ 0.58∗∗ 1
4 Job strain 0.33∗∗ 0.22∗∗ 0.27∗∗ 1
5 Control coping –0.16∗ 0.06 –0.03 –0.12 1
6 Escape coping 0.35∗∗ 0.14∗ 0.33∗∗ 0.12 0.03 1
–0.21∗∗ –0.15∗ –0.18∗∗ –0.29∗∗

f
7 Social support 0.07 –0.11 1

roo
∗ p < 0.05, ∗∗ p < 0.01. Two-tailed.

352 intracluster correlation with country as a random fac- effect was found between job strain and escape cop- 389

353 tor (Table 3). ing on anxiety symptoms (Table 3). Figure 2 shows 390

the predicted value of anxiety symptoms on the dif- 391

rP
354 3.1. Effect of coping resources on depression ferent levels of job strain (z-score) when interacting 392

355 symptom with escape coping. When workers used less escape 393

coping, the adverse effect of job strain on anxiety 394

356 The first column of findings in Table 3 displays symptom was minimal. However, the effect gradu- 395

the results of the regression performed to examine ally increased as the usage of escape coping increased 396

tho
357

358 associations with depression symptoms. Job strain (Fig. 2). 397

359 and escape coping were positively associated with

360 depression symptoms. No significant associations 3.3. Effect of coping resources on stress symptom 398

361 were found between control coping, social support,

and depression symptoms.
Au
362 The third column of findings in Table 3 displays 399

363 A significant interaction effect was found between the results of the regression performed to determine 400

364 job strain, escape coping, and social support on associations with stress symptoms. Job strain and 401

365 depression symptoms; however, control coping nei- escape coping were positively associated with stress 402

366 ther interacted with job strain nor with other coping symptoms. However, no interaction effect was found 403

resources (Table 3). Figure 1 shows the predicted between job strain and stress symptoms.
d

367 404

368 value of depression symptoms on the different levels

of job strain (z-score) when interacting with escape
cte

369

370 coping and social support. An interaction effect was 4. Discussion 405
371 distinctive when higher job strain presented. Specifi-
372 cally, workers using higher levels of escape coping This study investigated the interaction effect of 406
373 (mean z score +1 SD), as well as those receiving coping resources (coping strategies and social sup- 407
rre

374 greater levels of social support were more likely port) on the relationship between job strain and 408
375 to report greater depression symptoms than other psychological distress in office workers from two 409
376 workers with less social support from work. In con- different cultural settings. While these two cultural 410
377 trast, workers with low levels of escape coping were groups differed in two of their psychological distress 411
expected to have less depression symptoms with least
co

378 measures (depression and anxiety, Table 1), these 412

379 interaction effect of social support on the relationship differences were explained by job strain and coping 413
380 between job strain and depression symptom. resource factors, not cultural setting with negligible 414

clustering in any of the mixed models with country 415

Un

381 3.2. Effect of coping resources on anxiety as a random factor. This may infer generalizabil- 416

382 symptom ity between the cultural groups with respect to the 417

association between job strain, coping resources, and 418

383 The second column of findings in Table 3 displays psychological distress despite differences in distress 419

384 the results of the regression performed to exam- levels between cultural settings. 420

385 ine associations with anxiety symptoms. Job strain Overall, an increase in job strain was shown to be a 421

386 was positively associated with anxiety symptoms. significant predictor of higher levels of all three ele- 422

387 No significant associations were found between cop- ments of psychological distress (depression, anxiety, 423

388 ing resources, and anxiety symptoms. An interaction stress) in our sample of office workers (Hypothesis 424
 Table 3
Mixed effects linear regression analysis of job strain and coping resources with psychological distress symptoms (n = 220)
Explanatory variables Outcome variables
Depression Anxiety Stress
Coeff. 95% CI P value Coeff. 95% CI P value Coeff. 95% CI P value

Un
Age (per 10 years) –0.29 –0.60 to 0.03 0.07 –0.35 –0.67 to –0.03 0.03 –0.28 –0.72 to 0.14 0.20

D. Jun et al. / Job strain and psychological distress in office workers

Gender –0.43 –1.05 to 0.20 0.28 –0.27 –0.89 to 0.35 0.39 0.16 –0.69 to 1.02 0.71
BMI –0.02 –0.10 to 0.06 0.63 –0.06 –0.14 to 0.02 0.15 –0.04 –0.15 to 0.07 0.46
Drinking (>1–2 standard drink per day) 0.57 –0.66 to 1.80 0.37 1.51 0.28 to 2.73 0.02
Job insecurity

co
Satisfaction with physical work environment
0.16
–0.30
0.00 to 0.31
–0.85 to 0.25
0.047
0.29
0.11
–0.51
–0.04 to 0.26
–1.06 to 0.04
0.16
0.07

rre
Hours of computer work per week NI NI NI 0.04 –0.02 to 0.10 0.22
Break hours (>2 hrs per day) –0.64 –1.2 to –0.4 0.04 NI NI NI NI NI NI
Self-rated perceived health –0.54 –0.82 to –0.26 0.00 NI NI NI

cte
Physical activity levels (high activity level) NI NI NI
Moderate activity level 0.43 –0.25 to 1.10 0.21
Low activity level 0.95 0.05 to 1.84 0.04
Job strain 0.55 0.21 to 0.88 0.00 0.50 0.18 to 0.82 0.00 0.75 0.32 to 1.18 0.00
Control coping
Escape coping
Social support
Interaction effects
Job strain × Escape coping
–0.19
0.79
0.22 d
–0.50 to 0.11
0.46 to 1.12
–0.11 to 0.54
0.21
0.00

Au
0.19
Job strain x Escape coping × Social support
0.14
0.26
0.01
–0.17 to 0.44
–0.04 to 0.55
–0.31 to 0.32
0.38
0.09
0.97
Job strain × Escape coping
0.12
0.88
–0.21
–0.30 to 0.54
0.47 to 1.29
–0.64 to 0.22
No interaction effect
0.58
0.00
0.34

tho
0.38 0.05 to 0.71 0.03 0.20 –0.09 to 0.50 0.17
Job strain × Social support 0.11 –0.21 to 0.44 0.50 NI NI NI
Escape coping × Social support 0.15 –0.15 to 0.46 0.33 NI NI NI
Job strain × Escape coping X Social support 0.25 –0.01 to 0.51 0.06 NI NI NI
Adj. R-squared 0.35

rP 0.19 0.17
Interaction model effects: Interaction models were developed with three-way interaction terms (job strain × escape coping × social support) for depression symptoms and two-way interaction
terms (job strain × escape coping) for anxiety symptom model. An interaction model was not applied for stress symptoms due to non-significant effects of interaction terms. NI: explanatory
variables not included in the model.

roo
Note: The X-axis ranges from –3 to 3 by z-score transformed scale of job strain (originally range from –74 to –2). One standard deviations above the mean (z-score) and below the mean were
created for escape coping and social support to simulate four combination of coping strategies: 1) lower escape coping use(lower E)+less social support (less S), 2) higher escape coping use (higher

f
E)+less social support (less S), 3) lower escape coping use (lower E)+greater social support (greater S), 4)higher escape coping (higher E)+greater social support (greater S).

7
 8 D. Jun et al. / Job strain and psychological distress in office workers

425 1). However, interaction effects of coping resources have found when only low levels of job strain existed 475

426 impacted the relationship between job strain and [6, 16]. 476

427 these three elements of psychological distress dif- Although social support did not directly predict 477

428 ferently (Hypothesis 2). For depression symptoms, distress symptoms, it interacted with job strain and 478

429 the enhancing effect of job strain on depression escape coping to predict depression symptom. When 479

430 symptoms increased as the levels of escape coping lower levels of job strain existed, the workers reported 480

431 increased, while the enhancing effect reduced as the similar levels of depression symptoms, irrespective 481

432 levels of social support increased. On the other hand, of the levels of social support. However, the dif- 482

f
433 the enhancing effect of job strain on anxiety symp- ference between workers with depression symptoms 483

roo
434 toms was only moderated by escape coping. increased as levels of job strain increased. This 484

buffering effect of social support was consistent with 485

435 4.1. Comparisons with other studies previous studies [5, 34]. In contrast, a reverse buffer- 486

ing effect of social support on distress symptom was 487

436 This is the first study to examine the effect of reported when it interacted with escape coping. The 488

rP
437 coping resources on psychological distress in office buffering effect of social support was not presented 489

438 workers using a work-oriented coping assessment when higher levels of escape coping were used. This 490

439 scale. Our findings were consistent with previous finding may infer that social support would not work 491

440 studies which reported that job strain and escape anymore as a stress buffer when it interacts with 492

tho
441 strategies were positively associated with the level emotional coping, but stress catalyst. An exceptional 493

442 of psychological distress [6, 9, 13, 15]. However our case for the buffer hypothesis of social support with 494

443 findings did not support previous findings that con- regards to persistent fatigue at work was previously 495

444 trol coping [11, 15] and social support [31, 32] are reported by Vanroelen et al. [35] in that strong solidar- 496

445 negatively related to psychological distress. This dis- ity in team work could encourage workers to provide 497
Au
446 crepancy may be due to methodological differences other coworkers the same or higher amount of social 498

447 in studies. Previous studies did not incorporate both support as they were given, which in return resulted 499

448 elements of coping resources (coping strategies and in increased responsibility. Above all, this antagonis- 500

449 social support) and other confounders within their tic effect is not in line with what has been reported 501

450 analysis. Instead, in our multivariable analyses the from the general buffer hypothesis [36]. 502

effect of control coping and social support was elim- The observed effect of control coping on distress
d

451 503

452 inated by the addition of other strong predictors such symptoms are inconsistent in previous studies. In 504
cte

453 as age, job strain, and escape coping. These results the present study, control coping neither interacted 505

454 highlight the necessity of incorporating both elements with job strain nor other coping resources to predict 506

455 of coping resources when investigating the relation- distress symptoms. This was consistent with previ- 507

456 ship between job strain and psychological distress. ous studies that reported control coping to not have 508

457 This study found out that each distress symptom a buffering effect on the relationship between work 509
rre

458 was predicted by different interactive composition stressors and burnout [9, 37]. In contrast, other stud- 510

459 of coping resources and stress factors. Of the two ies had found an interaction effect of control coping 511

460 aspects of coping resources evaluated, escape cop- on distress symptoms [38–40]. Further research to 512

461 ing was revealed as a core coping strategy interacting understand factors that may underpin these observed 513
co

462 with other coping resources, as well as job strain. inconsistencies between studies is warranted. 514

463 Escape coping interacted with job strain and social

464 support when predicting depression symptoms while 4.2. Implications 515

465 it interacted with job strain when predicting anxiety

Un

466 symptom. The highest depression and anxiety symp- The findings of this study infer practical impli- 516

467 toms were found when escape copings and job strain cations for workers to assist in the management 517

468 were also high. This finding is consistent with pre- of work-related psychological distress. Intervention 518

469 vious studies which reported the role of emotional programs that include training to control work-related 519

470 oriented coping negatively interacted with work stres- stress have shown effectiveness in the management of 520

471 sor to predict psychological distress [16, 33]. In stress [41]. This study suggests office workers may 521

472 addition, this study also found a minor buffering benefit from reducing their dependence on escape 522

473 effect of higher levels of escape coping on depres- coping in the work environment, as the buffering 523

474 sion and anxiety symptom, which few studies also effect of social factors and control coping seemed to 524
 D. Jun et al. / Job strain and psychological distress in office workers 9

525 be ineffective when escape coping was utilized. These Furthermore, the results indicated the moderating 572

526 findings are of relevance to both industry and research effect of coping resources may influence the three ele- 573

527 settings. In particular, it is suggested that assessment ments of psychological distress (stress, depression, 574

528 of office workers should consider a comprehensive and anxiety) differently. A key implication from this 575

529 battery of measures that include job strain, coping study is that coping resources should be considered 576

530 resources, as well as psychological distress. Lastly, when planning interventions for mitigating distress 577

531 this study reinforces the need to carefully consider job among office workers. 578

532 strain in office workers as it is likely to have a direct

f
533 detrimental effect on their psychological wellbeing.

roo
Institution and Ethics approval and informed 579
534 4.3. Limitation consent 580

535 Some limitations of this study should be acknowl- All participants provided written consent to partic- 581
536 edged. Psychological distress is multifactorial ipate and ethical approval for the study was granted

rP
582
537 incorporating physical, mental and psychosomatic by the institute Human Research Ethics Committee 583
538 factors [41, 42]. Although the present study rep- in the University of Queensland (Approval number; 584
539 resents an advancement in the field through the 2014000308). 585
540 examination of coping among a range of other fac-

tho
541 tors, there may be other factors influencing distress
542 mechanisms that were beyond the scope of this study.
Funding 586
543 Additionally, the coping strategies addressed in this
544 study were only considered from a work perspective
This study was supported by the Grant sponsor: 587
545 without consideration of coping strategies outside
Office Ergonomic Research Committee, Minneapo-
Au 588
546 of the work environment. Social factors outside of
lis, USA; Grant number: none. The authors wish to 589
547 the workplace (e.g., marital status, dependents) were
thank the committee for the great support by financing 590
548 also not included. While this was intentional to focus
instruments, and participation reimbursement. SMM 591
549 specifically on workplace factors it could represent a
is supported by the Grant sponsor: National Health 592
550 study limitation. The inclusion of two cultural groups
and Medical Research Council of Australia Fellow- 593
may be considered a strength of the study; however,
d

551
ship; Grant number: APP1090440. 594
552 findings may not be generalizable to dissimilar cul-
cte

553 tures or other types of occupational environments. It

554 is important to note that the findings are based on a
555 relatively small sample size for a study of this nature. Conflict of interest 595

556 This study was also restricted to pain-free healthy

557 office workers and observed relationships between The authors declare no conflicts of interest. 596
rre

558 factors considered in this study may be different in the

559 presence of pain or illness [43]. However, this study
560 design was appropriate for successfully addressing Acknowledgments 597

561 the stated research aims to examine the relationship

co

562 between job strain, coping, and psychological distress This research was supported by Kyungsung Uni- 598

563 in office workers without the confounding effect of versity Research Grants in 2019. 599

564 pain or illness.

Un

References 600
565 5. Conclusion
[1] Janwantanakul P, Sitthipornvorakul E, Paksaichol A. Risk 601
566 This study has provided new evidence highlighting factors for the onset of nonspecific low back pain in office 602
567 the direct relationship between job strain and psycho- workers: A systematic review of prospective cohort studies. 603

568 logical distress in office workers across two different Journal of Manipulative and Physiological Therapeutics. 604
2012;35(7):568-77. 605
569 cultures. However, the results also indicated that [2] Stansfeld S, Candy B. Psychosocial work environment and 606
570 office workers’ coping resources (coping strategies mental health–a meta-analytic review. Scandinavian Journal 607
571 and social support) may moderate this relationship. of Work, Environment & Health. 2006;32(6):443-62. 608
 10 D. Jun et al. / Job strain and psychological distress in office workers

609 [3] Theorell T, Hammarstrom A, Aronsson G, Traskman Bendz [21] Karasek R, Theorell T. Healthy Work: Stress, Productivity, 674
610 L, Grape T, Hogstedt C, et al. A systematic review including and the Reconstgruction of Working Life. New York, USA: 675
611 meta-analysis of work environment and depressive symp- Basic Books Inc.; 1990. 676
612 toms. BMC Public Health. 2015;15:738. [22] Landsbergis PA, Schnall PL, Warren K, Pickering TG, 677
613 [4] Karasek RA. Job demands, job decision latitude, and mental Schwartz JE. Association between ambulatory blood pres- 678
614 strain: Implications for job redesign. Administrative Science sure and alternative formulations of job strain. Scandinavian 679
615 Quarterly. 1979;24(2):285-308. Journal of Work, Environment & Health. 1994;20(5):349- 680
616 [5] Choi B, Ostergren PO, Canivet C, Moghadassi M, Lindeberg 63. 681
617 S, Karasek R, et al. Synergistic interaction effect between [23] Eum K-D, Li J, Jhun H-J, Park J-T, Tak S-W, Karasek R, 682
618 job control and social support at work on general psycho- et al. Psychometric properties of the Korean version of the 683

f
619 logical distress. International Archives of Occupational and job content questionnaire: Data from health care workers. 684

roo
620 Environmental Health. 2011;84(1):77-89. International Archives of Occupational and Environmental 685
621 [6] Day AL, Livingstone HA. Chronic and acute stressors Health. 2007;80(6):497-504. 686
622 among military personnel: Do coping styles buffer their [24] Latack JC. Coping with job stress: Measures and future 687
623 negative impact on health? Journal of Occupational Health directions for scale development. The Journal of Applied 688
624 Psychology. 2001;6(4):348-60. Psychology. 1986;71(3):377-85. 689
625 [7] Greenglass ER, Burke RJ, Ondrack M. A gender-role [25] Havlovic SJ, Keenan JP. Coping with work stress: The influ- 690

rP
626 perspective of coping and burnout. Applied Psychology. ence of individual differences. Journal of Social Behavior 691
627 1990;39(1):5-27. and Personality. 1991;6(7):199. 692
628 [8] Monat A, Lazarus RS. Stress and Coping: An anthology. [26] Crawford JR, Henry JD. The Depression Anxiety Stress 693
629 New York, U.S.A: Columbia University Press; 1977. Scales (DASS): Normative data and latent structure in a 694
630 [9] Fortes-Ferreira L, Peiro JM, Gonzalez-Morales MG, Martin large non-clinical sample. The British Journal of Clinical 695
631 I. Work-related stress and well-being: The roles of direct Psychology. 2003;42(Pt 2):111-31. 696

tho
632 action coping and palliative coping. Scandinavian Journal [27] Group I. Guidlines for data processing and analysis of 697
633 of Psychology. 2006;47(4):293-302. the International Physical Activity Questionnaire (IPAQ). 698
634 [10] Holton MK, Barry AE, Chaney JD. Employee stress man- Available from: http://www.ipaq.ki.se/scoring.pdf2005. 699
635 agement: An examination of adaptive and maladaptive [28] Tibshirani R. Regression shrinkage and selection via the 700
636 coping strategies on employee health. Work (Reading, lasso. Journal of the Royal Statistical Society Series B 701
637 Mass). 2015;53(2):299-305. (Methodological). 1996;58(1):267-88. 702
Au
638 [11] Grossi G. Coping and emotional distress in a sample of [29] Akaike H. A new look at the statistical model identification. 703
639 Swedish unemployed. Scandinavian Journal of Psychology. IEEE Transactions on Automatic Control. 1974;19(6):716- 704
640 1999;40(3):157-65. 23. 705
641 [12] Hobfoll SE, Dunahoo CL, Ben-Porath Y, Monnier J. Gen- [30] Brambor T, Clark WR, Golder M. Understanding interaction 706
642 der and coping: The dual-axis model of coping. American models: Improving empirical analyses. Political Analysis. 707
643 Journal of Community Psychology. 1994;22(1):49-82. 2006;14(1):63-82. 708
644 [13] Smari J, Arason E, Hafsteinsson H, Ingimarsson S. Unem- [31] Bultmann U, Kant IJ, Van den Brandt PA, Kasl SV. Psy- 709
d

645 ployment, coping and psychological distress. Scandinavian chosocial work characteristics as risk factors for the onset 710
646 Journal of Psychology. 1997;38(2):151-6. of fatigue and psychological distress: Prospective results 711
cte

647 [14] Bonde JP. Psychosocial factors at work and risk of from the Maastricht Cohort Study. Psychological Medicine. 712
648 depression: A systematic review of the epidemiologi- 2002;32(2):333-45. 713
649 cal evidence. Occupational and Environmental Medicine. [32] Wang J. Perceived work stress and major depressive 714
650 2008;65(7):438-45. episodes in a population of employed Canadians over 18 715
651 [15] Lian Y, Gu Y, Han R, Jiang Y, Guan S, Xiao J, et al. Effect of years old. The Journal of Nervous and Mental Disease. 716
652 changing work stressors and coping resources on psycho- 2004;192(2):160-3. 717
rre

653 logical distress. Journal of Occupational and Environmental [33] Patterson GT. Examining the effects of coping and social 718
654 Medicine. 2016;58(7):e256-63. support on work and life stress among police officers. Jour- 719
655 [16] Bhagat RS, Allie SM, Jr F, L D. Coping with stressful life nal of Criminal Justice. 2003;31(3):215-26. 720
656 events: An empirical analysis. Washington, USA: Taylor & [34] Sanne B, Mykletun A, Dahl AA, Moen BE, Tell GS. Test- 721
657 Francis; 1995. ing the job demand-control-support model with anxiety 722
co

658 [17] Rick J, Guppy A. Coping strategies and mental health in and depression as outcomes: The hordaland health study. 723
659 white collar public sector employees. European Work and Occupational Medicine (Oxford, England). 2005;55(6): 724
660 Organizational Psychologist. 1994;4(2):121-37. 463-73. 725
661 [18] Lazarus RS. Toward better research on stress and coping. [35] Vanroelen C, Levecque K, Louckx F. Psychosocial work- 726
662 The American Psychologist. 2000;55(6):665-73. ing conditions and self-reported health in a representative 727
Un

663 [19] Cote P, van der Velde G, Cassidy JD, Carroll LJ, Hogg- sample of wage-earners: A test of the different hypothe- 728
664 Johnson S, Holm LW, et al. The burden and determinants ses of the Demand–Control–Support–Model. International 729
665 of neck pain in workers: Results of the bone and joint Archives of Occupational and Environmental Health. 730
666 decade 2000-2010 task force on neck pain and its associ- 2009;82(3):329-42. 731
667 ated disorders. Journal of Manipulative and Physiological [36] Karasek RA, Triantis KP, Chaudhry SS. Coworker and 732
668 Therapeutics. 2009;32(2 Suppl):S70-86. supervisor support as moderators of associations between 733
669 [20] Guzman J, Hurwitz EL, Carroll LJ, Haldeman S, Cote P, task characteristics and mental strain. Journal of Occupa- 734
670 Carragee EJ, et al. A new conceptual model of neck pain: tional Behaviour. 1982;3(2):181-200. 735
671 Linking onset, course, and care: The bone and joint decade [37] Leiter MP. Coping patterns as predictors of burnout: The 736
672 2000-2010 task force on neck pain and its associated disor- function of control and escapist coping patterns. Journal of 737
673 ders. Spine. 2008;33(4 Suppl):S14-23. Organizational Behavior. 1991;12(2):123-44. 738
 D. Jun et al. / Job strain and psychological distress in office workers 11

739 [38] De Rijk AE, Blanc PML, Schaufeli WB, De Jonge [41] Talbot F, Nouwen A. A review of the relationship between 752
740 J. Active coping and need for control as moderators depression and diabetes in adults: Is there a link? Diabetes 753
741 of the job demand–control model: Effects on burnout. Care. 2000;23(10):1556-62. 754
742 Journal of Occupational and Organizational Psychology. [42] Zapf D, Dormann C, Frese M. Longitudinal studies in orga- 755
743 1998;71(1):1-18. nizational stress research: A review of the literature with 756
744 [39] Koeske GF, Kirk SA, Koeske RD. Coping with job stress: reference to methodological issues. Journal of Occupational 757
745 Which strategies work best? Journal of Occupational and Health Psychology. 1996;1(2):145-69. 758
746 Organizational Psychology. 1993;66(4):319-35. [43] Katon W, Lin EH, Kroenke K. The association of depres- 759
747 [40] Ree E, Odeen M, Eriksen HR, Indahl A, Ihlebaek C, Hetland sion and anxiety with medical symptom burden in patients 760
748 J, et al. Subjective health complaints and self-rated health: with chronic medical illness. General Hospital Psychiatry. 761

f
749 Are expectancies more important than socioeconomic sta- 2007;29(2):147-55. 762

roo
750 tus and workload? International Journal of Behavioral
751 Medicine. 2014;21(3):411-20.

rP
tho
d Au
cte
rre
co
Un