You are here

Article Text

Article menu

Download PDFPDF

Workplace
Original article
The impact of job stress due to the lack of organisational support on occupational injury
  1. Mireia Julià1,2,
  2. Carlos Catalina-Romero3,
  3. Eva Calvo-Bonacho3,
  4. Fernando G Benavides1,4
  1. 1Centre for Research in Occupational Health (CiSAL), Universitat Pompeu Fabra, Barcelona, Spain
  2. 2Health Inequalities Research Group (GREDS-EMCONET), Universitat Pompeu Fabra, Barcelona, Spain
  3. 3Ibermutuamur (Mutual Insurance Company for Occupational Accidents and Diseases, Social Security no. 274), Madrid, Spain
  4. 4CIBER of Epidemiology and Public Health, Barcelona, Spain
  1. Correspondence to Mireia Julià, Centre for Research in Occupational Health (CISAL), Universitat Pompeu Fabra, C/ Doctor Aiguader 88, Primera planta, 08003 Barcelona, Spain; mireia.julia{at}upf.edu

Abstract

Objectives The aim of this study was to analyse the association between job stress and occupational injuries.

Methods A prospective cohort study in a sample of 10 667 workers belonging to the insured population of the Mutual Insurance Company in Spain. Job stress was assessed with the Spanish version of the Job Stress Survey. A 1-year follow-up of the workers’ clinical records was conducted to determine the incidence of occupational injuries, and the incidence rate per 1000 workers-year was calculated. The associations between the incidence of occupational injuries, job stress and job stress components (job pressure (JP) and lack of organisational support (LOS)) were assessed calculating the rate ratio (RR) and its CI of 95% using Poisson regression models.

Results After adjusting for confounders, a significant association between LOS and increased incidence of occupational injuries was found. Such an association was observed for the LOS index (RRa=3.11, 95% CI 1.53 to 6.31), LOS severity (RRa=2.64, 95% CI 1.31 to 5.33) and LOS frequency (RRa=2.67, 95% CI 1.32 to 5.38) scales in women. There was no significant association between job stress or its components and the incidence of occupational injuries among men.

Conclusions This prospective study found evidence of an association between the LOS and the incidence of occupational injuries in women, with potential implications for the prevention of accidents at work.

https://doi.org/10.1136/oemed-2012-101184

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    What this paper adds

    • Previous research suggests that job stress can increase the risk of occupational injury.

    • Scientific evidence is inconclusive due to the methodological limitations of previous studies (cross-sectional designs, self-reporting measures to assess occupational injuries).

    • The current study finds evidence of a significant association between lack of organisational support, one of the components of job stress, and the incidence of occupational injuries in women.

    • The results were obtained through a longitudinal study of a large sample with a wide representation of different occupations, and by assessing occupational injuries using data from workers’ clinical records.

    • Job stress and lack of organisational support should be included in the assessment and prevention of occupational injury.

    Introduction

    Occupational injuries (OI) represent a major public health problem.1 In the European Union, the annual incidence of OI leading to at least 3 days of sick leave among people who work or have worked in the previous year amounts to 32 per 1000, or 6.9 million people per year.2 In 2010 in Spain, there were 627 876 instances of OI with at least 1 day of sick leave, reflecting an incidence rate (IR) of 37.7 per 1000 workers registered with the social security system.3

    Research on the causes of OI has identified many risk factors at the individual level (eg, age or work experience), job level (eg, occupation, tasks or tools and equipment) and organisational level (eg, working environment, social support or the demands of the job).4 There is, however, less research that identifies job stress as a risk factor for OI.5 By contrast, job stress has been studied extensively in relation to a variety of health problems, such as cardiovascular disease,6 musculoskeletal illnesses,7 and mental disorders, such as depression.8 ,9

    One important study of the relationship between job stress and OI was a systematic review conducted over 15 years ago, which highlighted the existence of a positive association between stress and OI. It also pointed out the difficulties involved in extrapolating results from research conducted using different designs and assessment methods and including a very limited range of occupations.10 Studies carried out following publication of this review have shown an increase in the risk of suffering an OI due to various stress factors, such as working hours,11 time pressure,12 high levels of stress at work,13 ,14 high demands,15 ,16 low social support,16–18 ambiguity of roles,14 low decision latitude,19 low skill discretion,19 monotonous work,19 interpersonal conflicts at work,19 ,20 a highly variable workload 16 or insecurity about the job future.16 Several of these studies have also identified gender differences in the association between stress and OI, but the specific pattern of association among gender, job stress and OI has not yet been clarified.18 ,19 ,21

    As reported in the 1995 review, much of the research that has been conducted presents limitations that make it difficult to draw definitive conclusions. Most of the studies have employed a cross-sectional design.16 ,17 ,22 ,23 Both OI and job stress have frequently been assessed using self-reporting methods, collecting retrospective data after the injury has occurred.13 ,14 ,16–18 This presents a threat to internal validity, since workers who have suffered an OI may systematically report higher stress levels.23 Some studies have focused on very different cultural and economic contexts,20–22 and have generally tended to concentrate on specific sectors rather than exploring the association in a wider sample in which all occupations are represented.16 ,18 ,20 Other studies have presented limitations, such as small sample size,12 or a low response rate.15 Lastly, we have been unable to find studies that used reliably validated instruments for assessing job stress that have been adapted for use in standard professional practice, and that facilitate comparison with the results obtained in different countries, and application of the results to real-life contexts.16–18 ,20 ,22

    The aim of the present study was to assess the association between job stress and risk of OI in a sample of the Spanish working population in which both sexes, of all relevant age levels and all occupational sectors were widely represented, employing a prospective design and using a method for assessing job stress that has been validated and adapted for use in standard professional practice.

    Methods

    Study design

    This was a prospective cohort study. Subjects were workers selected when attending an occupational health check-up at the Ibermutuamur Prevention Society between May 2007 and May 2010. This study formed part of a larger research programme aimed at assessing exposure to psychosocial risks at work among the Spanish working population and the implications for health and business. Although this larger programme has employed stratified random sampling (with strata defined according to the variables sex, age, autonomous region in Spain, economic activity sector and occupational level), for the present analysis we used the sample that was employed to validate the scales in the Spanish version of the Job Stress Survey (JSS),24 which consisted of the first 10 672 workers recruited for the study for whom all the relevant data were complete. We considered that this sample size would enable us to detect an increase of one percentage point or more in the incidence of OI among workers exposed to high levels of job stress compared with non-exposed workers with a confidence level of 95% and a statistical power of 80%.

    An automated procedure was used to select subjects who were identified as eligible to participate in the study, that is, those whose medical examination admission data were complete in the multicentre electronic medical record. If a worker’s characteristics coincided with the parameters of a stratum sample pending completion, an electronic alert was activated and the field researcher provided information about the study. Subjects were selected if (1) their knowledge of the Spanish language allowed them to respond reliably to the self-administered questionnaires, (2) they voluntarily accepted to participate in the study and (3) they provided informed consent. The study protocol was approved by the clinical research ethics committee of the Parc Salut Mar in Barcelona. The study was conducted in strict compliance with the Declaration of Helsinki.

    Data collection

    The selected workers completed a self-administered questionnaire which collected demographic data and information about the company and the job, job stress and any sick leave taken in the previous 12 months. The questionnaire response rate was 70.5%. As soon as the subjects had been selected and had filled out the questionnaire, a computer programme was activated which monitored their medical history for the following 365 days, making it possible to determine whether or not the subject suffered an OI requiring heathcare and sick leave in the year following inclusion in the study.

    Variables

    The dependent variable was the first OI with sick leave of at least 1 day in addition to the day when the injury happened, which occurred during the follow-up year and which was reported to the Mutual Insurance Company. The independent variables included the different dimensions of job stress assessed by the Spanish version of the JSS.24 This questionnaire, which has been validated for use with the Spanish working population, lists the 30 most stressful situations that can arise at work, and assesses the degree of distress associated with each of them for the individual (severity) and the frequency with which they have occurred over the previous 6 months. The JSS gives scores for the overall level of job stress experienced by the worker (Job Stress) and for two subscales derived from factor analysis: the Job Pressure subscale assesses occupational stressors that reflect aspects of the job's structure, design or duties, such as working overtime, meeting deadlines and excessive paperwork, while the Lack of Organisational Support subscale reflects events involving other people and organisational policies or procedures, rather than specific aspects of the job itself, such as difficulties getting along with supervisors, poorly motivated coworkers and lack of opportunity for promotion.25 Three different scores were obtained for the Job Stress scale and the Job Pressure and Lack of Organisational Support subscales: Severity, Frequency and Index (Severity × Frequency) scores. The term index is used to refer to the result of considering the severity and frequency of stressful events in conjunction. For example, the Lack of Organisational Support Severity subscale assessed the average level of perceived severity of the 10 JSS stressor events that most directly relate to lack of organisational support (eg, low participation in policy-making decisions, difficulties with fellow workers or supervisors); the Lack of Organisational Support Frequency subscale assessed how often such events occurred during the last 6 months; and the Lack of Organisational Support Index assessed the amount of occupational stress (combined severity and frequency) that can be attributed to lack of organisational support.25

    The Spanish version of the JSS has shown that the test is reliable and valid. The internal consistency indices for the JSS scales and subscales were always higher than 0.85. Moreover, validation studies have shown that the JSS scores predict job satisfaction and self-reported episodes of sick leave.24 We considered the three occupational groups that had been used in the process of validation of the Spanish version of the JSS: manual (skilled agricultural and fishery workers, craftsmen, skilled manufacturing, construction and mining workers and unskilled workers in general), intermediate skilled (workers in the catering, personal, protective and sales services, and plant and machinery operators and assemblers), and office occupations (company and public administration managers, science and engineering professionals and technicians, support technicians and professionals and clerical workers). The rationale for using these occupational groups is that they significantly differ and can be categorised according to their responses in the JSS. That is, the same raw score obtained by two workers belonging to different occupational groups should be interpreted in a different way. A standard score was calculated for each worker:Embedded Image

    Standard scores were the result of subtracting the occupational group mean from a worker's raw score and then dividing the difference by the occupational group standard deviation. Then, standard scores were dichotomised, setting the cut-off at 1, that is, on the basis of this standard score, the levels were defined as high, when greater than or equal to 1, and low, when less than 1.

    Sex was used as a stratification variable, and both individual and job characteristics were considered as possible confounding variables. Individual characteristics included age (<24 years, 25–34 years, 35–44 years, 45–54 years, ≥50 years), educational level (elementary/primary, secondary, university), marital status (single, married/civil union, separated/divorced/widowed), children (yes/no), nationality (Spanish/foreign), body mass index (BMI) (<25 kg/m2; 25–29.99 kg/m2; ≥30 kg/m2) and episodes of sick leave in the 12 months prior to inclusion in the study (yes/no), while job characteristics included occupation (manual, not manual), type of contract (permanent, temporary), length of time in post (<6 months, 6 months–2 years, 2–5 years, 5–10 years, >10 years), economic activity sector (agriculture, livestock and fisheries; manufacturing; construction; services) and company size (<9 workers; 10–49 workers; 50–249 workers; 250–499 workers; ≥500 workers).

    Statistical analysis

    Subjects were excluded from the analysis if their data lacked information on any of the variables studied: date of accident (5 workers excluded), sex (35), age (15), educational level (126), marital status (79), children (165), nationality (5), BMI (117), episodes of disability in the past year (68), economic activity (199), company size (95), occupation (525) and length of time in post (65). The final sample studied included 10 667 workers.

    For the descriptive analysis, we calculated workers-day on the basis of the time elapsed during follow-up from entry into the cohort until the first OI with sick leave, where it existed. From this we calculated the IR per 1 000 workers-year according to the independent and confounding variables.

    To measure the association between job stress and the IR of OI, we calculated the rate ratio (RR) and its CI of 95%, using Poisson regression models to estimate average rates, controlling for overdispersion of data, and stratifying the analyses by sex. Poisson regression is a valid method for the analysis of survival data in occupational health, providing an alternative to proportional hazard models.26 As the denominator in the Poisson models we use workers-day for each variable. RR calculation was done in a longitudinal manner. Crude and adjusted models were obtained for all the confounding variables which were significantly associated with the incidence of OI in the bivariate analyses, to determine whether colinearity existed or not. Where it did exist, we incorporated the variable that had most significance in the bivariate analysis into the multivariate analysis. To know if effect modification existed, we calculated the interaction between sex and the different dimensions of job stress. Analyses were performed using SPSS V.15 and STATA SE, V.9.0.

    Results

    In the sample studied, 212 OI with sick leave were reported (table 1). The IR was 20.4 OI per 1000 workers-year, and this was significantly higher in men (24.1) than in women (12.3). Similarly, the lower the level of education, the higher the IR (44.2 OI per 1000 workers-year among workers with an elementary/primary education compared with 7.8 among workers with a university education), and a relationship was also observed with BMI, such that as BMI increased, the IR also tended to increase (17.5 in people with normal weight compared with 30.5 in obese people). Regarding nationality, foreign workers showed a much higher IR (42.3) than Spanish workers (19.8), and the IR was also much higher for workers who had had at least one episode of sick leave during the year prior to inclusion in the study (35.6) compared with those who had not taken any sick leave (17.0).

    View this table:
    Table 1

    Incidence rate of non-fatal occupational injury with sick leave per 1000 worker-years in a cohort of 10 672 workers during a 1-year follow-up Ibermutuamur, 2007–2010. Spain.

    With regard to occupational variables, the economic activities with the highest IR were construction (33.4) and manufacturing (25.6), compared with services (15.3). Companies with more than 500 workers had a lower IR (13.2) than companies with 50–249 workers (25.9). Manual workers had a higher IR (33.9) than non-manual workers (14.9). Regarding the type of contract, temporary workers had a much higher IR (26.8) than workers on a permanent contract (18.6).

    When stress levels were analysed (table 2), the highest IR in men were observed when Lack of Organisational Support Severity, Lack of Organisational Support Index and Job Stress Frequency were high (32.9, 32.3 and 32.1, respectively). High IR was observed among women when the Lack of Organisational Support Index, Frequency and Severity scores were high (27.1, 25.6 and 24.0, respectively).

    View this table:
    Table 2

    Job stress dimensions. Incidence rate of non-fatal occupational injury (OI) with sick leave per 1000 worker-years and its 95% CI in a cohort of 10 667 workers during a 1-year follow-up. Ibermutuamur, 2007–2010. Spain

    When the crude relationship between the stress dimensions and the incidence of OI was analysed, no significant associations were found for men between any of the JSS scores. However, we did find significant associations for women when levels of lack of organisational support were high. After the model had been adjusted, there were once again no significant associations between any of the JSS scores and the incidence of OI during the follow-up year for men (table 3). Among women, the significant association between Lack of Organisational Support and OI was maintained in all three cases: the Lack of Organisational Support Index (RRa=3.11, 95% CI 1.53 to 6.31) and the Lack of Organisational Support Severity (RRa=2.64, 95% CI 1.31 to 5.33) and the Lack of Organisational Support Frequency (RRa=2.67, 95% CI 1.32 to 5.38) subscales. Effect modifications were found in case of the Lack of Organisational Support Index (p=0.022) and the Lack of Organisational Support Frequency (p=0.033).

    View this table:
    Table 3

    Relationship between stress dimensions and risk of non-fatal occupational injury in men and women

    Discussion

    The results of this study show an association between lack of organisational support, one of the components of job stress, and the incidence of severe OI which require healthcare and sick leave. On the contrary, the association between job pressure and OI was not significant. In other words, the risk of such injuries is related primarily to organisational policies and culture or to support from colleagues and supervisors rather than to structural aspects of the job in terms of the amount or pace of work. This association was significant for women, but not for men. These results are consistent with previous studies that found an increased incidence of OI in workers exposed to stress factors associated with low levels of social support or interpersonal conflict at work.13 ,14 ,16–20

    Our findings are also consistent with previous research on gender differences in the association between job stress and OI. Several studies have provided evidence for the existence of such differences,18 ,19 ,21 ,27 although the data are much more contradictory about the specific pattern of association between the various stressors and the incidence of OI as a function of gender. For example, some authors have suggested that OI are related to sets of stressors which are partially different for each sex,16 ,18 ,19 whereas others have reported a relationship between job stress and OI in women but not in men.21 ,27 This latter, a study based on the Canadian Community Health Survey, used data from the largest sample analysed to date in this field, and was highly representative from the point of view of occupations included.27 Other studies have indicated an increased risk among women exposed to scant social support at work or in the home, or who experience interpersonal conflict at work.13 ,16 The heterogeneity of the results is not surprising given the vast differences between studies in regards to defining, operationalising and measuring the concepts of job stress and OI, the research design employed and the population studied.10

    Several hypotheses have been suggested regarding the causes of gender differences in the impact of job stress on health. On the one hand, there is the association between family demands and poorer health, which particularly affects women with a lower socioeconomic status.28 ,29 On the other, it has been suggested that women may be more vulnerable to stress than men, which could help explain epidemiological differences in psychiatric disorders, such as anxiety and depression.30 In light of this point of view, one might ask whether the mechanism responsible for the association between job stress and OI in women might be increased anxious-depressive symptomatology. The answers to these questions have important implications for prevention, and future research will be required to provide an answer.

    To our knowledge, only one study has examined the relative importance of the severity of the distress associated with job stress and the frequency with which the worker is exposed to such stress, finding a close relationship between the parameter severity and the incidence of OI in petrochemical workers in Taiwan.22 By contrast, the present results suggest that both the intensity of the distress associated with job stress and the frequency with which the latter is experienced are relevant. Again, it should be noted that the two studies have focused on very different populations from a cultural and occupational point of view, and have employed different methods for assessing job stress.

    The strengths of the present study include the fact that the size of the sample studied was larger than that which has been used in most previous research, and that all sectors and occupations were amply represented. An additional strength of the study is its prospective design, very rare in research on the possible association between job stress and OI. Furthermore, the use of clinical records to evaluate OI has enabled us to avoid recall bias, as well as other forms of bias related to common-method variance, since part of the association between stress and OI reported by many previous studies could be explained as being the result of using self-reporting methods to assess exposure and effect variables. Finally, in addition to Spain, the JSS has been validated in many other countries which facilitates comparison with the results obtained by different research teams worldwide, and is a tool designed for practical application in real-life professional contexts, thus facilitating the transfer of scientific knowledge beyond a strictly research context.24 ,25 ,31

    The most important limitation of the study concerns the possibility of selection bias, since all workers attended the medical examination voluntarily. Of these, we analysed all those who agreed to complete the questionnaire, and we do not know the possible differences which may have existed regarding exposure to job stress between the total number of workers who could have participated, all of whom, in principle, were covered, and those who agreed to participate. In fact, the IR observed in our study (20.4 per 1000) is lower than the overall rate for Spain in the same year (37.7 per 1000), using the same inclusion criteria and year. The composition of the sample was compared with the whole of the Spanish working population provided by the Spanish Labour Force Survey for the first quarter of 2006.32 Young workers (under 25 years old) and the agriculture, livestock and fishery sectors were slightly under-represented in our sample. Based on this supplementary information, we could hypothesise that workers from our sample are working in less dangerous workplaces than the average worker of the Spanish workforce, and it is possible that the association between job stress and OI could be stronger in a more representative sample. Another limitation of the current study, frequent in research about health outcomes of job stress, relates to the measurement of chronic versus acute stress. Despite the fact that JSS assesses worker exposure to sources of job stress during the former 6 months, it seems possible that the assessment of a longer period of exposure would be needed to increase the risk of suffering an OI. Thus, it could be important to assess job stress at least twice to identify workers with chronic exposure. Some studies have reported associations between chronic stress and health outcomes, like cardiovascular diseases, using such an approach.33 Furthermore, our study only included OI requiring healthcare and sick leave. Further research is needed to investigate the relationship between job stress and less severe OI (ie, injuries that do not require sick leave). Finally, the study design did not include psychological variables, like Eisenberger's concept of Perceived Organisational Support, that could mediate the effect of the exposure to stressful events related to a lack of organisational support and the incidence of OI.34

    Future research should determine the generalisability of the results to more disadvantaged or less health-conscious groups of workers, to identify the variables that potentially moderate the impact of organisational support, and finally it should test if organisational-level intervention programmes oriented to increase organisational support for employees are an effective strategy to prevent OI.

    References

    1. ↵
      1. Hamalainen P,
      2. Takala J,
      3. Saarela KL
      . Global estimates of occupational accidents. Saf Sci 2006;44:137–56.
      OpenUrlCrossRef
    2. ↵
      European Commission Employment, Social Affairs and Equal Opportunities (EUROSTAT). Health and safety at work in Europe (1999–2007). 2010. http://epp.eurostat.ec.europa.eu/portal/page/portal/product_details/publication?p_product_code=KS-31–09–290 (accessed 11 Sep 2012).
    3. ↵
      Instituto Nacional de Seguridad e Higiene en el Trabajo. Informe anual de accidentes de trabajo en España 2010. http://www.oect.es/Observatorio/Contenidos/InformesPropios/Siniestralidad/Ficheros/INFORME_ANUAL_AT2010.pdf (accessed 11 Sep 2012).
    4. ↵
      1. Khanzode VV,
      2. Maiti J,
      3. Ray P
      . Occupational injury and accident research: a comprehensive review. Saf Sci 2012;50:1355–67.
      OpenUrlCrossRef
    5. ↵
      1. Smith TD,
      2. DeJoy DM
      . Occupational injury in America: an analysis of risk factors using data from the General Social Survey (GSS). J Safety Res 2012;43:67–74.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Backe EM,
      2. Seidler A,
      3. Latza U,
      4. et al
      . The role of psychosocial stress at work for the development of cardiovascular diseases: a systematic review. Int Arch Occup Environ Health 2012;85:67–79.
      OpenUrlCrossRefPubMed
    7. ↵
      1. van Rijn RM,
      2. Huisstede BMA,
      3. Koes BW,
      4. et al
      . Associations between work-related factors and specific disorders of the shoulder—a systematic review of the literature. Scand J Work Environ Health 2010;36:189–201.
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Clays E,
      2. De Bacquer D,
      3. Leynen F,
      4. et al
      . Job stress and depression symptoms in middle-aged workers—prospective results from the Belstress study. Scand J Work Environ Health 2007;33:252–9.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Siegrist J
      . Chronic psychosocial stress at work and risk of depression: evidence from prospective studies. Eur Arch Psychiatry Clin Neurosci 2008;258(Suppl 5):115–9.
      OpenUrlCrossRefPubMed
    10. ↵
      1. Johnston JJ
      . Occupational injury and stress. J Occup Environ Med 1995;37:1199–203.
      OpenUrlCrossRefPubMed
    11. ↵
      1. Kirkcaldy BD,
      2. Trimpop R,
      3. Cooper CL
      . Working hours, job stress, work satisfaction, and accident rates among medical practitioners and allied personnel. Int J Stress Manag 1997;4:79–97.
      OpenUrlWeb of Science
    12. ↵
      1. Greiner BA,
      2. Krause N,
      3. Ragland DR,
      4. et al
      . Objective stress factors, accidents, and absenteeism in transit operators: a theoretical framework and empirical evidence. J Occup Health Psychol 1998;3:130.
      OpenUrlCrossRefPubMed
    13. ↵
      1. Wilkins K,
      2. Beaudet MP
      . Work stress and health. Health Rep 1998;10:47–62.
      OpenUrlPubMed
    14. ↵
      1. Hemingway MA,
      2. Smith CS
      . Organizational climate and occupational stressors as predictors of withdrawal behaviours and injuries in nurses. J Occup Organ Psychol 1999;72:285–99.
      OpenUrlCrossRef
    15. ↵
      1. Swaen GM,
      2. van Amelsvoort LP,
      3. Bultmann U,
      4. et al
      . Psychosocial work characteristics as risk factors for being injured in an occupational accident. J Occup Environ Med 2004;46:521–7.
      OpenUrlCrossRefPubMedWeb of Science
    16. ↵
      1. Nakata A,
      2. Ikeda T,
      3. Takahashi M,
      4. et al
      . Impact of psychosocial job stress on non-fatal occupational injuries in small and medium-sized manufacturing enterprises. Am J Ind Med 2006;49:658–69.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Trimpop R,
      2. Kirkcaldy B,
      3. Athanasou J,
      4. et al
      . Individual differences in working hours, work perceptions and accident rates in veterinary surgeries. Work Stress 2000;14:181–8.
      OpenUrlCrossRefWeb of Science
    18. ↵
      1. Kim HC,
      2. Min JY,
      3. Min KB,
      4. et al
      . Job strain and the risk for occupational injury in small- to medium-sized manufacturing enterprises: a prospective study of 1,209 Korean employees. Am J Ind Med 2009;52:322–30.
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Salminen S,
      2. Kivimaki M,
      3. Elovainio M,
      4. et al
      . Stress factors predicting injuries of hospital personnel. Am J Ind Med 2003;44:32–6.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Ghosh AK,
      2. Bhattacherjee A,
      3. Chau N
      . Relationships of working conditions and individual characteristics to occupational injuries: a case-control study in coal miners. J Occup Health 2004;46:470–80.
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. Murata K,
      2. Kawakami N,
      3. Amari N
      . Does job stress affect injury due to labor accident in Japanese male and female blue-collar workers? Ind Health 2000; 38:246–51.
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      1. Li CY,
      2. Chen KR,
      3. Wu CH,
      4. et al
      . Job stress and dissatisfaction in association with non-fatal injuries on the job in a cross-sectional sample of petrochemical workers. Occup Med 2001;51:50–5.
      OpenUrlAbstract
    23. ↵
      1. Rundmo T
      . Perceived risk, safety status, and job stress among injured and noninjured employees on offshore petroleum installations. J Safety Res 1995; 26:87–97.
      OpenUrlCrossRefWeb of Science
    24. ↵
      1. Spielberger CD,
      2. Vagg PR
      . (adaptación española, Catalina C). JSS Cuestionario de Estrés Laboral. Madrid: TEA Ediciones, 2010.
    25. ↵
      1. Spielberger CD,
      2. Vagg PR
      . Job Stress Survey (JSS). Professional manual. Odessa, FL: Psychological Assessment Resources, 1991.
    26. ↵
      1. Callas PW,
      2. Pastides H,
      3. Hosmer DV
      . Empirical comparisons of proportional hazards, poisson, and logistic regression modeling of occupational cohort data. Am J Ind Med 1998;33:33–47.
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      1. Wilkins K,
      2. Mackenzie SG
      . Work injuries. Health Rep 2007;18:25–42.
      OpenUrlPubMed
    28. ↵
      1. Artazcoz L,
      2. Borrell C,
      3. Benach J
      . Gender inequalities in health among workers: the relation with family demands. J Epidemiol Community Health 2001; 55:639–47.
      OpenUrlAbstract/FREE Full Text
    29. ↵
      1. Artazcoz LL,
      2. Borrell C,
      3. Benach J,
      4. et al
      . Women, family demands and health: the importance of employment status and socio-economic position. Soc Sci Med 2004;59:263–74.
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      1. Sandanger I,
      2. Nygård JF,
      3. Sørensen T,
      4. et al
      . Is women's mental health more susceptible than men's to the influence of surrounding stress? Soc Psychiatry Psychiatr Epidemiol 2004;39:177–84.
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      1. Spielberger CD,
      2. Saranson IG
      , eds. Stress & emotion: anxiety, anger and curiosity. Washington, DC: Taylor & Francis, 1999.
    32. ↵
      Instituto Nacional de Estadística. Encuesta de Población Activa 2006. http://www.ine.es/jaxi/menu.do?L=0&type=pcaxis&path=/t22/e308_mnu&file=inebase (accessed 9 Mar 2013).
    33. ↵
      1. Chandola T,
      2. Britton A,
      3. Brunner E,
      4. et al
      . Work stress and coronary heart disease: what are the mechanisms? Eur Heart J 2008;29:640–8.
      OpenUrlAbstract/FREE Full Text
    34. ↵
      1. Rhoades L,
      2. Eisenberger R
      . Perceived organizational support: a review of the literature. J Appl Psychol 2002;87:698–714.
      OpenUrlCrossRefPubMedWeb of Science

    Footnotes

    • Contributors All the authors jointly participated in the data analysis and drafting of the manuscript, and all are in agreement with the text and findings. The paper has been read and approved by all authors.

    • Funding Fellowship in the Center for Research in Occupational Health (CiSAL), Universitat Pompeu Fabra.

    • Competing interests None.

    • Patient consent Obtained.

    • Ethics approval Committee of the Clinical Research Ethics, Parc Salut Mar (Barcelona).

    • Provenance and peer review Not commissioned; externally peer reviewed.