Introduction
he World Health Organization (WHO) estimates that 740 million out of 1.17 billion people with disabilities worldwide are in the working-age group and also, 2%-4% of them have functional difficulties [1]. These people constitute approximately 15% of the total population of each country [2]. In developing countries, such as Iran, the unemployment rate for people with disabilities of working age is 80%-90% [3].
Employment is positively correlated with workers’ health and well-being [4, 5]. Each profession should be matched with the abilities of the individual; otherwise, he may feel disappointed and somehow rejected [6, 7]. Physically Disabled People (PhDP) face more socioeconomic inequalities, communication restrictions [8], and physical barriers [9] by entering the job market, and also they experience more stress [10] and health disparities [8, 11 , 12, 13] than healthy people. Thus, employment as a mechanism to partially compensate for the physical and social deficiencies of disabled people is highly important for them because of reconstructing daily life [14] and promoting their quality of life.
Studies on the effect of employment type on health in healthy people suggested that part-time employments are associated with poorer health for both genders, whereas they are linked to more adverse consequences for women [4, 16]. Ross and Mirowsky [5] considered both social causation and selection effect for employment respecting the health of people. They stated that both genders with full-time jobs had better health than others [5]. However, little is known about the health impact of Employment Status (ES) and Occupation Type (OT) on disabled employees. 
Occupational Stress (OS) can be considered as the accumulation of stressors and job-related situations, in which most people agree that they are stressors [17]. It is well-known that OS plays a significant role in exacerbating workers’ depression and lowering their General Health (GH) level [18, 19]. However, an important issue regarding disabled people is that on the one hand, disability can be a stressor and on the other hand, OS can be a risk factor for disability [20]. Therefore, these people experience an increased risk for physical and psychological complications of OS [10]. However, few studies have explored the relations between OS and health status among disabled employees. 
Several studies have attempted to determine the factors affecting different aspects of disabled people’s health. Ali et al. noted that social problems had a positive association with anxiety and depression of intellectually disabled people [21]. In addition, it has been shown that miscomprehension of disabled employees’ special needs might result in higher rates of job burnout and stress [6] and lower GH [6, 22, 23].
Disabled people are in a worse situation in terms of stress, health, and socio-demographic indicators than healthy people [8, 11 ,12, 13]. Considering that little knowledge is available about the impact of working conditions of this group, such as employment status and occupation type on their GH and OS, especially in employees with physical disabilities, we investigated the effect of employment status, occupation type, and other factors in predicting GH. We also examined the relations between OS and GH while adjusting for demographic factors among physically disabled employees in vocational centers of the State Welfare Organization of Tehran.
Materials and Methods 
This research was a descriptive-analytical study with a cross-sectional design, which was done from April 2019 to September 2019. According to the latest data (2015), there were 23 vocational centers for PhDP in Tehran [24], of which nine centers were selected for this study. They covered 800 PhDP. According to the Morgan table, selecting 260 samples from this statistical population has sufficient power [25]. Therefore, with respect to the probability of losing samples during the study, 273 samples were selected through the convenience sampling technique. These workrooms were under the supervision of the State Welfare Organization of Iran (SWOI) to empower these people, providing supportive employment [26]. People who had any type of physical disabilities who aged more than 20 years old with more than six months of work experience were included in the study. Having work experience of less than six months and mental retardation or severe neurological and psychiatric disorders were exclusion criteria. 
The health of subjects was checked with Goldberg’s (1979) 28-item General Health Questionnaire (GHQ) [27]. The GHQ-28 comprises four subscales of somatic symptoms, anxiety and insomnia, social dysfunction, and depression symptoms. A high score indicates poor GH and vice versa. Its validity and reliability were assessed in a study done by Tabatabaei (2003) on Iranian PhDP (Cronbach’s alpha=0.96) [7].
Moreover, we used the Occupational Stress Index (OSI) developed by Belkic (1991) based on the cognitive ergonomics concept with 58 questions [28]. In addition to career information, it incorporates five key potential stressors: physical condition of the workplace, accidents and injuries, time pressure, workplace problems, and decision-making situations [28]. A score higher than 43 indicates high stress and vice versa. The validity and reliability of the Persian version of this scale were evaluated by Tabatabaei et al. (Cronbach’s alpha=0.78) [29]. Socio-demographic attributes, such as gender, marital status, ES, and OT were also asked by another questionnaire designed by the authors. 
We conducted a t-test and one-way ANOVA to analyze the differences in OSI and GHQ scores of respondents respecting their gender, OT, marital status, and ES. Furthermore, backward stepwise multiple linear regression analyses were done on each subscale to test the relationships between the GHQ’s subscales (dependent variables) and the OSI’s subscales (independent variables) while adjusting for demographic variables. OT is in four categories, which are entered in the model with three dummy variables. If only one of the variables were significant, all were kept in the model. Assumptions of normality and linearity were confirmed using Q-Q and partial plots. The independence of errors was confirmed using the Durbin-Watson test. Variance Inflation Factor (VIF) less than 5 confirmed collinearity and was not an issue in the model. Moreover, using *ZRESID and *ZPRED plots, homoscedasticity was confirmed. The R2 defines the proportion of variance of the response that can be explained by the independent variables. Analyses were performed using SPSS (IBM Corp., Armonk, NY, Ver. 24.0) at the significance level of P<0.05.
Results 
The Mean±SD age of the PhDP was 33.65±3.76 years and most of them aged 30 to 39 years and were (55.2%), men (60.4%), married (60.4%), engaged in productive jobs (30%) and had part-time employment (67%). Also, 23.8% of them were employed in art-related jobs, 23.1% in industrial jobs and the remaining were simple workers (23.1%). Kolmogorov-Smirnov test revealed that assumptions for parametric tests could be met (P<0.05). Therefore, two independent samples t-test and one-way ANOVA were then conducted.
The results of Table 1 revealed that there was a significant difference (P<0.05) between mean scores of some of the subscales of GH and most of OS subscales with respect to the gender and marital status of PhDP without adjusting for other variables. 


Women had more somatic symptoms, social dysfunction, and OS. Given the higher mean values in married subjects, they experience more OS, while they were found with a lower level of somatic symptoms and depression than the single subjects. Moreover, there were significant differences in GH and OS subscales regarding ES. 
Table 2 indicates a significant difference (P<0.001) between mean scores of GHQ and OSI subscales in different OTs.


Then, using Tukey’s post hoc test, a pair-wise comparison was done for these subscales in different occupation groups (Table 3).


Somatic symptoms in industrial jobs employees had a significant difference with other groups (P<0.001). Art-related jobs were found with a higher rate of anxiety and social dysfunction (P<0.05). The level of depression was different in art-related, industrial, and productive groups, significantly (P<0.001). Furthermore, the post hoc test showed that the overall GH status of PhDP in industrial and art-related jobs was better than employees of productive jobs and simple workers (P<0.001). However, more somatic symptoms were observed in these groups. In addition, people who were engaged in industrial jobs had a higher level of OS than other occupations (P<0.001). Also, the OS level was higher in simple workers than in cases with productive jobs (P<0.001).
Multiple linear regression analyses were then performed on data to identify GH predictors (Table 4). 


Coefficients estimated changes in the score of related subscales associated with each independent variable. For example, in somatic symptoms, with a 1-year increase in age, the mean score enhances by 0.16, considering other variables constant. However, this increase in age does not make a significant change in the other subscales of GH. Disabled women had a higher score of somatic symptoms than men (β=1.97). Marital status had no significant effect on GH subscales. Furthermore, the results showed that part-time employment is directly associated with somatic symptoms and depression symptoms (β=1.26 and 1.51). In addition, productive jobs were significantly associated with a higher level of somatic symptoms, anxiety and insomnia, and depression symptoms (β=1.28, 1.53, and 1.95, respectively). Therefore, being female, employing as a part-time worker, and working in productive jobs are significant predictors of somatic and depressive symptoms in physically disabled employees. Regarding the effect of OSI subscales, the physical condition of the workplace, workplace problems, and decision-making situations were strongly associated with somatic symptoms. Overall, the results suggest that the best-fit model was for somatic symptoms (R2=49%), which means that 49% of the variance in somatic symptoms could be explained by independent variables (Table 4).
Discussion and Conclusion 
This study focused on employees with physical disabilities with the aim of examining the relationship between their demographic factors and OS and GH levels and also investigating the predictive effect of ES and OT. Our results demonstrated that females had a higher level of somatic symptoms than males. Accordingly, disabled women were more active in the workplace and exposed more to improper physical conditions, and faced more decision-making situations than men. These two variables predicted somatic symptoms, positively. This combination leads to more OS among women than men, which decreases the level of GH. These results are in close agreement with those obtained by Kozak et al. [7], Tsai [15], and Rajati et al. [30]. Disabled women suffer a double disadvantage because of both their gender and disability [12, 31, 32]. Our findings contradict the results of Tsai [15] who reported no significant differences in health subscales between genders. A possible explanation for this contradiction can be the different nature of the work done by the employees in our study and the mentioned study [15]. Our participants were employed in workrooms of the SWOI [26] doing industrial, productive, and art-related jobs or working in cleaning affairs, whereas Tsai’s [15] sample group were employed in an electronic manufacturing company with monotonous assembly tasks.
Despite higher OS in married PhDP, we found that marital status had no significant effect on GH. This result is inconsistent with another study [15], which showed better GH in married subjects. This contradiction might be due to scaling the GH, which was considered as a subscale in the mentioned study [15] whereas, a separate scale containing four subscales in our study. According to the regression analysis (Table 4), marital status had a weaker effect on GH in comparison with other independent variables and then, was eliminated from the model.
Additionally, part-time employees are more likely subjected to harmful physical/ergonomic factors in the working environment and time pressure, and subsequently, more OS than full-time workers. Consequently, they had a higher rate of somatic and depressive symptoms. These results are consistent with those of Ali et al. [21], who showed a positive relationship between lower employment opportunities, anxiety, and depression. These findings were also similar to those of Kozak et al. [7], Lysaght et al. [23], and Salguero et al. [33]. Pagan [34] found that disabled people in Europe prefer to be employed as part-time employees to keep their balance between work and health. Similarly, most of our participants were part-time employees (67%). However, the difference between our study and Pagan’s research is that part-time is not a selection done by our participants. This contradiction might be attributed to the socio-economic factors specific to developing countries, such as Iran, where there are limited job opportunities and more competition in occupying job positions. Thus, our part-time participants may force themselves to have more involvement in their work, representing their qualifications. However, as recommended by Pagan [34, 35], increasing part-time and self-employment opportunities should be encouraged by policy-makers to provide better accommodation between disability and work while improving their employment rate and levels of health.
Moreover, the results indicated that participants working in productive jobs, such as sewing due to fewer movements required to do their works, were more exposed to improper physical conditions, which is positively related to their higher rate of somatic symptoms. PhDP doing the art-related works, such as making handicrafts and potteries and coloring them, were in the second rank with respect to anxiety. In addition, exposure to more anxiety and interruption in social functions may cause more OS in this group, influencing their GH. This effect is attributable to their work nature, which is based on customers’ tact. Depending on the number and variety of orders, they experience more work pace, insufficient time, and problems related to conflicting work. More depression in simple workers may be derived from different types of tasks they have to do. They are more isolated from others, having lower social relationships, whereas experiencing noticeable injuries. Similarly, Niedhammer and Chastang [36] suggested that high psychological demands and low social support are related to the first depressive episode.
Based on our results, higher levels of OS had a direct association with higher scores of GH subscales, supporting the general assumption that somatic symptoms and anxiety can be predicted through stress. Other studies have supported this exposure-response relationship [15, 18]. This finding, which reflects the differences in the somatic symptoms and health of PhDP by their type of occupation, could in some way be complementary to the causes of the inequalities in the disability-induced health outcomes among the different groups of disabled people who were studied by Rowland et al. [37]. 
Our study had some limitations. Its design was cross-sectional, which only allows assessing the associations. Moreover, the workrooms were dispersed and the exact number of the covered workers was not accessible precisely. Participants had different types of impairment; hence, some of them had difficulty in filling the questionnaires and refused to participate in the study. Furthermore, they considered unnecessarily to include the effects of other socio-demographic factors, such as residing in the city or countryside, income [15, 32], and impairment type [12] on their health and employment opportunities.
Therefore, it seems that providing full-time employment opportunities and considering PhDP’ specific work nature is critical in the vocational rehabilitation process through job redesign interventions, which address their unique characteristics. Because this study was done on PhDP in workrooms of the SWOI, caution should be taken about applying the results for other PhDP who are employed in other parts of the labor market.
ES and OT are significant predictors of OS, causing GH disparities among PhDP. Regarding a higher level of somatic symptoms and depression in part-time employees, it seems that changing their ES to full-time, can reduce their stress load and improve their health. In addition, women and married subjects experience more OS and have more somatic symptoms. This indicates the need of accommodating their mental and physical workload with their capabilities. Identifying these risk factors enables rehabilitation and ergonomics professionals to apply related principles through designing appropriate interventions to modify the work situations of this special group leading to optimizing their health and well-being.

Ethical Considerations
Compliance with ethical guidelines
This study was approved by the Ethics Review Committee of Tehran University of Medical Sciences (IR.TUMS.SPH.REC.1395.18.67). Prior to the study, all participants were informed about the aim of the study and signed the consent form. In addition, the confidentiality of the personal and research data was ensured.

Funding
This research did not receive any grant from funding agencies in the public, commercial, or non-profit sectors.

Authors' contributions
Conceptualization: Seyed Abolfazl Zakerian, Saeed Ghaneh-Ezabadi and Kianoush Abdi; Methodology: Seyed Abolfazl Zakerian, Saeed Ghaneh-Ezabadi and Leila Nezamabadi-Farahani; Validation: Seyed Abolfazl Zakerian, Saeed Ghaneh-Ezabadi, Leila Nezamabadi-Farahani and Kianoush Abdi; Data analysis: Saeed Ghaneh-Ezabadi and Leila Nezamabadi-Farahani; Investigation: Seyed Abolfazl Zakerian, Saeed Ghaneh-Ezabadi, Leila Nezamabadi-Farahani and Maral Saadat; Resources: Saeed Ghaneh-Ezabadi, Maral Saadat and Kianoush Abdi; Initial draft preparation: Saeed Ghaneh-Ezabadi and Leila Nezamabadi-Farahani; Editing and review: Saeed Ghaneh-Ezabadi, Kianoush Abdi and Seyed Abolfazl Zakerian; Project administration and supervision: Seyed Abolfazl Zakerian and Kianoush Abdi.

Conflict of interest
The authors declare that there is no conflict of interests.


References

1. World Health Organization. World report on disability [Internet]. 2011 [Updated 14 December 2011]. Available from: https://www.who.int/disabilities/world_report/2011/en/
2. Hosseini A. [Traditional rehabilitation in the village case: disabled and aged persons in the villages of tehran province (Persian)]. Social Development and Welfare Planning. 2011; 2(2):33-15. https://qjsd.atu.ac.ir/article_3632_6a50b3b216b43b38a2d5aeb0316562c2.pdf
3. United Nations. Factsheet on Persons with Disabilities [Internet]. 2007 [Updated 2007]. Available from: https://www.un.org/development/desa/disabilities/resources/factsheet-on-persons-with-disabilities.html
4. Rodriguez E. Marginal employment and health in Britain and Germany: Does unstable employment predict health? Social Science & Medicine. 2002; 55(6):963-79. [DOI:10.1016/S0277-9536(01)00234-9]
5. Ross CE, Mirowsky J. Does employment affect health? Journal of Health and Social Behavior. 1995; 36(3):230-43. [DOI:10.2307/2137340] [PMID]
6. Tabatabaei S. Study of self esteem and mental health of orthopaedically handicapped in relation to their employment status in Iran. Chandigarh, India: Panjab University; 2003.
7. Kozak A, Kersten M, Schillmöller Z, Nienhaus A. Psychosocial work-related predictors and consequences of personal burnout among staff working with people with intellectual disabilities. Research in Developmental Disabilities. 2013; 34(1):102-15. [DOI:10.1016/j.ridd.2012.07.021] [PMID]
8. Mithen J, Aitken Z, Ziersch A, Kavanagh AM. Inequalities in social capital and health between people with and without disabilities. Social Science Medicine. 2015; 126:26-35. [DOI:10.1016/j.socscimed.2014.12.009] [PMID]
9. Abdi K, Arab M, Rashidian A, Kamali M, Khankeh HR, Farahani FK. Exploring barriers of the health system to rehabilitation services for people with disabilities in Iran: A qualitative study. Electronic Physician. 2015; 7(7):1476-85. [DOI:10.19082/1476] [PMID] [PMCID]
10. Merz MA, Bricout JC, Koch LC. Disability and job stress: Implications for vocational rehabilitation planning. Work. 2001; 17(2):85-95. https://content.iospress.com/articles/work/wor00175
11. Ayoubi Avaz K, Hashemi O, Karami S, Rassafiani M, Hatami R. [Comparison of general health and depression in the people covered with the Community-Based Rehabilitation program (CBR) With the Non-Covered (Persian)]. Archives of Rehabilitation. 2018; 19(1):54-63. [DOI:10.21859/jrehab.19.1.54]
12. Kavanagh AM, Krnjacki L, Aitken Z, LaMontagne AD, Beer A, Baker E, et al. Intersections between disability, type of impairment, gender and socio-economic disadvantage in a nationally representative sample of 33,101 working-aged Australians. Disability and Health Journal. 2015; 8(2):191-9. [DOI:10.1016/j.dhjo.2014.08.008] [PMID]
13. Reichard A, Stolzle H, Fox MH. Health disparities among adults with physical disabilities or cognitive limitations compared to individuals with no disabilities in the United States. Disability and Health Journal. 2011; 4(2):59-67. [DOI:10.1016/j.dhjo.2010.05.003] [PMID]
14. Donnell CM, Mizelle ND, Zheng Y. Consumers of vocational rehabilitation services diagnosed with psychiatric and substance use disorders. Journal of Rehabilitation. 2009; 75(3):41-9. https://web.b.ebscohost.com/abstract?
15. Tsai SY. Health-related quality of life association with work-related stress and social support among female and male disabled employees. Women Health. 2016; 56(8):957-76. [DOI:10.1080/03630242.2016.1145172] [PMID]
16. Pirani E, Salvini S. Is temporary employment damaging to health? A longitudinal study on Italian workers. Social Science & Medicine. 2015; 124:121-31. [DOI:10.1016/j.socscimed.2014.11.033] [PMID]
17. Quick JC, Quick JD, Nelson DL, Hurrell Jr JJ. Stress in organizations. Preventive stress management in organizations. Washington, DC, US: American Psychological Association; 1997. [DOI:10.1037/10238-001]
18. Khamisa N, Oldenburg B, Peltzer K, Ilic D. Work related stress, burnout, job satisfaction and general health of nurses. International Journal of Environmental Research and Public Health. 2015; 12(1):652-66. [DOI:10.3390/ijerph120100652] [PMID] [PMCID]
19. Zhong JIE, You J, Gan Y, Zhang Y, Lu C, Wang H. Job stress, burnout, depression symptoms, and physical health among Chinese university teachers. Psychological Reports. 2009; 105(3_suppl):1248-54. [DOI:10.2466/PR0.105.F.1248-1254] [PMID]
20. Szymanski EM. Disability, job stress, the changing nature of careers, and the career resilience portfolio. Rehabilitation Counseling Bulletin. 1999; 42(4):279-89. https://eric.ed.gov/?id=EJ600816
21. Ali A, King M, Strydom A, Hassiotis A. Self-reported stigma and symptoms of anxiety and depression in people with intellectual disabilities: Findings from a cross sectional study in England. Journal of Affective Disorders. 2015; 187:224-31. [DOI:10.1016/j.jad.2015.07.046] [PMID]
22. Lustig DC, Strauser DR, Donnell C. Quality employment outcomes: Benefits for individuals with disabilities. Rehabilitation Counseling Bulletin. 2003; 47(1):5-14. [DOI:10.1177/00343552030470010201]
23. Lysaght R, Fabrigar L, Larmour-Trode S, Stewart J, Friesen M. Measuring workplace social support for workers with disability. Journal of Occupational Rehabilitation. 2012; 22(3):376-86. [DOI:10.1007/s10926-012-9357-1] [PMID]
24. Center IIS. Number of rehabilitation services units of Welfare Organization of Tehran province during 2011-2015. Tehran: Intelligent Information Solutions Center; 2015.
25. Krejcie RV, Morgan DW. Determining sample size for research activities. Educational and Psychological Measurement. 1970; 30(3):607-10. [DOI:10.1177/001316447003000308]
26. Welfare MoCLaS. The state welfare organization of Iran Tehran, Iran [Internet]. 2020 [Updated 2020 April 13]. Available from: https://www.mcls.gov.ir/fa/sazemanhayetabehe
27. Goldberg DP, Hillier VF. A scaled version of the General Health Questionnaire. Psychological Medicine. 1979; 9(1):139-45. [DOI:10.1017/S0033291700021644] [PMID]
28. Belkic K, Savic C. The Occupational Stressor Index (OSI): A comprehensive model derived from cognitive ergonomics for clinical practice. Job stressors and mental health. Singapore: World Scientific Publishing Co. Pte. Ltd.; 2013. [DOI:10.1142/8914]
29. Tabatabaei S, Khani Jazani R, Kavousi A, Bahreini M. Relationship between ergonomic workstations with musculoskeletal disorders and job stress among staff of Bank in Tehran city. In: G. DV, Lightner N, editors. Advances in Human Factors and Ergonomics in Healthcare and Medical Devices. Los Angeles, California, USA: Springer; 2017. [DOI:10.1007/978-3-319-60483-1_39]
30. Rajati F, Ashtarian H, Salari N, Ghanbari M, Naghibifar Z, Hosseini SY. Quality of life predictors in physically disabled people. Journal of Education and Health Promotion. 2018; 7:61. [DOI:10.4103/jehp.jehp_115_17] [PMID] [PMCID]
31. Kavanagh AM, Krnjacki L, Beer A, Lamontagne AD, Bentley R. Time trends in socio-economic inequalities for women and men with disabilities in Australia: Evidence of persisting inequalities. International Journal for Equity in Health. 2013; 12(1):73. [DOI:10.1186/1475-9276-12-73] [PMID] [PMCID]
32. Mondéjar-Jiménez J, Vargas-Vargas M, Meseguer-Santamaría M-L, Mondéjar-Jiménez JA. Impact of social factors on labour discrimination of disabled women. Research in Developmental Disabilities. 2009; 30(6):1115-23. [DOI:10.1016/j.ridd.2009.07.013] [PMID]
33. Salguero A, Martínez-García R, Molinero O, Márquez S. Physical activity, quality of life and symptoms of depression in community-dwelling and institutionalized older adults. Archives of Gerontology & Geriatrics. 2011; 53(2):152-7. [DOI:10.1016/j.archger.2010.10.005] [PMID]
34. Pagan R. Part-time work among older workers with disabilities in Europe. Public Health. 2009; 123(5):378-83. [DOI:10.1016/j.puhe.2009.02.010] [PMID]
35. Pagán R. Self‐employment among people with disabilities: Evidence for Europe. Disability & Society. 2009; 24(2):217-29. [DOI:10.1080/09687590802652504]
36. Niedhammer I, Chastang JF. Psychosocial work factors and first depressive episode: retrospective results from the French national SIP survey. International Archives of Occupational and Environmental Health. 2015; 88(7):835-47. [DOI:10.1007/s00420-014-1013-y] [PMID]
37. Rowland M, Peterson-Besse J, Dobbertin K, Walsh ES, Horner-Johnson W, Andresen E, et al. Health outcome disparities among subgroups of people with disabilities: A scoping review. Disability and Health Journal. 2014; 7(2):136-50. [DOI:10.1016/j.dhjo.2013.09.003] [PMID]