- Top of page
- SUBJECTS AND METHODS
The psychological and health benefits of employment have long been recognized in addition to its economic benefit. Work is an important source of identity, self esteem, and competence (1–6). Loss of employment has been associated with increased depression and anxiety, reduced physical functioning, and mortality (7–13). Among individuals with arthritis, there is ample evidence linking the pain and disability of the disease to giving up employment (14–24). The ramifications of this are considerable given that arthritis is the leading cause of long-term physical disability in adults, that the number of people with arthritis disability is expected to nearly double by 2020, and that half of this increase is expected to occur among individuals who are in their highest earning years (ages 45–65 years) (25–28).
A focus on job loss has meant that relatively little is known about the number and range of work transitions that people with arthritis use to remain employed and the perceptions, expectations, and work factors associated with these job changes. Specifically, few studies have examined occasional lost work hours (such as using sick days or vacation time) to manage arthritis, changes to the type and nature of work (such as changing one's job or forgoing a promotion), and permanent changes to work hours (such as going from full-time to part-time work). What research exists finds that people with rheumatoid arthritis (RA) and osteoarthritis (OA) often report comparable indirect costs of illness (29), that there is large variability in the number of sick days reported (30), that employers rarely initiate workplace accommodations (24, 31), and that the factors important to remaining at work may relate less to disease and more to age and work characteristics (32).
This article examines work changes that people attribute to their arthritis. Of interest is the nature and number of work transitions people report and how they relate to people's arthritis workplace activity limitations, their perceived capacity, future employment expectations, and their decision whether or not to disclose their arthritis to their employer. Previous studies, drawing on theories of behavior change and adaptation, have found that people's perceptions of their current capacity and their expectations for the future are important in understanding decisions and behavior changes (33–42). People assess potential outcomes associated with their decisions and attempt to maximize their gains and minimize their losses (33, 34). Moreover, variables associated with behavior change, particularly those assessing people's perceptions, may interact with one another to jointly predict change.
For example, work changes may require an employer's endorsement before they can be implemented. As a result, employer knowledge of a person's arthritis is expected to be associated with more work transitions. The potential gain is that, by disclosing to their employer, people may be able to make work changes that would enable them to remain working. The potential risks are that an employer will not make job accommodations, that perceptions of the person as a “good” employee will change, or that disclosing one's condition will contribute to having to ultimately forgo employment. For this reason, individuals may decide to disclose their arthritis only when they perceive it is necessary. This includes when they believe arthritis has significantly changed their capacity to perform activities and meet goals and when they have concerns that they may not be able to remain employed in the future unless changes are implemented (43, 44). From a methodologic point of view, it suggests that disclosure to an employer, perceptions of current capacity, and future employment expectations will not be independently associated with work changes, but will interact with one another to predict work transitions.
As noted, research suggests that people will view employment as an important role and will make efforts to remain working. If individuals attempt to minimize work-related losses due to arthritis, a greater percentage would report occasional lost work hours to deal with arthritis (such as sick days, extended breaks at work, and vacation time) than they would permanent changes to work hours or to the type or nature of work. However, having arthritis may put limits on people's capacity to minimize their losses. Hence, arthritis-related workplace activity limitations are expected to be associated with work transitions. Finally, given the importance of the work role, depressive affect is expected to relate to work changes, with more work changes being associated with greater depression.
- Top of page
- SUBJECTS AND METHODS
Table 1 presents sample characteristics for the study. On average, participants were 51 years old and more than three-quarters were women. More than half the respondents were married and most were relatively well educated with annual incomes exceeding $40,000. One-third of participants reported inflammatory arthritis (n = 165). Of these, 145 (88%) had RA, the remainder having other types of inflammatory arthritis, such as psoriatic arthritis. More than half the sample had OA and 10% had both OA and RA. Nearly 30% of respondents reported not taking any prescription medication for their arthritis, with most of these individuals having OA (81.1%). On average, respondents had arthritis for 9 years and most had utilized health services for their condition in the past year. More than two-thirds of respondents worked in business, finance, health, science, art, or sports fields. The remainder worked in sales, service, trades, transportation, or equipment operating occupations. Approximately one-quarter of respondents had not told their employer/supervisor about their condition and 11.0% anticipated leaving their job in the next year because of their arthritis.
Table 1. Sample characteristics (n = 492)*
| Female||383 (77.8)|
| Male||109 (22.2)|
|Age, mean ± SD years||50.9 ± 9.3|
|Marital status|| |
| Married/living as married||298 (60.6)|
| Divorced/separated/widowed||118 (24.0)|
| Single||76 (15.4)|
| Secondary school or less||85 (17.3)|
| Some postsecondary||113 (23.0)|
| Postsecondary||197 (40.0)|
| Postgraduate||95 (19.3)|
|Household income, $|| |
| <39,999||91 (18.5)|
| 40,000–69,999||147 (29.9)|
| 70,000–99,999||119 (24.2)|
| ≥100,000||93 (18.9)|
|Physician diagnosis|| |
| Inflammatory||165 (33.5)|
| Osteoarthritis||278 (56.5)|
| Both||49 (10.0)|
|Duration, mean ± SD, years||9.2 (8.8)|
| DMARD||125 (25.4)|
| NSAID||283 (57.5)|
| Corticosteroids||42 (8.5)|
| Biologics||11 (2.2)|
| Pain medication||51 (10.4)|
|Health utilization|| |
| Family physician||397 (80.7)|
| Rheumatologist||203 (41.3)|
| Surgeon||115 (23.0)|
| Other medical specialist||64 (12.8)|
| Physical therapist||147 (29.9)|
| Chiropractor||78 (15.9)|
| Naturopath||37 (7.3)|
| Other (e.g. Arthritis Society, RMT, acupuncturist)||73 (12.8)|
|Job sector|| |
| Business, finance, administration||164 (33.3)|
| Health, science, art, sport||175 (35.6)|
| Sales and service||102 (20.7)|
| Trades, transportation, equipment operator||50 (10.2)|
|Employer knowledge of arthritis|| |
| Knows||334 (67.9)|
| Does not know||108 (22.0)|
|Future job expectations|| |
| Anticipate leaving job||54 (11.0)|
| Do not anticipate leaving job||425 (86.4)|
Table 2 presents percentages of respondents reporting the 7 work transition behaviors. On average, participants reported 1.6 work changes (SD 1.6) with >70% of participants reporting at least 1 arthritis-related work change. Approximately 28% reported 1 work transition, 18.3% reported 2 work transitions, and 24.6% reported ≥3 work transitions.
Table 2. Type and percentage of work transitions reported in the past 6 months due to arthritis*
|Occasional loss of work hours|| |
| Work interrupted ≥20 minutes||188 (38.2)|
| Days absent, mean ± SD||4.56 ± 12.48|
| 0||298 (60.6)|
| 1||25 (4.9)|
| 2||37 (7.5)|
| ≥3||131 (26.8)|
| Use vacation days||47 (9.6)|
|Change to type/nature of work|| |
| Change type of work||99 (20.1)|
| Unable to seek promotion or job transfer||48 (9.8)|
| Unable to take on extra projects or responsibilities||109 (22.2)|
|Permanent change of work hours|| |
| Change work hours||105 (21.3)|
|Respondents reporting no work transitions||144 (29.3)|
|Total work transition score, mean ± SD||1.60 (1.56)|
More respondents reported occasional lost work hours in the past 6 months than changes to the type of work or permanent reductions in work hours. Specifically, both work interruptions and at least 1-day absenteeism were reported by ∼40% of respondents. Among those who reported absenteeism (n = 193), 67.9% reported being away from work 3 or more days in the past 6 months. The mean number of days absent among those reporting arthritis-related sick days was 11.6 days (SD 17.8).
Approximately 1 in 5 respondents made changes to their type of work at some point in their work life, reported being unable to take on extra projects or work responsibilities in the past 6 months, or reported reducing their work hours in the past 6 months. Using vacation days to manage arthritis and being unable to seek a promotion or job transfer in the past 6 months was reported by 10% and 9.8% of respondents, respectively. However, for both items there was a high percentage of “not applicable” responses (13.6% and 31.7%).
Bivariate regression analyses examined the relationship of total work transition scores to demographic, work limitation, and psychological variables (see Table 3). The results show that older respondents and those with OA reported fewer work transitions than younger respondents and those with inflammatory arthritis. Respondents with both inflammatory arthritis and OA reported significantly more work transitions than individuals with inflammatory arthritis alone. In addition, participants reporting greater workplace limitations, whose employers knew about their arthritis, who reported greater changed capacities and goals, had greater depression scores, and who anticipated leaving their job in the next year reported more work transitions.
Table 3. Bivariate unstandardized (b) and standardized (beta) regression coefficients for explanatory variables by total transitions score
|Marital status||Married or living as married|| || || |
| Widowed, divorced, separated|| ||−0.04||−0.01||0.79|
| Never married|| ||0.18||0.04||0.38|
|Education||Postgraduate|| || || |
| Secondary or less|| ||−0.41||−0.10||0.08|
| Some postsecondary|| ||−0.32||−0.09||0.14|
| Postsecondary|| ||−0.11||−0.03||0.58|
|Household income, $||≥100 000|| || || |
| ≤39,999|| ||−0.02||−0.01||0.93|
| 40,000–69,999|| ||−0.07||−0.02||0.74|
| 70,000–99,999|| ||0.16||0.05||0.47|
|Diagnosis||Inflammatory|| || || |
| Noninflammatory|| ||−0.41||−0.13||0.01|
| Both|| ||0.50||0.10||0.05|
|Occupational field||Business, finance, administration|| || || |
| Health, science, art, sport|| ||−0.19||−0.06||0.45|
| Sales and service|| ||0.16||0.04||0.43|
| Trades and transportation|| ||0.18||0.06||0.48|
|Employer knows about arthritis||No||0.60||0.16||0.00|
|Job involves physical activity|| ||0.05||0.04||0.37|
|Workplace limitations|| ||0.17||0.48||0.00|
|Changed capacities and goals|| ||0.14||0.48||0.00|
|Future job expectations||No||1.24||0.25||0.00|
Standard multivariate linear regression analysis examined the association of the predictor variables to total work transitions. Variables significant at the P < 0.10 level in the bivariate analyses were included in the multivariate analysis. The final model accounted for 38% of the variance (see Table 4). Younger respondents and those who reported greater workplace limitations were more likely to make work changes. Work changes were also associated with increased reports of depression. Diagnosis was not significantly associated with work transitions, although there was a trend for people with both OA and inflammatory arthritis to report more changes to work than those with inflammatory arthritis alone (P < 0.06).
Table 4. Multivariate unstandardized (b) and standardized (beta) regression coefficients for explanatory variables by total work transitions score
|Diagnosis||Inflammatory|| || || |
| Noninflammatory|| ||−0.05||−0.02||0.70|
| Both|| ||0.44||0.08||0.06|
|Employer knows about arthritis||No||0.90||0.25||0.18|
|Workplace limitations|| ||0.08||0.22||0.00|
|Changed capacities and goals|| ||0.10||0.33||0.00|
|Future job expectations||No||4.67||0.98||0.08|
|Future job expectations × employer knows about arthritis|| ||−5.54||−1.10||0.08|
|Future job expectations × changed capacities and goals|| ||−0.17||−1.05||0.09|
|Employer knows about arthritis × changed capacities and goals|| ||−0.04||−0.28||0.19|
|Future job expectations × employer knows about arthritis × changed capacities and goals|| ||0.23||1.30||0.05|
There was a significant main effect for changed capacities and goals (P < 0.001), as well as trends in the 2-way interactions: employer awareness and future expectations (P < 0.08) and changed capacities and goals and future expectations (P < 0.09). However, a significant 3-way interaction necessitates a closer examination of this effect. To facilitate understanding, the interaction was graphed (see Figures 1 and 2). Changed capacities and goals scores were dichotomized on the mean (25, >25). Figure 1 depicts the relationship of work transitions, employer awareness of arthritis, and people's perceived changed capacities and goals when respondents expected to remain employed in the future.
Figure 1. Three-way interaction of mean number of work transitions by employer awareness of arthritis and perceptions of changed capacity and goals when employees expect to remain employed.
Download figure to PowerPoint
Figure 2. Three-way interaction of mean number of work transitions by employer awareness of arthritis and perceptions of changed capacity and goals when employees expect to give up employment.
Download figure to PowerPoint
The results revealed that people who perceived their arthritis had a large impact on their capacity, yet also believed that they would be able to remain employed, made approximately the same level of work transitions, regardless of whether or not they had told their employer about their condition. Those believing arthritis had less of an impact made fewer changes overall. However, among these individuals, people who had told their employers about their condition reported somewhat more changes than those not disclosing to their employer.
A different pattern emerged when respondents anticipated leaving their job. The greatest number of work changes was reported by individuals who perceived a big impact of arthritis on their capacity and who had told their employer about their condition. Among those who anticipated leaving, but who perceived their arthritis had little impact, the opposite was true. They made more work transitions when they had not told their employer than when their employer was aware of their condition.
- Top of page
- SUBJECTS AND METHODS
This study extends research on arthritis and employment by looking at the ways people deal with the demands of their condition and their job by making work changes. Previous epidemiologic studies have focused on giving up employment and often have been unable to directly link work changes to arthritis. Studies using clinical samples of RA or ankylosing spondylitis patients have also emphasized work loss. The findings of the present study revealed that most respondents had made work transitions and >40% had made >1 change. The findings also point to the utility of including a greater range of personal and psychological variables to understand the impact of arthritis on employment in addition to disease and work context variables.
To frame the research, we drew on theories of behavior change and adaptation and anticipated that people would try to minimize workplace losses from arthritis (33). Support for this hypothesis was found in that respondents were more likely to report occasional lost work hours than permanent decreases in hours or changes to the type of work. Sick days and extended breaks are often available and less likely to have an impact on one's career and financial situation than other work changes. Participants may also prefer the flexibility afforded by sick days and breaks because they can draw upon them when needed and avoid them at other times. At the same time, more than one-quarter of participants noted absences from work of >3 days in the past 6 months. The interview may have captured a time of more intense disease activity among some respondents. However, if it is a common occurrence, it may not be sustainable in the long term. Longitudinal research needs to assess whether occasional lost work hours is related to other types of work transitions, including giving up employment. It would also be helpful to know whether the respondents lost wages as a result of occasional lost work hours.
As noted, previous studies have linked employment to people's well-being. In this study, respondents who reported more work transitions reported more depressive symptoms. Because the study is cross-sectional, the direction of the findings is not clear. However, other research finds that, after controlling for initial mental health, people who make work changes show more negative psychological wellbeing (12). The findings are also in keeping with research on RA in which depression was associated with work disability (50). It suggests that greater attention needs to be paid to the effects of work transitions on the psychological well-being of people with arthritis. If employment-related depression is significant, it becomes an additional factor to deal with in efforts to help people remain at or return to work.
Age was also associated with work transitions. Younger workers reported more arthritis-related work changes than older workers. In other arthritis research, younger workers were better able to remain employed (32). This may be because younger individuals were more motivated, having had fewer years of experience in the workplace and wanting to maintain a sense of identity as productive. The financial incentive for younger workers to remain working is also likely to be substantial. Older workers may perceive that they are more limited in workplace options to deal with arthritis and, instead, may pursue options such as early retirement. Older workers may also view arthritis as age normative and make fewer coping efforts to deal with it (43).
Not surprisingly, the more difficulty respondents reported with workplace activities, the more work changes they made. Occupational field, overall level of physical activity at work, and diagnosis were not significantly related to work transitions in the multivariate analyses. To understand work changes, these findings point to the importance of looking at the fit between a person's job demands and their physical capacity. By asking respondents about the amount of difficulty they experienced with specific job tasks (such as reaching, standing, sitting for long periods, scheduling demands, and the pace of work), we gained insight into the fit of the job and the individual. The concept of job fit has also been discussed in other theories of work behavior in terms of the match between vocational demands and personality (5).
The findings also suggest that to understand work changes, we need to assess people's perceptions, expectations, and decisions to disclose arthritis. The relationships among these variables were complex and did not point simply to benefits or costs to disclosing one's health to an employer. For example, most respondents expected to remain employed in the next year. Among them, more work transitions were made by people who perceived their arthritis had a substantial impact on their capacity and goals, regardless of whether their employer knew about their condition. This suggests that many respondents had the flexibility to make some work changes when their arthritis threatened their capacity, without concern that they would suffer future job recriminations. Interestingly, individuals who believed they would remain employed and whose arthritis had little impact reported more changes when they had told their employer about their condition than when they had not. This may reflect people's perception that it was safe to disclose the information; it may also reflect proactive efforts to maintain employment being made by employees and employers.
Among those expecting to leave their jobs, more work changes were reported when people believed their arthritis substantially affected their capacity and when they had told their employer. The greater number of work changes suggests that difficulties managing arthritis and employment may have reached a critical threshold in which participants or their employers did not see employment as viable in the future. Additional research needs to examine whether this was related to disease severity or specific contextual features of the workplace, including the nature of the job, workplace support, or workplace policies.
Finally, differences emerged in work changes related to future expectations among those who had not told their employers about their arthritis and who perceived little impact from their condition. On average, those who expected to remain employed made fewer work changes than those expecting to leave. The findings need to be replicated. It may be that individuals who made work changes in the absence of a substantial perceived impact of arthritis and without telling their employer about their condition were more likely to be viewed by others as malingering, poor team players, or as not committed to their jobs because they had made work changes in the absence of information that would explain their decisions. This may relate to their expectation that they would be leaving their job in the future. Research examining workplace climate and interpersonal relationships would be useful to illuminate these findings.
Several limitations in this research need to be addressed in future studies. First, although our recruitment strategy enabled us to generalize beyond a clinical sample and was comparable to other samples in terms of the age, education, marital status, income, and occupation of participants, the sample was a purposive one. The type and extent of work transitions should be replicated in other studies and patterns of work changes examined in longitudinal studies. Examining patterns of work transitions would enable us to learn about the relationship of occasional and permanent changes to work hours and of the type of work to employment loss. In addition, the results should be replicated in samples with more men and that include a greater range of health status and arthritis symptom severity measures. Also important would be to look at other changes and accommodations people with arthritis or their employers make to help individuals remain employed. These include coping efforts, such as modifying the way tasks are performed, help from coworkers, use of assistive devices, and ergonomic changes to workplace environments.
Having acknowledged some of its limitations, this study extends research on arthritis and employment by looking at ways people change their work to manage their condition and employment, as well as factors associated with work changes. The findings highlight the dynamic interplay among the arthritis limitations people report, their perceptions, and the role that others like employers play in changes to employment. They also suggest the utility of including a greater range of personal and psychological variables in addition to disease and work context variables to understand the impact of arthritis on employment.