Examination of arthritis-related work place activity limitations and intermittent disability over four-and-a-half years and its relationship to job modifications and outcomes




To examine the type, degree, and episodic nature of arthritis-related work place activity limitations and the consistency of the relationship of activity limitations to job modifications and work place outcomes.


Using an interviewer-administered structured questionnaire, individuals with osteoarthritis (OA) or inflammatory arthritis (IA) were interviewed at 4 time points, 18 months apart. At baseline, all participants (n = 490; 381 women, 109 men) were employed. Respondents were recruited using community advertising and from rheumatology and rehabilitation clinics. The Workplace Activity Limitations Scale (WALS) assessed arthritis-related disability with job tasks. Job modifications/accommodations (e.g., scheduling changes), work place outcomes (e.g., absenteeism), demographics, illness, and work context were also measured. Repeated cross-sectional logistic regressions examined levels of WALS disability with job modifications and outcomes at each time point.


Similar levels of activity limitations were found comparing OA and IA with fewer difficulties with global aspects of work (e.g., scheduling) than with specific tasks (e.g., working with hands). Three-quarters of the participants had episodic or intermittent WALS difficulty over time. Medium and high levels of work place activity limitations were significantly associated with job modifications, and high WALS difficulty was consistently related to negative work outcomes.


Many individuals with arthritis report some difficulty with work place activities. However, these difficulties are often intermittent and may not result in changes to work productivity until they are consistently high. This is important for designing work place interventions and for employers, insurers, and the government to understand to avoid viewing individuals with arthritis as a permanent drain on work place and health resources.


There is considerable interest in understanding the work ing experiences of individuals with arthritis. Evidence links different types of arthritis to work loss and highlights that disability and productivity costs may be 2–4 times the direct costs of managing these conditions (1–10). Research also finds that arthritis-related work loss is only partly attributable to clinical features and disease symptoms. Health factors combine with work context variables (e.g., job type), the environment (e.g., work place access), interpersonal relationships (e.g., work place support), and psychological factors (e.g., job stress) to predict work transitions such as absenteeism, reducing work hours, changing jobs, and giving up work (11–18). For example, individuals working with arthritis can experience numerous stressors at work. Some relate to managing symptoms of the disease such as fatigue and pain. Others relate to worries about remaining employed and career planning; decisions whether to disclose one's condition to others; balancing work, health, and family; symptom unpredictability; and difficulties with work place tasks and the scheduling or pace of work (12, 13, 19–27). Emerging from this research is the need to better identify the types of work place activities with which individuals living with arthritis report difficulties in order to inform work place interventions and prevent at-work disability and job loss (6, 13, 19, 26, 28, 29).

Also important to understand is the unpredictable and potentially variable course of arthritis at work. Many individuals with arthritis note periods of wellness interrupted by episodes of more serious impairment and disability that can last for considerable time. As such, arthritis can be characterized as a chronic episodic condition. The intermittent nature of arthritis-related activity limitations and intraindividual variability in these activities has not been examined at work. We also know little about the relationship of work place activity limitations to the use of self-management and job accommodations. One hypothesis is that episodic activity limitations will result in people refraining from using job modifications or asking for help in the hopes that difficulties will be alleviated over time without recourse to work place accommodations (i.e., a “wait and see” approach). Therefore, it may only be when workers report high levels of activity limitations that they use accommodations, and this may place them at risk for negative work outcomes such as absenteeism, job interruptions, and difficulties with colleagues.

Recently, a number of measures have been developed examining diverse aspects of working with chronic diseases, and studies have assessed their reliability, validity, and responsiveness in individuals with arthritis (12, 13, 29–38). However, few have examined disability (i.e., activity limitations) at work. One exception is the Workplace Activity Limitations Scale (WALS), which measures limitations related to lower extremity mobility, upper extremity functioning, the pace and scheduling of work, and concentration at work, and has been used with individuals reporting osteoarthritis (OA), rheumatoid arthritis (RA), and lupus (24, 30, 33, 39). Results in 3 samples found good internal consistency, validity, and responsiveness to perceived changes in work ability (24, 30, 33, 39). Cross-sectional comparisons showed that greater work place activity limitations were significantly associated with being more likely to disclose one's condition to others, difficulty balancing health and work roles, more job stress, lost productivity, and self-management behaviors (6, 12, 13, 24, 25, 33, 39). Longitudinal research found that, controlling for a range of factors, greater work place activity limitations were associated with subsequently reporting more depression and job disruptions (14, 40). However, intraindividual variability in work place activity limitations and the level or degree of work place activity limitations that is generally tolerated by individuals without making job modifications compared to that which is consistently related to modifications and outcomes has not been examined.

This study examines the types and degree of activity limitations reported by individuals with OA and inflammatory arthritis (IA) at 4 separate time points spanning 4.5 years, and investigates whether at-work activity limitations are relatively stable or intermittent over time. The association of different levels of work place activity limitations to the use of job modifications (e.g., scheduling changes, assistive devices) is examined in order to explore whether individuals report using job modifications at the first signs of activity limitations or whether they report their use only when activity limitations are more pronounced. Also examined is the level at which activity limitations are consistently related to outcomes such as absenteeism, lost time at work, and at-work interpersonal difficulties.



Individuals with OA or IA residing in southwestern Ontario, Canada, were interviewed 4 times (T1–T4), each 18 months apart (2000–2006), as part of a longitudinal study examining coping and adaptation efforts that persons with arthritis use to remain employed. The sample was purposive to ensure diversity across occupations and that individuals receiving fewer health care services were not systematically excluded. Respondents were recruited from community advertisements, with additional respondents from rheumatology and rehabilitation clinics and The Arthritis Society, Ontario Division. Eligibility criteria were a physician diagnosis of OA or IA, arthritis duration of at least one year, paid employment at baseline (T1), no comorbid conditions causing physical disability (e.g., multiple sclerosis), and fluency in English.


Participants were screened for eligibility using a telephone-screening questionnaire. Eligible participants were administered a face-to-face, in-depth structured questionnaire lasting ∼2 hours at their home or a location of their choice. All interviewers completed standardized training supplemented by regular meetings. A small honorarium (Canadian $20) was provided at each time point. Ethics approval was received and informed written consent was obtained at each interview.


Work place activity limitations.

The 12-item WALS asked about arthritis-related activity difficulties at work. Previous studies have reported on the 11-item WALS (24). Participants were asked, in general, how much difficulty they had carrying out activities “without help from another person or without the help of a special gadget or piece of equipment.” Responses were on a 4-point scale (where 0 = no difficulty and 3 = not able to do). Participants indicating that an activity was not applicable to their job (e.g., crouching) were assumed not to be currently limited by this activity at work and were assigned a score of 0 (no difficulty) for that activity. Responses were summed (score range 0–36). Internal consistency of the WALS using Cronbach's alpha in samples with OA, IA, and lupus have ranged from 0.78 in community samples to 0.87 in patient populations (24, 30, 33, 39).


Data on age, sex, marital status, and education were collected.

Arthritis type and duration.

Respondents provided the time since diagnosis. Arthritis type was coded in 3 categories: IA (e.g., RA, psoriatic arthritis), OA, or both.

Symptom severity.

Respondents were asked about pain and fatigue over the preceding month (14, 41). Pain was assessed with, “How often did you have severe pain from your arthritis?” and fatigue with, “How often have you felt fatigue as a result of your arthritis?” Responses were from 1 = no days to 5 = all days.

Employment status and occupation.

Categories for employment status were full time, part time, leave of absence/short-term disability, long-term disability, unemployed (i.e., looking for work), not employed (i.e., not looking for work), retired, homemaker, and other. They were collapsed into “employed” (full- or part-time employment) or “not in the labor force” (all other categories). Occupation was classified using the Human Resources Development Canada National Occupation Classification Matrix 2001 (42). Occupations were collapsed into 1) business, finance, and administration; 2) health, science, arts, and teaching; 3) sales and service; and 4) trades, transportation, and equipment operation.

Job modifications and accommodations.

Four categories of arthritis-related job modifications were assessed. At T1, participants were asked about accommodations in the previous year. Subsequent interviews asked about modifications/changes “since your last interview.” Scheduling changes (yes/no) included at least one arthritis-related change to a respondent's work schedule from a list that included more breaks, a shortened work day, and changes to the time the work day began or finished. Participants were asked if they used any gadgets, assistive devices, or furniture or equipment (e.g., special keyboard, stools; yes/no). In response to WALS difficulties, participants were asked whether they received help from others (yes/no). Participants were asked about consultations with work place professionals (e.g., ergonomic specialist, employee assistance plan personnel; yes/no).

Work place outcomes.

Seven arthritis-related work disruptions/transitions were assessed (no/yes). Except for absenteeism, outcomes were asked about in the previous year (T1) or “since your last interview” (T2–T4). Participants were asked about absenteeism in the previous 6 months. Lost time at work included leaving work early, arriving late, or taking an extended lunch hour. Work interruptions asked about interruptions during the work day of at least 20 minutes. Participants were also asked whether arthritis resulted in interpersonal difficulties with their manager or coworkers, being unable to attend meetings/business trips, or being unable to take on extra projects or responsibilities and whether arthritis resulted in permanently reducing work hours.

Statistical analysis.

Means, SDs, and percentages described the sample and work place activities. T-tests and chi-square analyses compared respondents at T1 and T4 and by diagnosis type. Internal consistency of the WALS (i.e., Cronbach's alpha) was assessed at each time point. Sample distributions revealed that 3 WALS levels existed corresponding approximately to sample tertiles and representing a consistent pattern of WALS scores at each time point (i.e., low = 0–4, medium = 5–8, and high = ≥9). To explore the episodic nature of work place activity limitations, intraindividual variation in the 3 WALS levels at T1–T4 were examined with intraclass correlation coefficients (ICCs) (43).

Repeated cross-sectional logistic regressions examined whether, and at what level, job modifications and work outcomes (independent variables) were associated with work place activity limitations (dependent variable). Repeated analyses were adopted because the study aimed to examine the consistency of relationships over time. Analyses controlled for age, sex, arthritis diagnosis, disease duration, and occupation type, as these variables have been associated with work outcomes in previous studies (education was not included in analyses because of its significant association with occupation). Analyses used SAS, version 9.0.


At T1, respondents were on average 51 years old and 78% were women (n = 490) (Table 1). More than half of the sample had OA (57%), 33% had IA, and 10% had both OA and IA. The mean arthritis duration was ∼9 years, with one-third of the sample reporting arthritis pain most or all days and 49.8% reporting fatigue most or all days. Eighty-two percent of the participants were employed full time and 18% were employed part time. At T4, 26.9% of the participants reported not being in the labor force, with most (62.5%) reporting their decision was related to their arthritis. Other common reasons for not working included the demands of work (57.5%) and stress levels (55%) (data not shown). Across T1–T4, demographic and occupational characteristics remained similar, although more respondents reported having both IA and OA at T4 than at T1 (18.3% versus 10%; P < 0.001). Given the nature of arthritis symptoms and the study duration, it is not surprising that pain and fatigue were variable. At T4, participants reported somewhat lower frequency of severe pain and fatigue (P < 0.05). At each time point, 84.3–93.8% of the sample was retained, with an overall response rate of 71% over 4.5 years (n = 349 at T4) (Figure 1). Reasons for loss to followup were being unable to locate participants and participant withdrawal from the study.

Table 1. Sample characteristics of respondents at time 1 and time 4*
 Time 1 (n = 490)Time 4 (n = 349)
  • *

    Values are the number (percentage) unless otherwise indicated. IA = inflammatory arthritis; OA = osteoarthritis.

  • Seventy-nine percent of participants with IA reported a diagnosis of rheumatoid arthritis.

Age, no./mean ± SD years487/51.1 ± 9.3346/56.6 ± 9.0
 Male109 (22.2)78 (22.4)
 Female381 (77.8)271 (77.7)
Marital status  
 Married/living as married297 (60.6)215 (62.1)
 Divorced/separated/widowed117 (23.9)84 (24.3)
 Never married76 (15.5)47 (13.6)
 Elementary and secondary85 (17.4)59 (17.0)
 Some postsecondary112 (22.9)74 (21.3)
 Postsecondary196 (40.2)143 (41.2)
 Postgraduate95 (19.5)71 (20.5)
Arthritis diagnosis  
 IA163 (33.3)108 (31.0)
 OA278 (56.7)177 (50.7)
 Both IA and OA49 (10.0)64 (18.3)
Arthritis duration, mean ± SD years9.2 ± 8.7 
Pain severity in last month  
 No days86 (17.6)74 (21.2)
 A few days120 (24.5)113 (32.4)
 Some days117 (23.9)80 (22.9)
 Most days109 (22.2)51 (14.6)
 All days54 (11.0)31 (8.9)
 No days76 (15.5)58 (16.6)
 A few days72 (14.7)84 (24.1)
 Some days95 (19.4)76 (21.8)
 Most days121 (24.7)74 (21.2)
 All days123 (25.1)57 (16.3)
Not in the labor force0 (0.0)94 (26.9)
Job sector  
 Business/administration162 (33.1)88 (34.5)
 Health/science/teaching175 (35.8)93 (36.5)
 Sales/services102 (20.9)51 (20.0)
 Trades/transportation50 (10.2)23 (9.0)
Figure 1.

Flow chart of sample recruitment and interviewing over 4 time points, each 18 months apart.

Table 2 provides percentages of participants reporting at least some difficulty with work place activities at T1–T4. WALS items are shown in ascending order based on percentages of respondents reporting the activity as difficult. Percentages are shown for IA, OA, and the total sample. Individuals with both IA and OA were collapsed into the IA group. Respondents had fewer difficulties with global aspects of their jobs such as scheduling and getting to and from work than specific tasks such as sitting for long periods, crouching, or kneeling. Working with the hands was consistently reported as significantly more difficult for those with IA than OA. Otherwise, there were few significant differences in activity difficulties between the groups. The percentage of respondents reporting difficulties with individual activities as well as the overall mean WALS scores were relatively similar at T1–T4 (WALS scores at T1 were based on 11 items; WALS scores at T4 were based on 12 items). In addition, sample sizes vary over time. Internal consistency of the WALS using Cronbach's alpha ranged from 0.75 to 0.84 at T1–T4 by diagnosis.

Table 2. Frequency of respondents reporting difficulties with job activities on the WALS at 4 time points by diagnosis of IA or OA*
WALS itemsTime 1Time 2Time 3Time 4
IA (n = 207)OA (n = 267)Total (n = 474)IA (n = 159)OA (n = 194)Total (n = 353)IA (n = 144)OA (n = 152)Total (n = 296)IA (n = 134)OA (n = 121)Total (n = 255)
  • *

    Participants with both IA and OA were collapsed into the IA group. Numbers of participants vary from time 1 to time 4 and include only those respondents who were employed. WALS = Workplace Activity Limitations Scale; IA = inflammatory arthritis; OA = osteoarthritis.

  • Significant difference between the IA and OA groups (P < 0.05).

  • Not included at time 1.

Reported at least some difficulty with, %            
 1. Scheduling or hours for work/job34.430.632.229.427.628.428.521.124.735.127.331.4
 2. Getting to and from work36.426.230.623.122.122.527.819.723.626.124.825.5
 3. Meeting current job demands35.431.333.136.336.736.525.029.627.436.631.434.1
 4. Getting around work place38.938.138.435.042.939.338.233.635.838.840.539.6
 5. Concentrating on work36.339.337.929.234.932.131.342.136.5
 6. Pace of work/job required40.639.239.841.935.738.538.937.538.242.533.938.4
 7. Reaching35.426.330.236.329.132.334.727.030.736.638.837.6
 8. Working with hands67.043.953.966.943.954.263.946.154.765.752.959.6
 9. Sitting for long periods of time48.149.548.955.059.557.554.259.957.159.067.863.1
 10. Standing for long periods of time59.958.559.146.953.650.652.861.257.167.963.665.9
 11. Lifting, carrying, or moving62.651.356.
 12. Crouching, bending, or kneeling59.060.559.857.265.361.762.571.166.971.676.974.1
WALS score, mean ± SD7.0 ± 4.66.0 ± 4.26.4 ± 4.46.8 ± 5.06.6 ± 4.66.7 ± 4.86.7 ± 4.66.9 ± 4.76.8 ± 4.67.7 ± 5.17.7 ± 4.67.7 ± 4.8
Observed range0–200–200–200–260–210–260–230–190–230–200–190–20
Cronbach's alpha0.790.750.770.840.810.820.810.810.810.840.780.81
WALS score levels, %            
 Low (0–4)32.941.237.637.739.738.836.
 Medium (5–8)30.934.833.133.325.829.131.928.330.130.625.628.2
 High (≥9)

WALS sample distributions were examined at each time point. Three levels of WALS difficulties were explored that did not overlap across time and that approximated sample tertiles. Between 31% and 38.8% of respondents reported low WALS difficulty for at least 1 of the 4 time points (score 0–4). This indicates that these respondents ranged from having no arthritis-related difficulties at work to some difficulties with up to one-third of WALS items. Another 25.8–34.8% reported WALS difficulty scores of 5–8. Scores of 9 or more signified that participants had ongoing difficulty with two-thirds of their work place tasks or that they were unable to perform some work activities. The percentage of participants with high WALS difficulty increased over time (T1 = 29.4%, T4 = 40.8%). Chi-square analyses comparing WALS levels related to OA and IA were significantly different at T1 only (P < 0.01), with those having OA reporting lower WALS scores. Further analyses found that similar percentages of respondents with IA and OA had given up employment. This suggests that the absence of significant differences between the conditions from T2–T4 was not explained by changes in sample composition related to remaining employed. Because the type of arthritis was largely unrelated to WALS levels, remaining analyses collapsed across diagnosis.

To investigate the consistency of WALS scores, we examined participants who remained employed at all time points (n = 214) (Table 3). Only 26.6% of participants (n = 57) reported the same level of difficulty (low score [0–4], medium score [5–8], or high score [≥9]) from T1–T4. The remainder (73.4%, n = 157) reported variable levels of difficulty with work place activities. Among them, 25.2% reported at least one increase in WALS level difficulty over T1–T4 (e.g., low to medium, medium to high, or low to high). Decreases in WALS difficulty (e.g., medium to low, high to medium, or high to low) were reported by 16.4% of participants. Nearly one-third of participants (31.8%) reported both increases and decreases in WALS level difficulty over 4.5 years. The raw-score WALS ICC was 0.61, indicating that 61% of the variance in WALS was related to differences among respondents and 39% was related to within-individual differences (43). This pattern was similar across IA and OA.

Table 3. Intraindividual consistency and variability in levels of WALS difficulty from time 1 to time 4*
 Total (n = 214)
  • *

    Percentages only include participants who remained employed across all time points. WALS = Workplace Activity Limitation Scale.

Respondents reporting a consistent level of WALS difficulty, no. (%) 
 Low difficulty (WALS score 0–4)34 (15.9)
 Medium difficulty (WALS score 5–8)3 (1.4)
 High difficulty (WALS score ≥9)20 (9.3)
 Total57 (26.6)
Respondents reporting variable levels of WALS difficulty, no. (%) 
 Increase in level of WALS difficulty54 (25.2)
 Decrease in level of WALS difficulty35 (16.4)
 Both increase and decrease in level of WALS difficulty68 (31.8)
 Total157 (73.4)

Percentages of respondents reporting job modifications and work place outcomes at T1–T4 showed variability over time (Table 4). Scheduling changes and consultations with work place professionals were reported infrequently. At T1, 8.4% of respondents reported modifying their work schedule and 14.9% consulted a work place professional. These percentages remained under 17% of respondents from T2–T4. In contrast, gadgets/assistive devices/furniture and receipts of help were reported by 47.1% and 46.7% of the sample, respectively (T1), and were consistently reported by approximately half of the respondents from T2–T4. Absenteeism and work place interruptions of 20 minutes or more were the most frequently reported arthritis-related work place outcomes. Permanently reducing work hours was reported least frequently.

Table 4. Percentages of respondents reporting job modifications/accommodations and work place outcomes from time 1 to time 4
 Time 1 (n = 474)Time 2 (n = 353)Time 3 (n = 296)Time 4 (n = 255)
Job modifications/accommodations    
 Scheduling changes (e.g., more breaks, changes to work arrival/departure time)8.413.916.216.9
 Gadgets/assistive devices/furniture/equipment47.158.152.449.4
 Help from others (e.g., coworkers, manager)46.756.752.756.1
 Work place professionals consulted14.914.412.28.6
Work place outcomes    
 Work interruptions ≥20 minutes38.438.232.432.2
 Lost time (e.g., arrive late/leave early)15.311.98.411.4
 Difficulties with manager/coworkers13.511.99.511.4
 Unable to attend meetings/business trips12.211.314.210.2
 Unable to take on extra projects/responsibilities22.
 Permanently reduced work hours10.211.310.17.8

Repeated regression analyses examined consistency in the relationship of levels of work place activity limitations with job modifications and work outcomes, controlling for age, sex, arthritis diagnosis, disease duration, and occupation type (Table 5). Compared to those with low WALS difficulty, the odds of using gadgets/assistive devices/furniture and reporting help at work were increased significantly and consistently among those with medium WALS difficulty. Medium WALS difficulty was less consistently associated with scheduling changes and consulting work place professionals. High levels of WALS difficulty were consistently related to increases in the likelihood of all 4 types of work place modifications.

Table 5. Logistic regression analyses comparing low/medium and low/high work place activity limitations with job modifications and work place outcomes at 4 time points*
 Time 1 (n = 474), OR (95% CI)Time 2 (n = 353), OR (95% CI)Time 3 (n = 296), OR (95% CI)Time 4 (n = 255), OR (95% CI)
WALS 5–8 (n = 157)WALS ≥9 (n = 139)WALS 5–8 (n = 103)WALS ≥9 (n = 113)WALS 5–8 (n = 89)WALS ≥9 (n = 100)WALS 5–8 (n = 72)WALS ≥9 (n = 104)
  • *

    Reference group for Workplace Activity Limitation Scale (WALS) is low difficulty (score 0–4; n = 178 at time 1, n = 137 at time 2, n = 107 at time 3, and n = 79 at time 4); analyses compare low/medium and low/high scores only. All logistic regression analyses controlled for age, sex, diagnosis, and occupation. OR = odds ratio; 95% CI = 95% confidence interval.

  • P < 0.05.

  • P < 0.001.

  • §

    P < 0.01.

Job modifications        
 Scheduling changes1.5 (0.6–3.5)2.3 (1.0–5.2)18.4 (4.0–84.7)28.7 (6.3–130.1)2.3 (0.8–6.7)7.6 (3.0–19.7)2.1 (0.8–5.5)2.4 (1.0–6.1)
 Gadgets, assistive devices, furniture, equipment use3.6 (2.3–5.8)5.1 (3.1–8.3)3.9 (2.2–6.9)8.9 (4.8–16.4)8.9 (4.5–17.8)17.3 (8.3–36.1)3.5 (1.7–7.0)4.4 (2.3–8.6)
 Help from others2.9 (1.8–4.6)6.8 (4.15–11.3)8.3 (4.4–15.4)9.0 (4.9–16.4)5.4 (2.8–10.6)14.3 (7.0–29.3)4.5 (2.2–9.3)9.4 (4.6–19.0)
 Work place professionals consulted3.5 (1.6–7.8)§6.2 (2.8–13.6)2.6 (1.0–6.8)5.6 (2.4–13.4)1.2 (0.4–3.8)6.0 (2.3–16.2)1.9 (0.4–9.2)5.6 (1.4–21.5)
Work place outcomes        
 Absenteeism1.4 (0.8–2.2)3.0 (1.9–4.9)4.0 (2.0–7.7)6.2 (3.2–11.8)2.3 (1.1–4.9)6.1 (3.1–12.2)2.5 (0.9–6.7)10.6 (4.3–26.2)
 Lost time (e.g., arrive late/leave early)1.9 (1.0–3.8)2.9 (1.5–5.7)§1.2 (0.4–3.2)3.4 (1.5–8.0)§3.9 (1.0–15.3)6.8 (1.8–25.6)§1.6 (0.4–6.6)4.8 (1.4–16.2)
 Work interruptions (20 minutes or more)3.5 (2.1–5.9)7.0 (4.1–11.9)3.5 (1.5–8.1)14.5 (7.4–28.4)5.5 (2.4–12.9)20.9 (8.9–49.1)2.1 (0.6–7.6)12.0 (5.0–28.3)
 Difficulties with manager, coworkers4.6 (1.5–14.2)§22.6 (7.7–66.4)1.4 (0.4–4.5)9.0 (3.5–23.1)6.1 (1.2–30.1)12.5 (2.7–57.5)§1.9 (0.3–12.6)15.7 (3.2–76.3)
 Unable to attend meetings or business trips2.4 (0.9–6.2)8.4 (3.5–20.0)2.6 (0.9–7.6)6.5 (2.5–17.0)2.6 (0.8–8.0)8.5 (3.0–23.8)2.4 (0.2–27.0)22.9 (2.9–178.8)§
 Unable to take on extra projects/responsibilities6.1 (2.7–13.6)15.8 (7.2–34.8)3.5 (1.5–8.1)§7.9 (3.6–17.1)4.7 (1.2–17.9)15.3 (4.4–53.9)2.1 (0.6–7.6)6.6 (2.1–20.4)
 Permanently reduced work hours1.9 (0.8–4.5)3.9 (1.7–8.7)3.7 (1.2–11.6)9.5 (3.3–27.2)1.0 (0.3–3.0)2.9 (1.1–7.6)5.3 (0.6–48.5)13.0 (1.6–105.4)

Overall, a medium level of WALS difficulty was not consistently related to work place outcomes (Table 5). However, the odds of reporting work interruptions and being unable to take on extra projects significantly increased at 3 of 4 time points among those with medium WALS difficulties. In contrast, losing time because of arriving late or leaving early, being unable to attend meetings or business trips, and permanently reducing hours was only significantly related to WALS difficulty at scores of 9 or more.


People living with arthritis often characterize their condition as variable and unpredictable, with periods of relative wellness interrupted by episodes of more serious impairment and disability. This variability and its implications for at-work disability, use of job modifications, and relationship to work outcomes have not been fully examined despite evidence that the unpredictability of arthritis is stressful (13, 26). This research examined the work place activity limitations of individuals with IA and OA across 4.5 years, intraindividual variability in at-work disability, and the consistency of the associations of different levels of activity limitations with job modifications and work place outcomes. The findings show that, although there was relative stability in work place activity limitations at a group level, within-individual comparisons revealed considerable fluctuation. Moreover, a relatively consistent pattern emerged in the levels of limitations related to the use of job accommodations and work place outcomes.

One-third to three-quarters of participants reported some difficulties with work place activities. At each time point, there were few significant differences related to arthritis diagnosis, with the exception of tasks involving the hands. This is in keeping with existing research examining activities such as computer use, where those with RA report more difficulty at work than those with OA (21). The finding of a comparable impact of arthritis type on job activities is important in understanding the overall burden of arthritis. OA is one of the top 10 causes of disability worldwide and is projected to increase in prevalence, not only because of the aging of the population, but also because of lifestyle factors such as obesity and lack of physical activity (44–48). Currently, there is an abundance of research on RA, despite the fact that many individuals with OA develop disease symptoms in their prime earning years (ages 45–65 years). This suggests the need for additional research on OA and employment.

Of interest is the considerable within-individual variability in reports of work place activity limitations over time. Nearly three-quarters of participants reported activity limitation levels that either increased, decreased, or both increased and decreased over time. Currently, there is little understanding of the implications of variability in work place activity limitations for managing arthritis or predicting job outcomes. Characterizing arthritis as a chronic episodic condition with intermittent disability is an important message for employers, human resource professionals, insurers, and the government to ensure that people with arthritis are not inevitably viewed as a cost or drain on work place and health resources. At the same time, individuals may benefit from better coordination and tailoring of job accommodations and benefits to meet their intermittent disability needs. This variability highlights the need for additional research to better understand the course of at-work disability and its implications.

Also explored were different levels of activity limitations. Examining different thresholds of at-work disability is advantageous because the wide range of job modifications and work outcomes that are important to arthritis are unlikely to be associated with one single activity limitation cutoff value. In this study, WALS scores of 9 or more were designated as “high” because they were equivalent to having ongoing difficulties with two-thirds of work place activities or with being unable to perform some activities because of arthritis. However, although a wide range of scores was exhibited by the sample (0–20 of a possible score of 36), the full range of scores was not reported, likely for two reasons. First, diversity across jobs means that not all activities are relevant to all employees (e.g., standing for long periods). Second, we focused on employed individuals who must be able to perform many of these activities in some capacity to remain working. Individuals who are no longer employed might report higher WALS difficulties.

To examine the meaningfulness and consistency of the levels of activity limitations, their relationship to job modifications was examined. Health and work place professionals often promote the use of job accommodations before substantial disability occurs so that early management can prevent negative work place outcomes such as absenteeism, job disruptions, and giving up employment. In this study, gadgets/assistive devices and help from others were reported at medium levels of WALS difficulty. However, scheduling changes and help from work place professionals were reported infrequently and were only consistently reported at high levels of disability. This variability is similar to previous research (24, 28, 39, 49) finding a higher prevalence of informal work changes (help from coworkers) compared to structural changes to work and use of work place benefits. Although additional research is needed, the findings suggest that many individuals might wait until ongoing and relatively substantial problems with job activities exist before making changes (i.e., WALS score >9). This may be too late to prevent negative job outcomes, highlighting the need for education about the value of early intervention.

Further evidence for the potential value of early intervention comes from examining changes at work made by those employed with arthritis. Consistent and significantly more absenteeism, lost work time, work interruptions, interpersonal difficulties, being unable to attend meetings or take on extra projects, and reduced work hours were found among those with WALS scores of 9 or more compared to low WALS difficulty. Although not as consistent, work interruptions and being unable to take on extra projects also were reported at a medium level of WALS difficulty. Adaptations at work may ultimately reduce WALS difficulties and be of benefit to people with arthritis. However, many work place outcomes also can have long-term implications for career planning, and some are linked to giving up employment (14). A WALS score >9 may be an important warning sign of the potential for more serious work outcomes and a medium level of WALS difficulty may be useful to signal the need for self-management and work place strategies to be implemented. At the same time, additional research is needed, given that some modifications and outcomes were rare and odds ratios varied widely.

Several limitations to this research should be considered. An 18-month timeline was chosen to document working experiences. Although it has the advantage of being long enough to gauge changes in employment, it may not disentangle some employment events. For example, a job accommodation may have occurred in the previous 9 months, but activity limitations would have been assessed generally over 18 months, making it difficult to capture the impact of accommodations at any one time. Moreover, variables other than activity limitations might be relevant to accommodations and outcomes (e.g., work place support). Also, although the sample was comparable to other work place studies, more in-depth research is needed to understand differences in work place activity limitations related to job type, changing jobs, fluctuating symptoms of disease (e.g., pain, fatigue), and other factors. Finally, we focused on the experiences of individuals with arthritis who remained working. Additional research is needed to examine work place activity limitations among those who leave the labor force.

In conclusion, this study highlights the level and intermittent nature of work place activity limitations in those with arthritis. Many individuals report at least some difficulty with at-work activities. However, these difficulties are often episodic and may not result in changes to work productivity until they are consistently high. This is important for researchers to understand when designing work place interventions, as well as employers, insurers, and the government, who may incorrectly view individuals with arthritis as a permanent drain on work place and health resources.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Gignac had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Gignac.

Acquisition of data. Gignac.

Analysis and interpretation of data. Gignac, Cao, Tang, Beaton.