Psoriatic arthritis (PsA) is an inflammatory arthropathy that affects about 0.2–1% of the population.1 2 Women and men are considered to be equally affected. Psoriasis is usually present several years before the onset of arthritis. Comparative studies have indicated that patients with PsA have a burden of illness which is comparable with that of patients with rheumatoid arthritis (RA) or ankylosing spondylitis.3 Both the skin and joint components of the disease have a negative impact on quality of life, and the patients may experience more pain and more role limitations due to emotional problems than patients with RA,4 and mortality has been reported to be increased compared with the general population.5

Several previous studies have addressed gender aspects in disease characteristics and employment status in RA,6^–10 but not in PsA. Our aim was to compare health status and work disability (WD) between males and females with PsA in a younger patient group aged 18–45 years with less expected influence of comorbidity than in an older age group. Further, we wanted to compare health status and demographic variables between being work disabled and not work disabled for each gender, and finally to examine independent associations between WD and demographic variables, levels of health status and disease activity measures.

MATERIALS AND METHODS

Setting

We used data collected in the Norwegian Disease Modifying Antirheumatic Drug (NOR-DMARD) study which is a phase IV longitudinal observational study.11 The study is conducted in five Norwegian rheumatology departments. Adult patients (⩾18 years) with inflammatory arthropathies are included consecutively when starting a new disease-modifying antirheumatic drug (DMARD) regimen, including also biological agents. Patients are registered as a new case when they switch to another DMARD regimen. Demographic variables are recorded at baseline, and patients are assessed at baseline, after 3, 6 and 12 months, and then yearly, with core measures of disease activity and health status measures. By May 2006, 5597 DMARD regimens were included in the database.

The status of individual patients was included in the current analyses, and data from the last inclusion were used when an individual patient had been included with several DMARD prescriptions. Each patient signed a written informed consent form before participation, and the study had been evaluated by the Norwegian Data Inspectorate, the regional ethical committee and the drug regulatory authorities.

Patients

The patient group fulfilling the eligibility criteria consisted of 102 women and 169 men. Patients were eligible for the present analyses if they were diagnosed with PsA with peripheral arthritis. Among the women, 81 (80.2%) had fewer than five swollen joints, and the corresponding number among the men was 121 (72.0%) (p = 0.13). Only two of the patients were registered with spinal arthritis, but both of them also had peripheral arthritis. The patients had either been diagnosed by a rheumatologist previously, or they were diagnosed by the treating rheumatologist.

Three patients had missing data for employment status and education. Mean (SD) age in women was 35.5 (6.0) years and in men 36.5 (6.1) years, and mean disease duration was 6.9 (7.1) and 5.6 (6.6) years, respectively (table 1).

Assessments

In these cross-sectional analyses we used data collected at inclusion in the study by the treating rheumatologist and a study nurse.

Patients who received a permanent national WD pension, either full or partial (at least 50%), were classified as work disabled (dependent variable). Information about any WD pension was registered by the doctor or a study nurse. A person who receives a WD pension in Norway has been working (full or part time) and has had to stop or reduce work because of, for example, a chronic disease, first by sick leave, followed by a permanent pension. Patients under the age of 26 years who have been too ill ever to have worked may have a WD pension without having worked before.

Formal education was registered as categorical variables and condensed into three groups according to the Norwegian school system: up to and including 10 years (elementary school), 11–12 years (college) and ⩾13 years (university).

Several health status measures were used. Fatigue, pain and patient’s global assessment were measured on 100 mm visual analogue scales (VAS) (100 = worst score). The Health Assessment Questionnaire (HAQ) is commonly used in clinical trials of PsA.12 The Modified Health Assessment Questionnaire (MHAQ) is a shortened, modified version of the HAQ with a score from 0 to 3 (3 = worst disability).13 The Medical Outcome Study 36-item short form health survey (SF-36)14 is a commonly used health status measure which has also been validated in PsA and has shown significant differences between patients with PsA and the general population.15 It contains 36 questions measuring health across eight different dimensions (physical functioning, role limitations due to physical health problems, bodily pain, vitality, social functioning, role limitation due to emotional problems, mental health and general health). A score is computed within each dimension with a value from 0 to 100 (best possible health state).

The disease activity score-28 (DAS-28) was computed based on 28 tender and swollen joint counts, patient’s global assessment on a 100 mm VAS and the erythrocyte sedimentation rate (ESR). The joint counts were performed partly by rheumatologists and partly by trained study nurses.

Information about erosive disease was based on a routine judgement of x ray pictures of the hands and feet by a radiologist, and erosions were registered as present or not present.

Analyses

Group comparisons (females vs males, work disabled vs not work disabled) were performed with the Mann–Whitney U test for continuous variables and χ² tests for categorical variables.

A multiple logistic regression model was built up to identify factors associated with WD. The associations between independent variables (demographics, health status and disease activity) and WD were first examined by univariate logistic regression analyses. Selection of variables for the multivariate model was based on clinical judgement and findings in previous RA studies, and variables that explained most of the variations were included in the multivariate model when several variables captured similar health constructs. In the logistic regression model, duration of education was categorised into three groups (⩽10 years, 11–12 years and ⩾13 years). In the final multiple model,adjustments were performed for duration of education, disease duration, age, erosive disease, disability score (MHAQ (0–3)), the SF-36 score for mental health, and gender. Presence of erosive disease was applied as a surrogate measure of cumulative inflammation,16^–18 MHAQ as a measure of physical disability and SF-36 score of mental health as a measure of the psychological domain.

Data were analysed using the Statistical Package for the Social Sciences, version 14.0 (SPSS, Chicago, Illinois, USA). The levels of significance were set to 0.05 unless stated otherwise.

RESULTS

The basic characteristics of males and females are shown in table 1, and the basic characteristics of work-disabled and non-work-disabled males and females are shown in table 2.

Females had worse health status than males for MHAQ score, fatigue VAS, and SF-36 physical function and vitality, but no significant gender differences were observed for other variables (table 1).

Out of 101 females, 33 (32.7%) were work disabled. Corresponding numbers for males were 29 out of 167 (17.4%) (p = 0.004). Work-disabled patients had generally worse health status than patients who were not work disabled. These differences seemed to be more pronounced for males than for females (table 2). The MHAQ score was significantly worse for the work-disabled compared with the non-work-disabled females, but not between the male groups. The following variables were statistically significantly different between the work disabled and the not work disabled for both genders: SF-36 physical function and general health, VAS pain and the 28-SJC.

In the multiple logistic regression model (table 3), low educational level, increasing level of MHAQ, presence of erosive disease, female gender and disease duration were independently associated with WD. The strength of association of the model were: Cox and Snell R² 0.19, Nagelkerke R² 0.30.

The educational level, disability level and presence of erosive disease had the strongest associations with WD. However, the females had a twofold independent risk of being work disabled compared with the males, odds ratio (OR) 2.09 (95% CI 1.03 to 4.02) (table 3).

DISCUSSION

In this cross-sectional study we found that low educational level, increasing disability level (MHAQ), presence of erosive disease, female gender and disease duration were independently associated with WD in patients with PsA between 18 and 45 years of age who were starting DMARDs and/or biological treatment. The most important predictors of WD in RA have been reported to be demographic variables such as age, occupation and level of education, but also self-reported functional status measured by, for example, the HAQ or MHAQ.19

In the current study of patients with PsA, educational level was independently associated with WD, with a fivefold increased risk for both the low and the medium levels compared with the highest educational level (table 3). This observation is in accordance with findings in RA studies where a high level of education is recognised as a protective factor against WD,8 19 probably because educational level is strongly related to the type of job a person performs. Unfortunately, the type of job, the percentage of patients working immediately prior to disease onset, and onset of WD were not recorded in the present study, which are obvious study limitations. The level of education was similar between the genders, which is in accordance with the general educational level among men and women aged 18–45 years in Norway. However, the work-disabled patients had a lower educational level than the non-work-disabled patients, indirectly indicating that the work-disabled patients might have had more physically demanding jobs.

Disability level (MHAQ) was independently associated with WD in the current study (table 3), and this finding supports the reasoning that reduced physical functioning is also associated with WD in patients with PsA. This is in accordance with RA studies which have reported that HAQ disability levels predict WD.19 The females had on average worse scores than the males for MHAQ and SF-36 physical functioning (table 1), and the work disabled of both genders had worse scores than the working patients (table 2).

The female patients had a twofold increased risk of WD compared with male patients of the same age (table 3). Female gender has been reported to be an independent risk factor for new WD in RA,6 8 and our data support that females also are at a higher risk of WD in PsA. A limitation of the current study was that women who were not able to stay in work because of the PsA disease and had chosen to be homemakers without receiving a WD pension were not registered as work disabled, and the actual number of work-disabled women might even be higher than we have registered. In Norway a large number of women participate in the labour force, and studies from other areas have shown that female patients with RA more often have no paid job compared with the male patients in countries with a high labour force participation among women.20

RA more frequently starts early in females,21 22 indicating that females are exposed to the inflammation for longer than males. In the current PsA patient cohort we did not find any difference in mean disease duration or mean patient age between the genders or between the work disabled and not work disabled of both genders. However, in the multiple regression model, disease duration had a weak independent association with WD. In RA studies, disease duration has been reported to have an inconsistent association with WD.19 The patient cohort in this study was under 45 years of age, which may partly explain the weak association between disease duration and WD.

Different modifications of the DAS, including DAS-28, have been shown to be responsive and to discriminate between active treatment and placebo in patients with PsA.23 However, exclusion of the feet and distal interphalangeal joints from joint counts may lead to underestimation of disease activity in individual patients, and it is still not clear what a certain level of DAS or DAS-28 means in PsA.23 Because of this uncertainty, we did not include DAS-28 in the multivariate analyses, but used the presence of erosive disease as a surrogate marker for cumulative inflammatory activity,16^–18 and an independent association with work disability was demonstrated (table 3). In RA studies, structural damage has been inconsistently associated with WD.19 A Canadian study has shown that the radiographic severity in the hands and feet of patients with PsA was comparable with that of patients with RA,18 and erosive disease is common among patients with PsA.18 However, at present, none of the commonly used radiographic methods attempts to capture all of the radiographic abnormalities in PsA, and a new radiographic classification system that accounts for the deposition of new bone and the damage of the surrounding articular and periarticular structures should be elaborated.18

We found that both work-disabled males and females with PsA had overall a worse health status than working patients, and this difference seemed to be more pronounced in males than in females (table 2). Previous studies have addressed the relationship between health status, gender and work status in patients with RA,6^–10 but not in PsA. De Roos et al found that work-disabled men with RA were more likely to have single marital status and higher helplessness scores than working men, whereas these differences were not present in women.6 Self-reported pain was more strongly correlated to WD in women than men.6 These differences between men and women may be a consequence of how men and women perceive their health,24 and females in this study also had generally worse scores for patient-reported outcomes than males (table 1).

Qualitative research suggests that fatigue is a barrier for employees with RA in carrying out their work.25 Females reported higher fatigue intensity than males in the current study (table 1), and the females who were work disabled had borderline elevated fatigue levels compared with the working patients (table 2).

The data collected in the current study did not include information about whether WD was due to the rheumatic disease or not, which is a study limitation. An additional limitation is that we do not know to what extent the patients were affected by psoriatic skin disease and how this might contribute to WD. Psoriatic skin disease has been found to affect social functioning, and to lead to decreased work efficiency and subjective distress at work.26

Patients with RA are at risk of WD from the very onset of their disease. Prospective cohorts of patients with early RA have shown that 20–30% become permanently work disabled during the first 2–3 years of the disease.27 The current study indicates that patients with PsA are also at increased risk of WD, even at an early age. Longitudinal observational studies are needed to study the effects of new treatment options on working status. A recent study demonstrated a link between treatment effect and reduced WD in patients with RA who received treatment with DMARDs and antitumour necrosis factor (TNF) drugs.28 A Norwegian study of patients with PsA showed that anti-TNF treatment was superior to methotrexate in improving bodily pain, vitality, role physical and general health scores, but did not have power to examine effects on WD.29

Further longitudinal studies are needed to estimate the risk of WD in PsA and especially to examine how this outcome can be influenced by effective treatments. The current study showed that low educational level, disability score, erosive disease, female gender and disease duration were independently associated with WD in PsA. These findings need confirmation. As suggested for patients with RA,20 30 also working patients with PsA should have a very careful analysis of their work, workplace, work commitments and their plans and aspirations at a very early stage after onset of disease. Clinicians should also be particularly aware of the gender risk in management programmes which focus on reductions in WD in PsA.