Predictors of functional deterioration in Chinese patients with Psoriatic arthritis: a longitudinal study

Psoriatic arthritis (PsA) is an inflammatory arthritis associated with psoriasis. It affects young adults in their working ages. It has deleterious effects on patients and joint deformities and disease progression have been shown to develop over time[1‐4]. We and other investigators have shown that the physical function among patients with psoriatic arthritis is lower than the normal population[5‐8]; and comparable with rheumatoid arthritis[9, 10]. However, data were derived from cross sectional studies that reflected the physical function at a certain point of time which may be episodic, short lasting or long term disability. It is important to understand variables that are associated with deterioration in physical function in psoriatic arthritis over time. There were only a few studies on the longitudinal course of physical function in PsA from Caucasian countries[11, 12], while physiotherapy was shown to improve physical function in spondyloarthropathies (SpA)[13‐15]. There are cultural and ethnicity differences among Asian that affect disease manifestation in PsA and SpA[16, 17], which may also affect disease progression and physical function. We hypothese that disease activities at baseline determine the physical function over time in patients suffering from PsA. In this study, we aim to describe the variables associated with a change in physical function in Chinese patients with psoriatic arthritis over a 6-year period, and to determine the importance of joint damage and disease activity in the deterioration of physical function.

A total of 97 patients completed the follow up survey, after a mean (±SD) follow up time of 6.2 (±0.7) years. PsA affect adults in their middle age and affected individual had poor physical function and consistently poorer health related quality of life (HRQoL) compared to normal population over time. There were 56.7%, 59.8% and 35.1% who had clinically damaged joints, radiographic erosion of hand/ wrist joints and radiographic sacroiliitis at baseline respectively. At baseline, high percentages (41%) of patient have used DMARD therapy; while utilization of biological agents was low due to the non-reimbursement of medical expense. Table 1 summarizes the characteristics of the patients who completed both survey. There was no significant difference in the baseline characteristics between the 28 patients who did not participated in the follow up survey. (data not shown).

There were 41.7% of patients who had persistent minimal disability by HAQ categories (HAQ 0–0.49) over time, while 14.6% had persistent moderate disability (HAQ 0.5-1.50) (Table 2). There were 33% who had improvement in disability status and 10.3% had deterioration in disability status. There were 17 patients who had deterioration in physical function as defined by an increment of HAQ score of more than 0.2 at the follow up survey.

Independent variables associated with physical function at follow up as measured by HAQ included years of education (p = 0.042), number of tender joint (p = 0.033), number of damaged joint (p = 0.008), ESR (p = 0.031), pain score (p < 0.0001), physicians’ global assessment (p = 0.004), the PCS (p <0.0001) and MCS (p = 0.033) of SF36. In multivariate analysis, age, baseline physical function as measured by the PCS, and the number of damaged joint were related to HAQ at follow up (Table 3). In the binary logistic regression models using 1) deterioration of HAQ disability status; and 2) increment of HAQ > 0.2, no variable was found to have significant association.

This is the first longitudinal study on the course of PsA in Chinese over time, showing deterioration of physical function over time. One-fifth of patients had deterioration of physical function more than the reported MCID. Age, poorer physical function at baseline significantly associated with poorer physical function, while the number of damage joint was also related to physical function over a 6-year peroid. Educational level, markers of disease activities (pain level, number of tender joint count, ESR and physician’s global assessment of disease activity) were also associated with poorer physical function over time; but the association became insignificant after adjustment of other factors. This is consistent of our previous cross sectional study, demonstrating the significant social and economic impact of PsA in Chinese, and damaged joint were associated with poor functional status[6].

There have been a few studies evaluating the longitudinal course of physical function in PsA from Caucasian countries[11, 12]. However, differences in clinical manifestation among Asian patients with PsA have been described[16], which limit direct extrapolation of results. These include the exceptionally low prevalence of psoriasis and PsA among Japanese[33], high prevalence of spinal involvement among Koreans[34], high percentage of asymptomatic radiographic sacroiliitis among Chinese[6, 35], and high risk among Indians with psoriasis to develop PsA as compared to Chinese[35, 36]. To our knowledge, this is the first longitudinal study in PsA cohort in Chinese. Psoriasis is less prevalent in China than in Europe[37], and the prevalence of PsA in China seem to be lower from a population-based study[38]. This may explain the small sample size in our secondary/tertiary center with a lengthy recruitment period. We believe that this is a representative sample of PsA in the region for two reasons. First, our primary health care system was not established to provide long term care for PsA patients; and second, with the financial reimbursement system and low utilization rate of anti-TNF, it is also unlikely that patients with PsA were solely managed in the dermatology setting without being referred to rheumatology setting. Nevertheless, we demonstrated the non-benign course of PsA. Two-third of patients had clinically damaged joints and radiographic erosions, while another third had sacroiliitis on X-rays at baseline. There were poor physical function and HRQoL among PsA patients comparing to normal population at baseline and over time. Over a 6-year period, one-fifth of PsA patients experienced deterioration in physical function. Husted et al.[11] evaluated the transition of three 3 physical function status according to HAQ scores (0–0.49, 0.5–1.5, and 1.51–3) in patients with PsA over a ten-year period, 72% experienced either enduring disability or moving between disability states. Comparing to Husted’s cohort, a higher percentage of our patients seem to be resistant to deterioration in physical function. There were 42% of our patients who had persistent minimal disease activity (HAQ 0–0.49) and 15% had zero HAQ score at both time points over the follow up period. However, given a shorter follow up period and only having two assessment time points, it would be difficult to draw a definite conclusion.

We demonstrated age, educational level, disease activities (pain, tender joint count, ESR and physicians’ global assessment), disease damage (damaged joint count) and baseline physical function were associated with HAQ over time. In general this is consistent with the Western literature[11]. Advancing age and disability level at baseline are associated with progression of disability among elderly populations[39, 40], elderly with arthritis[41], patients with rheumatoid arthritis (RA)[42] as well as in patients with PsA[11]. Sheer at al[42] reported that age was the single most powerful predictor of subsequent disability in RA, but was also partially dependent on its interrelationship with comorbidities and fragility. Female gender was found to be associated with progression of disability in PsA[11], but not observed in our study. Joint inflammation has been proposed as the main driver of physical disability in the early stage of PsA; while joint damage is considered the major determinant at later stage. The detection of swollen joints was shown to precede the development of radiological joint damage in PsA[43]. Husted at al[12] investigated the differential effects of disease activity and damage on physical function, and the positive effects of actively inflamed joints on HAQ score decreased over increasing duration of PsA; and there was less evidence to suggest that the positive effect of joint damage on the HAQ score increased over time. Similar findings were revealed in our study. Damaged joint was significantly associated with HAQ in our previous cross sectional study and the association became marginal in the longitudinal dataset. This could be the result of changes in disease activity, response to drugs, coping strategies, and adaptation to the disease. However, the baseline physical function remained a significant factor associated with HAQ over time, regardless of treatment. SpA patients on anti-TNF are expected to have good functional outcomes[44], and a short delay in diagnosis, male gender and preserved physical function were predictors of minimal disease activity (MDA) at the 5-year follow-up in a early PsA cohort[45]. However, we were not able to demonstrate the effect of anti-TNF in modulating the physical function at 6 year. This is probably related to the low utilization rate of anti-TNF in our cohort and some patients had stop anti-TNF prematurely due to financial constrain. Taking together with evidence of literature, our findings support the treatment of joint inflammation in alleviating long term disability in patients suffering from PsA.

The strength of our study is the representative sample of PsA in the region with reasonable sample size. Our center is the only secondary and tertiary referral center for PsA in the region, while there is essentially no primary health care other setting that provide long term care for PsA. The response rate at baseline was high and there was an acceptable response rate (77.6%) at the 6-year follow up survey with high completion of dataset. We used standardized and mostly validated outcome measures according to standardized protocol and clinical assessments at baseline were limited to three clinicians experienced in PsA clinical trials. The main limitation of our study is the small sample size. It is the main reason why we were unable to demonstrate predictive variables for deterioration of HAQ using two definitions. Finally, we only have physical function data at 2 time points and thus unable to tell whether functional statuses were fluctuating or gradual declining with time.

We demonstrated Chinese patients with PsA had poor physical function and quality of life. One fifth of patient experienced deterioration of physical function over time. Advancing age, baseline physical function and joint damage were important factors associated with poor physical function in PsA over time.