Article Text

Download PDFPDF

Concise report
Limitations in daily activities are the major determinant of reduced health-related quality of life in patients with hand osteoarthritis
  1. W Y Kwok1,
  2. T P M Vliet Vlieland1,
  3. F R Rosendaal2,
  4. T W J Huizinga1,
  5. M Kloppenburg1
  1. 1Department of Rheumatology, Leiden University Medical Center, Leiden, the Netherlands
  2. 2Department of Clinical Epidemiology, Leiden University Medical Center, Leiden, the Netherlands
  1. Correspondence to Mrs Wing-Yee Kwok, Department of Rheumatology, C1-R, Leiden University Medical Center, Albinusdreef 2, PO Box 9600, 2300 RC Leiden, the Netherlands; w.y.kwok{at}


Objective To determine the impact of limitations in daily activities and pain on quality of life (QoL) in patients with osteoarthritis (OA) visiting a rheumatologist.

Methods Patients diagnosed by the rheumatologist with primary hand, knee or hip OA were consecutively included from August 2005 to April 2009. QoL was assessed by Short Form-36, with the physical component summary score (PCS), calculated using data from a norm-based population. Self-reported pain and function in patients with hand OA was assessed by the Australian/Canadian OA hand index (AUSCAN) pain (range 0–20) and AUSCAN function (range 0–36). Linear regression analyses were performed to investigate associations between PCS and demographic characteristics, and between PCS and pain and function in patients with OA.

Results Hand OA was diagnosed in 95% of 460 included patients (89% women, mean age 61 years). PCS was lowered in patients with OA. Patients with hand OA reported a considerable amount of pain (mean 9.5 (SD 4.3)) and disability (mean 16.5 (SD 8.6)). AUSCAN function was associated with PCS (adjusted β=−0.3, 95% CI –0.4 to –0.2), but AUSCAN pain was not.

Conclusions Hand OA was the most common OA subtype in secondary care. Health-related QoL is decreased in patients with OA and is associated with limitations in daily activities.

View Full Text

Statistics from


Knowledge concerning osteoarthritis (OA) results mainly from studies in the general population,1 2 in which many participants have only radiographic OA with no or mild complaints.3 Data in symptomatic patients with OA are scarce and deal almost exclusively with lower extremity OA.4 Knowledge about hand OA is limited and research in patients with hand OA is mostly performed in selected patient populations.5 6

Our objectives were to describe the phenotype of patients with OA in rheumatology practice, to compare patients with OA with the general population, to investigate their health-related quality of life (HRQoL) and to assess their most important problem (pain or impaired function).

Patients and methods

Patient population

This study was performed at the rheumatology outpatient clinic of the Leiden University Medical Center, Netherlands from August 2005 to April 2009. Patients diagnosed by the rheumatologist with primary hand, knee or hip OA were referred to the clinical nurse specialist and consecutively included. Clinical diagnoses of primary OA were verified by the medical chart.

Demographic characteristics

Collection of demographic and anthropometric data was performed by standardised questionnaires. Lower education was defined as those people who did not receive education, went to primary school only or received lower vocational education.

Random digit dialling population

Middle-aged controls (n=345, mean age 57 years, Leiden region) were recruited by random sampling of the population by telephone—random digit dialling (RDD).7 The control group was originally frequency matched to another case group in a previous study and relatively more men (36%) were included.8 Therefore all analyses are adjusted for age and sex.

Radiographic diagnosis of OA

Osteophytes (OST) and joint space narrowing (JSN) were scored by the OARSI scoring method.9 Radiographic hand OA was defined as OST or JSN grade ≥ 1 in the distal, proximal, interphalangeal thumb joint (DIPJ, PIPJ, IPJ, respectively), metacarpophalangeal joint and first carpometacarpal joint (first CMCJ).10 Erosions were scored by the Verbruggen–Veys scoring method and were defined as having eroded or remodelled subchondral plates (R-phase) in DIPJs, PIPJs or IPJs.11 Radiographs were scored by WYK, blinded for clinical and demographic data. To calculate intraclass correlation coefficients, a random sample of 10% was scored twice. The intraclass correlation coefficient (95% CI) for OST and JSN scores were 0.93 (0.81 to 0.97) and 0.89 (0.76 to 0.95), respectively. The intraobserver reliability of erosions, expressed by κ statistics, was 0.94.

Health-related QoL

HRQoL of patients with OA was measured by summary component scores for physical health (PCS) and mental health (MCS) in the Short-Form 36. Scores of a Dutch general population were used to standardise our scores to apply the norm-based scoring since no information about HRQoL was available in RDD controls.12 All scores were standardised to a mean of 50 with a SD of 10.13 Lower scores represent worse health status.

Self-reported pain and function in hands

Self-reported pain and function in patients with hand OA were measured with the disease-specific questionnaire Australian/Canadian OA hand index (AUSCAN) Likert scale 3.1. containing five items for pain, one for stiffness and nine for physical functioning.14 Each item is scored from 0 (best) to 4 (extreme). AUSCAN subscales range from 0 to 20 for pain, 0 to 36 for function and 0 to 60 for total.

Statistical analysis

Data were analysed by SPSS, version 16 (SPSS, Chicago, Illinois, USA). Multivariate logistic regression analyses were used for comparison of demographic characteristics between patients with OA and RDD controls. Results were presented as odds ratio (OR) with a 95% CI, with adjustments when appropriate.

Linear regression analyses were performed for continuous outcomes in patients with OA (dependent variables: PCS, MCS; independent variables: AUSCAN total score, function and pain). Results were presented as β-estimates (95% CI), with adjustments when appropriate.


Population of patients with OA

The clinical nurse specialist included 487 patients with OA in the study. After verification of the medical chart 27 patients were excluded owing to concomitant musculoskeletal disorders (eg, rheumatoid arthritis (RA), haemochromatosis, psoriatic arthritis, acromegaly).

Comparison of patients with OA with RDD controls

Four-hundred sixty patients were included, of whom the majority were middle-aged and women (table 1). More patients in the OA population were overweight, married and had paid employment than controls, not only adjusted for age and sex, but also for all other demographic characteristics (eg, employment is adjusted in addition to age and sex, also for body mass index, martial status, low education and smoking). Categorisation of cohabitating patients with married patients did not change the results.

Table 1

Baseline characteristics of patients with OA and RDD controls

OA phenotypes

Monoarticular joint site involvement (mono OA) was seen in 244 patients; 94% had hand involvement. OA in more than one joint site (poly OA) was present in 216 patients; 97%, 43% and 11% had hand, knee and hip OA, respectively.

Of all patients with hand OA (n=439), 7.7% reported pain in first CMCJs only, 41.2% in DIPJs and PIPJs only and 42.8% in first CMCJs with DIPJs/PIPJs.

Radiographic hand OA

Hand radiographs were obtained in 247 (56%) of 439 patients with hand OA, showing radiographic OA in the DIPJs, PIPJs, IPJs or first CMCJs in 244 (99%) patients. At least one erosion in DIPJs, PIPJs or IPJs was seen in 61/247 patients (25%), 41 patients had two or more erosions. No differences in demographic characteristics, self-reported pain and function were seen between the groups with or without radiographs (data not shown).

Quality of life

Patients with OA reported a lower PCS than the norm-based population (mean 43, figure 1). MCS was similar to that of the norm-based population. The PCS score was positively (representing better physical QoL) associated with marital status and negatively (representing worse physical QoL) with overweight. Patients with mono OA reported a better PCS (β=2.5, 95% CI 0.7 to 4.3) than patients with poly OA (supplementary table S1).

Figure 1

Mean Short-Form 36 PCS and MCS in patients with OA, in comparison with the general Dutch population using norm-based scoring. Bold middle line = mean score of 50—the average score for PCS and MCS in the general population. Upper and lower lines = a standard deviation of 10 (SD) of PCS and MCS in the general population. *Significantly different from the average score in the general population. MCS, mental component summary score of Short-Form 36; OA, osteoarthritis; PCS, physical component summary score of Short-Form 36.

Self-reported pain and disability

Patients with hand OA (n=439) reported means (SD) of 28.0 (2.6), 9.5 (4.3) and 16.5 (8.6) on the AUSCAN total, pain and function subscales, respectively. When comparing patients with and without first CMCJ involvement, PCS was 2.0 (95% CI −3.9 to −0.1) lower for patients with involvement, adjusted for the number of symptomatic hand joints (supplementary table S2).

Association between self-reported pain and disability with HRQoL

Self-reported disability was associated with lower HRQoL (adjusted β=−0.3, 95% CI −0.4 to −0.2). If patients reported more disability, they reported worse HRQoL. No associations were seen between self-reported pain and HRQoL (Table 2).

Table 2

Association between AUSCAN with SF-36 PCS and AUSCAN with SF-36 MCS (n=439)


Most patients with OA in rheumatology practice have hand OA, with or without involvement of other joint sites. The majority of these patients are women, more often overweight, married and having employment than controls from the general population. HRQoL is lowered in patients with hand OA and is associated with disability, but not with pain. Clear focus on improvements of hand function seems relevant in treatment of these patients.

The predominance of hand OA in rheumatology practice reflects the referral policy in the Netherlands. Patients with hand OA visit rheumatologists, especially when there is doubt about the inflammatory or degenerative origin of disease. Patients with hip and knee OA will be referred to orthopaedic surgeons.

Physical HRQoL was lowered in all patients with OA. This result was in line with an earlier study reporting a lower HRQoL in 190 female patients with hand OA than in healthy controls.5 15 In these patients, worse mental health was also seen, which was not confirmed by us. Since our patient population represents the daily clinical practice in rheumatology and included consecutive patients (including men), it was possible to generalise the results to all patients with hand OA in secondary care.

Van der Kooij et al16 studied HRQoL in patients with RA using the same norm-based data. Patients with RA have lower HRQoL at the beginning of their disease, but if disease activity after 2 years is reduced by treatment, HRQoL in patients with RA is better than in our patients with hand OA. This study emphasises the importance of the lower HRQoL in patients with hand OA.

Limitations in daily activities and pain are major problems in hand OA. Recently, Bijsterbosch et al reported the clinical burden of hand OA in different subgroups.6 17 Both studies were performed in patients who were selected with familial OA. In our study, we investigated HRQoL, pain and function in a less selected population and confirmed the previous findings. Patients who visit the rheumatologist score even worse, which indicates the severity of patients in secondary care.

Interestingly, a higher score on the AUSCAN function subscale in our study was associated with a lower HRQoL, but the AUSCAN pain subscale was not associated with HRQoL. It might be that pain is not the major problem causing patients to visit rheumatologists. Another explanation might be that pain fluctuates over time (eg, with inflammation) and is absent at the moment the clinical nurse specialist is visited.

A study limitation is that diagnosis of hand OA was based on rheumatologist opinion and not American College of Rheumatology (ACR) criteria.18 Diagnosis by rheumatologists reflects clinical reality. Unfortunately, not all radiographs from patients were available. It represents the course of daily clinical practice and is in line with ACR criteria stating that hand OA is a clinical diagnosis. However, available radiographs in patients with hand OA showed that most structural damage in hands was compatible with hand OA. No differences were seen in demographic and clinical characteristics between people with or without a hand radiograph.


View Abstract


  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.