Article Text

Download PDFPDF

Concise report
Thumb base involvement in symptomatic hand osteoarthritis is associated with more pain and functional disability
  1. Jessica Bijsterbosch1,
  2. Willemien Visser1,
  3. Herman M Kroon2,
  4. Tanja Stamm3,
  5. Ingrid Meulenbelt4,
  6. Tom W J Huizinga1,
  7. Margreet Kloppenburg1
  1. 1Department of Rheumatology, Leiden University Medical Centre, Leiden, The Netherlands
  2. 2Department of Radiology, Leiden University Medical Centre, Leiden, The Netherlands
  3. 3Department of Internal Medicine III, Division of Rheumatology, Medical University of Vienna, Vienna, Austria
  4. 4Department of Molecular Epidemiology, Leiden University Medical Centre, Leiden, the Netherlands
  1. Correspondence to Jessica Bijsterbosch, Leiden University Medical Centre, Department of Rheumatology, C1-R, PO Box 9600, 2300 RC Leiden, The Netherlands; j.bijsterbosch{at}


Objective To assess the impact of different subsets of symptomatic hand osteoarthritis (OA) on pain and disability.

Methods From 308 patients with hand OA a group with carpometacarpal joint (CMCJ) symptoms only (group I, n=20) was identified as well as groups with symptoms at the interphalangeal joints (IPJs) only (group II, n=138), and symptoms at both sites (group III, n=150). Hand pain and function, assessed with the AUSCAN, were compared between groups using linear mixed models. Radiological OA was assessed using the Kellgren–Lawrence grading scale.

Results Mean (SD) AUSCAN scores for groups I, II and III were 23.1 (11.7), 18.3 (11.9) and 26.4 (12.5), respectively. After adjustment for age, gender, body mass index, family effects and number of symptomatic hand joints, significant differences in AUSCAN scores of 7.4 (95% CI 1.8 to 13.0) between groups I and II, and 5.7 (95% CI 2.7 to 8.6) between groups II and III were found. AUSCAN scores were 5.8 (95% CI 3.1 to 8.6) higher for patients with versus patients without CMCJ symptoms. Kellgren–Lawrence scores did not differ between groups.

Conclusion In symptomatic hand OA, CMCJ OA contributes more to pain and disability than IPJ OA. Hence, treatment of CMCJ OA should be emphasised, even if it coincides with IPJ OA.

Statistics from

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.


Hand osteoarthritis (OA) is a common musculoskeletal disorder, leading to variable degrees of pain and disability.1 It typically affects the distal interphalangeal joints (DIPJs), followed by the proximal interphalangeal joints (PIPJs), and the first carpometacarpal joints (CMCJs).1 2

Different subsets of hand OA have been proposed based on different risk factors, associations and outcomes, although evidence is limited.3 4 Recognised subsets are interphalangeal joint (IPJ) OA (with/without nodes) and CMCJ OA. Articular hypermobility was positively associated with CMCJ OA, while it was found to be protective for IPJ OA.5 6 In addition, IPJ OA was found more often in the dominant hand, whereas CMCJ OA was found more often in the non-dominant hand.7 Few data are available on health outcomes in these subsets.8

The impact of functional limitations in the IPJs can differ from that in CMCJs, because IPJ OA causes limitations in movement of the fingers, whereas CMCJ OA affects closure of the first web. Therefore, different treatment strategies may be required. Current EULAR recommendations state that treatment of hand OA should be individualised according to its localisation.9

In this study we take advantage of the presence of different subsets of symptomatic hand OA in a relatively large cohort. A group of patients with CMCJ OA only was identified as well as patients with IPJ OA only and patients with OA at both joint sites. We compared the pain and disability in between these subsets, which may have implications for the importance of treatment for each joint group. This study can contribute to the further distinction between subsets of hand OA and recommended management strategies.

Patients and methods

Study design and patient population

The Genetics, Arthrosis and Progression (GARP) study is a cohort study aimed at identifying determinants of OA susceptibility and progression.10 A total of 192 Caucasian sib pairs with OA at multiple sites in the hands or in two or more sites–namely, hand, knee, hip, or spine, were included after giving informed consent. Details on the recruitment and selection have been published elsewhere.10 The study was approved by the medical ethics committee.

Patients were eligible for this study if they (a) fulfilled the American College of Rheumatology (ACR) criteria for clinical hand OA11 or if they (b) had hand pain or stiffness on most of the days of the preceding month in addition to multiple bony swellings in the selected joints from the ACR criteria, or a Kellgren–Lawrence score ≥2 in any hand joint.

A standard diagram of the hand joints was used to identify painful and stiff joints. Based on the location of these self-reported symptoms patients were assigned to three groups: group I with CMCJ symptoms only, group II with IPJ symptoms only and group III with symptoms at both sites. The number of symptomatic joints (maximum 30) identified by this method was used for analysis.

Disease characteristics

Self-reported hand pain and function were assessed with the pain (five items) and physical functioning (nine items) subscales, as well as the total score (15 items) of the Australian/Canadian Osteoarthritis Hand Index LK 3.0 (AUSCAN) on a five-point Likert scale (0=none to 4=extreme).12

Hand radiographs (dorsal-volar) were obtained by a single radiographer, employing a standard protocol. Radiological hand OA was evaluated by an experienced radiologist (HMK) using the Kellgren–Lawrence grading scale.13 Intrareader reproducibility was high.10

Statistical analysis

Data were analysed using SPSS, version 14.0 (SPSS, Chicago, Illinois, USA). Demographic characteristics, AUSCAN and Kellgren–Lawrence scores were compared between the three groups using one-way ANOVA for normally distributed variables, the Kruskal–Wallis test for not normally distributed variables and χ2 test for proportions. For post hoc analysis the Bonferroni test and Mann–Whitney U test were used. All tests were two-tailed and p values <0.05 were considered statistically significant.

Hand pain and function measured by the AUSCAN were compared between groups using linear mixed models, adjusting for age, gender, body mass index (BMI) and number of symptomatic hand joints. A random intercept was used to adjust for family effects, meaning resemblance between siblings of one family. First the initial three groups were compared, followed by comparison of patients with CMCJ symptoms (groups I + III) and those without CMCJ symptoms (group II). Estimates of fixed effects are reported with 95% CI.


Population description

Of the 308 eligible patients 20 (6.5%) were assigned to group I (CMCJ symptoms only), 138 (44.8%) to group II (IPJ symptoms only), and 150 (48.7%) to group III (symptoms at both sites). The mean age was 60 years, and the majority of patients were women and fulfilled the ACR criteria for clinical hand OA (table 1). Group III consisted of significantly more women than groups I and II. Other demographic characteristics did not differ between the groups. The mean (SD) AUSCAN total score for the whole population was 22.5 (12.8). AUSCAN was positively associated with the number of symptomatic joints.

Table 1

Demographic characteristics, AUSCAN, and Kellgren–Lawrence scores of 308 patients with symptomatic hand osteoarthritis (OA)*

Hand pain and function

Mean (SD) AUSCAN total scores were 23.1 (11.7) for group I, 18.3 (11.9) for group II, and 26.4 (12.5) for group III (table 1). Multivariate analysis showed differences in AUSCAN total scores of 7.4 (95% CI 1.8 to 13.0) between groups I and II, and 5.7 (95% CI 2.7 to 8.6) between groups II and III. Differences between group I and III were not significant. AUSCAN pain and function scores showed the same pattern.

A comparison of patients with and without CMCJ symptoms (groups I + III vs group II) showed that AUSCAN total scores were 5.8 (95% CI 3.1 to 8.6) higher for patients with CMCJ symptoms than for patients without CMCJ symptoms; AUSCAN pain scores were 2.1 (95% CI 1.2 to 3.1) higher and AUSCAN function scores were 3.6 (95% CI 1.8 to 5.5) higher.

Radiological damage

Median Kellgren–Lawrence scores for the total hand did not differ between the groups (table 1). Consideration of the CMCJs showed that group II had lower scores than groups I and III (p<0.01).


In this study it was found that symptomatic CMCJ OA contributes substantially to the level of self-reported pain and disability in patients with symptomatic hand OA. Patients with IPJ symptoms only, reported the lowest levels of pain and disability, followed by patients with CMCJ symptoms only. Patients with symptoms at both sites experienced the highest levels of pain and disability. After adjustment for the number of symptomatic joints, which was associated with pain and disability, the levels of pain and disability reported by patients with CMCJ symptoms remained significantly higher than for patients without CMCJ symptoms. This suggests that treatment aiming at CMCJ symptoms in patients with symptomatic hand OA is important, even if it coincides with IPJ symptoms.

This is one of the first studies comparing patients with symptomatic CMCJ OA with patients with symptomatic IPJ OA. Spacek et al8 compared disability and perceived handicap in hand OA between patients with predominantly thumb base symptoms and patients with predominantly IPJ symptoms. They found that disability and perceived handicap levels were comparable between the groups. However, they classified patients based on the location with most severe symptoms. Thus, patients in the thumb base group could experience IPJ symptoms and vice versa. This classification may be the reason why no differences between the groups were found. The classification criteria used in this study were stricter, resulting in a more pronounced distinction between the groups. In general, no classification criteria for subsets of hand OA are available. We chose self-reported symptoms as classification criteria because symptomatic hand OA is considered the disease of clinical and public health interest.

Several limitations of this study have to be considered. The first, is the small number of patients in the group with CMCJ symptoms only. However, this small number may reflect clinical reality where isolated symptomatic CMCJ OA is not very prevalent. Second, patients in this study had familial OA at multiple sites. Whether the results can be generalised to patients with hand OA only, in a less selected population, has to be investigated.

Based on these results it seems that CMCJ OA adds more to pain and disability in symptomatic hand OA than IPJ OA alone. This may be explained by the prominent role of the thumb in hand functioning. CMCJ symptoms may, therefore, be perceived as more severe and as having more impact on functioning than symptoms at the IPJs. Although no cut-off values are available for the AUSCAN, differences on the function subscale between those with and without CMCJ symptoms seem clinically relevant.14

The findings of this study suggest that treatment of CMCJ symptoms may substantially reduce levels of pain and disability, even if there is concurrent IPJ involvement. The results support expert opinion on the use of intra-articular corticoids and thumb orthosis for CMCJ OA.9 Occupational factors involving repetitive thumb use or heavy load on the thumb are modifiable factors that can contribute to CMCJ OA. Therefore, they should be taken into account when education and lifestyle advice are considered.15 Future research should aim at elucidating the efficacy of interventions targeted at the CMCJ in symptomatic hand OA.



  • Funding The GARP study was supported by Pfizer Inc, Groton, Connecticut, USA and the Dutch Arthritis Association.

  • Ethics approval This study was conducted with the approval of the medical ethics committee of the Leiden University Medical Centre.

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

Linked Articles

  • Miscellaneous
    BMJ Publishing Group Ltd and European League Against Rheumatism
  • Miscellaneous
    BMJ Publishing Group Ltd and European League Against Rheumatism
  • Miscellaneous
    BMJ Publishing Group Ltd and European League Against Rheumatism