Background Although hand pain and function are important complaints in patients with hand osteoarthritis (OA), their determinants are unclear. Comorbidity is shown to be associated with complaints in knee and hip OA.
Objectives Therefore, we studied the association of comorbidities with hand pain and function in hand OA patients.
Methods Cross-sectional data were used of the HOSTAS (Hand OSTeoArthritis in Secondary care) study, which included consecutive patients diagnosed by their treating rheumatologist with primary hand OA from 2009 to 2015. Self-reported comorbidity was assessed by a 17-item list (modified Charlson index). Additionally, presence of knee and/or hip OA (poly OA) was determined by fulfilling the ACR criteria and presence of obesity by body mass index (BMI) ≥30 kg/m2. Number of comorbidities (max 19) and groups of comorbidities (figure) were studied. Self-reported hand pain (0–20) and hand function (0–36) were assessed by Australian/Canadian Hand OA Index (AUSCAN, Likert scale), where higher scores mean worse health. Multivariable linear regression analysis was used to associate comorbidities with hand pain and function, adjusting for age, sex and education (3 categories, as proxy for socioeconomic status). To assess clinical relevance, we compared mean differences between absence and presence of specific comorbidities, with the minimal clinically important improvement (MCII, published by Bellamy et al.), which was 1.6 (95%CI 1.0 to 2.0) for pain and 1.4 (0.1 to 2.2) for function.
Results 538 patients were included with a mean (SD) age of 61 (9) years, mean BMI of 27 (5) kg/m2, 86% women and 91% fulfilling the ACR hand OA criteria. 27% had a low education level. 38% of patients reported ≥1 comorbidity (of 17), 24% had poly OA (mainly knee) and 23% had obesity. Mean (SD) pain and function were 9.3 (4.3) and 15.6 (8.5), where women had worse scores than men. Presence of any comorbidity was associated with worse pain and function; mean difference (95%CI) pain 1.1 (0.3 to 1.9) and function 1.8 (0.2 to 3.3). Also the number of comorbidities resulted in worse complaints (regression coefficient (95%CI) pain 0.8 (0.4 to 1.1) and function 1.4 (0.8 to 2.1)). Presence of poly OA or lung disease was associated with both pain and function, while presence of cardiovascular disease was associated with pain only and osteoporosis with function only. Other comorbidities, like obesity, were not associated (figure). For pain, only the presence of cardiovascular disease was similar to the MCII, while all associations (i.e. presence of any comorbidity, poly OA, osteoporosis and lung disease) were clinically relevant for function. For example, presence of lung disease resulted in 3.4 points higher (worse) score for function (figure), while MCII was 1.4, so this difference could be considered clinically relevant.
Conclusions Presence as well as number of comorbidities showed a clinically relevant association with self-reported hand pain and function in patients with hand OA. Remarkably, as in knee OA, not only musculoskeletal comorbidity was associated with hand symptoms, but also non-musculoskeletal comorbidities (lung and cardiovascular disease). Further studies should investigate the role of comorbidities in clinical burden in hand OA and how these can be modified.
Disclosure of Interest None declared