The association between comorbidity and physical activity levels in people with osteoarthritis: secondary analysis from two randomised controlled trials

Abstract

Purpose: Osteoarthritis (OA) is highly prevalent in patients with multiple long-term chronic morbidities. Clinical guidelines recommend physical activity (PA) in the form of general aerobic and strengthening exercises for people with OA irrespective of comorbidity to improve health. However, the relationship between comorbidity and PA behaviour in adults with OA, is unknown. The objectives of this study are to investigate whether; 1) comorbidity presence is associated with PA levels 2) the number of comorbid is conditions associated with PA levels and 3) different types of comorbidity are associated with PA levels. Methods: Adults aged ≥45 with knee pain attributed to OA were recruited to the BEEP trial (ISRCTN93634563) and the MOSAICS trial (ISRCTN06984617) respectively between 2009 and 2014 from UK general practices and NHS physiotherapy services. Comorbidities including; respiratory (asthma, bronchitis, chronic obstructive pulmonary disease), cardiovascular (angina, heart failure, stroke, heart disease, heart attack), depression, type 2 diabetes and obesity (BMI ≥30) were identified from Self-report baseline questionnaire (BEEP) and medical record review (MOSAICS). Self-report PA levels were measured using the Physical Activity Scale for the Elderly (PASE - ranging from 0-400+ with higher scores indicating higher PA levels). Linear regression models were used to estimate the adjusted mean change (95% confidence interval) in PASE score in the presence of; any comorbidity; increased number of comorbidities (0 (reference), 1, 2, 3+ comorbidities), and; presence of specific comorbidity types. Results were adjusted for covariates (demographics; age, gender, relationship status and clinical measures; Western Ontario and McMaster Osteoarthritis Index (function subscale) and anxiety (Generalised Anxiety Disorder 7). Results: The BEEP and MOSAICS trials recruited 514 and 525 participants respectively. Participant mean ± standard deviation age was 62.9±9.8; 67.3±10.5, and 51%; 60% were female (BEEP; MOSAICS). Over half of the participants had at least one comorbidity (BEEP: n=324 (66%); MOSAICS: n=271 (52%)). Mean PASE scores were 176.9±83.5 (BEEP) and; 142.7±80.3 (MOSAICS). Comorbidity presence was associated with a decrease in PASE score in BEEP however, no statistically significant association was observed in MOSAICS (Table 1). In the comorbidity count model, each additional comorbidity was associated with an incrementally lower PASE score in BEEP (adjusted models displayed in Table 1). This pattern was similar in MOSAICS, but with a plateau in association from two comorbidities onward. The specific types of comorbidity had different strengths of associations with PASE within and between trials Table 1.). In BEEP and MOSAICS, respiratory and cardiovascular comorbidity were associated with the largest reduction in PASE scores respectively. A greater number of comorbidity types were found to be associated with PASE score in BEEP than in MOSAICS. Conclusions: This is the first study to investigate and describe in detail the association between comorbidity and PA in people with OA. Our findings suggest a dose-response relationship with activity levels decreasing as the number of comorbidities increase. Presence of respiratory disease, cardiovascular disease and Type II diabetes were most strongly associated with lower PA levels. A strength of this analysis was the use of two datasets and two methods for recording comorbidity. A limitation was that only self-report PA was measured within the two trials, there was no objectively measured PA. Although the directions of associations were generally consistent across the two trials there was some difference in the magnitude of associations. Since our analysis is cross-sectional, we cannot infer causation and it is possible that a reciprocal relationship exists where comorbidities may be both risk factors for, or a result of, reduced PA. Future explanatory research could help inform our understanding of the relationship between comorbidity and PA behaviour in people with OA. This information could then be used to inform future interventions aimed at increasing PA in people with OA and comorbidity.