S Swain1, C Coupland2, V Strauss3, C Mallen4, C F Kuo5, A Sarmanova6, S M A Bierma-Zeinstra7, M Englund8, D Prieto-Alhambra3, M Doherty9, W Zhang10. 1. Academic Rheumatology, School of Medicine, University of Nottingham, UK; Nuffield Department of Primary Care Health Sciences, University of Oxford, UK. 2. Division of Primary Care, School of Medicine, University of Nottingham, UK. 3. Centre for Statistics in Medicine, NDORMS, University of Oxford, UK. 4. School of Medicine, Keele University, UK. 5. Academic Rheumatology, School of Medicine, University of Nottingham, UK; Division of Rheumatology, Allergy and Immunology, Chang Gung Memorial Hospital, Taiwan. 6. Musculoskeletal Research Unit, Bristol Medical School, Translational Health Sciences, University of Bristol, UK. 7. Department of General Practice, Department of Orthopaedics and Sports Medicine, Erasmus MC - University Medical Center Rotterdam, the Netherlands. 8. Clinical Epidemiology Unit, Orthopaedics, Department of Clinical Sciences Lund, Lund University, Sweden. 9. Academic Rheumatology, School of Medicine, University of Nottingham, UK; Pain Centre Versus Arthritis, University of Nottingham, UK. 10. Academic Rheumatology, School of Medicine, University of Nottingham, UK; Pain Centre Versus Arthritis, University of Nottingham, UK. Electronic address: weiya.zhang@nottingham.ac.uk.
Abstract
OBJECTIVE: To examine the clusters of chronic conditions present in people with osteoarthritis and the associated risk factors and health outcomes. METHODS: Clinical Practice Research Datalink (CPRD) GOLD was used to identify people diagnosed with incident osteoarthritis (n = 221,807) between 1997 and 2017 and age (±2 years), gender, and practice matched controls (no osteoarthritis, n = 221,807) from UK primary care. Clustering of people was examined for 49 conditions using latent class analysis. The associations between cluster membership and covariates were quantified by odds ratios (OR) using multinomial logistic regression. General practice (GP) consultations, hospitalisations, and all-cause mortality rates were compared across the clusters identified at the time of first diagnosis of osteoarthritis (index date). RESULTS: In both groups, conditions largely grouped around five clusters: relatively healthy; cardiovascular (CV), musculoskeletal-mental health (MSK-MH), CV-musculoskeletal (CV-MSK) and metabolic (MB). In the osteoarthritis group, compared to the relatively healthy cluster, strong associations were seen for 1) age with all clusters; 2) women with the MB cluster (OR 5.55: 5.14-5.99); 3) obesity with the CV-MSK (OR 2.11: 2.03-2.20) and CV clusters (OR 2.03: 1.97-2.09). The CV-MSK cluster in the osteoarthritis group had the highest number of GP consultations and hospitalisations, and the mortality risk was 2.45 (2.33-2.58) times higher compared to the relatively healthy cluster. CONCLUSIONS: Of the five identified clusters, CV-MSK, CV, and MSK-MH are more common in OA and CV-MSK cluster had higher health utilisation. Further research is warranted to better understand the mechanistic pathways and clinical implications.
OBJECTIVE: To examine the clusters of chronic conditions present in people with osteoarthritis and the associated risk factors and health outcomes. METHODS: Clinical Practice Research Datalink (CPRD) GOLD was used to identify people diagnosed with incident osteoarthritis (n = 221,807) between 1997 and 2017 and age (±2 years), gender, and practice matched controls (no osteoarthritis, n = 221,807) from UK primary care. Clustering of people was examined for 49 conditions using latent class analysis. The associations between cluster membership and covariates were quantified by odds ratios (OR) using multinomial logistic regression. General practice (GP) consultations, hospitalisations, and all-cause mortality rates were compared across the clusters identified at the time of first diagnosis of osteoarthritis (index date). RESULTS: In both groups, conditions largely grouped around five clusters: relatively healthy; cardiovascular (CV), musculoskeletal-mental health (MSK-MH), CV-musculoskeletal (CV-MSK) and metabolic (MB). In the osteoarthritis group, compared to the relatively healthy cluster, strong associations were seen for 1) age with all clusters; 2) women with the MB cluster (OR 5.55: 5.14-5.99); 3) obesity with the CV-MSK (OR 2.11: 2.03-2.20) and CV clusters (OR 2.03: 1.97-2.09). The CV-MSK cluster in the osteoarthritis group had the highest number of GP consultations and hospitalisations, and the mortality risk was 2.45 (2.33-2.58) times higher compared to the relatively healthy cluster. CONCLUSIONS: Of the five identified clusters, CV-MSK, CV, and MSK-MH are more common in OA and CV-MSK cluster had higher health utilisation. Further research is warranted to better understand the mechanistic pathways and clinical implications.