Identifying subgroups of community-dwelling older adults and their prospective associations with long-term knee osteoarthritis outcomes

Ishanka P. Munugoda, Feng Pan, Karen Wills, Siti M. Mattap, Flavia Cicuttini, Stephen E. Graves, Michelle Lorimer, Graeme Jones, Michele L. Callisaya, Dawn Aitken

Research output: Contribution to journalArticleResearchpeer-review

1 Citation (Scopus)


Objectives: To identify subgroups of community-dwelling older adults and to assess their longitudinal associations with long-term osteoarthritis (OA) outcomes. Methods: 1046 older adults aged 50–80 years were studied. At baseline, body mass index (BMI), pedometer-measured ambulatory activity (AA), and Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) determined knee pain and information on comorbidities were obtained. Tibial cartilage volume and bone-marrow lesions (BMLs) were assessed using MRI at baseline and 10 years and total knee replacements (TKR) by data linkage to the Australian Orthopaedic Association National Joint Replacement Registry. Latent class analysis was used to determine participant subgroups, considering baseline BMI, AA, pain and comorbidities, and linear mixed-effects or log-binomial models were used to assess the associations. Results: Three subgroups/classes were identified: subgroup 1 (43%): Normal/overweight participants with higher AA, lower pain and lower comorbidities; subgroup 2 (32%): Overweight participants with lower AA, mild pain and higher comorbidities; subgroup 3 (25%): Obese participants with lower AA, mild pain and higher comorbidities. Subgroup 3 had greater cartilage volume loss (β − 60.56 mm3, 95% CI − 105.91, − 15.21) and a higher risk of TKR (RR 3.19, 95% CI 1.75, 5.81), compared to subgroup 1. Subgroup 2 was not associated with cartilage volume change (β 13.06 mm3, 95% CI − 30.87, 57.00) or risk of TKR (RR 1.16, 95% CI 0.56, 2.36), compared to subgroup 1. Subgroup membership was not associated with worsening BMLs. Conclusions: Our findings suggest the existence of homogeneous subgroups of participants and support the utility of identifying patterns of characteristics/risk factors that may cluster together and using them to identify subgroups of people who may be at a higher risk of developing and/or progressing OA. Key Points• Complex interplay among characteristics/factors leads to conflicting evidence between ambulatory activity and knee osteoarthritis.• Distinct subgroups are identifiable based on ambulatory activity, body mass index, knee pain, and comorbidities.• Identifying subgroups can be used to determine those who are at risk of developing/progressing osteoarthritis.

Original languageEnglish
Pages (from-to)1429-1437
Number of pages9
JournalClinical Rheumatology
Issue number5
Publication statusPublished - May 2020


  • Bone-marrow lesions
  • Cartilage volume
  • Latent-class analysis
  • Osteoarthritis
  • Total knee replacement

Cite this