Symptom assessment in knee osteoarthritis needs to account for physical activity level

Grace H. Lo*, Timothy E. McAlindon, Gillian A. Hawker, Jeffrey B. Driban, Lori Lyn Price, Jing Song, Charles B. Eaton, Marc C. Hochberg, Rebecca D. Jackson, C. Kent Kwoh, Michael C. Nevitt, Dorothy D. Dunlop

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

21 Scopus citations


Objective Pain is not always correlated with severity of radiographic osteoarthritis (OA), possibly because people modify activities to manage symptoms. Measures of symptoms that consider pain in the context of activity level may therefore provide greater discrimination than a measure of pain alone. We undertook this study to compare discrimination provided by a measure of pain alone with that provided by combined measures of pain in the context of physical activity across radiographic OA severity levels. Methods This was a cross-sectional study nested within the Osteoarthritis Initiative (OAI). The population was drawn from 2,127 persons enrolled in an OAI accelerometer monitoring substudy, including those with and those without knee OA. Two composite pain and activity knee symptom (PAKS) scores were calculated as the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain score (plus 1) divided by a physical activity measure (step count for the first PAKS score [PAKS1 score] and activity count for the second PAKS score [PAKS2 score]). Symptom score discrimination across Kellgren/Lawrence (K/L) grades was evaluated using histograms and quantile regression. Results A total of 1,806 participants (55.5% of whom were women) were included (mean ± SD age 65.1 ± 9.1 years, mean ± SD body mass index 28.4 ± 4.8 kg/m2). The WOMAC pain score, but not the PAKS scores, exhibited a floor effect. The adjusted median WOMAC pain scores by K/L grades 0-4 were 0, 0, 0, 1, and 3, respectively. The adjusted median PAKS1 scores were 24.9, 26.0, 32.4, 46.1, and 97.9, respectively, and the adjusted median PAKS2 scores were 7.2, 7.2, 9.2, 12.9, and 23.8, respectively. The PAKS scores had more statistically significant comparisons between K/L grades than did the WOMAC pain score. Conclusion Symptom assessments incorporating pain and physical activity did not exhibit a floor effect and were better able to discriminate radiographic severity than an assessment of pain alone, particularly in milder disease. Pain in the context of physical activity level should be used to assess knee OA symptoms.

Original languageEnglish (US)
Pages (from-to)2897-2904
Number of pages8
JournalArthritis and Rheumatology
Issue number11
StatePublished - Nov 2015

ASJC Scopus subject areas

  • Immunology and Allergy
  • Rheumatology
  • Immunology


Dive into the research topics of 'Symptom assessment in knee osteoarthritis needs to account for physical activity level'. Together they form a unique fingerprint.

Cite this