Background It has been suggested that the heterogeneous population of persons with knee osteoarthritis (OA) consists of different phenotypes or subgroups, which may have important implications for clinical care and research. In a recent paper by Knoop et al. 2011 five distinct knee OA phenotypes were identified.

Objectives To identify and compare different knee OA phenotypes in a Norwegian OA cohort according to the method suggested by Knoop et al. 2011.

Methods Data from 106 persons in the “Musculoskeletal pain in Ullensaker STudy” (MUST), an ongoing population-based study of individuals with self-reported knee OA, was included. Phenotypes were identified by K-means cluster analysis based on four clustering variables: severity of radiographic knee OA, lower extremity muscle strength (the 30-second Chair Stand Test), body mass index (BMI), and “emotional problems” (a single COOP/WONCA item). The phenotypes were compared on clinical outcome variables: self-reported pain and function, and the Six-Minute Walk Test.

Results The mean age (sd) was 65 (8.9) years, and there were 65% females (n=69). Mean BMI was 28.4 (4.5), and for 60% (n=64) the ACR criteria for knee OA were fulfilled. Five different phenotypes were identified in the cluster analysis (Table 1). Comparing the phenotypes on clinical outcome variables there was a gradient from the first, strong and little affected phenotype, to the fifth showing higher levels of pain and disability, and shorter walked distance. Contrary to the findings by Knoop et al, it was not the fifth phenotype that was characterised with “emotional problems”.

Conclusions The phenotype characteristics identified in the present study were similar to those described by Knoop et al. Thus, the present study gives empirical support to the hypothesis that tailoring the treatment of knee OA to phenotype, might optimize the clinical effect.

1. Knoop J et al. 2011

Disclosure of Interest None Declared