Background/Purpose

Knee instability in the setting of osteoarthritis (OA) encompasses a spectrum of symptoms and phenomena, including a feeling of low overall confidence in the knees, low confidence that the knees will not buckle or give way (buckling confidence), actual episodes of buckling, and excessive frontal plane motion.  Current treatment for knee OA does little to address instability.  Given the central role of the knee in weightbearing activity, confidence and buckling in particular may influence nature and intensity of activity, and could be important proximal factors in a chain of events leading to function decline and disability.  It is unclear whether these factors are more important to outcome than instability objectively measured during gait.  We hypothesized that overall confidence, buckling confidence, buckling, and excessive frontal plane motion during gait are each associated with poor 2-year function outcome.     

Methods

Persons with OA in at least one knee were queried at baseline about overall knee confidence using the KOOS question (how troubled are you by the lack of confidence in your knees, higher worse), buckling confidence (i.e., confidence that knees will not buckle or give way, higher better), and any knee buckling in the past 3 months, and underwent quantitative gait analysis (3-dimensional knee kinematics and kinetics during ambulation recorded using an 8-camera Digital Real-Time Eagle motion analysis system, and 6 AMTI force plates).  Physical function was assessed using the Late-Life Function Instrument – Advanced Lower Extremity Domain scaled score; quintiles were used to categorize these scores into groups.  Poor outcome was defined as moving into a worse function group or remaining in the 2 worst function groups between baseline and 2 years.  Logistic regression was used to evaluate the relationship between baseline instability measures, knee confidence and poor baseline-to-2-year function outcome, adjusting for potential confounders.

Results

212 persons [163 (77%) women, mean age 65 (10, SD), BMI 28.5 (5.7)] comprised the sample.

98 (46%) had a poor outcome.  As shown in the Table, buckling in the past 3 months, buckling confidence, and overall knee confidence, but neither varus-valgus excursion nor maximal varus-valgus angular velocity during gait were associated with the outcome in univariate analyses.  In fully-adjusted models, these findings persisted (except for buckling confidence which approached significance).  Age and self-efficacy were also consistently associated with the outcome. 

Conclusion

Worse baseline overall knee confidence and recent buckling, but neither varus-valgus excursion nor angular velocity using quantitative gait analysis were associated with greater risk of poor 2-year function outcome.  Interventions to address confidence and buckling may improve outcome in knee OA.