Determining optimal diet/exercise treatment assignment for patients with symptomatic knee osteoarthritis using baseline gait forces.

Objective: We examined whether precision medicine models to determine the optimal treatment regimen for participants in an 18-month diet (D), exercise (E), and D ​+ ​E trial for knee osteoarthritis (KOA) could be further improved with the addition of baseline gait forces (ground reaction, muscle, compressive, and shear forces). Methods: We used data from 286 participants in the Intensive Diet and Exercise for Arthritis trial (IDEA). Four machine learning models were used to develop individualized treatment rules for change in outcomes: weight, WOMAC (Western Ontario and McMaster Universities Osteoarthritis Index) pain/function/stiffness, tibiofemoral compressive forces, plasma Interleukin-6 levels, and SF-36 physical component score. We selected the optimal model for each outcome and compared it to the optimal fixed treatment model as well as the optimal model excluding gait forces. Results: We found no statistically significant differences between estimated values of any zero order models (ZOMs) and optimal precision medicine models (PMMs) with gait, nor between optimal PMMs with and without gait. For several outcomes, the optimal PMM without gait performed slightly better than the optimal PMM with gait. The only outcome for which the PMM resulted in a higher estimated value than both the ZOM and the PMM without gait was WOMAC function change. Conclusions: PMMs exhibited no statistically significant differences in estimated values for 18-month change in outcomes when including gait forces. Although potentially underpowered, these results suggest that gait forces, although involved in the KOA phenotype, may not meaningfully influence diet/exercise treatment outcomes for individuals with obesity and symptomatic KOA.