Effects of a Physiotherapist-delivered Dietary Weight Loss Program in People with Knee OA Who Have Overweight/obesity

Clinical guidelines for the management of knee OA recommend exercise as a core treatment for all patients as well as losing weight for those patients who also have overweight/obesity. However, health care systems are frequently under strain often experiencing limited numbers of, and access to, clinicians with specialist weight loss and health behavior change skills. New models of care for OA are needed that support expanded practice roles for clinicians. Physiotherapists are key providers of exercise therapy for knee OA but the effectiveness of physiotherapists to deliver a dietary weight loss program for patients with knee OA is not known. This study is a randomized controlled trial for which the aim is to evaluate the effects of a physiotherapist-delivered dietary weight loss program in addition to exercise, on weight loss and other clinical outcomes among people with knee OA who have overweight or obesity. The research question is: Is a 6-month physiotherapist-delivered dietary weight loss plus exercise program more effective for improving clinical outcomes than a physiotherapist-delivered exercise program alone in people with knee OA who have overweight or obesity? Clinical practice guidelines for knee OA recommend patients who have overweight or obesity should lose at least 5-7.5% of body weight. Therefore the study is powered to detect a conservative between-group difference in weight loss of 5% of body weight assuming no change in weight in the control group (exercise only) based on previous research. A total of 88 participants with chronic knee pain and a clinical diagnosis of knee OA will be recruited from the community. Participants will be enrolled into the study following informed consent and completion of baseline questionnaires and laboratory-based measures. Each participant will be randomly allocated to receive a: a) diet plus exercise program or; b) exercise program alone, over 6 months. The randomization schedule will be computer generated and prepared by the biostatistician (permuted block sizes 6 to 12) stratified by physiotherapist and sex of participant. The schedule will be stored on a password-protected website maintained by a researcher not involved in either participant recruitment or administration of outcome measures. Group allocation will be revealed to the physiotherapists and the participants by this same researcher following randomisation. The physiotherapists will undertake comprehensive training prior to being allocated trial participants. Physiotherapists will be trained in weight management, the ketogenic very low calorie diet and trial procedures. Study participants in both groups will visit a physiotherapist in-person for six individual sessions over 6 months. Participants will choose the therapist according to location. The same therapist will undertake all consultations with any given participant. Those in the diet plus exercise group will undertake a ketogenic very low calorie diet (VLCD) which has been demonstrated as a safe and effective means of achieving rapid weight loss in the adult population with overweight/obesity. They will receive meal replacements (maximum 2 per day) for up to 14 weeks from the start of the study as well as educational resources. Both groups will also undertake a home-based lower limb muscle strengthening exercise and physical activity program. A biostatistician will analyse blinded data. Comparative analyses between groups will use intention-to-treat. Multiple imputation will be used to account for missing data if the proportion of missing data is \>5%. For the primary outcome, the difference in mean percentage change in body weight will be compared between groups using mixed linear regression model adjusting for baseline weight and the stratification variables, with random intercepts for treating physiotherapist. Similar analyses will be conducted for continuous secondary outcomes. We will also calculate the proportion of participants achieving ≥5% and ≥10% loss of body weight in both groups. For binary outcomes, logistic regression models will be fit using generalized estimating equations, with risk differences and 95% confidence intervals calculated.