Author Block: J. Most, H.M. Eijking, I. Dorling, C. Geurts, B. Boonen;
Zuyderland Medical Center, Geleen, Netherlands.

Abstract:
Background and aims: Patients with osteoarthrosis are disproportionally affected by metabolic diseases, likely due to low physical activity caused by pain in activity. The present study was designed to assess the efficacy of total knee arthroplasty on functional capacity, physical activity and ultimately metabolic parameters.
Materials and methods: Within an ongoing randomized controlled trial testing the clinical efficacy or robot-assistance in total knee arthroplasty, we assessed metabolic, functional and activity parameters to investigate the effect of knee replacements on metabolic health and its determinants. In 81 patients, data was collected on vitals, fasting metabolic parameters, body composition using bioimpedance, patient-reported outcome measures on arthrosis symptoms, mobility, function and strength by clinical testing (De Morton-mobility-index, stand-up-and-go, 2-minute-walking, sit-to-stand, handgrip strength) and physical activity (at least 3 observation days with >18 observation hours and 1000 steps) using hip-worn accelerometry for 7 days. At the time of the conference, we will have continuous pre- and post-surgery glucose monitoring data of ~20 patients. Pre- and postoperative data was compared by paired t-tests using SPSS.
Results: The present cohort consists of 81 patients (48% female, 69±8 years, 46% with obesity, 12% self-reported diabetes). Based on HbA1c-levels (>39 and >46), 46% and 12% of patients are defined as pre-diabetes and diabetes, respectively. Diabetes was poorly controlled, as diabetes-defining hyperglycemia (HbA1c>47 mmol/mol) was still observed in >75% of self-reported diabetes patients, and in 90% of patients who took diabetes medication.

Three months after surgery, symptoms have improved mildly (Oxford Knee Score [0 worst, 48 best], 26±7 to 32±7, p<0.001). Mobility (DEMMI) was not yet improved after surgery (NS). Fasting glucose concentrations decreased significantly (6.0±1.4 to 5.7±1.2 mmol/l, P=0.001), whereas HbA1c, fasting triglycerides and HDL-cholesterol were unchanged (NS). Physical activity was not improved after 3 months (180±5458 steps/day), but changes showed large interindividual variations, possibly associated with season. Patients were sedentary 85% of the observation period. Leg strength significantly increased (8.8±3.0 to 9.6±2.9 sit-to-stand transfers during 30sec, p=0.008), while endurance remained unchanged (two-minute walking distance, 137±31 to 139±29m, NS). On average, patients lost 1.7±2.6 kg body mass (p<0.001), at unchanged body composition.
Conclusion: The present study demonstrates for the first time an effect of arthroplasty on diabetes risk. The observed effect size associates with a 10% reduced mortality due to hyperglycemia. Future research should aim to understand the reduction and etiology of the improvement in glucose metabolism, eg by use of continuous glucose monitoring.