Project Summary/Abstract In 2016, 47,000 individuals who initiated dialysis in the United States (~42% of all incident dialysis patients that year) had obesity, with a body mass index (BMI) of ≥ 30 kilograms per meters squared. Across the BMI spectrum, individuals with kidney disease commonly lose weight after initiating dialysis treatment. However, whereas body weight typically stabilizes after the first several months of dialysis among patients without obesity, those with obesity often continue to lose weight. Both people with and without obesity who are on dialysis may lose weight due to muscle wasting and malnutrition, and recent studies have identified weight loss as a risk factor for death among people on dialysis, independent of BMI. Yet, some of the weight loss observed among obese dialysis patients may also reflect deliberate attempts to improve health, mobility, or access to kidney transplantation. Currently, there are no guidelines to help clinicians to differentiate between healthy and high-risk weight loss among people with obesity on dialysis. Further, typical obesity management paradigms are not easily transferrable to obese people with end-stage kidney disease, given factors such as chronic malnutrition, inflammation, and sarcopenia in this population that may modify the risks and benefits of different weight loss strategies. Therefore, the overarching goal of this five-year research proposal is to define healthy and high-risk weight loss phenotypes among people with obesity who are on dialysis, and to provide clinically feasible tools to improve obesity management in the setting of end-stage kidney disease. We will accomplish this goal by conducting three distinct but interrelated studies. In the first study, we will qualitatively determine patient-prioritized endpoints of weight loss, in addition to patient, physician and other stakeholder perspectives on the key factors that differentiate healthy from high-risk weight loss on dialysis. In the second study, we will leverage a national dataset of 23,000 obese dialysis patients and apply constructs of high and low physiologic reserve to derive healthy and high-risk weight loss phenotypes. We will then develop a weight- loss risk calculator tool that predicts the risks of hospitalization and death that are associated with each weight loss phenotype, using dynamic predictive joint models and machine learning techniques. In the third study, we will enroll 250 obese dialysis patients in a prospective, longitudinal study across five regions in the United States to evaluate the association between nutritional, inflammatory, and hemodynamic biomarkers and measures of health trajectory that are not typically captured in registry data, such as sarcopenia, dynapenia, body composition, and patient-prioritized endpoints such as quality of life. In accomplishing its aims, this research will provide urgently needed knowledge and tools that will improve the medical management of tens of thousan...