Multiple myeloma (MM) is a lethal neoplasm and a common hematologic malignancy. MM is uniformly preceded by monoclonal gammopathy of undetermined significance (MGUS). Unlike MM, patients with MGUS are asymptomatic. The current management for MGUS is watchful waiting for disease progression. A marked racial disparity in this disease area is long-established with a 2 to 3-fold increased risk in African Americans (AAs) compared to Caucasians. Moreover, obesity is a risk factor for MM independent of race. Obesity is prevalent in U.S. adults, and particularly more prevalent among AAs than Caucasians. As a result, without any intervention, racial disparities in this disease will continue to worsen. Metformin, a widely-used, safe, welltolerated, and inexpensive medication, induces weight loss and has been found to be more effective in glycemic control in AAs compared to Caucasians. It has also been used in prospective trials for non-diabetes indications and solid tumor malignancies. We therefore hypothesize that metformin use in MGUS patients will prevent MM and reduce MM disparities. This project plans to focus on the precursor condition of MM – MGUS. The findings from the proposed project will inform biological mechanism studies and MGUS/MM prevention trials. The long-term goal is to identify intervention strategies to prevent the progression of MGUS to MM, reduce the overall burden of MM, and reduce MM disparities. We plan to identify whether high body mass index (BMI) and/or significant change in BMI over the life course are risk factors for MGUS by race (Aim 1), utilizing linked databases with nearly lifelong follow-up of BMI and other health measures ,as well as utilizing artificial intelligence, i.e., machine learning approaches, to perform big data analyses. We will then assess racial differences in the M-protein trajectory after MGUS diagnosis in metformin versus non-metformin users in a subgroup of MGUS patients diagnosed with diabetes mellitus (Aim 2.1) and the association of metformin use with the progression of MGUS to MM (Aim 2.2). Last, we will assess racial differences in the M-protein trajectory in the subgroup of MGUS patients without diabetes mellitus (Aim 3). This project is significant in its capability to 1) identify perhaps the only modifiable risk factor (high BMI) to inform interventions to prevent MM; 2) identify a dynamic marker for disease progression by race (M-protein concentration), where these biomarkers can be a surrogate for MM diagnosis in future prevention trials; and 3) reduce MM health disparities by a) identifying race-specific biomarkers for MGUS and MGUS progression, available through clinical encounters (as opposed to expensive genetic testing); and b) exploring metformin use as a chemopreventive measure. It is innovative in its 1) focus on MM prevention rather than treatment and 2) utilization of artificial intelligence to analyze big data. Successful completion of this study will provide evidence for a paradigm shift in current clinical practice of MGUS management and help prevent MM, an incurable and costly disease. More importantly, it will provide evidence to guide interventions to reduce MM disparities.