Diversity Supplement: Urine cadmium and risk of fracture and bone loss Abstract We propose to leverage the parent grant, which uses existing samples and data from the two largest prospective US osteoporosis cohort studies to perform an in depth study of urine cadmium (U-Cd), bone loss, and fractures that will include long-term follow-up of up to 20 years using: 1) The Osteoporotic Fractures in Men (MrOS) Study and 2) the Study of Osteoporotic Fractures (SOF). The parent grant uses an efficient case-cohort study design to investigate the prospective association between U-Cd and incident fractures in 1,321 MrOS men and 1,578 SOF women who will be sex matched with 1500 randomly selected persons from each respective cohort. Cd, creatinine, osmolality, and cotinine are being analyzed in urine. The parent grant also evaluates the prospective association between U-Cd and rate of loss of total hip bone mineral density (BMD) in men and women from the subcohorts, and in incident fracture cases; and will utilize markers of bone metabolism and structure (PINP, CTX, and HR-pQCT) to provide novel insights into the cellular and structural mechanisms by which Cd may adversely affect bone. To this parent project, we are adding two Aims in this Supplement. Aim 1. Evaluate the prospective association between U-Cd and muscle mass in men and women from the subcohorts, and in fracture cases. This will be a follow-up to a recent cross-sectional study which reported an association between U-Cd and muscle loss. Aim 2. Explore whether mixtures of metals and nutrients are associated with fracture in the parent case- cohort study, and whether these mixtures are associated with bone loss, and muscle loss in the subcohorts. Shedding insights on risks from mixtures has been a desire of epidemiologists for decades; we will be using cutting edge statistical learning methods to analyze associations between mixtures and risk of incident fractures, bone loss, and muscle loss. These methods, including Bayesian kernel machine regression (BKMR) and weighted quantile sum (WQS) regression, have the advantages over standard regression models containing interaction terms of avoiding problems associated with multiple testing, interpretation of higher-order interaction terms, and limited statistical power. A large fraction of older US men and women have documented Cd exposure. Highly accurate measures of Cd exposure linked to powerful epidemiological cohorts provide a unique, cost effective approach to this important public health issue. These studies will transform the field of Cd-musculoskeletal research and have high potential to impact policy decisions in the US and globally.