AS740 - Fracture risk assessment among older adults with diabetes

Investigator Names and Contact Information

Introduction/Intent

Fracture risk among older adults with type 2 diabetes mellitus (DM) is up to 50% higher for non-vertebral fractures and up to 2-times higher for hip fracture, compared to those without DM.^1-7 Fractures in those with DM are associated with greater post-operative complications, significant morbidity (e.g., heart and renal failure), greater physical disabilities and chronic pain, decreased quality of life, and increased mortality.^8-19 Given these significant adverse outcomes, it is paramount to identify individuals at high-risk for fracture in this population. In current clinical practice, identification of persons with increased fracture risk is primarily based on bone mineral density (BMD) measured by dual x-ray absorptiometry (DXA) at the hip and lumbar spine and the calculated risk by the Fracture Assessment tool (FRAX), which provides the estimated 10-year risk of both major osteoporotic fractures and hip fractures, determined by bone-related clinical risk factors, such as age, sex, use of corticosteroids, history of previous fractures, and femoral neck BMD.^20,21 However, the fracture risk associated with DM paradoxically occurs despite a higher BMD by DXA, and the FRAX tool significantly underestimates frac-ture risk among older adults with DM.^2,22 Additional assessments of fracture risk among older adults with DM are needed to identify at-risk persons more accurately and thereby precisely target fracture prevention interven-tions to reduce morbidity and mortality in this increasing population. Therefore, to address this gap, this application will:

o Develop measures of fracture risk specific to older adults with DM that can be readily translated to clinical practice and research for risk stratification and intervention targeting;

o Add DM-related and novel factors to the current fracture assessment tool (i.e., FRAX) to better predict fracture risk among older adults with DM;

o Build on preliminary data which have identified several DM-related and novel factors associated with fracture risk in this population;

o Include metabolites identified using the metabolomics approach, which has previously identified novel risk factors in metabolic disorders

o Utilize existing data and biospecimens from 3 well-established cohorts to increase rigor and reproducibility

Because metabolites fall downstream of genetic, proteomic, and environmental risk factors, comprehensive metabolic profiling, or “metabolomics”, provides integrated measure of phenotype for complex metabolic diseases such as DM and osteoporosis. DM-related fracture has specific pathophysiology distinct in metabolic bone disease. The underlying mechanisms for this increased fracture risk have not been fully identified, but given DM’s multiple systemic effects, there are likely multiple contributing factors. Metabolomics allows for the discovery of biomarkers that predict disease incidence, severity, and progression, and sheds light on underlying mechanistic abnormalities. For example, in work by co-Investigator Dr. Kiel, metabolomics identified several lipid metabolites that were negatively associated with BMD, as well as amino acid metabolites that were associated with an increased risk of osteoporosis, including threonine and serine²³. Similarly, collaborator Dr. Bain identified the association between the levels of the branched chain amino acids and DM-related factors such as insulin resistance, obesity, and dysglycemia, through the use of metabolomics technology.^24-28

Likewise, our preliminary evidence, collected during Principal Investigator Dr. Lee’s successful K23, identified metabolites utilizing targeted metabolomics that were associated with incident fracture among older adults with DM. In targeted metabolomics, isotope-labeled reference standards are used and, unlike non-targeted metabo-lomics assays, can thereby quantify the absolute concentrations of the metabolites in the biospecimen. Additionally, we analyzed the metabolomics data using principal component analysis, a data-driven statistical procedure that defines independent factors or “components”. In this study, the metabolomic component consisting of serine, arginine, asparagine/aspartate and glutamine/ glutamate, similar to those amino acid metabolites identified by Dr. Kiel, was significantly associated with incident fracture. Therefore, with the expertise of our team, we will evaluate in the current application, these metabolic profiles using targeted metabolomics within 3 existing, well-characterized cohorts and integrate fracture risk assessment in the setting of other bone-related factors.