Perfluoroalkyl substances and incident type 2 diabetes in a US population: A metabolome-genome investigation

The prevalence of type 2 diabetes (T2D) is increasing, and etiological factors promoting the T2D epidemic are not fully understood. Growing experimental evidence shows that exposures to endocrine-disrupting chemicals (EDCs), such as per- and polyfluoroalkyl substances (PFAS), promote T2D development, likely in synergy with known risk factors such as genetic variations. PFAS are ubiquitous and persistent chemicals that perturb metabolism. However, few prospective studies examined the association between PFAS and T2D risk, and those were in small and not fully representative samples. Previous studies also lacked clinically ascertained T2D diagnosis, investigated only a few of the many potentially hazardous PFAS, and did not examine potential effects of PFAS mixtures or gene–PFAS interactions. State-of-the-art integrated omics approaches can overcome these limitations to advance the field. We propose the first integrated metabolome–genome approach to (1) characterize the associations between PFAS concentrations (individual PFAS and mixtures) in prediagnostic plasma samples and incident T2D risk and potential effect modification by genetic predisposition to T2D using polygenic risk scores as an innovative solution for studying interactions, (2) identify underlying dysregulated metabolic pathways, and (3) identify metabolic signatures in prediagnostic plasma samples defined by EDC exposures and endogenous metabolites associated with T2D risk. We will perform a nested case–control study leveraging BioMe, an ongoing electronic health record-linked biobank with >55,000 participants enrolled while seeking primary care at Mount Sinai Hospital (NY) since 2007. Incident T2D cases are matched (1:1) to BioMe T2D-free controls (N = 1,700), with ~6 years average time between blood draw and T2D diagnosis. We will use prediagnostic plasma to measure PFAS and metabolic pathways using state-of-the-art high-resolution metabolomics (HRM) approaches. We will replicate findings among incident T2D cases and matched controls from an existing population-based study in Los Angeles and Hawaii with extant genome data and prediagnostic plasma concentrations of PFAS and HRM measured at the same lab as BioMe samples. In contrast to prior studies, we incorporate a wide suite of legacy and emerging PFAS, exposure-mixture effects, and gene–environment interactions by leveraging state-of-the-art metabolome–genome approaches and a rigorous discovery–replication design in two unique, well-phenotyped US cohorts with prediagnostic plasma samples to identify early biomarkers associated with T2D. This research relies on a multidisciplinary team of seasoned investigators with expertise in environmental/genetic epidemiology, PFAS and T2D research, and state-of-theart HRM, genomics, and biostatistical exposure–mixture methods. Findings will inform precision medicine approaches for T2D prevention and treatment in US populations.