Circulating bile acids and HOMA-IR: cross-sectional results from the RoCAV population-based study

Sarah Grossi¹Emanuele Maria Giusti¹Giovanni Veronesi¹Simone Delaiti²Andrea Mancini²Laura Migliaccio¹Selene Genova^1,3Simona Costanzo^1,4Kieran Michael Tuohy⁵Marco Mario Ferrario¹Francesco Gianfagna^1,6*

• ¹Universita degli Studi dell'Insubria Dipartimento di Medicina e Chirurgia, Varese, Italy
• ²Fondazione Edmund Mach Centro Ricerca e Innovazione, San Michele all'Adige, Italy
• ³ASL NO, Novara, Italy
• ⁴NEUROMED Dipartimento di Epidemiologia e Prevenzione, Pozzilli, Italy
• ⁵University of Leeds, Leeds, United Kingdom
• ⁶Aziende Socio Sanitarie Territoriale dei Sette Laghi, Varese, Italy

Background: Circulating bile acids (cBAs) function as signaling molecules that activate the farnesoid X receptor (FXR), promoting the secretion of fibroblast growth factor 19 (FGF-19), a gut-derived hormone involved in bile acid (BA) synthesis, glucose metabolism, and insulin sensitivity. However, the relationship between fasting cBAs, FGF-19, and insulin resistance—as estimated by HOMA-IR—remains unclear. This study explored these associations in an elderly population from Northern Italy. Material and methods: We examined a subsample of 1080 subjects (aged 60–75 years, 1:1 male-to-female ratio) from the RoCAV population-based study (2013–2016). Fasting blood samples were analyzed for 33 cBAs using UHPLC-MS/MS, of which 23 met quality criteria. FGF-19 levels were also measured. After excluding individuals with missing data or fibrate therapy, 1049 participants were included. Associations between cBAs, FGF-19, and HOMA-IR were assessed via linear regression, adjusting for age, sex, BMI, diet, alcohol intake, hypertension, and dyslipidemia. ROC curve analysis evaluated the ability of cBAs and FGF-19 to discriminate T2DM cases. Results: After excluding participants with missing anthropometric and clinical data or in fibrate treatment, data from 1049 subjects (mean±SD age 68.6±4.5 years, males 49.5%, T2DM 10.6%) were analysed. FGF-19 showed a positive correlation with primary cBAs (Spearman's ρ=0.33, p<0.001). Adjusted for covariates, both primary and secondary cBAs were positively associated with HOMA-IR (β=0.07, p=510-5; β=0.9, p=410-7) while FGF-19 was not (β=-0.02, p=0.31). In the mutually-adjusted model - including primary and secondary cBAs, FGF-19, and covariates - the β coefficients for cBAs were attenuated but remained significant (primary: β=0.06, p=0.005; secondary: β=0.07, p=0.0003), and FGF-19 retained an inverse association (β=–0.05, p=0.009). When total cBAs were used in the FGF-19–adjusted model, the association with HOMA-IR was the strongest (β=0.19, p=5×10⁻¹⁹). ROC curve analysis indicated that the inclusion of primary and secondary cBAs and FGF-19 improved model discrimination for T2DM (ΔAUC=0.03, 95% Confidence Interval: 0.01-0.06; Net Reclassification Improvement=0.54; 95%CI: 0.30-0.75). Conclusions: In this elderly Italian population, primary and secondary cBAs were positively associated with insulin resistance, after adjusting for each other, whereas FGF-19 negatively. These markers may enhance T2DM risk stratification and may give insights on bile acid–glucose metabolism links.