Factors associated with insulin responses to oral sugars in a mixed‐breed cohort of ponies

Background Serum insulin concentration at 60 min (InsulinT60) during an oral sugar test (OST) indicates future laminitis risk and insulin dysregulation (ID). Associations between InsulinT60 and physical and owner‐reported variables may help clinicians select individuals for ID testing. Associations between InsulinT60 and other metabolic markers may help elucidate ID pathophysiology. Objectives To describe associations between (A) season, physically‐apparent and owner‐reported factors and binary InsulinT60 interpretation (initial models) and (B) variables included in the initial models, other metabolic markers and continuous InsulinT60 (full models). Study design Prospective longitudinal. Methods Non‐laminitic ponies were examined and OSTs (0.3 mL/kg Karo syrup) performed every 6 months (autumn and spring) for ≤4 years. Factors associated with InsulinT60 were determined using mixed‐effects models with physical, owner‐reported, season and serum/plasma markers as fixed effects and pony and premises identifiers as random effects. Autumn and spring data were analysed separately for full models. Results One thousand seven hundred and sixty‐three OSTs from 367 ponies were included. High‐risk InsulinT60 (>153 μIU/mL) was independently associated with (odds ratio, 95% confidence interval [CI]): age (1.36, 1.16–1.60), body condition score (BCS) (2.38, 1.21–4.69), and bulging supraorbital fatpads (7.25, 2.1–24.98). However, the initial models provided little explanatory power (Nakagawa R2 = 0.1–0.27). LoginsulinT60 was independently associated with (effect estimate, 95% CI): age (0.02, 0.01–0.04), Welsh/Welsh X breed (0.22, 0.05, 0.39), sex (gelding = −0.2, −0.34 to 0.06), BCS (0.16, 0.08–0.23), plasma adiponectin (−0.02, −0.02 to 0.01) and basal insulin (0.01, 0.01–0.01) in spring, and: age (0.03, 0.02–0.04), BCS (0.17, 0.08–0.26), bulging supraorbital fatpads (0.37, 0.2–0.54), turnout score (0.05, 0.02–0.09), plasma adiponectin (−0.01, −0.02 to 0.01), ACTH (per 10 pg/mL) (0.01, 0.00–0.01), triglycerides (0.28, 0.07–0.49) and InsulinT0 (0.01, 0.01–0.01) in autumn. Main limitations Only non‐laminitic ponies in one region were included. Conclusions Owner‐reported and physically‐apparent data were limited InsulinT60 predictors. InsulinT60 is a complex trait, independently associated with numerous variables, some with seasonal interactions.