Serum adipokines are useful for predicting hyperinsulinemia secondary to corticosteroid treatment in horses

Objective: To evaluate pretreatment biomarkers for predicting the severity of corticosteroid-induced hyperinsulinemia. Methods: Oral glucose tests were performed before (day 1) and after (day 8) a dexamethasone course (0.08 mg/kg, q 48 h) in 24 healthy adult Standardbreds. Associations between day 8 oral glucose test insulin parameters (baseline [T0insulin], maximal [Cmaxinsulin], and area under the curve [AUCinsulin]) and day 1 serum markers (insulin, glucose, adiponectin, leptin, and their ratios) were examined using Spearman correlation, log-linked generalized linear models. Relative risk of exceeding specific insulin thresholds was evaluated for horses with pretreatment markers above and below reference ranges. Results: Day 8 insulin increased by (mean) 11.5-fold (range, 2 to 37; T0insulin), 6.5-fold (range, 2 to 17; Cmaxinsulin), and 6.6-fold (range, 2 to 15; AUCinsulin) over day 1 values (P < .001). Day 1 adiponectin had the strongest correlation with day 8 Cmaxinsulin (ρ = -0.67; P < .001) and AUCinsulin (ρ = -0.68; P < .001). Day 1 leptin-to-adiponectin ratio had the strongest correlation with day 8 resting insulin (ρ = 0.68; P < .001) and was also the best predictor of day 8 T0insulin, Cmaxinsulin, and AUCinsulin on generalized linear model analysis (P < .0001). Horses with a leptin-to-adiponectin ratio > 1 were 4.9 times (range, 1.5 to 18.2) more likely to have day 8 T0insulin > 100 µIU/mL (P = .01). Conclusions: Pretreatment leptin, adiponectin, and leptin-to-adiponectin ratio are predictors of corticosteroid-associated hyperinsulinemia. Clinical Relevance: Adipokines warrant clinical evaluation as biomarkers of risk for corticosteroid-associated hyperinsulinemia and laminitis.