Fasting length and hay type effects on metabolic parameters in the horse

Abstract

Several metabolic disorders in the horse have been associated with post-prandial fluctuations of blood insulin and glucose concentrations in response to the ingredient and/or chemical composition of feedstuffs consumed. The objectives of this study were to 1) quantify changes in plasma glucose, insulin and cortisol concentrations in response to different fasting intervals, and 2) evaluate the effects of consumption of warm-season and cool-season hays on those same parameters. Six mature geldings were utilized in a 2 × 3 factorial design experiment in which they were fed either warm-season (WS) bermudagrass hay (Cynodon dactylon) or cool-season (CS) tall fescue hay (Lolium arundinaceum) during each fasting treatment. The three fasting treatments consisted of no fast (NF) in which the horses were offered hay ad libitum hay throughout the night, a short fast (SF) in which the animals were offered hay until 2200 h with hay re-introduced at 0600 h, and a long fast (LF) in which the animals were offered hay until 2200 h and not re-offered hay until 0700 h. All horses received concentrate and additional hay at 0700 h. Following a 6-d adaptation period, on d 7, a 7-h serial blood draw was conducted in conjunction with the morning feeding. Area under the curve (AUC) was calculated by the trapezoidal method for plasma glucose, insulin and cortisol. Glucose, insulin and cortisol AUC were evaluated using PROC GLM. No significant differences in glucose, insulin or cortisol AUC were detected due to hay type, fasting length, or the interactions between the two. Peak concentration and time to peak concentration of metabolites were also analyzed using PROC GLM, with no significant differences for glucose. There was a significant difference for time to peak insulin concentration between WS-LF and both the WS-NF (P = 0.024) and WS-SF (P = 0.003) groups. There were no differences among treatments in peak insulin concentration. There were no differences among treatments for cortisol peak values, but CS-SF and CS-LF differed (P = 0.04) for time to peak cortisol concentration. Individual differences were observed for cortisol and insulin for both AUC and peak concentrations. There was also a significant difference for time to peak and peak glucose concentrations between animals.