Low starch diet alters microbiome composition and decreases gastric disease and ridden horse pain ethogram scores in performance horses

Background: Gastric disease is common in performance horses, and can lead to behavioural, performance and welfare issues (Sykes et al. 2015). It is often attributed to high-starch and sugar feeds and reduced forage, however evidence for diet-induced changes on the faecal microbiome, and its impact on the behaviours of the ridden horse, is currently lacking.

Objectives: To determine the impact of diet changes on gastric disease scores, behaviours and pain at exercise and faecal microbiome composition in elite show jumping horses.

Study design: Prospective observational cohort study.

Methods: Nine showjumpers from the same stable, in training and regularly competing, had concentrate pelleted diets changed from a high sugar and starch (>30%) feed to a cooked, muesli-type low-starch (11%) feed. Feeding frequency was increased from three to four times a day. Prior to and 12 weeks post-diet change, horses underwent a standardized exercise test (SET), that was video-recorded, and blood was taken. Gastroscopies and faecal samples were performed the day after. Equine gastric disease (EGD) was blindly graded using proprietary scoring systems - EGD scores were calculated by adding lesion scores for equine squamous gastric disease (ESGD) and equine glandular gastric disease (EGGD), and microbiomes profiled using 16S rRNA gene amplicon sequencing. A ridden pain score (Ridden Horse pain Ethogram, RHpE) (Dyson 2022) was calculated by two blinded observers by watching the videos. Results were analysed with Wilcoxon and Spearman tests. Microbiome community structure changes were determined using PERMANOVA. Firmicute to Bacteroidota (F/B) ratio, a common metric of dysbiosis (Park et al. 2021), was compared to gastric scores and blood markers using linear mixed models.

Results: After 12 weeks of a low-starch diet, EGD (t1,8 = −6.17, p < 0.001; T0: 4[IQR:1.5], T12: 1[IQR:1]), squamous (t1,8 = −5.32, p < 0.001; T0: 1[IQR:2], T12: 0[IQR:0]) and glandular (t1,8 = −2.53, p = 0.04; T0: 2.5[IQR:1], T12: 0[IQR:1]) disease, and RHpE scores (6.9 ± 2.9 at T0 vs 2.9 ± 2.0 at T12, p < 0.001) improved significantly. Squamous, glandular and EGD scores were positively correlated with RHpE scores (respectively r = 0.747, p < 0.001, r = 0.742, p < 0.001 and r = 0.867, p < 0.001). Diet change did not impact faecal microbiome community structure (PERMANOVA: F1,16 = 1.37, p = 0.15, r2 = 0.08), however faecal F/B ratios declined (t1,8 = −3.13, p = 0.01; T0: 2.07 ± 0.21 vs. T12: 1.29 ± 0.14). Lower EGD scores were associated with reduced F/B ratios (ChiSq1,8 = 3.83, p = 0.05), GGT (ChiSq1,7 = 28.82, p < 0.0001) and CK (ChiSq1,8 = 4.29, p = 0.04), and elevated vitamin E (ChiSq1,6 = 8.02, p < 0.01).

Conclusion: There is a positive correlation between gastric disease and pain scores in ridden horses. Transition to a low-carbohydrate diet did not change faecal microbiome composition, but successfully reduced F/B ratio and improved gastric and pain ethogram scores in elite showjumpers during a competitive season without medication. It is possible to mitigate gastric ulcers and to improve the comfort of working equine athletes through dietary adjustments.