Amanda M. de Mestre

Determining whether endometrial microbial isolates are pathogens, contaminants, or even part of the “normal” microbiome is extremely complex, particularly given the absence of “gold standard” tests for endometritis. Population-level benchmarking and temporal monitoring can provide novel insights and a wider context to improve understanding. This study aimed to (i) estimate the prevalence of endometrial isolates from swabs of Thoroughbred broodmares in Newmarket, UK between 2014 and 2020; and (ii) evaluate the effects of year, mare age, and cytology findings on isolate prevalence. Generalised linear mixed models with a logit link, both null models and models using year of sampling, mare age, or cytology findings as predictors, were fitted to estimate isolate prevalence. Over the 7-year period, data were available from 18,996 endometrial-swab samples from 6050 mares on 290 premises. The overall isolate prevalence was 35.5% (95% confidence interval (CI) 33.0–37.9), and this varied significantly between years. The most prevalent isolates were β-hemolytic Streptococcus (17.9; 95% CI: 17–19) and E. coli (10.3%; 95% CI: 9.0–11.6). Isolate prevalence increased with mare age except for E. coli isolates, and with increasing category of cytology findings except for α-hemolytic Streptococcus isolates. The results provide novel estimates of isolate prevalence and highlight knowledge gaps around potential complexities in the interpretation of findings.

Background Up‐to‐date estimates of early‐life morbidity and mortality in Thoroughbreds are lacking. Methods A birth cohort was established on Thoroughbred stud farms across the UK and Ireland. All veterinary interventions for disease or injury between birth and 18 months of age or leaving the study were recorded. Multilevel Poisson regression models with farm and foal as random effects were fitted to estimate incidence rates. Results Data were available for 3328 foal‐months at risk for 275 foals on seven farms. The overall rates of disease and injury requiring veterinary intervention and mortality were 11.9 cases/100 foal‐months at risk (95% confidence interval [CI] 8.6–16.2) and 0.2 cases/100 foal‐months at risk (95% CI 0.1–0.4), respectively. Almost half ( n = 133/273, 49%, 95% CI 43–55) of the live‐born cohort required veterinary intervention for musculoskeletal disease or injury, equating to 5.8 cases/100 foal‐months at risk (95% CI 4.1–8.2), predominantly reported as developmental orthopaedic disease (DOD). Limitations Convenience sampling of participants may affect the generalisability of the findings. Conclusions Rates of musculoskeletal disease and injury, in particular DOD, on Thoroughbred stud farms were high. Further work to identify modifiable risk factors and further understanding of the economic impact of these conditions and long‐term consequences for musculoskeletal health and performance is required.