Mixed-Effects Modelling of the Risk Factors Associated with Multiple Pregnancies in Thoroughbred Mares

Abstract

Multiple pregnancies (MPs) are commonly diagnosed during breeding management of mares. Whilst some studies have reported on factors associated with the risk of MPs, few have utilised multivariable data analysis to control for confounding variables. A prospective cohort study of Thoroughbred broodmares was conducted with information collected on 27 factors. Mixed-effects logistic regression was used to determine risk factors for MPs. Mare, stallion, stud, and veterinarian were evaluated as random effects. The prevalence of MPs in 1754 mares and 2245 pregnancies was 16.06% (95% confidence interval [CI] = 14.54, 17.58). Multiple ovulations (OR = 15.57, 95% CI = 11.88, 20.53) and treatment with cloprostenol (OR = 1.35, 95% CI = 1.015, 1.80) were associated with increased odds of MPs following multivariable analysis. Mares that foaled at the start of the breeding season (OR = 0.66, 95% CI = 0.47, 0.94), conceived at the second or more oestrus cycles (OR = 0.60, 95% CI= 0.43, 0.84), or identified with a uterine cyst (OR = 0.63, 95% CI = 0.40, 0.97) were at reduced odds of conceiving MPs. Mare, stallion, stud, and veterinarian were not associated with MPs. These findings provide possible explanations as to why the prevalence of MPs but not MOs have increased over the last decade.