Weight prediction from linear measures of growing Thoroughbreds

Equine Vet J. 2004 Mar;36(2):149-54. doi: 10.2746/0425164044868585.

Abstract

Reason for performing study: Monitoring weight of foals is a useful management practice to aid in maximising athletic potential while minimising risks associated with deviations from normal growth.

Objective: To develop predictive equations for weight, based on linear measurements of growing Thoroughbreds (TBs).

Methods: Morphometric equations predicting weight from measurements of the trunk and legs were developed from data of 153 foals. The accuracy, precision and bias of the best fitting equation were compared to published equations using a naive data set of 22 foals.

Results: Accuracy and precision were maximised with a broken line relating calculated volumes (V(t + l)) to measured weights. Use of the broken line is a 2 step process. V(t + l) is calculated from linear measures (m) of girth (G), carpus circumference (C), and length of body (B) and left forelimb (F). V(t + I) = ([G2 x B] + 4[C2 x F]) 4pi. If V(t + l) < 0.27 m3, weight is estimated: Weight (kg) = V(t + l) x 1093. If V(t + l) > or = 0.27 m3: Weight (kg) = V(t + l) x 984 + 24. The broken line was more accurate and precise than 3 published equations predicting the weight of young TBs.

Conclusions: Estimation of weight using morphometric equations requires attention to temporal changes in body shape and density; hence, a broken line is needed. Including calculated leg volume in the broken line model is another contributing factor to improvement in predictive capability.

Potential relevance: The broken line maximises its value to equine professionals through its accuracy, precision and convenience.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Animal Nutritional Physiological Phenomena
  • Animals
  • Biometry
  • Body Weight / physiology*
  • Female
  • Horses / anatomy & histology*
  • Horses / growth & development*
  • Male
  • Mathematics
  • Predictive Value of Tests
  • Reproducibility of Results
  • Sensitivity and Specificity