• horse;
  • colic;
  • epidemiology;
  • prospective;
  • risk factors;
  • logistic regression


A1 year prospective study was conducted on 31 horse farms to identify risk factors for equine colic. Farms were randomly selected from a list from 2 adjacent counties of Virginia and Maryland, USA. The association between colic and farm or individual horse risk factors related to management, housing, pasture, use, nutrition, health and events was first examined by univariate statistical analysis. Individually significant (P<=0.25 for farm factors, P<=0.10 for horse factors) variables were used in a stepwise multivariable forward logistic regression to select explanatory factors (P<=0.05). Analysis was conducted at 2 levels: farm and individual horse with farm specified as a random effects variable. No farm-level variables were significant Significant horse-level variables included: age, odds ratio (OR)=2.8 for horses age 2–10 years compared to <2 years; history of previous colic, OR=3.6 relative to no colic; changes in concentrate feeding during the year (1 per year, OR=3.6, more than 1, OR=2.2) relative to no changes; more than 1 change in hay feeding during the year, OR=2.1 relative to no changes; feeding high levels of concentrate (>2.5 kg/day dry matter, OR=4.8, >5 kg/day dry matter, OR=6.3) relative to feeding no concentrate; and vaccination with monocytic ehrlichiosis vaccine during the study, OR=2.0 relative to no vaccination. Feeding a whole grain with or without other concentrate components reduced risk, OR=0.4, relative to feeding no whole grain. Results of the study suggest that diet and changes in diet are important risks for colic in a population of horses on farms.