Interday variation in vertical ground reaction force in clinically normal Greyhounds at the trot

To determine the effect of interday variation on vertical ground reaction force variables in dogs.

52 clinically normal Greyhounds of either sex weighing between 22 and 35 kg.

Dogs were led at a trot across a floor-mounted force platform to determine vertical ground reaction force variables (peak [PFz] and impulse [IFz]) from hind limbs. Data were collected from each dog on 3 consecutive days. Variance components were estimated, using maximal likelihood to evaluate contributions of interday variation within dogs and variation attributable to dogs and repetitions. An ANOVA was used to test significance of interday variation within dogs and day within dog interactions.

PFz, IFz, or both differed significantly from day to day for 29 of 52 dogs. Only PFz differed significantly among days for 16 dogs, and only IFz differed among days for 5 dogs. The PFz and IFz differed significantly from day to day in 8 dogs. Using ANOVA, the difference for PFz and IFz among days within dogs was significant.

Effect of day within dog variation (interaction) should be considered as a component in statistical models in which data from 1 day are evaluated against data from the same subject on another day. We propose a statistical model that incorporates an accommodation for interday variation. Investigators should determine the factors that affect their studies, including the extent of interday variation, and compensate for the variation attributable to each factor in the statistical models used to analyze their data.