Leash attachment and device position on the collar does not interfere with monitoring dog activity using FitBark 2...

To evaluate the significance of leash attachment and device positioning along the collar of a commercial triaxial accelerometer (FitBark 2) on physical activity measurements during leashed walks in dogs.

10 medium- to large-breed healthy dogs (1 to 11 years old) were enrolled for a 3- to 14-day period. Dogs were fitted with 2 collars with 2 devices on each (dorsolateral leash [LL], dorsolateral no leash [L-NL], ventral leash [VL], ventral no leash [V-NL]) that were worn during 6 to 11 walks where the leash was only attached to 1 designated collar. Device outputs (ie, total and average BarkPoints) were analyzed using Pearson correlation and mixed models.

There was a very high correlation between all devices. Using mixed models, activity outputs did not differ between devices on leashed collars compared to their nonleashed counterparts (LL vs L-NL, VL vs V-NL) or between devices on the nonleashed collar (L-NL vs V-NL). Total and average BarkPoints differed significantly between devices on the leashed collar (LL vs VL). The total BarkPoints difference between LL and VL remained unchanged with increasing walk duration, and thus the percentage difference decreased over time.

Triaxial accelerometry can be used to monitor the activity of dogs during routine walks without significant interference of the leash attachment or the device position along the collar.

Leash attachment and device orientation are important variables to validate accurate long-term physical activity measurements.