Martin, Kyle W., authorOlsen, Anastasia M., authorDuncan, Colleen G., authorDuerr, Felix M., authorBioMed Central, publisher2017-03-172017-03-172017-02-10Martin, Kyle W., Anastasia M. Olsen, Colleen G. Duncan, and Felix M. Duerr, The Method of Attachment Influences Accelerometer-based Activity Data in Dogs, BMC Veterinary Research 13:48 (2017): 6 pages. http://dx.doi.org/10.1186/s12917-017-0971-1http://hdl.handle.net/10217/180034Background: Accelerometer-based activity monitoring is a promising new tool in veterinary medicine used to objectively assess activity levels in dogs. To date, it is unknown how device orientation, attachment method, and attachment of a leash to the collar holding an accelerometer affect canine activity data. It was our goal to evaluate whether attachment methods of accelerometers affect activity counts. Eight healthy, client-owned dogs were fitted with two identical neck collars to which two identical activity monitors were attached using six different methods of attachment. These methods of attachment evaluated the use of a protective case, positioning of the activity monitor and the tightness of attachment of the accelerometer. Lastly, the effect of leash attachment to the collar was evaluated. For trials where the effect of leash attachment to the collar was not being studied, the leash was attached to a harness. Activity data obtained from separate monitors within a given experiment were compared using Pearson correlation coefficients and across all experiments using the Kruskal-Wallis Test. Results: There was excellent correlation and low variability between activity monitors on separate collars when the leash was attached to a harness, regardless of their relative positions. There was good correlation when activity monitors were placed on the same collar regardless of orientation. There were poor correlations between activity monitors in three experiments: when the leash was fastened to the collar that held an activity monitor, when one activity monitor was housed in the protective casing, and when one activity monitor was loosely zip-tied to the collar rather than threaded on using the provided metal loop. Follow-up, pair-wise comparisons identified the correlation associated with these three methods of attachment to be statistically different from the level of correlation when monitors were placed on separate collars. Conclusions: While accelerometer-based activity monitors are useful tools to objectively assess physical activity in dogs, care must be taken when choosing a method to attach the device. The attachment of the activity monitor to the collar should utilize a second, dedicated collar that is not used for leash attachment and the attachment method should remain consistent throughout a study period.born digitalarticlesengaccelerometeractivity monitorresearch toolsTextThis article is open access and distributed under the terms and conditions of the Creative Commons Attribution 4.0 International (CC BY 4.0).https://dx.doi.org/10.1186/s12917-017-0971-1