Accelerometer activity tracking in horses and the effect of pasture management on time budget.

Abstract

BACKGROUND: Accelerometry is an accepted means of quantifying human physical activity. Quantitative physical activity tracking could be beneficial for studies into equine health and disease prevention, for example in relation to obesity management. OBJECTIVES: Validate accelerometer use in grazing horses, determine between-day repeatability, and assess the effects of pasture size on time budget (i.e. duration in each activity category). STUDY DESIGN: Proof of concept. METHODS: Accelerometers (ActiGraph) were positioned at the poll. Horses underwent 5&#xa0;min of observed activity in three categories: standing, grazing and ambulating. Receiver-operating characteristic curve analysis, used on ten second data epochs, calculated cut points between the activities. A 20-day study was then undertaken on 6 horses at pasture. Time in each category (per day) was deduced; a Mann Whitney U test was performed to compare standard vs. small paddock and day vs. night turn out. RESULTS: Cut-off values with the optimum sensitivity (94.7-97.7%) and specificity (94.7-96.8%) were found to be <127.6 counts for standing, 127.6-702.7 counts for grazing and >702.7 counts for ambulating. Repeatability was analysed descriptively: Median (IQR) of the between-day difference in minutes standing, grazing and ambulating were 46.9 (21.3-87.9), 77.3 (40.2-124.5) and 15.6 (6.8-40.2) respectively. Median times standing and ambulating were significantly different between standard and small paddocks: standing: 8.7 vs. 10.3&#xa0;h (P<0.001); ambulating: 55.7 vs. 39.6&#xa0;min (P&#xa0;=&#xa0;0.002). There was no significant difference in the median time spent grazing. There were significant differences between day and night: standing: 32.95% vs. 50.97% (P&#xa0;=&#xa0;0.001), grazing: 60.81% vs. 46.77% (P<0.001) and ambulating: 4.57% vs. 2.40% (P<0.001). MAIN LIMITATIONS: Small sample size and lack of cross-validation of cut-off points on independent, 'unseen' data. CONCLUSIONS: Accelerometry can differentiate standing, grazing and ambulating in horses. Our proof-of-concept study demonstrates modifying pasture size influences activity budgets; opening avenues into studying obesity management.