Evaluation of a digitally integrated accelerometer-based activity monitor for the measurement of activity in cats

To assess the correlation between data generated by an accelerometer-based activity monitor and the distance moved in cats. Prospective experimental study. Three, four-year-old, male, purpose-bred research cats, weighing between 5.1 and 5.9 kg. Part I: Collar and harness mounted accelerometers were evaluated in three cats, comparing simultaneously collected accelerometer data with movement data from computer-analyzed video. Part II: Cats wore collar and harness mounted accelerometers, and data were recorded for 4 weeks to evaluate day-to-day and week-to-week variation in activity. Part I: 432 hours of simultaneous video and accelerometer data were collected. The correlation between accelerometer counts and distance moved was 0.82 overall. Agreement between collar and harness mounted accelerometers was excellent with only 6% of the differences in measurements lying outside the mean difference ± 2 standard deviations. The adjusted R2 for harness accelerometer output and 6% mobility was 0.75; for movement 0.84; and for mean velocity 0.83. Evaluation of video indicated eating, grooming and scratching created high accelerometer counts with little effect on movement. Part II: There was a significant effect of day on harness (p < 0.001) and collar (p < 0.002) counts, with counts being lowest at the weekend. There was a significant effect of week on harness-mounted accelerometer counts (p < 0.034), but not on collar-mounted accelerometer counts. Harness accelerometer counts were lowest in week 1. Output from an acceleration-based digitally integrated accelerometer correlated well with distance moved and mobility in freely moving cats provided the mobility threshold in the analysis software was ≥6%. Acceleration-based activity monitors may allow for objective measurement of improved mobility following analgesic treatment for conditions such as osteoarthritis.