Commercially available wearable health monitor in dogs is unreliable for tracking energy intake and expenditure

Abstract

Abstract OBJECTIVE Assess the accuracy of predicted daily energy requirement (pDER) reported by a triaxial accelerometer and activity monitor for dogs (FitBark 2; FitBark Inc) and determine whether the activity monitor accurately estimates the observed daily energy requirement (oDER). We hypothesized that the activity monitor would accurately estimate oDER in dogs and meet standards established for human devices. ANIMALS 23 dogs between the ages of 1 and 10 years and variable sex, breed, and body weight were enrolled from May 5, 2021, through July 23, 2021. METHODS Dogs were weighed before and after the study period to ensure stable body weights. Owners recorded their dogs’ daily caloric intake for the entire 28-day study period while the device monitored physical activity and calculated pDER. oDER was defined as the reported caloric intake required to maintain a stable body weight over a 28-day period. pDER and oDER were compared using Bland-Altman graphs, Passing-Bablock analysis, and Lin’s Concordance correlation analysis. P ≤ .05 was considered significant. RESULTS 23 apparently healthy dogs completed the study. There was no significant difference between starting body weights and ending body weights (P= .5). The activity monitor overpredicted 28-day pDER compared to 28-day oDER in the majority (18/23, 78.3%) of dogs. Based on Bland-Altman analysis, Passing-Bablok regression, and Lin’s concordance correlation analysis, there was poor agreement between the pDER and oDER. CLINICAL RELEVANCE The activity monitors consistently reported inaccurate pDER compared to oDER. Its usability for estimating pDER is of limited clinical and research utility based on the results of this study.