Background/Purpose: Pet dogs are a promising translational model for naturally occurring multiple joint osteoarthritis (MJOA). We characterized the impact of symptomatic (sx) MJOA, over the non-MJOA phenotype, on weekday physical activity (PA) patterns in pet dogs with OA pain using accelerometer-based data.

Methods: Full-body radiographs and orthopedic assessments were collected from pet dogs with evidence of OA (N&#3f99). MJOA was defined as radiographic OA (rxOA) in 3+ unique appendicular joint sites (bilateral joints=single site) with at least one painful rxOA joint. Missing radiographs were allowed if N of OA joint sites from available images was >=3. Burden of OA was defined only for dogs with complete radiographs as the total number of joints with both OA and pain. Dogs wore a collar-mounted activity monitor (Actigraph GT3X) for at least 7 days. Per minute average vector magnitudes were recorded on weekdays and used to create composite 24-h PA profiles. Function on scalar regression (FoSR) was used to examine the association of 24-h PA (outcome) with MJOA and with burden of OA (total number of joints with both OA and pain) adjusting for age and body condition score (BCS). Coefficients and 95% confidence intervals (CIs) are interpreted analogously to a simple linear regression but estimated per minute. To provide an intuitive interpretation, we created the “averaged” coefficients and 95% CIs by integrating the beta and CI curves over time.

Results: Of 99 dogs, 89 had sufficient data to define MJOA and were included (48% male, 52% female, mean age 9 years). Fifty-nine (66%) dogs had MJOA; distribution of joints is shown in Fig1. All dogs with rxMJOA were symptomatic (sxMJOA). Age, BCS, and MJOA had negative effects on PA patterns (averaged βs [95% CIs]: -0.40 [-0.44, -0.35], -0.52 [-0.67, -0.37], and -0.86 [-1.19, -0.54]). Fig2A shows 24-h average activity. Dogs without MJOA (red line) showed higher activity during peak hours. The CIs were below 0 from 3 am to 2 pm and from 6 pm to ~11 pm, suggesting that dogs with MJOA (blue line) are less active than those without in these time windows. The coefficient function and CI for MJOA and age were below 0 during most of the 24-h period, suggesting older dogs and those with MJOA are less active (Fig2B). A short period of PA around and after midnight in older dogs and those with MJOA (curves and CIs above the green line) suggests pain-related sleep interference. Similarly, dogs with higher OA-pain burden (n=82 dogs with complete radiographs) displayed less PA during daytime hours (Fig3). These associations were not seen when data were collapsed to average per-minute activity (averaged β [95% CIs]: -0.01 [-0.09, -0.07]).

Conclusion: Analysis of 24-h PA patterns clearly demonstrates the negative influence of MJOA on weekday PA, as well as the greater negative influence on PA patterns when more OA joints are painful. If similar effects are seen in humans with symptomatic MJOA, our results would further support the use of the naturally occurring pet dog model of OA pain, and of PA patterns as a clinically relevant outcome measure. The FoSR approach to evaluating PA patterns identified timeframes with lower PA in dogs with MJOA which were not evident if sum scores of PA were used.

Radiographic OA and painful joints and joint sites in client canine population.

Comparison of mean activity distributions between dogs with and without multiple joint osteoarthritis (MJOA) over time. The green lines correspond to beta=0, which indicates that there is no difference between groups at the corresponding time point.

Estimated mean activity in dogs with different N of joints with rxOA and pain ( < 4 and >=4) and estimated coefficient function for OA burden derived from the function on scalar regression model of accelerometer data on continuous N of joints adjusted forage and BCS. The green line corresponds to beta=0, which indicates that there is no effect of OA burden on PA at the corresponding time point.

« Back to ACR Convergence 2025

ACR Meeting Abstracts - https://acrabstracts.org/abstract/time-series-analysis-of-continuous-accelerometry-data-to-determine-the-impact-of-symptomatic-multiple-joint-osteoarthritis-mjoa-in-the-naturally-occurring-canine-model/