Skip to main content
NCBI home page
Journal of Animal Science logoLink to Journal of Animal Science
. 2021 Jul 3;99(8):skab204. doi: 10.1093/jas/skab204

Physical activity patterns of free living dogs diagnosed with osteoarthritis

Anne H Lee 1, Katelyn B Detweiler 1, Tisha A Harper 2, Kim E Knap 2, Maria R C de Godoy 1,3, Kelly S Swanson 1,2,3,✉
PMCID: PMC8355611  PMID: 34216471

Abstract

Osteoarthritis (OA) affects about 90% of dogs > 5 yr of age in the United States, resulting in reduced range of motion, difficulty climbing and jumping, reduced physical activity, and lower quality of life. Our objective was to use activity monitors to measure physical activity and identify how activity counts correlate with age, body weight (BW), body condition score (BCS), serum inflammatory markers, veterinarian pain assessment, and owner perception of pain in free-living dogs with OA. The University of Illinois Institutional Animal Care and Use Committee approved the study and owner consent was received prior to experimentation. Fifty-six client-owned dogs (mean age = 7.8 yr; mean BCS = 6.1) with clinical signs and veterinary diagnosis of OA wore HeyRex activity collars continuously over a 49-d period. Blood samples were collected on day 0 and 49, and dog owners completed canine brief pain inventory (CBPI) and Liverpool osteoarthritis in dogs (LOAD) surveys on day 0, 21, 35, and 49. All data were analyzed using SAS 9.3 using repeated measures and R Studio 1.0.136 was used to generate Pearson correlation coefficients between data outcomes. Average activity throughout the study demonstrated greater activity levels on weekends. It also showed that 24-h activity spiked twice daily, once in the morning and another in the afternoon. Serum C-reactive protein concentration was lower (P < 0.01) at day 49 compared to day 0. Survey data indicated lower (P < 0.05) overall pain intensity and severity score on day 21, 35 and 49 compared to day 0. BW was correlated with average activity counts (P = 0.02; r = −0.12) and run activity (P = 0.10; r = −0.24). Weekend average activity counts were correlated with owner pain intensity scores (P = 0.0813; r = −0.2311), but weekday average activity count was not. Age was not correlated with total activity count, sleep activity, or run activity, but it was correlated with scratch (P = 0.03; r = −0.10), alert (P = 0.03; r = −0.13), and walk (P = 0.09; r = −0.23) activities. Total activity counts and activity type (sleep, scratch, alert, walk, and run) were not correlated with pain scored by veterinarians, pain intensity or severity scored by owners, or baseline BCS. Even though the lack of controls and/or information on the individual living conditions of dogs resulted in a high level of variability in this study, our data suggest that the use of activity monitors have the potential to aid in the management of OA and other conditions affecting activity (e.g., allergy; anxiety).

Keywords: canine exercise, joint health, pet health

Introduction

Osteoarthritis (OA) is a progressive, chronic condition associated with joint dysfunction, pain, and stiffness due to joint deterioration and it affects about 20% of all adult dogs and > 90% of dogs > 5 yr of age in North America (Johnston, 1997; Servet et al., 2006; Roush et al., 2010). OA is a debilitating disease that leads to exercise intolerance and an unwillingness or inability to climb or jump due to pain, lowering the overall quality of life for the animal (Lawler et al., 2005; Gupta et al., 2012; Comblain et al., 2016; Vassalotti et al., 2017; Musco et al., 2019; Johnson et al., 2020). This condition requires a lifelong management from its onset, and the goal of treatment is to delay the progression of the disease as well as proper pain management.

Proper pain assessment is needed to provide adequate treatment; however, pain assessment is challenging as animals are unable to communicate their intensity of pain. Traditionally, this assessment was done primarily by veterinarians, based on the knowledge of an animal’s “normal” and “abnormal” behaviors. This kind of assessment, however, introduces bias in a clinical setting because an unfamiliar environment can introduce anxiety and stress in the animal (Wiseman et al., 2001; Pollmeier et al., 2006). Instruments such as force plate platforms may be used for gait analysis, but access to these instruments is limited and costly to perform as part of a routine examination. In recent years, owner questionnaires such as the Liverpool osteoarthritis in dogs (LOAD) and canine brief pain inventory (CBPI) were introduced and validated for pain assessment, and evaluation of response to treatments (Brown et al., 2008; Hercock et al., 2009). However, these data are also subjective, with wide variation from owner to owner. Owner survey data often show a poor correlation with peak vertical force analysis, as owners are more focused on the behavior of their animal rather than lameness (Brown et al., 2013).

In recent years, a variety of activity monitors such as accelerometers have become commercially available and tested to measure physical activity objectively (Belda et al., 2018; Mejia et al., 2019). Accelerometers have been used to objectively evaluate the response to therapy in companion animals by monitoring their activity in a research laboratory or home setting (Lascelles et al., 2007; Lascelles et al., 2008; Brown et al., 2010; Edwards and Gibson, 2012; Preston et al., 2012). These monitors are noninvasive, portable devices that can provide information regarding activity intensity, frequency, and duration, with the possibility of categorizing data into activity type. However, there is limited research establishing a correlation between pain evaluation and activity data in OA dogs to determine whether variations in activity can translate into clinically relevant information regarding the pain scale of the animals. Thus, the objective of this study was to characterize the general activity pattern of free-living dogs diagnosed with OA. In addition, we aimed to assess how the tracked activity correlated with age, body weight (BW), body condition score (BCS) (LaFlamme, 1997), circulating markers of inflammation (C-reactive protein; matrix metalloproteinase-3), veterinarian pain assessment, and owner perception of pain. We hypothesized that activity data would be negatively correlated with owner perception of pain, veterinarian assessment of pain, age, BW, BCS, and serum inflammatory markers.

Materials and Methods

Animals

A total of 72 adult dogs with OA were recruited through the University of Illinois Veterinary Teaching Hospital. Animals were selected based on their clinical history, clinical signs, radiographic evidence of OA, and diagnosis by a board-certified small animal orthopedic surgeon. Dogs that could not remain off pain relief medication such as corticosteroids and nonsteroidal anti-inflammatory drugs were excluded from the study. Dogs that were under 18.2 kg BW; had evidence of neurological disease affecting gait, neoplasia, or acute instability of the joint; and female dogs that were pregnant or planned on being bred during the study were also excluded. This study was approved by the Institutional Animal Care and Use Committee at the University of Illinois at Urbana-Champaign (Protocol #15253) and written owner consent was obtained prior to enrollment. The total duration of the study was 49 d.

Activity data

Voluntary physical activity levels were measured using HeyRex monitors (HeyRex Ltd., Wellington, New Zealand). Each dog enrolled into the study was assigned to a HeyRex monitor. Monitors were mounted to a collar that was dedicated to it because pilot testing of the monitor showed that connecting a leash to the collar during walks altered its position and the activity data collected. The collar was fitted to the dog’s neck with enough room for two fingers to fit between the collar and the neck and placed in a ventral position. Dogs were required to wear the activity monitor during the entire 49 d of study. HeyRex monitors have the capacity to store data for 7 d, with data being automatically uploaded to an online data server from a device-specific receiver. Each owner was provided with a receiver and instructed to connect it to an Ethernet port of their WiFi modem and router at home, in a location where the animal would spend significant time each day. This was to ensure that the recorded activity data was being uploaded at least every 24 h. Each monitor and receiver was registered and tested prior to use. Once a connection is detected, these devices automatically send stored data wirelessly to the router, which is then uploaded to the server, making it possible for investigators to view and assess data remotely. Data collected for 49 d were downloaded from the online server as an excel spreadsheet for analyses. Dogs with more than 4 d of activity data loss or data with an integrity score < 90% were excluded from analyses. Integrity score for each day of data collected was assigned based on the percentage of “quality activity data” that was recorded, excluding any data loss. At the end of the study, an average of the daily integrity score was assessed and used to determine overall data quality. HeyRex activity epoch length was set to 15 min. In addition to total activity measured, the following parameters were generated: running, walking, sleeping, scratching, and alert/moving (not running or walking). The delta-G, duration, and intensity (delta-G/duration = intensity) of each activity were provided. Delta-G expresses activity as the change in vector magnitude of all three accelerometer axes.

Owner survey

Dog owners completed the CBPI and LOAD surveys on day 0, 21, 35, and 49 of the study. CBPI is a validated survey for OA and osteosarcoma developed by researchers at the University of Pennsylvania (Brown et al., 2008). The survey consists of a total of ten questions, with the first nine rating the owner’s perception of their dog’s pain and how it interferes with the animal’s daily routine over the past 14 d. The rating scale for each question ranges from 0 (no pain/interference) to 10 (extreme pain/interference). The last question is a qualitative score of an owner’s perception of the dog’s quality of life, ranging from poor to excellent. The LOAD survey is a validated survey for canine elbow OA and gait analyses. This survey consists of 10 questions on the background and lifestyle information of the dog, followed by a series of 13 questions focused on the dog’s mobility on a 5-point scale with descriptive terms.

Blood collection, serum chemistry, and serum inflammatory marker analyses

On day 0 and 49 of study, overnight fasted blood samples were collected via jugular venipuncture by the University of Illinois small animal veterinary staff. On day 0, blood was collected while the animal was under sedation for radiographic assessment of OA. Blood was collected into BD Vacutainer SST tubes (Becton, Dickinson and Company, Franklin Lakes, NJ) and allowed to clot at room temperature. One of the tubes was sent to the University of Illinois Veterinary Diagnostic Laboratory (Urbana, IL) for serum chemistry analysis. The second tube was centrifuged at 1,210 × g for 15 min at 4 °C. Isolated serum was stored at −80 °C until serum inflammatory marker analysis. The following inflammatory markers were analyzed using enzyme-linked immunosorbent assays: canine C-reactive protein (Abcam, Cambridge, MA) and matrix metalloproteinase-3 (Cloud-Clone Corp., Houston, TX).

Statistical analysis

Data were analyzed with SAS (version 9.3; SAS Institute, Inc., Cary, NC) using the Mixed Models procedure with a repeated-measures design. Proc Univariate was used to evaluate data normality. Means were separated using Fisher-protected least significant difference with Tukey’s adjustment to control for experiment-wise error. Data correlation was performed with R Studio (version 1.0.136; RStudio, Inc., Boston, MA) using Pearson correlation coefficients. Data were reported as means ± SEM with statistical significance set at P < 0.05 and P < 0.10 considered a trend.

Results

Dog population

Seventy-two dogs were enrolled in the study. Three dogs were unable to complete the study due to secondary injury and inability to stay off medication. An additional thirteen dogs were excluded from analysis due to partial loss of activity data set (> 4 d of activity loss). This threshold was based on data integrity scorings provided by HeyRex and those with a data integrity scoring less than 90% were excluded from analyses. This loss of data was possibly due to poor connection between the monitor and the receiver connected at each residence. As a result, 56 dogs were considered for statistical analyses. Dog breeds included Alaskan Malamute (n = 1), American Bulldog (n = 1), American Pitbull terrier (n = 2), Australian Shepherd (n = 1), Bermese Mountain dog (n = 2), Border collie (n = 1), Boxer (n = 1), French Mastiff (n = 1), German shepherd (n = 3), Golden retriever (n = 3), Great Dane (n = 1), Keeshond (n = 4), Labrador retriever (n = 4), Mixed breed (n = 28), Pembroke Welsh Corgi (n = 1), Rottweiler (n = 1), and Siberian Husky (n = 1). The mean age of the dogs that completed the study were 7.8 ± 3.4 yr old with a mean BW of 31.4 ± 10.9 kg. The average BCS assigned by veterinarians was 6.2 ± 1.2. The BCS assigned by lab personnel was 6.1 ± 1.1, while average BCS assigned by owners was 5.5 ± 1.4.

Owner survey data

Owner responses to LOAD and CBPI questionnaires showed several significant changes over time. The results from the CBPI questionnaire for all dogs collected from day 0, 21, 35, and 49 are summarized in Table 1. Responses to average pain; pain as of right now; interference of general activity, interference with the ability to rise from a lying position; interference with the ability to run; interference with the ability to climb; pain severity score (PSS); and pain interference score (PIS) were lower (P < 0.05) on day 21, 35, and 49 compared to day 0. Owner response to pain at its worst was higher (P < 0.05) on day 0 compared to day 35 and 49. Response to pain at its least tended to be lower (P < 0.10) on day 49 compared to day 0, while pain at its worst was lower (P < 0.05) on day 35 and 49 compared to day 0. Interference of enjoyment of life was lower (P < 0.05) on day 21 compared to day 0.

Table 1.

CBPI scores for dogs on day 0, 21, 35, and 49 of study

Item Day 0 Day 21 Day 35 Day 49 SEM P-value
Pain at its worst 3.42b 2.97a,b 2.77a 2.75a 0.299 0.03
Pain at its least 1.14b 0.98a,b 1.03a,b 0.82a 0.178 0.10
Pain on average 2.21b 1.85a 1.78a 1.55a 0.225 <0.01
Pain as of right now 1.71b 1.37a 1.36a 1.19a 0.214 0.02
Interference of general activity 2.23b 1.67a 1.79a 1.80a 0.288 0.05
Interference of enjoyment of life 1.63b 0.93a 1.25a,b 1.29a,b 0.227 0.04
Interference with ability to rise from lying position 3.21b 2.15a 2.28a 2.07a 0.303 <0.0001
Interference with the ability to walk 1.89b 1.35a 1.54a,b 1.57a,b 0.252 0.05
Interference with the ability to run 2.80b 2.10a 2.15a 2.19a 0.336 0.02
Interference with ability to climb 3.43b 2.21a 2.34a 2.41a 0.359 <0.0001
Overall quality of life 4.04 4.09 4.01 4.14 0.109 0.52
PSS1 2.12b 1.79a 1.71a 1.58a 0.212 <0.01
PIS2 2.53b 1.73a 1.89a 1.89a 0.266 <0.0001
Open in a new tab

1PSS (Pain severity score), Arithmetic mean of four questions rating the dog’s pain intensity in the last 7 d.

2PIS (Pain interference score), Arithmetic mean of six questions rating how much the dog’s pain interferes with its normal activity.

a,

bWithin a row, means with different superscripts differ (P < 0.05).

Results of LOAD survey are represented in Figure 1 and Table 2. Figure 1 summarizes answers pertaining to background and exercise questions from 7 d before the start of the study (day 0). Responses indicated that the vast majority of dogs enrolled in the study had suffered from mobility problems for more than 6 mo (87%). According to owner response, 39% of dogs walked 0 to 1 km daily, 37% of dogs walked 1 to 2 km daily, and 24% walked > 2 km daily. Approximately 29% of responses indicated that dogs did not go outside for walks, 43% indicated that dogs were walked at least once a day, and 28% were walked > 2 times a day. Questions regarding exercise style indicated that a little over half of the dogs were walked while on a lead (34% always on lead; 23% mostly on lead), with 41% mostly or always off a lead. Table 2 reports the results from owner responses collected on day 0, 21, 35, and 49 of study. Response to question 9 showed that on day 35 and 49, dogs were limiting the amount of exercise more (P < 0.01) than when compared to day 0. At day 35 and 49, owners responses indicated that their dog’s mobility was better (P < 0.05) than that at day 0, but it was a small numerical change. Most owner responses to questions about dog activity, stiffness, and lameness were not altered over time, but responses to indicated that at day 49, the effect of exercise on their dog’s lameness was lower (P < 0.05) than that of day 0. Also, owners indicated that the effect of cold, damp weather on their dog’s ability to exercise was had a greater (P < 0.05) effect on day 35 and 49 than on day 0.

Figure 1.

Figure 1.

Open in a new tab

LOAD survey baseline data for background and lifestyle questions (% of responses). Mobility problems: How long has your pet been suffering with his/her mobility problems?; Exercise: In the last week, on average, how far has your dog exercised each day?; Walks: In the last week, on average, how many walks has your dog had each day?; Exercise type: What type of exercise is this?; Terrain: On what sort of terrain does your dog most often exercise?; How dog is handled at exercise: At exercise, how is your dog handled?

Table 2.

Owner responses to the LOAD survey for dogs on day 0, 21, 35, and 49 of study

Item Day 0 Day 21 Day 35 Day 49 SEM P-value
Q9 Who limits the extent to which your dog exercises?1 1.38a 1.45a,b 1.56b 1.53b 0.062 0.004
Q10 How is your dog’s mobility in general?2 2.30b 2.28a,b 2.13a 2.15a 0.092 0.050
Q11 How disabled is your dog by his/her lameness?3 2.17 2.08 2.07 1.98 0.088 0.094
Q12 How active is your dog?4 3.08 3.01 3.04 2.92 0.116 0.273
Q13 What is the effect of cold, damp weather on your dog’s lameness?5 1.96 1.99 1.96 2.04 0.110 0.779
Q14 To what degree does your dog show stiffness in the affected leg after a “lie down”?6 2.48 2.30 2.36 2.29 0.104 0.091
Q15 At exercise, how active is your dog?7 2.74 2.76 2.82 2.77 0.111 0.723
Q16 How keen to exercise is your dog?4 2.21 2.31 2.40 2.22 0.131 0.147
Q17 How would you rate your dog’s ability to exercise?2 2.21 2.15 2.29 2.19 0.110 0.372
Q18 What overall effect does exercise have on your dog’s lameness?5 2.49b 2.38a,b 2.34a,b 2.23a 0.101 0.043
Q19 How often does your dog rest (stop/sit down) during exercise?8 2.25 2.20 2.20 2.25 0.121 0.940
Q20 What is the effect of cold, damp weather on your dog’s ability to exercise?5 1.69a 1.80a,b 1.89b 1.93b 0.105 0.012
Q21 To what degree does your dog show stiffness in the affected joint after a “lie down” following exercise?6 2.62 2.42 2.41 2.40 0.109 0.056
Q22 What is the effect of your dog’s lameness on his/her ability to exercise?5 3.45 2.50 2.38 2.23 0.674 0.365
Open in a new tab

1Answer: 1) you; 2) your dog.

2Answer:1) very good; 2) good; 3) fair; 4) poor; 5) very poor.

3Answer: 1) not at all disabled; 2) slightly disabled; 3) moderately disabled; 4) severely disabled; 5) extremely disabled.

4Answer:1) extremely active/keen; 2) very active/keen; 3) moderately active/keen; 4) slightly active/keen; 5) not at all active/keen.

5Answer: 1) no effect; 2) mild effect; 3) moderate effect; 4) severe effect; 5) extreme effect.

6Answer: 1) no stiffness; 2) mild stiffness; 3) moderate stiffness; 4) severe stiffness; 5) extreme stiffness.

7Answer: 1) extremely active; 2) very active; 3) fairly active; 4) not very active; 5) not at all active

8Answer: 1) never; 2) hardly ever; 3) occasionally; 4) frequently; 5) very frequently.

Activity data

Physical activity was measured for 49 d, with an average of 48.6 d recorded and considered for analysis. Activity data measurements were reported in terms of delta-G force. The average 24 h activity pattern for all dogs is shown in Figure 2, with the average activity count for the entire study represented in Figure 3. Average weekly, weekday, and weekend activity counts for total activity, and by activity type (alert, run, walk, sleep, and scratch) are presented in Figure 4. Weekend total activity counts and counts in an alert state were higher than weekday counts. However, run, walk, sleep, and scratch activity counts were higher on weekdays than weekends.

Figure 2.

Figure 2.

Open in a new tab

Average 24-h daily activity pattern of dogs in delta-G.

Figure 3.

Figure 3.

Open in a new tab

Daily activity pattern of dogs throughout the entire study expressed in delta-G with contrast between weekdays and weekends (highlighted in gray).

Figure 4.

Figure 4.

Open in a new tab

Average weekly, weekday, and weekend energy expenditures (delta-G) for: (A) total weekly average; (B) alert; (C) sleep; (D) run; (E) walk; and (F) scratch activities.

Activity, owner pain assessment, age, BW, and BCS data correlations

Total activity measurements as well as activity measurement by type (alert, run, walk, sleep, and scratch) were correlated with dog age, BW, BCS (scored by lab personnel and a veterinarian), and owner assessment data obtained from LOAD and CBPI surveys (Figure 5). Age was negatively correlated with scratch (P = 0.03; r = −0.10) and alert (P = 0.03; r = −0.13) activity counts. BW was negatively correlated with total activity counts (P = 0.02; r = −0.12). BW of dogs with a BW > 30 kg (median BW) were positively correlated with alert activity counts (P = 0.01; r = 0.04). Age was negatively correlated with sleep (P = 0.02; r = −0.12) and alert (P = 0.004; r = −0.29) activity counts. BCS assigned by lab personnel was positively correlated with scratch activity counts (P = 0.06; r = 0.19), while being negatively correlated with sleep activity counts (P = 0.02; r = −0.26). BCS assigned by a veterinarian was positively correlated with scratch activity counts (P = 0.06; r = 0.07), while being negatively correlated with sleep activity counts (P = 0.02; r = −0.27). Total activity counts of dogs with BW < 30 kg were positively correlated with PIS (P = 0.04; r = 0.21), but negatively correlated with BW (P = 0.004; r = −0.12).

Figure 5.

Figure 5.

Open in a new tab

Correlations of physical activity (total; activity types) with owner pain assessment scores, age, BW, and BCS assessed by veterinarian and lab personnel. Plots include all dogs (n = 56) (A), dogs with BW > 30 kg (n = 30) (B), and dogs with BW < 30 kg (n = 26) (C); *P < 0.05.

Serum chemistry and serum inflammatory marker data

Serum chemistry data for all dogs are summarized in Table 3. All metabolites were within the reference ranges except for total alkaline phosphatase (ALP) and corticosteroid-induced alkaline phosphatase (CALP), which were already higher than the upper reference range at day 0. Serum total protein, albumin, globulin, calcium, potassium, total bilirubin, and cholesterol concentrations were greater (P < 0.05) on day 49 compared to day 0. Inversely, serum sodium, sodium:potassium ratio, chloride, and glucose concentrations were lower on day 49 compared to day 0 of study. Serum C-reactive protein concentrations were lower (P < 0.01) on day 49 than on day 0 (Table 4). Measured matrix metalloproteinase-3 concentrations were not altered over time.

Table 3.

Serum chemistry profiles for dogs on day 0 and 49 of study

Item Reference range1 Day 0 Day 49 SEM P-value
Creatinine, mg/dL 0.5-1.5 1.00 1.01 0.042 0.33
Blood urea nitrogen, mg/dL 6-30 16.68 16.77 0.910 0.88
Total protein, g/dL 5.1–7.0 6.03a 6.27b 0.065 <0.0001
Albumin, g/dL 2.5–3.8 3.08a 3.23b 0.039 <0.0001
Globulin, g/dL 2.7–4.4 2.95a 3.04b 0.049 0.05
Albumin:globulin ratio 0.6–1.1 1.05 1.07 0.020 0.18
Calcium, mg/dL 7.6–11.4 9.95a 10.13b 0.076 <0.0001
Phosphorus, mg/dL 2.7–5.2 3.82 3.86 0.089 0.67
Sodium, mmol/L 141–152 145.66b 144.52a 0.263 <0.0001
Potassium, mmol/L 3.9–5.5 4.41a 4.66b 0.040 <0.0001
Sodium:potassium 28–36 33.18b 31.14a 0.318 <0.0001
Chloride, mmol/L 107–118 114.20b 112.48a 0.325 <0.0001
Glucose, mg/dL 68–126 100.43b 92.27a 1.855 <0.01
ALP2, U/L 7–92 102.00 123.20 23.406 0.33
CALP3, U/L 0–40 64.05 80.91 19.781 0.02
ALT4, U/L 8–65 47.30 52.19 5.834 0.29
GGT5, U/L 0–7 3.36 3.04 0.266 0.16
Total bilirubin, mg/dL 0.1–0.3 0.24a 0.27b 0.015 0.04
Cholesterol, mg/dL 129–297 238.38a 259.84b 8.497 <0.0001
Triglycerides, mg/dL 32–154 85.84 98.61 13.336 0.42
Bicarbonate, mmol/L 16–24 20.64 20.98 0.321 0.35
Open in a new tab

1Reference ranges were provided from the University of Illinois Veterinary Diagnostic Lab.

2ALP, alkaline phosphatase.

3CALP, corticosteroid-induced alkaline phosphatase.

4ALT, alanine aminotransferase.

5GGT, gamma-glutamyl transferase.

a,

bWithin a row, means with different superscripts differ (P < 0.05).

Table 4.

Serum C-reactive protein and matrix metalloproteinase-3 concentrations for dogs on day 0 and 49 of study

Item Day 0 Day 49 SEM P-value
C-reactive protein, ng/mL 2957.51b 1520.14a 506.23 <0.01
Matrix metalloproteinase-3, ng/mL 8.72 8.99 0.549 0.59
Open in a new tab

a,

bWithin a row, means with different superscripts differ (P < 0.01).

Discussion

OA is a progressive chronic disease that degenerates the joints of an animal, not only bringing pain and disability to the affected area but also lowering the overall quality of life due to loss of function, reduced joint mobility, and increased pain. Currently, there is no cure for OA, but there are many potential treatment options aimed at managing pain, delaying disease progression, reducing joint instability and restoring near-normal joint function. Multimodal treatment approaches are often needed for patients suffering from OA, with most common treatment strategies including pharmacological interventions, rehabilitation, surgery, weight loss, nutraceuticals, and alternative methods such as acupuncture or laser therapy. Physical activity is also an important treatment for preventing the progression of OA and in aiding in weight loss to alleviate joint stress inflicted by excess BW. Therefore, there has been interest in using activity monitors for management of OA.

Voluntary physical activity in this study was assessed using HeyRex activity monitors, which have previously been shown to provide data highly correlated with that collected from Actical accelerometers (Mejia et al., 2019). Activity data varied greatly over time due to the heterogeneity of the free-living dog population studied. Observation of the average 24-h activity data indicated that activity levels peaked during early morning and mid-afternoon, which probably coincides with their feeding, walking and/or potty times. Several studies have evaluated activity level with varied feeding frequency in cats and mice and have reported increases due to food anticipatory activity (Deng et al., 2011, Luby et al., 2012; Kappen et al., 2014; de Godoy et al., 2015). However, those studies were conducted in research lab settings where there is little human interaction at other times during the day. Future studies should try to account for and measure the contribution of food anticipatory activity in the free-living dog population.

Interestingly, our data demonstrate greater physical activity on weekends compared to weekdays. This is unlike the pattern observed in a research laboratory setting, where cats have been shown to have reduced activity levels during weekends compared to weekdays due to reduced animal care procedures and lower traffic of animal caregivers (de Godoy and Shoveller, 2017). In free-living animals, we attribute the lower activity on weekdays to the owner working hours away from home, a time when there is reduced owner–dog interaction and/or animals are placed in a crate or a limited space in the home. This observed pattern likely indicates that in order to objectively evaluate physical activity of free-living dogs, at least 7 d of data collection are needed, which allows for the inclusion of both weekdays and weekends. In addition, the individual living environment (i.e., crated; free in house; free outdoors) should be considered when analyzing physical activity data.

Total activity and different activity types (run, scratch, alert, walk, and sleep) were correlated to PSS, PIS, pain scale provided by veterinarians, age, BW, and BCS. In humans, it has been well documented that joint pain, age, and body mass index are all factors that reduce physical activity level. In humans, at least two out of five adults with lower-extremity joint conditions do not engage in any moderate physical activity lasting 10 min over an entire week, which is attributed to pain (Fontaine et al., 2004; Shih et al., 2006; Dunlop et al., 2017). In addition, comorbidities such as abdominal obesity is commonly reported in OA patients when compared to healthy controls (63% vs. 38%), with limited physical activity being observed in these population (van Dijk et al., 2008; Puenpatom and Victor, 2009; Hawker, 2019). In this study, a discrepancy was observed in the BCS score evaluated by a veterinarian, lab personnel, and owners. Owner BCS averaged 5.5 ± 1.4, closer to an ideal body condition of 5, while veterinary (6.2 ± 1.2) and lab personnel (6.1 ± 1.1) scores were higher. A similar result was observed by Gerstner and Liesegang (2017), with owner scores being significantly lower than that of veterinarians, indicating a discrepancy in the owner’s perception of BCS of their pets. Another study comparing pet dog owners and sports dog owners reported that while both groups showed as much interest in the health and well-being of their dogs, pet dog owners were less confident in their knowledge of correct feeding amounts and were feeding more (i.e., ½ cup more per day) than sports dog owners (Kluess et al., 2021). This owner misconception demonstrates the need for specialists to guide owners on their pet’s body condition and proper nutrition and exercise regimes, which may help prevent the onset and progression of OA.

In the current study, total activity counts had a negative relationship with BW, but it was not clear whether this was due to the OA disease condition in the animal, as pain scores did not show any correlation with measured activity. Age and BW seem to be negatively correlated with total activity, alert activity, and scratching activity counts. However, because there was no correlation with pain scores, there is no indication that OA interfered with the activity levels in these dogs. In a client-owned dog study, owners reported a change in the distance and quality of the walk with the development of OA (Belshaw et al., 2020). Owners reported that after developing OA, their dogs walked more slowly and with more frequent stops than before developing OA, and on some bad days, dogs had decreased desire and ability to exercise (Belshaw et al., 2020).

Total activity count in dogs < 30 kg BW was the only parameter that was positively correlated with PIS (P = 0.04; r = 0.21) assigned by owner questionnaires. Because veterinarian pain scoring was not followed after initial diagnosis at the time of enrollment, we were unable to confirm such a relationship from professionals. Regardless, this correlation suggests that activity monitors may provide useful information to owners and veterinarians and serve as a tool to assess the type and amount of physical activity that might be adequate for each individual patient.

A study conducted in humans suggested that supervised aerobic exercise helped to reduce joint pain when performed at least three times a week (Juhl et al., 2014). Even though they are limited, dog studies have shown similar benefits. In a study conducted by Nganvongpanit et al. (2014), 55 dogs were assigned to one of three groups: 1) OA with swimming (n = 22); 2) healthy controls with swimming (n = 18); and 3) healthy controls with no swimming (n = 15). Dogs were allowed to swim for a total of 8 wk (twice a week, three cycles of 20-min intervals with 5-min rest). Blood was collected every 2 wk to measure OA biomarkers, including chondroitin sulfate epitope WF6 (CS-WF6) and hyaluronan. Results at the end of the study showed that dogs with OA that swam had lower (P < 0.05) circulating CS-WF6 concentrations and lower (P < 0.05) pain on palpation compared to pre-exercise. Overall, the effects of swimming twice a week over a period of 8 wk showed significant improvement of OA joint function. However, these therapies are not readily accessible to all dog owners and are expensive, so it is not a treatment given to all dogs with OA, especially to those with a mild diagnosis. In humans, studies have evaluated different forms of exercise such as running and walking that can help reduce pain while improving the range of motion of the knee just as effectively as aquatic therapy (Wang et al., 2011; Farrokhi et al., 2017). These types of exercise are more accessible as dogs are usually walked outside. However, a detailed exercise prescription based on the animal’s OA severity conditions might be needed from a specialist. For that, an activity collar may be a useful tool for a veterinarian to understand the lifestyle pattern of the animal and owner, which may help make exercise recommendations.

It should be noted that although pain relief medications and anti-inflammatory drugs were not allowed during the study, some of the owner-assessed pain and mobility scores improved over time. The potential reasons for these changes are unknown. Because the CBPI and LOAD surveys are subjective and rely on owner recall of their dog’s pain and mobility levels, changes over time may have been due to human error. It is possible that dog owners became accustomed to their dog’s condition as they went through the study and/or the use of activity monitors motivated owners to increase their dog’s activity levels over time and improvements in pain were a consequence of these changes. Likewise, many serum chemistry markers were altered and C-reactive protein decreased over time. These changes were also unexpected and not deemed to be clinically relevant because they remained within reference ranges. Although owners were asked to maintain dietary and supplement intake over time, diet records were not collected in this study. Therefore, any unknown changes to diet may have altered these measures. Future studies are suggested to control or more closely monitor diet and supplement intake to avoid or better explain such changes over time.

In conclusion, our data show that daily physical activity tended to peak early in the morning and again in late afternoon, likely coinciding with feeding, walking, or potty times. Total and alert activity counts were higher on weekends than weekdays. Body weight was negatively associated with total activity counts and age was negatively associated with alert and sleep activity counts, but pain assessment was not strongly associated with activity. Our data suggest that activity monitors may provide useful information and serve as a tool for OA management, but a greater level of control or understanding of the general living conditions and daily activity patterns of owners and dogs may have allowed for a more accurate interpretation of the data in this study. Future research should not only continue to test the usefulness of activity monitors in OA management, but also incorporate them into studies assessing pharmacological or nutraceutical treatments for canine OA. Given the discrepancy in BCS provided by owners compared to a veterinarian and lab personnel and changes in pain scores over time, this study also showed the need for more owner education on proper nutrition, weight management, and pain assessment. While CBPI is a useful survey to understand owner perceptions of their pet, it is a subjective scoring system. Owner education in regard to the health and pain assessment of their pet may allow these surveys to be more meaningful and interpretable.

Acknowledgment

The funding for this study was provided by USDA National Institute of Food and Agriculture (Hatch Grant #ILLU-538-937).

Glossary

Abbreviations

ALP

alkaline phosphatase

ALT

alanine aminotransferase

BCS

body condition score

BW

body weight

CALP

corticosteroid-induced alkaline phosphatase

CBPI

canine brief pain inventory

GGT

gamma-glutamyl transferase

LOAD

Liverpool osteoarthritis in dogs

OA

osteoarthritis

PIS

pain interference score

PSS

pain severity score

Conflict of interest statement

All authors have no conflicts of interest.

Literature Cited

  1. Belda, B., Enomoto M., Case B. C., and Lascelles B. D. X.. . 2018. Initial evaluation of PetPace activity monitor. Vet. J. 237:63–68. doi: 10.1016/j.tvjl.2018.05.011 [DOI] [PubMed] [Google Scholar]
  2. Belshaw, Z., Dean R., and Asher L.. . 2020. Slower, shorter, sadder: a qualitative study exploring how dog walks change when the canine participant develops osteoarthritis. BMC Vet. Res. 16:85. doi: 10.1186/s12917-020-02293-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Brown, D. C., Boston R. C., Coyne J. C., and Farrar J. T.. . 2008. Ability of the canine brief pain inventory to detect response to treatment in dogs with osteoarthritis. J. Am. Vet. Med. Assoc. 233:1278–1283. doi: 10.2460/javma.233.8.1278 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Brown, D. C., Boston R. C., and Farrar J. T.. . 2010. Use of an activity monitor to detect response to treatment in dogs with osteoarthritis. J. Am. Vet. Med. Assoc. 237:66–70. doi: 10.2460/javma.237.1.66 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Brown, D. C., Boston R. C., and Farrar J. T.. . 2013. Comparison of force plate gait analysis and owner assessment of pain using the Canine Brief Pain Inventory in dogs with osteoarthritis. J. Vet. Intern. Med. 27:22–30. doi: 10.1111/jvim.12004 [DOI] [PubMed] [Google Scholar]
  6. Comblain, F., Serisier S., Barthelemy N., Balligand M., and Henrotin Y.. . 2016. Review of dietary supplements for the management of osteoarthritis in dogs in studies from 2004 to 2014. J. Vet. Pharmacol. Ther. 39:1–15. doi: 10.1111/jvp.12251 [DOI] [PubMed] [Google Scholar]
  7. Deng, P., Grant R. W., and Swanson K. S.. . 2011. Physical activity level of adult cats with varied feeding frequency. Br. J. Nutr. 106Suppl 1:S166–S169. doi: 10.1017/S0007114511001863 [DOI] [PubMed] [Google Scholar]
  8. van Dijk, G. M., Veenhof C., Schellevis F., Hulsmans H., Bakker J. P., Arwert H., Dekker J. H., Lankhorst G. J., and Dekker J.. . 2008. Comorbidity, limitations in activities and pain in patients with osteoarthritis of the hip or knee. BMC Musculoskelet. Disord. 9:95. doi: 10.1186/1471-2474-9-95 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Dunlop, D. D., Song J., Lee J., Gilbert A. L., Semanik P. A., Ehrlich-Jones L., Pellegrini C. A., Pinto D., Ainsworth B., and Chang R. W.. . 2017. Physical activity minimum threshold predicting improved function in adults with Lower-Extremity symptoms. Arth. Care Res. 69:475–483. doi: 10.1002/acr.23181 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Edwards, J. D., and Gibson D. J. M.. . 2012. Novel technology for the remote monitoring of animals. Comp. Anim. Soc. Newsletter 23:56–59. [Google Scholar]
  11. Farrokhi, S., Jayabalan P., Gustafson J. A., Klatt B. A., Sowa G. A., and Piva S. R.. . 2017. The influence of continuous versus interval walking exercise on knee joint loading and pain in patients with knee osteoarthritis. Gait Posture 56:129–133. doi: 10.1016/j.gaitpost.2017.05.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Fontaine, K. R., Heo M., and Bathon J.. . 2004. Are US adults with arthritis meeting public health recommendations for physical activity? Arthritis Rheum. 50:624–628. doi: 10.1002/art.20057 [DOI] [PubMed] [Google Scholar]
  13. Gerstner, K., and Liesegang A.. . 2017. Survey: nutrition, body condition and activities of dogs in Switzerland. J. Anim. Physiol. Anim. Nutr. (Berl). 101Suppl 1:15–20. doi: 10.1111/jpn.12615 [DOI] [PubMed] [Google Scholar]
  14. de Godoy, M. R., and Shoveller A. K.. . 2017. Overweight adult cats have significantly lower voluntary physical activity than adult lean cats. J. Feline Med. Surg. 19:1267–1273. doi: 10.1177/1098612X17694252 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. de Godoy, M. R., Ochi K., de Oliveira Mateus L. F., de Justino A. C., and Swanson K. S.. . 2015. Feeding frequency, but not dietary water content, affects voluntary physical activity in young lean adult female cats. J. Anim. Sci. 93:2597–2601. doi: 10.2527/jas.2014-8574 [DOI] [PubMed] [Google Scholar]
  16. Gupta, R. C., Canerdy T. D., Lindley J., Konemann M., Minniear J., Carroll B. A., Hendrick C., Goad J. T., Rohde K., Doss R., . et al. 2012. Comparative therapeutic efficacy and safety of type-II collagen (uc-II), glucosamine and chondroitin in arthritic dogs: pain evaluation by ground force plate. J. Anim. Physiol. Anim. Nutr. 96:770–777. doi: 10.1111/j.1439-0396.2011.01166.x [DOI] [PubMed] [Google Scholar]
  17. Hawker, G. A. 2019. Osteoarthritis is a serious disease. Clin. Exp. Rheumatol. 37Suppl 120:3–6. [PubMed] [Google Scholar]
  18. Hercock, C. A., Pinchbeck G., Giejda A., Clegg P. D., and Innes J. F.. . 2009. Validation of a client-based clinical metrology instrument for the evaluation of canine elbow osteoarthritis. J. Small Anim. Pract. 50:266–271. doi: 10.1111/j.1748-5827.2009.00765.x [DOI] [PubMed] [Google Scholar]
  19. Johnson, K. A., Lee A. H., and Swanson K. S.. . 2020. Nutrition and nutraceuticals in the changing management of osteoarthritis for dogs and cats. J. Am. Vet. Med. Assoc. 256:1335–1341. doi: 10.2460/javma.256.12.1335 [DOI] [PubMed] [Google Scholar]
  20. Johnston, S. A. 1997. Osteoarthritis. Joint anatomy, physiology, and pathobiology. Vet. Clin. North Am. Small Anim. Pract. 27:699–723. doi: 10.1016/s0195-5616(97)50076-3 [DOI] [PubMed] [Google Scholar]
  21. Juhl, C., Christensen R., Roos E. M., Zhang W., and Lund H.. . 2014. Impact of exercise type and dose on pain and disability in knee osteoarthritis: a systematic review and meta-regression analysis of randomized controlled trials. Arthr. Rheumatol. 66:622–636. doi: 10.1002/art.38290 [DOI] [PubMed] [Google Scholar]
  22. Kappen, K. L., Garner L. M., Kerr K. R., and Swanson K. S.. . 2014. Effects of photoperiod on food intake, activity and metabolic rate in adult neutered male cats. J. Anim. Physiol. Anim. Nutr. (Berl). 98:958–967. doi: 10.1111/jpn.12147 [DOI] [PubMed] [Google Scholar]
  23. Kluess, H. A., Jones R. L., and Lee-Fowler T.. . 2021. Perceptions of body condition, diet and exercise by sports dog owners and pet dog owners. Anim. 11:1752. doi: 10.3390/ani11061752 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Laflamme, D. 1997. Development and validation of a body condition score system for cats: a clinical tool. Feline Pract. 25:13–18. [Google Scholar]
  25. Lascelles, B. D., Hansen B. D., Roe S., DePuy V., Thomson A., Pierce C. C., Smith E. S., and Rowinski E.. . 2007. Evaluation of client-specific outcome measures and activity monitoring to measure pain relief in cats with osteoarthritis. J. Vet. Intern. Med. 21:410–416. doi: 10.1892/0891-6640(2007)21[410:eocoma]2.0.co;2 [DOI] [PubMed] [Google Scholar]
  26. Lascelles, B. D., Hansen B. D., Thomson A., Pierce C. C., Boland E., and Smith E. S.. . 2008. Evaluation of a digitally integrated accelerometer-based activity monitor for the measurement of activity in cats. Vet. Anaesth. Analg. 35:173–183. doi: 10.1111/j.1467-2995.2007.00367.x [DOI] [PubMed] [Google Scholar]
  27. Lawler, D. F., Evans R. H., Larson B. T., Spitznagel E. L., Ellersieck M. R., and Kealy R. D.. . 2005. Influence of lifetime food restriction on causes, time, and predictors of death in dogs. J. Am. Vet. Med. Assoc. 226:225–231. doi: 10.2460/javma.2005.226.225 [DOI] [PubMed] [Google Scholar]
  28. Luby, M. D., Hsu C. T., Shuster S. A., Gallardo C. M., Mistlberger R. E., King O. D., and Steele A. D.. . 2012. Food anticipatory activity behavior of mice across a wide range of circadian and non-circadian intervals. PLoS One 7:e37992. doi: 10.1371/journal.pone.0037992 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Mejia, S., Duerr F. M., and Salman M.. . 2019. Comparison of activity levels derived from two accelerometers in dogs with osteoarthritis: implications for clinical trials. Vet. J. 252:105355. doi: 10.1016/j.tvjl.2019.105355 [DOI] [PubMed] [Google Scholar]
  30. Musco, N., Vassalotti G., Mastellone V., Cortese L., Della Rocca G., Molinari M. L., Calabrò S., Tudisco R., Cutrignelli M. I., and Lombardi P.. . 2019. Effects of a nutritional supplement in dogs affected by osteoarthritis. Vet. Med. Sci. 5:325–335. doi: 10.1002/vms3.182 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Nganvongpanit, K., Tanvisut S., Yano T., and Kongtawelert P.. . 2014. Effect of swimming on clinical functional parameters and serum biomarkers in healthy and osteoarthritic dogs. ISRN Vet. Sci. 2014:459809. doi: 10.1155/2014/459809 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Pollmeier, M., Toulemonde C., Fleishman C., and Hanson P. D.. . 2006. Clinical evaluation of firocoxib and carprofen for the treatment of dogs with osteoarthritis. Vet. Rec. 159:547–551. doi: 10.1136/vr.159.17.547 [DOI] [PubMed] [Google Scholar]
  33. Preston, T., Baltzer W., and Trost S.. . 2012. Accelerometer validity and placement for detection of changes in physical activity in dogs under controlled conditions on a treadmill. Res. Vet. Sci. 93:412–416. doi: 10.1016/j.rvsc.2011.08.005 [DOI] [PubMed] [Google Scholar]
  34. Puenpatom, R. A., and Victor T. W.. . 2009. Increased prevalence of metabolic syndrome in individuals with osteoarthritis: an analysis of NHANES III data. Postgrad. Med. 121:9–20. doi: 10.3810/pgm.2009.11.2073 [DOI] [PubMed] [Google Scholar]
  35. Roush, J. K., Dodd C. E., Fritsch D. A., Allen T. A., Jewell D. E., Schoenherr W. D., Richardson D. C., Leventhal P. S., and Hahn K. A.. . 2010. Multicenter veterinary practice assessment of the effects of omega-3 fatty acids on osteoarthritis in dogs. J. Am. Vet. Med. Assoc. 236:59–66. doi: 10.2460/javma.236.1.59 [DOI] [PubMed] [Google Scholar]
  36. Servet, E., Biourge V., and Marniquet P.. . 2006. Dietary intervention can improve clinical signs in osteoarthritic dogs. J. Nutr. 136(7 Suppl):1995S–1997S. doi: 10.1093/jn/136.7.1995S [DOI] [PubMed] [Google Scholar]
  37. Shih, M., Hootman J. M., Kruger J., and Helmick C. G.. . 2006. Physical activity in men and women with arthritis: national health interview survey, 2002. Am. J. Prev. Med. 30:385–393. doi: 10.1016/j.amepre.2005.12.005 [DOI] [PubMed] [Google Scholar]
  38. Vassalotti, G., Musco N., Lombardi P., Calabrò S., Tudisco R., Mastellone V., Grazioli R., Bianchi S., and Cutrignelli M. I.. . 2017. Nutritional management of search and rescue dogs. J. Nutr. Sci. 6:e44. doi: 10.1017/jns.2017.47 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Wang, T. J., Lee S. C., Liang S. Y., Tung H. H., Wu S. F. V., and Lin Y. P.. . 2011. Comparing the efficacy of aquatic exercises and land-based exercises for patients with knee osteoarthritis. J. Clin. Nurs. 20:2609–2622. doi: 10.1111/j.1365-2702.2010.03675.x [DOI] [PubMed] [Google Scholar]
  40. Wiseman, M. L., Nolan A. M., Reid J., and Scott E. M.. . 2001. Preliminary study on owner-reported behaviour changes associated with chronic pain in dogs. Vet. Rec. 149:423–424. doi: 10.1136/vr.149.14.423 [DOI] [PubMed] [Google Scholar]

Articles from Journal of Animal Science are provided here courtesy of Oxford University Press

RESOURCES