An Internet survey of risk factors for injury in North American dogs competing in flyball

Abstract

A survey was used to investigate injuries in dogs competing in flyball. Complete surveys were obtained from 272 respondents with 589 dogs. In the past year, 23.3% of dogs were injured, with 34.1% injured during their career to date. Common injury sites were paws/digits, back, shoulder, and iliopsoas muscle/groin. Injury in previous years, modified by weight:height ratio, was a significant risk factor for injury. Dogs > 2 y of age had increased risk of injury, as did dogs with best times < 4.0 s. Canadian dogs had increased risk of injury (30.7% injured) compared to United States dogs (20.1% injured). This relationship was modified by participation in other sports, which generally reduced risk of injury in Canadian dogs. Further investigation of risk factors should include differences in training and competition between the United States and Canada, as well as injury prevention strategies.

Introduction

Flyball is a popular canine sport in North America involving more than 6500 dogs (1). Teams of 4 dogs compete in a relay race in which each dog completes 4 jumps, retrieves a ball from a box, pushes off the box to make a 180° turn and returns over the same 4 jumps (1). Jump height is set a minimum of 5 inches (12.7 cm) below the withers height of the smallest dog on the team, with a minimum jump height of 7 inches (17.8 cm) and a maximum of 14 inches (35.6 cm) (1). The current North American team record is 14.433 s. The fastest dogs complete the 31-meter course in under 4 s (1,2).

Few studies have investigated injuries in flyball dogs. A 2010 British survey of 440 flyball dogs reported forelimb lameness, injuries of the footpads, muscle injuries, and injuries to the dewclaws as most common (3). A 2019 American survey of 375 dogs reported that 39% had at least 1 injury that occurred while competing in or training for flyball (4). The most common injury locations were the forepaw/pad/nail/digit, shoulder, back/neck/tail, iliopsoas/groin, carpus, and elbow. The only factors associated with increased injury risk were younger dogs and the use of protective wraps (4).

Significant injury levels have also been reported in other canine sports. Several studies report a high incidence of injury in agility, a sport involving jumps, tunnels, and other obstacles performed at high speed (5–9). In 1 survey, 32% of all dogs and 41% of border collies, had been injured during agility (6). Reported risk factors for agility injuries included border collie breeding, previous injury, the dog having ≤ 4 y agility experience, use of alternative treatments, less handler experience, and increasing dog age (5,7,8). An injury incidence of 22% of dogs was reported in canicross, a sport in which a person runs with 1 or 2 dogs in pulling harnesses attached to a waist belt (10). Risk factors for injury included Labrador retriever breed, running with a second dog, and participation in agility (10).

Unlike agility, in which courses vary and dogs typically turn in both directions, and in canicross, in which dogs run relatively straight lines, dogs in flyball always run the same course and nearly always turn the same direction on the box (3). Initial contact with the box may result in carpal hyperextension and substantial stress on the forelimbs (3,11). The dog pushes off the angled box at the end of the turn, typically with 1 pelvic limb higher on the angled box than the other. This may result in uneven forces between the 2 pelvic limbs and could predispose dogs to muscle strains (3). Each dog completes numerous runs at a single competition (2). Dogs participating in flyball may be predisposed to repetitive strain injuries, particularly with unidirectional turns (3,11).

The objectives of this study were to describe dog demographics and the frequency, types, and risk factors for injury in North American dogs competing in flyball. We hypothesized that there is a high incidence of injury, with the most common types being footpad injuries, carpal hyperextension, and muscle strains. We also hypothesized that breed, lack of warm-up prior to competition, and participation in other sports are risk factors for injury in competitive flyball dogs.

Materials and methods

Available information on injuries in other canine sports was used to guide design of an Internet-based survey (Fluid Surveys, Ottawa, Ontario). An exemption for ethical review was granted for this study by the University of Saskatchewan Behavioural Research Ethics Board. The survey was pre-tested to identify problems before distribution of the final version through North American Flyball Association (NAFA) region representatives (12). Responses were received between November 17, 2017 and March 12, 2018. The survey was restricted to a single response per Internet protocol address. The survey indicated that the investigators were veterinarians interested in injuries and injury prevention in sporting dogs and were surveying owners of both healthy and injured flyball dogs. It also stated that the survey was anonymous and that submission implied consent to use the participant’s information. Survey requirements included North American residency and an owner of at least 1 dog competing in flyball in the past year.

Owner demographic information included country of residence, NAFA region of residence, number of years of competitive flyball handling experience, and the number of dogs participating in competition in the past year. All dog-specific questions were repeated in the survey for each of the owner’s dogs (up to 10 dogs) that had competed in the previous year. Information collected for each dog included physical characteristics (sex, height, weight, age, and breed), and flyball performance (jump height, best time, direction of turn on the box, age at start of flyball training, number of years competing, number of days competing per year, and number of runs per day of competition). Questions on training were included, such as participation in other sports, time spent warming-up and cooling down, activities performed in warm-up and cool down, as well as time spent endurance training, strength training, and flyball training per week. Owners were asked if, in the past year, the dog had had any injuries or conditions that could be associated with flyball or that might be aggravated by flyball. Information collected for injuries that had occurred in the past year included injury location, side of the body affected, type of injury, who diagnosed the injury, and outcome of the injury. Owners were also asked to list any other flyball-related injuries that had occurred more than 1 y previously.

All responses were downloaded directly into Microsoft Excel 2010 (Microsoft, Redmond, Washington, USA). Any responses received from locations outside of North America or that were incomplete for each of the owner’s dogs were discarded.

Descriptive statistics were calculated using Microsoft Excel and Graphpad Prism, Version 8 (Graphpad, La Jolla, California, USA). Associations between the primary variable of interest (injury in the past year) and all other survey variables were initially screened using Chi-square and Fisher’s exact tests for categorical data, or unpaired Student’s t-test for continuous data. Variables with P-values ≤ 0.05 were then subjected to single variable screening using multilevel logistic regression, with injury in the past year as the outcome and dog owner as a random intercept, using reported methods (13). Variables with likelihood ratio test (LRT) P ≤ 0.20 were considered for inclusion in the final, multivariable model. Variables with LRT P ≤ 0.05 were kept in the final model. Biologically plausible, 2-way interactions were tested between final model variables and with non-significant variables considered for inclusion in the final model. Interactions were included if LRT P ≤ 0.05 and if the numbers of observations in each class of the cross-tabulations of injury and the 2 interacting variables were ≥ 10. Categories with numbers < 10 were kept if considered clinically relevant. Variables excluded from the final model were assessed for a confounding effect on final model variables. All logistic regression analyses were conducted in STATA15.1/IC (StataCorp, College Station, Texas, USA).

Results

Completed survey responses were received from 272 owners with a total of 589 dogs. Owner characteristics are summarized in Table 1. Most owners were from the United States (68%); the remainder were from Canada. Many owners had 1 to 3 competing dogs, with only 10.0% having 4 or more.

Table 1.

Selected characteristics of owners of North American Flyball dogs (N = 272).

	Number of owners (percent of owners)
Country of residence
Canada	87 (32.0)
United States	185 (68.0)
Mexico	0 (0.0)
Owner flyball experience (y)
< 1	7 (2.6)
1 to 2	23 (8.5)
3 to 4	52 (19.1)
5 to 10	74 (27.2)
>10	116 (42.6)
Number of owned dogs that competed in flyball in the past year
1	93 (34.2)
2	98 (36.0)
3	53 (19.5)
4	13 (4.8)
5	8 (2.9)
6 to 10	7 (2.6)

Selected characteristics of all dogs and of injured dogs are reported in Table 2. The border collie cross was the most common breed or breed cross comprising 150 of the 589 dogs (25.5%), followed by purebred border collies with 110 of 589 dogs (18.7%). No other breeds or breed cross comprised more than 4% of the total study population.

Table 2.

Number of dogs and number of injured dogs with selected characteristics.

Variable	Number of dogs (% of total dogs)	Number injured (% of dogs in row)
Country	N = 589	n = 137
United States	413 (70.1)	83 (20.1)
Canada	176 (29.9)	54 (30.7)
Dog age (y)	N = 589	n = 137
≤ 2	80 (13.6)	7 (8.8)
3 to 7	324 (55.0)	79 (24.1)
≥ 8	185 (31.4)	51 (27.6)
Best flyball time (s)	N = 589	n = 137
≥ 4.0	421 (71.5)	83 (19.7)
< 4.0	168 (28.5)	54 (32.1)
Dog weight:height ratio quartiles (lb:inch)	N = 495	n = 115
1st (0.73:1.38)	125 (25.3)	28 (22.4)
2nd (1.36:1.68)	123 (24.9)	20 (16.3)
3rd (1.68:2.08)	122 (24.7)	36 (29.5)
4th (2.08:3.11)	125 (25.3)	31 (24.8)
Dog participates in other sports	N = 589	n = 137
No	178 (30.2)	32 (18.0)
Yes	411 (69.8)	105 (25.5)
Border collie breeding	N = 589	n = 137
None	327 (55.5)	75 (22.9)
Purebred	109 (18.5)	27 (24.8)
Border collie cross	153 (26.0)	35 (22.9)

Owners reported that in the past year, 137 dogs (23.3% of 589 dogs) had been injured at least once. Twenty-one (15.3%) of the 137 dogs were injured twice in the past year, and 1 dog (0.7%) was injured 3 times in the past year. Fifty-five of the 137 dogs injured in the past year (40.1%) also had at least 1 injury in previous years. Of the 452 dogs not injured in the past year, 64 (14.2%) had been injured in previous years. Overall, 201 of 589 dogs (34.1%) had at least 1 reported injury over the duration of their flyball career. There were no significant differences in height or weight (unpaired t-test P = 0.4 and 0.3, respectively) between injured dogs (mean ± SD: 0.46 ± 0.10 m, 15.6 ± 6.7 kg) and uninjured dogs (0.45 ± 0.10 m and 15.0 ± 6.7 kg).

Anatomic locations of injuries are summarized in Table 3. The most common sites of injury were the paws/digits, back, shoulder, and iliopsoas muscle/groin. Characteristics of injuries are reported in Table 4. The most common types of injury, as reported by owners, were sprains/strains, pain, and tears. Most injuries (51.6%) were not diagnosed by a veterinarian. Although unilateral injuries were more commonly reported on the dog’s right side (37.7%) than the left side (19.5%), there was no significant difference (P = 0.4, Chi-square). There was no significant association between injury side and direction of turn on the box (P = 0.2, Fisher’s exact test).

Table 3.

Site of injuries incurred in dogs competing in flyball in the past year (n = 159 injuries in 137 dogs) and in previous years (n = 120 injuries in 119 dogs).

Area of injury	Number of injuries in the past year (%)	Number of injuries in previous years (%)
Paws/digits	38 (23.9)	39 (32.5)
Back	24 (15.1)	25 (20.8)
Shoulder	20 (12.6)	17 (14.2)
Iliopsoas/groin	14 (8.8)	7 (5.8)
Stifle	12 (7.5)	6 (5.0)
Hip	11 (6.9)	3 (2.5)
Neck	7 (4.4)	4 (3.3)
Open for hind limb	7 (4.4)	5 (4.2)
Open for forelimb	6 (3.8)	2 (1.7)
Hamstrings	4 (2.5)	0 (0.0)
Mouth/teeth	4 (2.5)	1 (0.8)
Elbow	3 (1.9)	2 (1.7)
Carpus	3 (1.9)	1 (0.8)
Other	3 (1.9)	6 (5.0)
Head	2 (1.3)	1 (0.8)
Tarsus	1 (0.6)	1 (0.8)

Table 4.

Characteristics of injuries incurred in dogs competing in flyball in the past year (n = 159 injuries in 137 dogs).

Number of injuries (%)
Type of injury (% of injuries)
Sprain/strain	45 (28.3)
Pain	30 (18.9)
Tear	16 (10.1)
Arthritis	11 (6.9)
Unknown	11 (6.9)
Other	10 (6.3)
Abrasion	10 (6.3)
Fracture	7 (4.4)
Intervertebral disc disease	6 (3.8)
Dislocation	5 (3.1)
Torn nail	5 (3.1)
Hyperextension	2 (1.3)
Tendonitis	1 (0.6)
Side of injury
Right	60 (37.7)
Left	31 (19.5)
Both	22 (13.8)
Unknown	10 (6.3)
Not applicable	36 (22.6)
Diagnosis of injury (% of injuries)
Veterinarian	77 (48.4)
Owner	39 (24.5)
Physiotherapist	16 (10.1)
Chiropractor	13 (8.2)
Trainer/handler/ instructor	5 (3.1)
Massage therapist	4 (2.5)
Rehabilitation therapist	3 (1.9)
Other	2 (1.3)
Outcome of injury
Returned to flyball	119 (74.8)
Ongoing rest/rehab	27 (17.0)
Retired	12 (7.5)
Other	1 (0.6)
Euthanasia	0 (0.0)

Dogs returned to flyball following 119 of 159 (74.8%) injuries that occurred in the past year (Table 4). Twenty-seven injuries (17.0%) resulted in ongoing rest or rehabilitation. Twelve injuries (7.5%) resulted in retirement from flyball. Four dogs retired due to intervertebral disc disease (IVDD), 2 with affected backs, and 2 with affected necks. One dog retired due to each of the following: shoulder osteochondrosis dissecans, shoulder pain, stifle arthritis, scrotal tear, mouth laceration, carpal pain after a previous episode of cervical IVDD, and iliopsoas pain after a previous episode of shoulder pain. No dogs were euthanized as a result of their injuries.

The following variables, significantly associated with injury in the past year (P ≤ 0.05), were reported after single variable analysis: country of origin, dog age category, best flyball time (< 4.0 s or not), direction of turn on the box, time for first warmup, dog participation in other sports, and dog being injured in previous years. Other variables considered for inclusion in the final model with P-values between 0.05 and 0.20 included dog weight:height ratio quartile categories, owner having more than 1 dog competing in flyball, jump height, and number of flyball runs per day of competition.

The main-effects final multivariable model included: country, dog age categories, best time, direction of turn on the box, time for first warm-up categories, dog weight:height ratio quartiles, and injury in previous years (Table 5). There were significant 2-way interactions between dog weight:height quartiles and injury in previous years and between country and dog competing in other sports. The final model also included owners having multiple dogs and handler experience categories as confounding variables. Once these were included, direction of turn on the box was no longer significantly associated with risk of injury (P = 0.14).

Table 5.

Results of the final multivariable logistic regression model for flyball injury that included a random intercept for dog owner (n = 493 responses included in final model).

Variable	OR (95% CI)	Wald P-value	LRT P-value
Country			< 0.01*
United States	Referent
Canada	7.51 (2.59 to 21.76)	< 0.01
Dog competes in other sports			0.01*
No	Referent	—
Yes	1.93 (0.93 to 4.00)	0.08
Interaction of country and dog competes in other sports			0.01*
United States * No	Referent	—
Canada * Yes	0.17 (0.05 to 0.56)	< 0.01
Dog age (y)			< 0.01*
≤ 2	Referent	—
3 to 7	4.08 (1.44 to 11.58)	0.01
≥ 8	4.51 (1.48 to 13.79)	0.01
Best time (s)			< 0.02*
≥ 4.0	Referent	—
< 4.0	1.95 (1.09 to 3.48)	0.02
Direction of turn			0.14
Both	Referent	—
Left versus both	0.17 (0.02 to 1.83)	0.14
Right versus both	0.13 (0.01 to 1.45)	0.10
Right versus left	0.74 (0.45 to 1.22)	0.24
Time for first warmup (min)			0.01*
< 1	0.34 (0.14 to 0.79)	0.01
1 to 2	0.70 (0.39 to 1.27)	0.24
3 to 5	Referent	—
6 to 15	1.31 (0.66 to 2.64)	0.44
16 to 30	0.14 (0.01 to 1.29)	0.08
Dog weight:height ratio (pounds:inches) quartiles			< 0.01*
1st (0.73:1.357)	Referent	—
2nd (1.356:1.6842)	0.70 (0.30 to 1.64)	0.41
3rd (1.6843:2.083)	0.91 (0.39 to 2.15)	0.84
4th (2.084:3.11)	1.11 (0.49 to 2.50)	0.80
Injury in previous years			< 0.01*
No	Referent	—
Yes	3.10 (0.93 to 4.00)	0.04
Interaction of weight:height quartiles and injury in previous years			< 0.01*
1st * No	Referent	—
2nd * Yes	0.25 (0.05 to 1.31)	0.10
3rd * Yes	4.55 (0.97 to 21.32)	0.05
4th * Yes	0.80 (0.19 to 3.31)	0.76
Multiple dogs in the house			0.11†
No	Referent
Yes	0.58 (0.30 to 1.13)	0.11
Handler experience (y)			0.58†
< 3	Referent	—
3 to 4	0.49 (0.17 to 1.44)	0.19
5 to 10	0.54 (0.20 to 1.48)	0.23
> 10	0.50 (0.19 to 1.34)	0.75
Random effect — owner	Estimates	P-value
Owner level SD (95% CI)	0.0093 (0 to 8)	0.999

^*Significant in the final model.

^†Confounding variable.

The time spent doing the first warm-up before competing in flyball was significant to the model; however, no distinct trends were observed. Dogs with < 1 min or 16 to 30 min of warm-up had significantly lower risk of injury than dogs with 3 to 5 min of warm-up.

The relationships between dog weight:height ratio quartiles (Q1–Q4) and previous flyball injury are shown in Figures 1 and 2. Generally, thinner dogs (Q1) and stockier dogs (Q3 and Q4) had higher risk of injury in the past year if they had had a previous injury, compared to those that did not have a previous injury (Figure 1). When comparing dogs that had previous injury in different weight:height quartiles, Q2 dogs had a lower risk of injury compared to stockier dogs (Q3 and Q4) (Figure 2). For dogs not injured in previous years, there were no significant differences in risk of injury in the past year for any weight:height quartiles (data not shown).

Figure 1.

Odds ratios (OR) and 95% confidence intervals (CI) for risk of flyball injury in the past year, shown for the interaction between injury in previous years and dog weight:height ratio quartile. Odds ratios > 1.0 in gray cells indicate that previously injured dogs had a significantly higher risk of injury in the past year than dogs without injury in previous years for that weight:height quartile comparison. White cells indicate comparisons that were not significant.

Figure 2.

Odds ratios (OR) and 95% confidence intervals (CI) for risk of injury in the past year shown for dogs injured in previous years by dog weight:height quartiles. Odds ratios > 1.0 in gray cells indicate that previously injured dogs in the row above had a significantly higher risk of injury in the past year than those in the weight:height quartile in the column on the left. White cells indicate comparisons that were not significant.

Owners reported that 28.5% of dogs achieved best times < 4.0 s. These dogs had a higher risk of injury than dogs with best times ≥ 4.0 s (OR: 1.95, P = 0.02) (Table 5). Age was also significantly associated with risk of injury. Middle-aged (3 to 7 y, OR: 4.08, P = 0.01) and older (≥ 8 y, OR: 4.51, P = 0.01) dogs all had a higher risk of injury than dogs 2 y of age or younger (Table 5).

Most (69.8%) dogs competed in 1 or more other sports including agility (36.5%), disc dog (22.4%), dock diving (21.7%), nosework (12.7%), obedience (12.2%), rally obedience (12.1%), and others (0.5 to 6.5%). Canadian dogs consistently had a higher risk of injury in the past year compared to United States (US) dogs, with 30.7% of Canadian dogs injured versus 20.1% of US dogs. Canadian dogs competing in other sports had a lower risk of injury than Canadian dogs that did not [OR: 0.32, 95% confidence interval (CI): 013 to 0.83]. There was no difference in injury risk between US dogs that did or did not compete in other sports (data not shown).

Discussion

This is the first study of flyball injuries in dogs that includes data from Canada and comparison with dogs in the United States. The 34.1% overall incidence of career injuries in North American dogs in our study was similar to those reported in previous studies of flyball (39%) and agility (32 and 33%), and higher than was reported in a small canicross study (22%) (4,6,8,10). It is important to note that definitions of injury differed among studies. We hypothesized that many injuries incurred in flyball would be repetitive strain injuries and that some chronic conditions (such as osteoarthritis or IVDD) might occur more frequently in or be aggravated by dogs participating in flyball. We chose a definition of injury encompassing multiple injury possibilities. In contrast, another recent flyball survey defined injury more narrowly (occurring during participation in flyball) (4). One agility survey defined injury as something causing the dog to leave practice or discontinue a competition (7). This definition might not adequately capture mild soft tissue or repetitive strain injuries not readily apparent during an event, particularly during short flyball run times.

The most common breed or breed cross in our study was the border collie cross. In many cases, these are purpose-bred crosses of various combinations of border collie, Jack Russell terrier, whippet, and Staffordshire terrier. Mixed or unknown breeds were most common in another recent flyball survey (14). Unlike in previous agility studies (5,8), in which border collies had increased risk of injury, our study found no increased injury risk in dogs with border collie breeding. Border collie breeds may be predisposed to agility injuries due to the requirements of the sport (numerous quick turns and changes of direction) in which they excel. The uniform nature of flyball courses may not exacerbate injury risks between breeds.

Sites of injury were similar to those reported previously, with the most common being digit or footpad injuries in both studies, followed by back, shoulder, and iliopsoas/groin injuries in this study compared to shoulder, neck/back/tail, and iliopsoas injuries in the other study (4). Carpal injuries were uncommon in our study, comprising only 1.9% of injuries. The most common types of injury were sprains/strains, pain, and tears. We recognize that overlap exists between these categories, as severe sprains and strains are tears. Although “pain” is not a specific diagnosis, we presumed that a specific diagnosis would not be known in all cases. Therefore, various terms were included in the survey so that owners could pick the term best describing their dog’s injury.

Rate of return to flyball after injury in the past year (74.8%) was similar to that for canicross (77.6%) (10), and higher than that for agility (67.4% after injury, and 65% after TPLO surgery) (15,16). An additional 17.0% of flyball-related injuries required rest or rehabilitation at the time of this survey; it is unknown how many of these dogs would eventually have returned to competition.

The duration of first warm-up was significantly associated with risk of injury in the past year, although no meaningful trend was discernable. In human athletes, it is well-recognized that warm-up protocols should be modified based on environmental conditions such as temperature and humidity (17). For example, a warm-up protocol that is appropriate in cold temperatures might cause overheating in hot conditions. Environmental conditions or other factors could affect the association between injury risk and duration of first warm-up. Although it was significantly associated with injury, first warm-up time may be associated with some other unmeasured variables.

Although not significant in the final model, it is notable that 5 of 6 dogs that turned both directions on the box had been injured in the past year. The survey did not identify if the dogs had always turned both directions or if they switched the direction of their turn after injury (such as after a unilateral injury, in order to avoid pain). If these dogs turned both directions prior to injury, this could indicate a problem with training. Once handler experience and owning multiple dogs were included in the model as confounding variables, direction of turn was no longer significant. This may indicate that handler experience or experience competing with multiple dogs had an influence on direction of turn, which could support the training-problem theory.

Dogs that had been injured in previous years generally had an increased risk of injury in the past year; however, this effect was modified by the interaction with weight:height ratio quartiles. Previous injury was also found to be a significant risk factor for injury in agility (5). It is interesting that weight:height Q1, Q3, and Q4 dogs with previous injury had increased risk of injury in the past year compared to Q2 dogs. Perhaps Q2 dogs are in ideal condition for flyball, having sufficient muscle mass for optimal strength, without being overconditioned. Note that weight:height quartiles are distinct from body condition scores, which might be another variable worth investigating in future studies. A previous study reported median body condition scores of 4 with a range of 1 to 6 on a 9-point scale in competitive flyball dogs (14); therefore, it may be challenging to find significant numbers of overweight and obese dogs for studies of typical competitive flyball populations.

Dogs with best times < 4.0 s had a higher risk of injury in the past year than those with slower times, which is not surprising as the fastest dogs may be hitting the box with greater force. These may also be very high-drive dogs that are less likely to show signs of pain with mild injuries and may continue to participate until such injuries become more severe and are noticeable. This finding was not consistent with a previous flyball study, in which no association between best time and injury risk was reported (4). These differences might be explained by differences in the populations of dogs studied or by differences in study design. We asked respondents to select the category of their dog’s best time, whereas Montalbano et al (4) asked respondents to list their dog’s best time, which would have allowed for generation of continuous data as opposed to the categorical data that we used.

Dogs ≤ 2 y old had fewer injuries in the past year than middle-aged and older dogs. These dogs may not yet have developed repetitive strain injuries or chronic conditions, or their clinical signs may not yet be significant enough to be noticed by owners. Montalbano et al (4) reported that age was significantly associated with injury risk, with the mean age of injured dogs and uninjured dogs being 4.7 and 6.2 y, respectively. It is important to note that Montalbano et al (4) looked at previous injury at any time during the dog’s career, whereas we looked at injury that occurred in the past year; therefore, the results are not directly comparable.

As Canadian dogs generally had a higher risk of injury in the past year than US dogs, there may be some Canada-specific factors affecting the injury risk. Due to the colder climate, Canadian dogs may train and compete more often indoors, typically on rubber mats over concrete flooring, compared to outdoor turf footing. Type of footing was not associated with increased injury risk in a recent survey of US dogs (4). Differences in climate between countries may also influence injury risk, with colder temperatures possibly predisposing dogs to more injuries. Despite the use of indoor locations in northern regions, some commonly used facilities, such as hockey arenas, have limited ability to control temperature.

Canadian dogs that competed in other sports had a lower risk of injury than those that did not. Conversely, there was no reduction in injury risk with participation in other sports in US dogs. Such differences may be attributable to differences in climate. For example, during Canadian winters, many dogs have limited outdoor exercise. Perhaps those dogs participating in other sports, which often take place indoors during the winter, maintain better condition with subsequently reduced injury risk. Conversely, in the warmer US climate, perhaps dogs spend more time outdoors and are better able to maintain their condition year-round without specific sports training.

A primary limitation of our study was the inability to discern if reported injuries were directly associated with flyball. Unlike agility, in which injuries associated with specific pieces of equipment are common (6), it is difficult to visualize injuries during a flyball run, except in cases of collisions. We used a very broad definition of injury to include those that are not immediately apparent during flyball, such as repetitive strain injuries. This may have resulted in the inclusion of injuries not directly associated with flyball. In the case of chronic conditions such as osteoarthritis or IVDD, it is unclear if flyball participation is associated with developing or exacerbating these conditions. Most reported injuries were diagnosed by owners or paraprofessionals and not by veterinarians; therefore, the different sources of injury diagnosis may be an additional source of bias. There is also a risk of selection bias towards owners with fewer dogs due to the time required to complete the survey for large numbers of dogs. Owners with injured dogs may have been more likely to complete the entire survey. There is also a chance of misclassification due to recall bias, particularly in the reports of injuries in previous years. Underreporting of previous injuries is also possible, as the reported number of injuries in all previous years was only 120 (compared to 159 reported injuries in the past year alone). Finally, as a cross-sectional study, the resulting odds ratios for explanatory variables with the outcome of injury in the past year are simply associations. It is not possible to infer a causal relationship between these variables and injury from this type of study.

This study provided new information on the demographics and risk factors for injury in North American flyball dogs and represents the first report of Canadian data. The incidence of injury in flyball dogs in our study was comparable to other reports. The most common injury sites were the paws/digits, back, shoulder, and iliopsoas/groin. There was no increased risk of injury with border collie breeding. Injury in previous years, modified by weight:height ratio, was a significant risk factor for injury in the past year. Dogs > 2 y of age had increased risk of injury compared to younger dogs. Dogs with best times < 4.0 s had increased risk of injury compared to dogs with times ≥ 4.0 s. Canadian dogs generally had increased risk of injury compared to US dogs; however, an interaction existed with participation in other sports, which reduced risk of injury in Canadian dogs. These results will help veterinarians target clinical examination of injured flyball dogs, as well as guide further investigations into the differences in flyball training, cross-training, and competition between the 2 countries. Future studies should also investigate other factors associated with increased injury risk, including possible injury prevention strategies.

Complete statistical methods, results, and survey questions are available on request from the corresponding author.