Gastric dilatation-volvulus is associated with Poodle breeds, increased body size, and male sex, but not primary diet type or anxiety in the Dog Aging Project cohort

Abstract

Objective

The purpose of this study was to update prevalence and incidence data for gastric dilatation-volvulus (GDV) in companion dogs in the US and evaluate potential risk factors based on data from the Dog Aging Project.

Methods

A nested case-control study was performed within data collected from the 47,444 dogs enrolled in the Dog Aging Project from December 16, 2019, to December 31, 2023. Gastric dilatation-volvulus cases were selected by identifying dogs for which the owner had reported “bloat with torsion/GDV” or a description of GDV in the free text; control dogs without reported GDV history were selected at a 2.2:1 ratio.

Results

170 dogs had an owner-reported GDV event. A total of 374 dogs were selected as a control population. The lifetime prevalence of GDV was 0.23%, and the incidence was 1.74/1,000 dog years at risk. Compared to controls, dogs at increased odds of GDV included those that were purebred, were male, had a body weight > 40 kg, had a low body condition score, and were Poodles/Poodle mixes. There was no increase in GDV odds associated with fearful/anxious behavior, neuter status, age at neuter, diet type, or number of meals fed per day.

Conclusions

Identification of Poodles as the breed with the highest occurrence of GDV was novel. Increasing body size, low body condition score, and male sex were confirmed as risk factors for development of GDV, but this analysis failed to identify increased anxiety or diet as risk factors.

Clinical Relevance

Early identification of dogs at highest risk for GDV facilitates conversation with owners about prophylactic gastropexy and clinical signs of GDV.

Gastric dilatation-volvulus (GDV) is a life-threatening condition characterized by stomach malposition, rapid accumulation of air in the stomach, increased intra-abdominal pressure, cardiogenic shock, and, sometimes, death.¹ Weakened intra-abdominal ligaments are hypothesized to allow the stomach to rotate 90° to 360°, preventing normal egress of gas from the cardia or pylorus.² Previous studies^3–5 have reported that GDV accounts for 0.13% to 0.76% of all veterinary hospital admissions, or an estimated 400,000 to 900,000 pet dogs/year in the US. Despite continued advances in medicine and surgery, GDV case fatality has been reported^1,3,5 to range from 28.6% to 49%. Gastropexy surgery has been shown to be > 95% effective in preventing GDV, both when performed prophylactically and when performed after GDV to prevent recurrence^6,7; a thorough understanding of risk factors could help veterinarians recommend this potentially life-saving surgery to high-risk patients.

Multiple previous studies, both in the US and Europe, have sought to identify risk factors for developing GDV. Published risk factors include breed (Great Dane most commonly cited, with Weimaraner, German Shepherd Dog, Akita, Irish Setter, and Bloodhound also identified in various studies^{1,3–5,8–11}), increasing age,^{1,3–5,9,10,12,13} increased fear/anxiety,^1,10,11,14 increased fat in the diet,² increased thoracic depth-to-width ratio,^1,13 low body condition score,^2,11,14 and history of GDV in a first-degree family member.^1,13,14 However, most of these studies were conducted with specific populations, namely purebred show dogs or military working dogs; few were conducted with large numbers of typical companion dogs in the US.

The Dog Aging Project (DAP) is a multi-institutional longitudinal study that currently encompasses approximately 50,000 companion dogs in the US.¹⁵ The DAP seeks to improve our understanding of canine diseases and how the risk and impact of those diseases change throughout a dog’s lifetime. Utilizing data from DAP participants, this study sought to provide a current, updated prevalence and incidence of GDV in companion dogs in the US. In addition, this work evaluated whether previously suggested risk factors (including age, sex and castration status, breed, behavior, and diet) increased the odds of GDV in the DAP population.

Methods

The DAP is a community science, long-term longitudinal study of the biological and environmental determinants of healthy aging in companion dogs in the US.¹⁵ Privately owned dogs of any age living in the US are eligible for enrollment in the DAP. Owners nominate their dogs through the DAP website. After giving informed consent, owners complete the 10-part Health and Life Experience Survey (HLES), which collects information about the dog’s signalment, behavior, environment, physical activity, diet, and health history. The participants are then considered members of the DAP Pack. On or near the anniversary of completing the HLES, DAP Pack members are asked to complete the Annual Follow-Up Survey (AFUS), a 10-part survey that updates all prior HLES responses, including capturing any new diagnoses over the past year. Owners are also asked to upload veterinary electronic medical records (VEMRs) annually, starting at the time of enrollment. Following the death of a DAP dog, the owners are invited to fill out the End of Life Survey (EoLS), in which the owners report on timing and causes of death, concurrent diseases, and more. The DAP survey instruments are freely available online.¹⁶

The University of Washington Institutional Review Board deemed that recruitment of dog owners for the DAP, and the administration and content of the DAP HLES, are human subjects research that qualifies for category 2 exempt status (Institutional Review Board Identification No. 5988, effective October 30, 2018). No interactions between researchers and privately owned dogs occurred; therefore, IACUC oversight was not required.

The case-control study reported here is nested within the larger DAP cohort study and utilized data collected from dogs enrolled in the DAP Pack from December 26, 2019, to December 31, 2023, included within the 2023 Curated Data Release. The DAP is an open data project, and data are housed on the Terra platform at the Broad Institute of the Massachusetts Institute of Technology and Harvard University.¹⁷

Case selection

Gastric dilatation-volvulus cases were selected from the HLES by identifying dogs in which the owners had indicated yes to medical history of the gastrointestinal system and selected bloat with torsion/GDV from the list of specific options. In addition, the free text associated with the option other was analyzed to determine whether any owners had written in descriptors of GDV including gastric/stomach torsion, flipped stomach, and stomach twist. The process was repeated for the AFUS and EoLS to determine whether any previously enrolled dogs developed a GDV following enrollment.

Control dogs were selected by first removing case dogs from the total population. From this population, controls were randomly selected without replacement by use of a 2.2:1 ratio of control-to-case, while also selecting the same ratio of dogs with and without uploaded VEMRs as was in the case population. The previously mentioned selection criteria was implemented to ensure that the selection of controls was in accordance with epidemiological recommendations^18,19 and produced an appropriately representative control population from randomization.^19,20

Variable selection and description

Potential risk factors for GDV evaluated in previous studies^{1,3–5,8–13} include increasing age, breed, increasing body size, low body condition score, increased fear/anxiety, feeding 1 meal/day, and increased fat in the diet. Data relevant to these characteristics were taken from the survey during which the GDV was reported.

Both age and life stage (according to American Animal Hospital Association guidelines²⁰) were collected for each dog. For all dogs, date of birth was collected from owner-reported dog demographics completed at the time of enrollment. If the exact date was not provided, but month and year of birth were known, date of birth was recorded as the first of the reported month. If the owner had reported an estimated age at the time of HLES completion, the year of birth was calculated by use of that estimated age and the date of birth was set as the 15th of the month during which the survey was completed. Age at the time of owner-reported GDV was calculated based on these exact or estimated birth dates. To capture breed, if purebred was indicated, owners were provided with a list of American Kennel Club–registered breeds to select from or a free-text option to write in non–American Kennel Club breeds. If mixed breed was indicated, owners were prompted to select the dog’s primary and secondary breeds. To capture body condition score, weight or “expected adult weight” (divided into 10-lb brackets), “ever been considered underweight,” and “ever been considered overweight” responses were collected for each dog. To capture fear or anxiety, results from the Fear and Anxiety section and Separation-Related Behavior section of the modified Canine Behavioral Assessment and Questionnaire portion of the DAP surveys were collected for each dog.²¹ To capture diet factors, the primary diet component was collected for each dog. The primary diet component types were grouped into the following: kibble (a commercially prepared extruded dry diet), home-prepared (included diets that were either cooked or raw, as well as those prepared at home from a commercial kit), canned (a commercially prepared high-moisture diet), freeze-dried (a commercially prepared freeze-dried diet), refrigerated raw (a commercially prepared diet that has not been cooked and requires refrigeration), refrigerated cooked (a commercially prepared diet that has been cooked and requires refrigeration), and semi-dry (a commercially prepared extruded diet with higher moisture content than kibble but less than canned). To capture feeding pattern, responders were asked whether their dog was fed once or more than once daily, in reference to a prior study²² that indicated feeding once daily as a risk factor for GDV.

Medical records review

The number of dogs with and without VEMRs uploaded by the owner was tallied. For case and control dogs with available VEMRs, records were reviewed by authors JR and TR for evidence of the presence (cases) or absence (controls) of diagnosis of GDV. Submitted VEMRs did not necessarily include the time period in which the GDV was reported or all possible clinics where a given dog may have received veterinary care. In such cases, GDV could neither be specifically confirmed nor refuted and owner-reported information (GDV or lack of GDV) was treated as correct. If the VEMRs provided direct evidence contrary to the owner’s report (evidence of an alternate diagnosis at the time of owner-reported GDV for cases, evidence of a GDV diagnosis at any time for controls), the dog was removed from the total for that group. For both cases and controls, VEMRs were also reviewed for evidence of splenectomy or gastropexy prior to the occurrence of GDV, so that these reported risk and protective factors, respectively, could be analyzed.

Statistical analysis

Summary statistics and general linear models were performed with RStudio (version 4.4.0; Posit Software PBC), dplyr (version 1.1.4; Wickham, François, Henry, et al),²³ and tidyverse (version 2.0.0; Posit Software PBC).²⁴ For the case-cohort population, the prevalence of GDV in HLES was calculated by taking the number of reported GDV’s in HLES divided by the total number of dogs in HLES and then multiplying by 100,000. To determine the incidence, only AFUS and EOLS reports were used. The total dog years at risk were calculated for all dogs with at least 1 follow-up survey. For the total population of dogs, the date of the latest AFUS completion was the end-date for each animal and the date of enrollment in DAP was the start date. Dogs whose survey date was “NA” were not included. For the dogs with reported GDV events, the owner-reported year and month of the GDV was the end date, with the 15th of each month assigned to be the “date” in each instance. Two dog owners had reported a GDV date at a follow-up survey that was prior to their enrollment date; these were excluded from the incidence calculation, but their data were used for the remainder of the analyses. The dog years at risk within the GDV population were divided by the dog years at risk in the total population without reported GDV events.

For the case-control population, a directed acyclic graph was utilized in identifying the most appropriate way to control for confounding variables in a causal analysis (Supplementary Figure S1). This was because some variables did not have identified confounders, while others had several, and it would be inappropriate to conduct multivariable analysis indiscriminately, as this introduces issues such as controlling for colliders, which can bias results to make an effect seem present where there is no true effect.

Several general linear models were run by use of logistic regression with the outcome of GDV status (yes or no) and the appropriate covariates within each model to determine the effect of sex, neuter status, breed status (purebred vs mixed breed), diet type, feeding frequency, previously overweight or underweight status, and behavior questions. Causal effects (covariates included) were sex (sex), breed status (breed status), neuter status (neuter status + breed status + sex), diet variables (diet variable + breed status + neuter status), overweight/underweight (variable + breed status + neuter status), and behavior variable (variable + breed status + diet type + neuter status + sex):

$$\log \left(\frac{P(GDV)}{1-P(GDV)}\right)={\beta }_0+{\beta }_1\left(covariate1\right)+\cdots$$

The coefficients of the estimates were then exponentiated and 95% CIs calculated. The behavior/anxiety P values were adjusted with the Benjamini-Hochberg method to account for multiple comparisons.

Results

Population

In the 2023 DAP data release, 47,444 dogs were enrolled in the DAP Pack (completed HLES between December 26, 2019, and December 31, 2023). A total of 170 dogs were reported to have a GDV event in total, with 114 identified on the initial HLES, 27 identified on AFUS, and 29 identified on EoLS. A total of 52 of the GDV dogs had a medical record uploaded and reviewed (30.6% of total GDV population). Upon medical record review, GDV was confirmed for 46 dogs (27% of total GDV population) and refuted for 6 dogs (3.5% of total GDV population). Thus, the total GDV population utilized was 164 dogs. From the HLES, the lifetime prevalence of GDV in the DAP population was 230/100,000 dogs (0.23%). Using AFUS and EoLS, the incidence rate for GDV in the DAP population was 1.7364/1,000 dog years at risk. A total of 374 dogs were randomly selected according to the previously described inclusion criteria to be the control population.

Prior medical history

Among the 46 GDV dogs with VEMRs, no dogs had evidence of a prior gastropexy and 2 dogs had evidence of a prior splenectomy. Among the control dogs, 1 dog had evidence of a prior gastropexy and no dogs had evidence of a prior splenectomy. No statistical analysis was conducted on these data.

Signalment

Dogs with a reported GDV diagnosis were more likely to be male (62%) than control dogs were (48%; OR, 1.76; 95% CI, 1.22 to 2.58; P < .05; Table 1). Spay/neuter status and age at spay/neuter was not different between dogs with and without a reported GDV event (P = .1; P = .857).

Table 1—

Selected demographic characteristics of owner-reported gastric dilatation-volvulus (GDV) cases (164 dogs) and controls with no owner-reported GDV event (374 dogs).

	No. of GDV dogs	No. of control dogs
Breed type
Purebreda	110 (37%)	187 (63%)
Mixed breed	54 (22%)	187 (78%)
Sex
Maleb	102 (36%)	181 (64%)
Female	62 (24%)	193 (76%)
Neuter status
Intact	17 (53%)	15 (47%)
Neutered	147 (29%)	359 (71%)
Dog weight range (kg)
0–9.9	7 (33%)	14 (67%)
10–19.9	13 (39%)	20 (61%)
20–29.9	60 (25%)	176 (75%)
30–39.9	49 (30%)	114 (70%)
> 40	35 (43%)	47 (57%)
Unknown	0 (0%)	3 (100%)
Previously underweight
Never underweight	99 (29%)	243 (71%)
Was underweight	39 (49%)	41 (51%)
Unknown	6 (15%)	33 (85%)
N/Ac	20 (26%)	57 (74%)
Previously overweight
Never overweight	115 (36%)	201 (64%)
Was overweight	23 (18%)	106 (82%)
Unknown	6 (38%)	10 (62%)
N/Ac	20 (26%)	57 (74%)
Dog primary purposed
Companion	157 (30%)	357 (70%)
Obedience	2 (67%)	1 (33%)
Show	0 (0%)	1 (100%)
Working	0 (0%)	3 (100%)
Service	0 (0%)	4 (100%)
Breeding	0 (0%)	1 (100%)
Unknown	5 (42%)	7 (58%)

Data as reported by owners from December 2019 to December 2023 in the Dog Aging Project.

N/A = Not applicable.

^aCases were significantly more likely to be purebred than controls were (P < .05).

^bCases were significantly more likely to be male than controls were (P < .05).

^cThe diet portion of the Health and Life Experience Survey was redesigned in 2022. Owners that completed the Health and Life Experience Survey before this revision were not asked about previous history of being overweight or underweight.

^dThere were no dogs in either the GDV or non- GDV group with a reported primary purpose of agility or hunting, although those were response options presented.

The weight brackets with the most GDV dogs were 20 to 29.9 kg (60 of 164 [36.6%]) and 30 to 39.9 kg (49 of 164 [29.9%]). The weight bracket with the highest percentages of GDV patients when compared to the weight-matched control group was > 40 kg (43% of all dogs > 40 kg). Seven GDV cases (4.3%) were reported in dogs < 10 kg and 13 GDV cases were reported in dogs < 20 kg (7.9%; Table 1). Dogs with a reported history of being underweight were more likely to have a GDV episode. Thirty-nine GDV dogs (23.8%) had a known history of being underweight, compared to 41 control dogs (11%). Dogs with a reported history of being overweight were less likely to have a GDV episode. Ninety-nine GDV dogs (60.4%) had a known history of being overweight, compared to 243 control dogs (65%).

The age at GDV event ranged from < 1 year (3 of 164 [1.83%]) to 16 years (2 of 164 [1.22%]), with a median of 9 years. The age at which the most GDV events occurred was 11 years (21 of 164 [12.8%]), followed by 9 years (16 of 164 [11.6%]; Figure 1). The dogs were not analyzed in life stage brackets (puppy, young adult, mature adult, senior²⁰), as the age definition of these brackets is not the same for all breeds and sizes of dogs.

Figure 1—

Age (years) at time of gastric dilatation-volvulus (GDV) episode for 164 dogs with owner-reported GDV events (Dog Aging Project, December 2019 to December 2023). *2 dogs had an incorrect date of GDV episode recorded prior to their date of birth. The true date of their GDV episode and age at episode are unknown.

Greater than 95% of the dogs included in this study had a primary purpose of being a companion. No dogs whose primary purpose was listed as agility or hunting were present in this study. No dogs whose primary purpose was listed as show, working, or service had a reported GDV event. Two dogs with a primary purpose listed as obedience had a reported GDV event. Overall, dog primary purpose was not tested for significance, as the groupings did not meet the criteria for a χ² test (Table 1).

Gastric dilatation-volvulus events were significantly more common among purebred dogs compared to mixed-breed dogs (OR, 2.05; 95% CI, 1.4 to 3.0; P < .05; Table 1). Among purebred dogs, Poodles were the breed with the highest number of GDV events (n = 26), followed by German Shepherd Dogs (n = 15). Excluding dogs that did not have a breed-matched control population for comparison, the breeds with the highest proportion of GDV were Poodles (26 of 35 [74%]) and Great Danes (8 of 12 [67%]). In the overall DAP population, Labrador Retrievers are the most common purebred represented (n = 2,723), followed by Golden Retrievers (2,472), German Shepherd Dogs (1,130), Australian Shepherds (844), and Poodles (822; Supplementary Table S1).

Among mixed-breed dogs, Poodle mixes had the highest number of GDV events (n = 15), followed by Labrador Retriever mixes (n = 11). Excluding dogs that did not have a breed-matched control population for comparison, the primary breeds with the highest proportion of GDV were Saint Bernard mix (2 of 3 [67%]) and Poodle mix (15 of 24 [60%]; Supplementary Table S2).

Fearful/anxious behavior

There was no difference in the median scores of GDV and non-GDV dogs in their frequency of hyperactivity, agitation, barking, chewing, or energy or their fear level when approached by unfamiliar people or unfamiliar situations or in response to loud noises. The non-GDV dogs had a higher median fear level when approached directly by unfamiliar dogs, but this difference was not statistically significant (P = .063; Table 2).

Table 2—

Median scores from fear/anxiety questions on the behavior survey for GDV and control dogs, as reported by their owners from December 2019 to December 2023 in the Dog Aging Project.

	Median score
	GDV dogs	Control dogs
Fear level
When approached directly by an unfamiliar person while away from your homea	0	0
When an unfamiliar person tries to touch or pet the dog	0	0
In response to strange or unfamiliar objects on or near the sidewalk	0	0
When approached directly by an unfamiliar dog	0b	1b
When first exposed to unfamiliar situations	1	1
When barked, growled, or lunged at by an unfamiliar dog	1	1
In response to sudden or loud noises	1	1
Frequency
Hyperactive, restless, has trouble settlingc	0	0
Restless, agitated, pacing	0	0
Barking or whining	0	0
Chewing or scratching at doors, furniture, etc	0	0
Active, energetic, always on the go	2	2
Total Score	5	6

^aFor fear-level prompts, owners had options 0 (no visible signs of fear) to 2 (mild-moderate fear/anxiety) to 4 (extreme fear).

^bNot a significant difference (P = .063).

^cFor frequency prompts, owners had options 0 (never), 1 (seldom), 2 (sometimes), 3 (usually), and 4 (always).

Diet

The most common diet type for both GDV and non-GDV dogs was kibble, representing over 75% of the total population of dogs (Table 3). There was not a significant increase in GDV odds in dogs with a primary diet of kibble (OR, 0.63; 95% CI, 0.25 to 1.67; P = .33). The majority of both GDV and non-GDV dogs were fed more than once daily (93.2% and 90.6%, respectively). Dogs eating 1 meal/day were not significantly more likely to have a GDV event (OR, 1.55; 95% CI, 0.72 to 3.66; P = .29).

Table 3—

Primary diet type and feeding schedule for case and control dogs, as reported by their owners from December 2019 to December 2023 in the Dog Aging Project.

Primary diet component	No. of GDV dogs	No. of control dogs
Kibble	128 (31%)	290 (69%)
Home-prepared	13 (30%)	31 (70%)
Canned	8 (40%)	12 (60%)
Freeze-dried	2 (22%)	7 (88%)
Refrigerated raw	7 (39%)	11 (61%)
Semi-dry	2 (33%)	4 (67%)
Refrigerated cooked	1 (33%)	2 (67%)
Unknown	3 (15%)	17 (85%)
Once daily	9 (24%)	28 (76%)
More than once dailya	153 (31%)	339 (69%)
Unknown	2 (22%)	7 (78%)

^aEncompasses dogs that were fed 2 to 6+ times daily.

Discussion

This nested case-control study was one of the largest evaluations of risk factors for GDV in companion dogs in the US, and it uniquely included data from a diverse group of owners (eg, those living in rural environments) that are often underrepresented in studies conducted at secondary or tertiary veterinary referral facilities. Poodles and Poodle mixes were newly identified as the breed with the highest occurrence of GDV. Purebred status, increased body size (> 40 kg), male sex, and a history of being underweight were confirmed as risk factors for development of GDV. In contrast, this analysis failed to identify increased GDV odds associated with fearful or anxious behavior, neuter status, age at neuter, primary diet type, or number of meals fed per day.

The lifetime prevalence of GDV in the DAP Pack, or proportion of dogs in this population that have experienced this particular condition at any point in their lives, was 0.23%. Previous studies^3–5 have reported GDV point prevalences, or proportions of individuals in a population that are experiencing a particular condition at a specific time, ranging from 0.13% to 0.76% of hospital admissions. It had previously been suggested that studies selecting participants from only emergency or referral hospitals could falsely elevate the prevalence of GDV.^4,5 This study, which included owner-reported GDV diagnoses and diagnoses from medical histories from primary veterinarians, emergency hospitals, and referral institutions, reported prevalence within the previously reported range.

As a longitudinal study, the DAP provides a unique opportunity to collect ongoing information from participants every year after enrollment. In the current study, 56 dogs developed a new GDV after initial enrollment in the DAP, resulting in an incidence rate of GDV of 1.74/1,000 dog years of risk. In the 2000 study by Glickman et al,¹ the incidence of GDV in high-risk large- and giant-breed dogs, respectively, was noticeably higher at 23 and 26 cases/1,000 dog years of risk.

Increasing body weight has previously been reported as a risk factor for developing GDV.^3,5,11 The results of this study agree with prior findings; the weight bracket with the highest percentage of GDV dogs when compared to controls was > 40 kg. Body condition score, however, has been debated as a possible risk factor for GDV development.^3,9,11,14,22 In the DAP population, GDV dogs were less likely to have a history of being overweight and more likely to have a history of being underweight than control dogs. It has been suggested that the increased odds of GDV in underweight dogs is associated with poor GI track health and altered GI motility.²² In overweight dogs, it is possible that the increased abdominal fat fills space in the abdomen, decreasing the ability of the stomach to rotate. It is possible that the overall increased odds with increased weight are therefore associated with specific breeds or specific body conformations, as increasing body mass alone does not increase odds of GDV. While it is beneficial to emphasize the importance of appropriate nutrition in growing large-breed dogs to prevent low body condition, it is important to note that obesity has many, significant consequences and weight gain should not be pursued as a preventative strategy for GDV.

Gastric dilatation-volvulus has historically been described as a disease of aging, with the risk of developing GDV increasing 20% to 33% with each year of age.^{1,3,5,9,11–13} In agreement with previous data, the most common ages at which GDV episodes occurred in the DAP population were 9 to 11 years, which would constitute the senior life stage for large- and giant-breed dogs. A smaller number of GDV cases were reported in dogs 12 to 16 years old, but few large- and giant-breed dogs live into their teen years. In addition, in this study, a high number of GDV episodes were reported in younger dogs. Forty-eight GDV episodes (48 of 164 [29%]) were reported in dogs < 5 years old, including 14 episodes in dogs < 2 years old (14 of 164 [8.5%]). Fifteen of these young GDV cases were confirmed by VEMR review, including 4 in dogs < 2 years old. Multiple studies^1,3,8,13,23 over the last 3 decades have reported small numbers of cases of GDV in “young” dogs, ranging from 5 months to 3 years old. The occurrence of GDV cases in juvenile dogs suggests a different etiology that is developmental rather than degenerative.

In the DAP Pack, males were at a significantly increased odds of developing GDV compared to females, in agreement with the majority of previous research.^{1,3,4,10,11,14} It has been historically debated whether neuter status is protective against GDV^5,10,11; in the current study, neuter status did not significantly alter the odds of developing GDV. It is possible that the increase in odds of GDV in males is hormonally mediated, as testosterone alters muscle density and metabolic rate.²⁴ However, it is also possible that the increased prevalence of GDV in male dogs is due to increased body size and body conformation; male dogs are often larger than females.

Multiple previous studies^3,5,11,13,25 have documented an increased risk of GDV in purebred dogs, ranging from 1.5 to 4.8 times the risk. Similarly, in the current study, purebred dogs had twice the odds of developing GDV than mixed-breed dogs. While it is possible that mixed-breed dogs may be protected by “hybrid vigor,” purebred dogs are also more likely to have an extreme size or shape.²⁵ Similar to previous data, the breeds at highest risk of GDV were large/giant-, deep-chested breeds, including Poodles, German Shepherd Dogs, and Great Danes. However, in contrast to earlier studies, Poodles had both the highest number of GDV events and highest proportion of GDV events when compared to control dogs. While Poodles have been mentioned on lists of breeds at risk of GDV, they have previously been near the bottom of those lists, representing a “lower” risk breed. German Shepherd Dogs have historically represented the dog breed with the most GDV events, as they are both a popular and high-risk breed.^4,10,11,26 Great Danes have historically represented the dog breed with the highest proportion of GDV events and the dog breed at highest risk of developing GDV when compared to control populations.^{1,3,5,7,11,12} Previous studies^3,5,7,8,26 have included Dogue de Bordeaux, Bloodhound, Gordon Setter, Grand Bleu de Gascogne, Otterhound, and Spinone Italiano as high-risk breeds for GDV; there were no dogs of any of these breeds with a reported GDV in the DAP Pack. Most of these dog breeds are uncommon, especially in the US.²⁷

To our knowledge, this was the first study that evaluated the odds of developing GDV in specific mixed-breed dogs with consideration to their primary breed composition. Similar to the findings in purebred dogs, Poodle mixes had the highest number of GDV events and a high proportion of GDV events when compared to the control population. The increased odds of Poodles and Poodle mixes developing GDV was a novel finding in this study. The high number of GDV cases in these dogs may be related to changes in breed popularity in the US in recent years/Poodles fell out of popularity in the early 2000s but reentered the American Kennel Club Top 5 Dog Breeds list in 2021 for the first time in over 10 years and have remained on that list²⁷ through 2023. The “Doodle” mixes, first introduced by guide dog organizations in the 1970s to 1980s, have continued to increase in popularity through the 2000s. In a 2022 analysis²⁸ of pet insurance claims by Nationwide, Doodle popularity had increased over 160% since 2013. It is also possible that the breed-specific prevalence of GDV has changed as GDV has become a more well-known medical condition. Owners of dogs with previously established risk factors may alter how they care for their dog or seek medical treatment more rapidly when signs begin, decreasing the occurrence of GDV in these dogs.

Previous studies^5,9–11,14 indicated that dogs described by their owners as fearful, anxious, or aggressive were more likely to develop GDV; it was theorized that increased anxiety was associated with increased gastrointestinal contraction or abnormal gastrointestinal motility. The Canine Behavior Assessment and Research Questionnaire has previously been validated as a reliable method of assessing behavior and temperament in dogs.²¹ Utilizing these prompts in the current study, there was not a significant difference in the fear and anxiety scores between GDV and non-GDV dogs.

It has historically been debated whether feeding a diet of exclusively dry, commercially prepared kibble increases risk of GDV; it was proposed that dry kibble could delay gastric emptying by distending and weighing down the stomach after feeding.^{2,9,10,14,22,29} In the current study, there was no difference in the primary diet type between GDV and non-GDV dogs. However, the majority of dogs in the DAP—over 75%—are fed kibble.³⁰ Furthermore, a previous study²² indicated that feeding a single meal per day increased the risk of a GDV episode. In the current study, there was no change in odds of a GDV episode between dogs that were fed once daily and those that were fed multiple times daily.

A previous study³¹ has identified history of splenectomy as a risk factor for GDV. In this study, only 1 control dog had evidence of a prior splenectomy identified in VEMRs. With this population, we are unable to comment on any correlation between splenectomy and GDV.

The current study identified a similar prevalence of GDV to previous studies, but a much lower incidence, and several previously identified GDV risk factors were not significant in this population. Previous studies were often conducted on smaller, specific populations of dogs (such as show dogs or working dogs) and, in particular, dogs that were already assumed to be at higher risk of GDV. The current study was conducted on a large, variable population of companion dogs. In addition, the DAP questionnaires are standardized and not tailored to a specific medical condition, decreasing the risk of survey bias and response framing. It is possible that the design of previous studies or populations available to analyze overemphasized small variances between GDV and non-GDV dogs.

There were several important limitations to this study. First, the DAP owner population underrepresents low-income households when compared to the overall population of the US.^30,32 As a survey-based study, the data were subject to recall and response bias. An owner may enroll their dog at any age, and therefore, they might remember past events incorrectly or fail to report them at the time of the survey. Furthermore, the results of this study were primarily reliant on dog owners correctly identifying a past medical event of their pet. Medical records were analyzed to confirm diagnosis of GDV, but medical records were only available for 32% of the total GDV population. Upon medical record review, GDV was refuted in 6 dogs, resulting in a known error percentage of 11.5%. The 6 refuted cases were excluded from further analysis, but both the confirmed and unconfirmed cases of GDV were included. Similarly, 2 dogs had a date of GDV reported prior to their date of birth, assumed to be an erroneous survey response. It is possible that additional GDV events also represent incorrect owner survey responses, rather than true GDV events. This study reported Poodles as the primary mixed breed with the highest odds of developing GDV. When completing the HLES survey, mixed-breed dog owners are instructed to select or guess their dog’s primary and secondary breed composition. It is possible that this survey design overemphasized Poodle mixes because, unlike many mixed-breed dogs with a difficult-to-distinguish breed composition, Doodles are mixed-breed dogs with known breeding.

This study reported an updated analysis on the current prevalence of, incidence of, and risk factors for development of GDV in companion dogs in the US. At this time, the exact cause of GDV is still unknown and, despite many studies over the last 50 years, the only consistent risk factors established are breed, larger body size, and male sex. It is unlikely a single factor is directly responsible for development of a complex condition, such as GDV.^9,13,33 Clear breed predispositions and episodes of GDV in juvenile dogs imply there is a genetic component involved in the development of GDV.³³ This theory is further supported by the previously established increased risk of GDV in first-generation family members.^9,11,13,33 However, it is more likely that the specific size, shape, and temperament of various dog breeds are all influential in the development of GDV. While Great Danes have historically been considered to be at the highest risk of GDV, this study showed that other, popular large-breed dogs, including Poodles, Labrador Retrievers, German Shepherd Dogs, and large mixed-breed dogs, are also at an increased risk of GDV development. In addition, this study identified cases of GDV in 20 small-breed dogs and 14 dogs < 2 years old. With these uncertainties in mind and the knowledge that the only consistently established risk factors for GDV are inherent and immutable, veterinarians should notify dog owners early about the risk and clinical signs of GDV and discuss prophylactic gastropexy as an effective preventative measure.

Supplementary Material

Supplementary materials are posted online at the journal website: avmajournals.avma.org.