Introduction

Primary dilated cardiomyopathy (DCM) is thought to have a genetic cause and is seen most commonly in large- and giant-breed dogs, such as Doberman Pinschers, Boxers, Great Danes, and Irish Wolfhounds.^1–6 Pit bull breeds (American Pit Bull Terriers, American Staffordshire Terriers, Staffordshire Bull Terriers, American Bulldogs, and American Bullies) and pit bull–type dogs (ie, not purebred or of unknown pedigree but considered to be a pit bull–type breed based on appearance) are not reported to be breeds commonly affected by primary DCM. A retrospective study² of 369 dogs with DCM included only 3 Staffordshire Bull Terriers (< 1%). Another retrospective study⁶ of 189 dogs with DCM included no pit bull–type breeds.

In addition to primary DCM, secondary forms of DCM can also occur in dogs as the result of drugs, toxins, nutritional deficiencies (eg, taurine and thiamine), or other diet-associated factors (eg, contamination of feed by monensin or heavy metals).^7–9 In 2018, the FDA issued an alert, followed by 3 updates, regarding new reports of what appears to be a secondary, diet-associated form of DCM.^10–13 This secondary DCM can affect dogs of breeds associated with primary DCM but also more atypical breeds.^10–17 In addition, it is associated with echocardiographic improvement and longer survival times after diet change compared to what is typically seen in primary DCM.^9–17 The specific cause of this diet-associated form of DCM is not currently known, but diets high in pulses (eg, peas, lentils, and chickpeas) and, to a lesser degree, potatoes and sweet potatoes, have been associated with secondary DCM; these ingredients are most commonly found in grain-free diets.^9–18

Recent studies on diet-associated DCM have included pit bull–type breeds.^12,14,16,17 For example, 1 retrospective study¹⁶ found that 3 of 43 (7%) dogs with DCM eating grain-free diets were pit bull–type breeds compared to none of the dogs with DCM eating grain-inclusive diets. A prospective study¹⁷ of dogs with DCM reported that 6 of 51 (12%) dogs with DCM eating nontraditional diets were pit bull–type breeds compared to none of the dogs with DCM eating traditional diets. It is unclear whether pit bull–type breeds are more susceptible to diet-associated DCM or are just more likely to be fed nontraditional diets. In addition, anecdotally, pit bull–type breeds with diet-associated DCM appear to have a good response to diet change if caught early, but there is not enough long-term follow-up on pit bull–type breeds yet.

Therefore, an objective of this study was to describe signalment, clinical signs, diet, echocardiographic findings, and outcome for pit bull–type breeds with DCM. A second objective was to determine whether echocardiographic improvement and survival are different between pit bull–type breeds with DCM eating nontraditional diets that changed diets, dogs eating nontraditional diets that did not change diets, and dogs eating traditional diets. The hypothesis was that pit bull–type breeds diagnosed with DCM eating nontraditional diets would show significant improvement with diet change.

Materials and Methods

For the purposes of this retrospective, multicenter study, pit bull–type breeds were defined as American Pit Bull Terriers, American Staffordshire Terriers, Staffordshire Bull Terriers, American Bulldogs, and American Bullies, as well as pit bull–type dogs (ie, not a purebred or if the dog was considered to be a pit bull–type breed by the owner or veterinarian). An online survey (Qualtrics XM) regarding DCM in pit bull–type breeds was posted to relevant Facebook groups and breed clubs and the American College of Veterinary Internal Medicine Cardiology listserv, asking cardiologists to notify owners of eligible dogs. The survey questions addressed information from the date of diagnosis, including dogs’ age, breed, weight, sex, clinical signs, echocardiographic measurements, diet, taurine concentrations, and cardiac medications, as well as information on diet changes, most recent echocardiographic measurements (if applicable), and outcome. The survey was available from May 31, 2022, through August 31, 2022, with 2 to 3 reminders emailed or posted to the original sources. Surveys were considered eligible for the study if the dog was diagnosed with DCM between 2015 and 2022 and had at least 1 echocardiogram with measurements at the time of diagnosis. Due to the limited number of owner-submitted cases, a request was posted to the American College of Veterinary Internal Medicine Cardiology listserv in July 2022, asking cardiologists to submit cases directly to the authors.

The study’s definition of DCM consisted of M-mode fractional shortening ≤ 25%, normalized left ventricular internal diameter in diastole (LVIDdN) ≥ 1.8, and normalized left ventricular internal diameter in systole (LVIDsN) ≥ 1.2.^14,15,17,19 Measurements of the left ventricle were available via M-mode only from some sites, M-mode and 2-D from some sites, and 2-D only from some sites. If > 1 measurement was available, M-mode measurements were preferentially used because they were the most common measurements available, except for the ratio of the left atrium to aorta in which the 2-D measurements were preferentially used. The same method was used for both echocardiograms in an individual dog. Dogs with tachycardia-induced cardiomyopathy and drug-induced cardiomyopathy (eg, doxorubicin and chloramphenicol) were excluded. Dogs with a diagnosis of DCM from the attending cardiologist based on echocardiography but whose echocardiographic measurements met 1 or 2 criteria but did not meet all 3 of the study’s echocardiographic criteria (fractional shortening ≤ 25%, LVIDdN ≥ 1.8, and LVIDsN ≥ 1.2) were enrolled in a separate group, called DCM-C.

Records were reviewed for each submitted case using a standardized form to collect data from the visit at which DCM was diagnosed for signalment, clinical signs, and echocardiographic measurements. The dog’s main diet (ie, the diet providing the majority of the dog’s calories) at the time of diagnosis and duration eating that diet were recorded, as were whether the diet was changed and each dog’s final main diet. For the purposes of the study, diets were classified as traditional when they were grain-inclusive extruded diets that did not contain pulses or potatoes as main ingredients (ie, the top 10 ingredients on the ingredient list). Nontraditional extruded diets were defined as those that were grain-free or contained pulses or potatoes as main ingredients. Diet change was defined as a change after the diagnosis of DCM or at the time heart disease was suspected but before DCM could be confirmed with an echocardiogram (and within 2 weeks of diagnosis of DCM). Taurine concentrations (plasma or whole blood), when measured, were recorded. Diet duration (ie, the duration of time dogs had been eating the main diet) was categorized as < 6 months, 6 months to 2 years, 2 to 4 years, or > 4 years. The last available echocardiographic measurements for each dog were also recorded. The date and cause of death, if not still alive, were recorded. Death was classified as sudden cardiac death if a dog’s death was “sudden and instantaneous as the result of unexpected collapse without evidence of dyspnea” or “when death occurred during sleep and no evidence of prior dyspnea was seen.”²⁰ Owners (for owner-submitted cases) or cardiologists (for cardiologist-submitted cases) were contacted when additional information was needed. If a dog’s outcome was unknown and no additional information was available, the dog was considered lost to follow-up.

Results

DCM group

One hundred twelve pit bull–type breed dogs were submitted for inclusion in this retrospective study, including 79 cases submitted by veterinary cardiologists, 29 cases submitted by owners, and 4 cases submitted by both owners and veterinary cardiologists. Twenty-one cases were excluded for the following reasons: did not meet the study echocardiographic criteria for DCM (n = 11; all 11 were included in the DCM-C group; see below), drug-induced DCM (3; doxorubicin [2], chloramphenicol [1]), dogs diagnosed with DCM before 2015 (2), dogs without echocardiographic measurements (2), tachycardia-induced DCM (2), and DCM associated with sepsis/systemic disease (1). Three dogs included in the DCM group had concurrent well-controlled diseases, as follows: hypothyroidism (n = 1), heartworm disease (1), and moderate pulmonic stenosis (1). One dog had no body weight available in the medical records but would have met the study’s echocardiographic criteria at any body weight. This dog was included in the analysis, but normalized left ventricular measurements could not be calculated.

A total of 91 DCM cases were analyzed (64 submitted by veterinary cardiologists, 24 submitted by owner, and 3 submitted by both). Most cases (n = 84) were from the US, with 18 states represented; cases were also submitted from Canada (6) and Germany (1). All DCM cases were diagnosed between January 2015 and August 31, 2022, as follows: 2015 (n = 1), 2016 (0), 2017 (2), 2018 (10), 2019 (22), 2020 (20), 2021 (21), and 2022 (through August 31; 15). At the time of DCM diagnosis, 76 dogs were eating commercial diets, 1 dog was eating a homemade diet, and diet was unknown or there was insufficient information available to determine whether the diet was nontraditional or traditional in 14 dogs (Figure 1). Most dogs were eating nontraditional commercial diets (64/76 [84%]), but 12 of 75 (16%) dogs were eating traditional diets. Diet duration was unknown in 43% of cases overall (Table 1).

View Full Size

Flow diagram illustrating baseline diet and diet change status for 91 pit bull–type breeds with dilated cardiomyopathy and whether dogs had a second echocardiogram (echo). The traditional diet group was analyzed as a single group due to the small numbers.

Citation: Journal of the American Veterinary Medical Association 261, 7; 10.2460/javma.23.01.0025

• Download Figure
• Download figure as PowerPoint slide

Table 1

Signalment, weight, and diet information for 91 pit bull–type breed dogs at the time of diagnosis of dilated cardiomyopathy. Data are reported as frequency (percentage of dogs within each column) for categorical variables and median (IQR) for continuous variables. The P value is for the comparison of the dogs eating traditional versus nontraditional diets. The diet for 14 dogs was unknown, and 1 dog was eating a homemade diet; these 15 dogs are not represented in the diet columns.

Variable	All dogs	Traditional diet	Nontraditional diet	P value
n	91	12	64	—
Age	7.5 (4.9–10.0)	6.9 (2.0–9.7)	7.5 (5.0–10.2)	.35
Sex				.83
Male	52 (57%; 44 neutered)	8 (67%; 7 neutered)	34 (53%; 28 neutered)
Female	39 (43%; 37 spayed)	4 (33%; 4 spayed)	30 (47%; 29 spayed)
Breed				.02
Pit bull–type dog	35 (39%)	6 (50%)	22 (34%)
American Pit Bull Terrier	31 (34%)	4 (33%)	22 (34%)
American Staffordshire Terrier	19 (21%)	0 (0%)	17 (26%)
American Bulldog	3 (3%)	0 (0%)	2 (3%)
American Bully	2 (2%)	2 (17%)	0 (0%)
Staffordshire Bull Terrier	1 (1%)	0 (0%)	1 (2%)
Weight (kg)	28 (23–33)	27 (22–29)	28 (23–32)	.48
Diet duration (years)				.94
Less than 6 mo	9 (10%)	1 (8%)	5 (8%)
6 mo–2 y	14 (15%)	2 (17%)	12 (19%)
2–4 y	11 (12%)	1 (8%)	10 (16%)
More than 4 y	18 (20%)	3 (25%)	13 (20%)
Duration unknown	39 (43%)	5 (42%)	24 (37%)
Changed diet				< .001
Yes	50 (55%)	3 (25%)	45 (70%)
No	18 (20%)	6 (50%)	4 (6%)
Unknown	23 (25%)	3 (25%)	15 (24%)

Overall, for the 91 dogs with DCM, the most common breeds represented were pit bull–type dogs, American Pit Bull Terriers, and American Staffordshire Terriers (Table 1; Supplementary Table S1). Male dogs predominated (57%), and the median weight was 28 kg (IQR, 23 to 33 kg). Overall, at the time of diagnosis, 63 of 91 (69%) dogs had congestive heart failure (CHF) and 47 of 91 (52%) had cardiac arrhythmias (supraventricular, 23/91 [25%]; ventricular, 37/91 [41%]; 14 dogs had both types of arrhythmias; Table 2; Supplementary Table S2). Twenty-three of the 91 (25%) dogs had plasma or whole blood taurine measured, with none documented to be deficient, although results for 1 dog were not available.

Table 2

Clinical findings and echocardiographic variables in 91 pit bull–type breed dogs at the time of diagnosis of dilated cardiomyopathy. Data are reported as frequency (percentage of dogs within each column) for categorical variables and median (IQR) for continuous variables. The P value is for the comparison of the dogs eating traditional versus nontraditional diets. The diet for 14 dogs was unknown and 1 dog was eating a homemade diet; these 15 dogs are not represented in the diet columns.

Variable	All dogs	Traditional diet	Nontraditional diet	P value
n	91	12	64	—
Congestive heart failure	63 (69%)	9 (75%)	43 (67%)	.59
Arrhythmias
Any arrhythmias	47 (52%)	9 (75%)	26 (41%)	.03
Supraventricular	23 (25%)	5 (42%)	12 (19%)	.06
Ventricular	37 (41%)	6 (50%)	22 (34%)	.22
Syncope	14 (15%)	2 (17%)	9 (14%)	.81
Presence of a cardiac murmur	66 (73%)	10 (83%)	45 (70%)	.54
Presence of a cardiac gallop	24 (26%)	3 (25%)	18 (28%)	.82
IVSd (cm)	0.95 (0.84–1.07)	0.98 (0.90–1.11)	0.95 (0.84–1.07)	.49
LVIDd (cm)	6.43 (5.42–6.93)	6.15 (5.00–6.76)	6.51 (5.45–7.09)	.32
LVWd (cm)	0.90 (0.80–0.98)	0.87 (0.74–1.04)	0.86 (0.78–0.96)	.79
IVSs (cm)	1.14 (0.90–1.30)	1.16 (0.94–1.41)	1.15 (0.89–1.30)	.79
LVIDs (cm)	5.45 (4.40–6.19)	4.77 (4.04–5.55)	5.74 (4.39–6.20)	.08
LVWs (cm)	1.10 (0.97–1.21)	1.17 (0.91–1.37)	1.08 (0.93–1.20)	.32
LVIDdN	2.33 (2.02–2.60)	2.16 (1.89–2.39)	2.40 (2.04–2.60)	.10
LVIDsN	1.86 (1.57–2.15)	1.69 (1.40–2.08)	1.94 (1.59–2.23)	.28
Fractional shortening (%)	14.80 (10.60–19.40)	19.25 (12.20–21.47)	13.53 (11.12–17.68)	.07
LA:Ao (2-D)	2.01 (1.68–2.39)	1.83 (1.33–2.37)	2.02 (1.68–2.43)	.32

IVSd/s = Interventricular septal thickness in diastole/systole. LA:Ao = Ratio of the left atrial to aortic diameters (2-D). LVIDd/s = Left ventricular internal diameter in diastole/systole. LVIDdN = Normalized left ventricular internal diameter in diastole. LVIDsN = Normalized left ventricular internal diameter in systole. LVWd/s = Left ventricular free wall in diastole/systole.

The only significant difference identified when comparing dogs with DCM eating nontraditional diets to those eating traditional diets was breed, with more American Staffordshire Terriers in the nontraditional diet group and more American Bully dogs in the traditional diet group (Table 1). In the nontraditional diet group, there was an approximately equal sex distribution (53% male and 47% female), while in the traditional diet group, 67% of dogs were male, but diet groups were not significantly different in sex distribution (P = .83). There were no other significant differences between diet groups in signalment, body weight, diet duration, or clinical signs. Dogs eating nontraditional diets had numerically larger LVIDdN (P = .10) and LVIDsN (P = .28) and lower fractional shortening (P = .07) compared to the traditional diet group, but these measurements did not reach statistical significance (Table 2). There were no significant differences in baseline characteristics or echocardiographic measurements when comparing dogs eating nontraditional diets that did or did not change diet (Supplementary Tables S1 and S2).

Over the course of their disease, dogs were treated with a variety of cardiac medications, including pimobendan (n = 83), furosemide (65), angiotensin-converting enzyme inhibitor (64), spironolactone (39), diltiazem (11), mexiletine (7), sotalol (7), amiodarone (3), digoxin (3), torsemide (3), hydrocodone (2), and 1 each of the following: atenolol, carvedilol, clopidogrel, and theophylline. There were no significant differences in cardiac medication administration between dogs eating nontraditional diets that changed diets and dogs eating traditional diets (dogs eating nontraditional diets that did not change diets were not compared because of their short survival time), except for sotalol, which was used more commonly in the traditional diet group (nontraditional group that changed diets, 3/45 [7%]; traditional group, 4/12 [33%]; P = .04). Dogs also received a variety of supplements over the course of their disease, including taurine (n = 61), carnitine (36), fish oil (28), coenzyme Q10 (3), and 1 each of the following: arginine, hawthorn, vitamin C, and a multinutrient supplement. There were no significant differences in the frequency of dogs receiving supplements between the nontraditional diet group that changed diet and the traditional diet group.

After the diagnosis of DCM, diet was changed in 50 of 91 (55%) dogs, while diet was not changed in 18 of 91 (20%) dogs; in 23 of 91 (25%) dogs, it was unknown whether diet was changed (Figure 1). Significantly more dogs in the nontraditional diet group changed diets (45/64 [70%]) compared to the traditional diet group (3/12 [25%], P < .001). Thirty-four of the 61 (56%) dogs that had both known baseline diet and known status of diet change had at least 1 additional echocardiogram after the time of diagnosis (traditional diet group, 7/12 [58%]; nontraditional diet group that changed diets, 27/45 [60%]; nontraditional diet group with no change, 0/4; P = .09).

There was a wide range in time between echocardiograms, as follows: traditional diet group, median, 251 days (IQR, 185 to 492 days); nontraditional diet group that changed diets, median, 395 days (IQR, 155 to 770 days; P = .38; overall range for both groups, 60 to 1,076 days). Compared to dogs in the traditional diet group, dogs in the nontraditional diet group that changed diets had a significantly greater decrease in LVIDdN (P = .02), LVIDsN (P = .048), and ratio of the left atrium to aorta (P = .002) and a significantly greater increase in fractional shortening (P = .02; Figure 2) between the first and second echocardiogram.

View Full Size

Changes in echocardiographic measurements in 34 of the 91 pit bull–type breeds with dilated cardiomyopathy (DCM) that had repeat echocardiograms (time between echocardiograms for all dogs ranged from 60 to 1,076 days). Dogs are divided into groups on the basis of their original diet and their diet change status, as follows: nontraditional diet group that changed diets (n = 27) or traditional diet group (7). P values are for comparisons between the diet groups. A—Change in normalized left ventricular internal diameter in diastole (LVIDdN; P = .02). B—Change in normalized left ventricular internal diameter in systole (LVIDsN; P = .048). C—Change in the left atrium-to-aorta ratio (LA:AO; P = .002); D—Change in fractional shortening (FS; P = .02).

Citation: Journal of the American Veterinary Medical Association 261, 7; 10.2460/javma.23.01.0025

• Download Figure
• Download figure as PowerPoint slide

At the time of analysis, 39 dogs were still alive, 33 were no longer alive (17 euthanized, 12 died, and 4 were known to be dead but the cause of death/euthanasia was unknown), and 19 were lost to follow-up. For dogs eating nontraditional diets that changed diets and were known to be deceased (n = 13), causes of death/euthanasia were sudden (7), worsening CHF (3), and cancer (2), with cause of death unknown in 1 dog. For dogs eating nontraditional diets that did not change diets and were known to be deceased (n = 4), causes of death/euthanasia were sudden (1) and worsening CHF (3). In the traditional diet group, 3 of 11 dogs had died and the cause of death/euthanasia was worsening CHF (n = 1) and unknown (2). There was no significant difference in cause of death between diet groups (P = .11). Median survival time was significantly longer in the nontraditional diet group that changed diets (median survival time not reached but ranged from 1 to 1,303 days; P < .001) and the traditional diet group (median survival time not reached but ranged from 1 to 675 days; P < .001) compared to the nontraditional diet group that did not change diet (median survival time, 1 day; range, 1 to 11 days; Figure 3). There was no significant difference in survival time between the nontraditional diet group that changed diets and traditional diet group (P = .99).

View Full Size

Kaplan-Meier survival curves in 61 pit bull–type breed dogs with dilated cardiomyopathy for which diet at the time of diagnosis could be classified as nontraditional or traditional, and for which a change in diet after diagnosis could be confirmed: 45 dogs eating nontraditional diets that had their diets changed after diagnosis (black solid line), 4 dogs eating nontraditional diets that did not have their diets changed after diagnosis (black dotted line), and 12 dogs eating traditional diets at the time of diagnosis (gray solid line). Survival time was significantly longer in dogs eating nontraditional diets that changed diets (P < .001) and in dogs eating traditional diets (P < .001) compared to dogs eating nontraditional diets that did not change diets. Survival time for dogs in the traditional diet group was not significantly different from the nontraditional diet group that changed diets (P = .99).

Citation: Journal of the American Veterinary Medical Association 261, 7; 10.2460/javma.23.01.0025

• Download Figure
• Download figure as PowerPoint slide

Discussion

Results of this study showed that pit bull–type breeds with DCM that had been eating nontraditional diets at the time of diagnosis and changed diets had a significant improvement in key echocardiographic measurements compared to dogs that had been eating traditional diets, as follows: larger reductions in left ventricular diameter and left atrial size, as well as a larger increase in fractional shortening. These results showing echocardiographic improvements were consistent with those of 5 published studies^9,14–17 on diet-associated DCM in which diet was changed in some (for retrospective studies) or all dogs (for prospective studies). Significant improvement in systolic function has been seen in 2 prospective studies,^9,17 but left ventricular and left atrial size improved significantly in all 5 studies.^9,14–17

Survival time was significantly longer in dogs with DCM eating nontraditional diets that changed diets than in dogs eating nontraditional diets that did not change diets. This is consistent with a previous retrospective study¹⁴ of dogs of multiple breeds with DCM and with cats with DCM.²¹ In the current study, it was possible that survival time for dogs in the nontraditional diet group that did not change diets was too short for there to be time to change the diet. However, survival time for dogs eating nontraditional diets that changed diets was not significantly different from dogs eating traditional diets. This result differed from a previous retrospective study¹⁶ in which dogs of multiple breeds eating grain-free diets had significantly longer survival compared to dogs eating grain-inclusive diets. However, the current study’s results were similar to those from a prospective study¹⁷ of dogs of multiple breeds with DCM in which there was no difference in survival time between dogs eating nontraditional compared to traditional diets. The lack of significant difference in survival time between dogs with DCM eating nontraditional diets that changed diets and dogs eating traditional diets could be related to the small percentage of dogs that were eating traditional diets (12/76 [16%] dogs for which baseline diet was known). It also could be due to the fact that previous diet history was not available for all dogs, so some dogs eating traditional diets at the time of diagnosis may have eaten nontraditional diets previously (so the traditional diet at the time of diagnosis may have represented a diet change). The range of survival times for dogs in the traditional diet group was narrower (longest survival at the time of analysis was 675 days) than for the nontraditional diet group that changed diets (longest survival at the time of analysis was 1,303 days), so the traditional diet group might have been followed for a shorter time period (the median survival time was not reached for either of these groups). It is also possible that DCM in pit bull–type breeds eating traditional diets behaves differently than in other breeds that develop DCM more commonly.

In the current study, the most common causes of death for dogs with DCM in both diet groups were sudden death and worsening CHF, and cause of death was not significantly different between diet groups. These results were similar to those from a retrospective study¹⁴ of diet-associated DCM in which 35% of dogs died suddenly and 58% of dogs were euthanized for CHF, while in a prospective study,¹⁷ 54% of dogs died suddenly and 33% of dogs were euthanized for CHF (there was no significant difference in cause of death between nontraditional and traditional diet groups for either study). These findings suggest that, despite a significant improvement in cardiac measurements, dogs with DCM phenotypes are at risk for arrhythmias and sudden cardiac death. Arrhythmias were present in many of the dogs in this retrospective study but there was no significant difference in ventricular arrhythmias between the nontraditional and traditional diet groups. In addition to affecting survival time, the presence of arrhythmias can limit the potential for echocardiographic improvement in dogs that had been eating nontraditional diets. Improvement in dogs with diet-associated DCM appears to be slow, typically occurring over months to years, so sudden death may occur before the heart has time to improve.

In addition to the echocardiographic improvements and longer survival times in dogs eating nontraditional diets that changed diets, this retrospective study also represented the largest report of DCM in pit bull–type breeds. Overall, both CHF and arrhythmias were common at the time of diagnosis. Thus, while pit bull–type breeds are not commonly considered breeds at risk for DCM, these results suggest that DCM can occur in this breed and should be considered in dogs with cardiac murmurs, arrhythmias, or clinical signs of CHF.

Taurine deficiency was not identified in any dogs in the current study of pit bull–type breed dogs with DCM. Taurine deficiency is a known cause of secondary DCM, but plasma and whole blood taurine deficiency have been uncommon in other studies^14–17 of diet-associated DCM except for 1 study⁹ of Golden Retrievers with taurine deficiency and DCM. While taurine was not measured in all dogs in the current study (25% of the 91 dogs), none of the dogs were found to be deficient. The fact that none of the dogs measured had taurine deficiency argues against taurine deficiency playing a primary role in the current episode of diet-associated DCM, but further investigation is needed given the limited percentage of dogs in this study in which taurine was evaluated. In addition, a recent study²² showed that whole blood taurine concentrations (and, in some dogs, plasma concentrations) may not be a sensitive measure of skeletal muscle depletion, and other measures of taurine status may be needed.

Other findings from the current study were similar to those of previous reports of dogs with diet-associated DCM, including the common findings of CHF and arrhythmias and few differences in clinical and echocardiographic variables between diet groups at the time of diagnosis.^14–17 The relatively common finding of arrhythmias could have contributed to sudden death in both of the diet groups and may be underestimated on the basis of the retrospective nature of the study and lack of Holter analysis in most dogs. While 3 dogs in the current study were diagnosed with DCM between 2015 and 2017, the majority of cases were diagnosed between 2018 and 2022, which is the period after the FDA’s first alert in 2018.¹⁰ Therefore, cardiologists may be more aware of a possible association between diet and dogs with DCM. However, we requested records from pit bull–type breeds with any form of DCM, so our cases should include both primary and diet-associated DCM. Whether the large number of dogs with DCM reported in the current study was related to diet (most were eating nontraditional diets) or increased awareness (or a combination of the two) is unclear. The frequency of pit bull–type breeds eating nontraditional commercial diets in the current study (84% for dogs in which the diet was known) was consistent with the high percentage of dogs in previous retrospective studies (64% to 79%)^11–13 and the 85% of dogs in a prospective study¹⁷ that were eating nontraditional diets.

In the current study, echocardiographic criteria for DCM were the same as those used in some previous studies of diet-associated DCM.^14–17 It is possible that, like certain breeds such as the Doberman Pinscher and Boxer, pit bull–type breeds may have some breed-specific echocardiographic abnormalities with DCM, so the echocardiographic criteria used may have been too stringent. However, we elected to err on the side of being conservative, so for the 11 dogs diagnosed by cardiologists with DCM but that did not meet all 3 of the echocardiographic measurements for this study (DCM-C group), we elected to evaluate these dogs separately, as was done in a recent prospective study.¹⁷ This group may reflect dogs with possible early stages of myocardial disease or breed-specific echocardiographic findings of DCM. While this DCM-C group was small (n = 11), these findings may be helpful for better understanding earlier stages of disease or provide information for future investigations as to whether breed-specific criteria for DCM are indicated for these breeds. Fewer than half of the dogs (5/11) had follow-up echocardiograms, but echocardiographic improvements were similar to those seen in the DCM group, including significant improvement in LVIDsN, left atrial size, and fractional shortening. These findings were similar to those from a prospective study¹⁷ of dogs with subclinical cardiac abnormalities.

This study had a number of important limitations. The retrospective nature of the study introduced variability in availability and timing of echocardiographic measurements. In fact, only 56% of dogs with full diet information had follow-up echocardiograms, and the timing ranged from 60 to 1,076 days after diagnosis. Variability was also introduced in the echocardiographic measurements since the study included dogs diagnosed with DCM by many different cardiologists. From other studies of dogs with diet-associated DCM and from the authors’ clinical experience, echocardiographic improvements can take many months, so the full extent of improvement may not have been achieved in many of the dogs. In addition, cardiac medications, such as pimobendan and furosemide, can be associated with mild decreases in left atrial and left ventricular size and mild increase in fractional shortening, so these medications may have influenced the changes in cardiac size and function in the current and previous studies. In addition, some dogs in the current study received taurine and other supplements that could impact the heart. However, there were no significant differences in medications or supplements between diet groups. Sample size was also an important limitation of the study since there were only 12 dogs in the traditional diet group and only 4 in the nontraditional diet group that did not change diets after diagnosis. This study included 5 breeds, but there is some controversy over which breeds should be included as “pit bull–type breeds.” Many of the dogs submitted for the study were classified as “pit bull” or “pit bull cross,” so the exact breed is unknown. Similarly, due to laws in some areas, owners of pit bull–type breeds may not list their dog as such (eg, instead identifying their dog as “mixed breed” or “Boxer cross” for medical records). Therefore, dogs listed in medical records in this way would not have been captured and submitted for this study, so there may be more cases that were not included. The lack of consistent cardiac troponin I and N-terminal pro–B-type natriuretic peptide measurement and infectious disease testing at the time of diagnosis was another limitation of this retrospective study and would have been useful information. Another limitation was that, for cardiologist-submitted cases, the exact diet, duration of diet, and whether diet was changed were often unknown, so there was less information on these cases compared to the owner-submitted cases. Similarly, for many of the cardiologist-submitted cases, only the diet at time of diagnosis was known, so some of these dogs may have been eating nontraditional diets before diagnosis. Also, some of the dogs in this study were adopted as adults or found as strays so prior diets were unknown. Some dogs in both diet groups were lost to follow-up. Therefore, the outcome for these dogs is unknown. This study only included cases that met the study’s specific echocardiographic criteria for DCM, so other than the 11 DCM-C dogs, dogs with possible earlier forms of disease (eg, systolic dysfunction, unexplained arrhythmias, and left ventricular enlargement) were excluded from the main results. Finally, the initial plan for the study was for dog owners to contribute all cases, but due to the limited number of owner-submitted cases, additional cases were solicited from cardiologists. It is unclear whether the intended “citizen science” approach was unsuccessful due to an insufficient recruitment time (3 months) or recruitment during the summertime, requests not reaching the owners of affected dogs, the requirement for echocardiographic measurements, or other factors, but this information might help to inform future studies attempting to recruit cardiology cases from pet owners.

This retrospective study of DCM in pit bull–type breeds provided information that may help clinicians and researchers better understand the disease in these breeds. In this study, dogs with DCM eating nontraditional diets that changed diets and dogs eating traditional diets had a significantly longer median survival time after diagnosis compared to dogs with DCM eating nontraditional diets that did not have a diet change, despite the common finding of CHF at the time of diagnosis; this information may be of prognostic value to owners of dogs of these breeds that are diagnosed with DCM. Results of this study showed that pit bull–type breeds with DCM eating nontraditional diets that changed diets had significant improvement in echocardiographic measurements compared to dogs eating traditional diets. Dogs with DCM-C may represent an earlier form or variation of DCM and showed similar improvement in echocardiographic measurements after diet change. Dilated cardiomyopathy in pit bull–type breeds was associated with nontraditional diets in this study, and echocardiographic indices improved with diet change. Diet history is extremely important in these breeds to ensure that diet change is recommended in dogs eating either grain-free or grain-inclusive diets that contain pulses or potatoes.

Supplementary Materials

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