Effect of a specially formulated diet on progression of heart enlargement in dogs with subclinical degenerative mitral valve disease

Abstract Background Previous studies in dogs with degenerative mitral valve disease (DMVD) have identified altered myocardial energy metabolism and oxidation, which might contribute to cardiac hypertrophy. Diets rich in medium chain fatty acids and antioxidants are a potential means of treatment. A previous clinical study found significantly smaller left atrial diameter (LAD) and left atrium‐to‐aorta diameter ratio (LA : Ao) in dogs with subclinical DMVD fed a specially formulated diet vs control diet for 6 months. Hypothesis/Objectives A specially formulated diet will slow or arrest left heart enlargement in dogs with subclinical DMVD over 365 days. Animals One hundred twenty‐seven dogs with unmedicated subclinical DMVD; 101 dogs in the per protocol cohort. Methods Randomized double‐blinded controlled multicenter clinical trial. Results The study's primary composite outcome measure was the sum of percentage change in LAD and left ventricular internal dimension at end‐diastole (LVIDd) at day 365. In the per protocol cohort, the outcome measure increased by 8.0% (95% confidence interval [CI], 2.9%‐13.1%) in dogs receiving the test diet vs 8.8% (95% CI, 5.1%‐12.5%) in dogs receiving control diet (P = .79). Neither component of the primary outcome measure was significantly different between groups (LAD, P = .65; LVIDd, P = .92). No difference was found in mitral valve E wave velocity (P = .36) or the proportion of dogs withdrawn from the study because of worsening DMVD and heart enlargement (P = .41). Conclusions and Clinical Importance Feeding a specially formulated diet for 365 days was not associated with a significantly different rate of change of left heart size in dogs with subclinical DMVD as compared to control.

Common measures of DMVD severity include the echocardiographic left atrial diameter (LAD), left atrium-to-aorta diameter ratio (LA : Ao), and the diameter of the left ventricle at end-diastole (LVDd). 1 The heart is a metabolically active organ and requires a constant supply of high energy phosphates. Long chain fatty acids (FAs) and glucose are the 2 most common myocardial energy substrates and undergo β-oxidation or glycolysis, respectively. 2 Limitations associated with β-oxidation of long chain FAs include reliance on a carnitine-mediated mitochondrial transport system and production of reactive oxygen species byproducts. 3 Ischemia, uncoupling of glucose uptake and oxidation, insulin resistance, and high angiotensin II concentrations impair glycolysis and can lead to an energy starved state. 4 Diet is a potential means of altering myocardial metabolism by selective avoidance or enrichment of certain ingredients. For instance, foods such as coconut and palm oil are rich in medium chain triglycerides (MCT), which are precursors of medium chain FAs. Medium chain FAs do not depend on mitochondrial carnitine transport and are easily oxidized as compared to long chain FAs. 5 Foods rich in antioxidants, such as vitamin E or taurine, or rich in long chain polyunsaturated FA, such as omega-3 FAs, similarly offer potential benefit. A previous controlled study 6 reported the effect of a custom formulated diet, enriched with antioxidants, MCT, and other metabolic precursors, on echocardiographic left atrial size in 19 dogs with subclinical DMVD. Dogs receiving the test diet experienced significant 3% and 4% decreases in LAD and LA : Ao, respectively, compared to increases of 7% and 11% in dogs receiving control diet. This study was performed in a sample of primarily adult Beagles housed in a research environment for 6 months. We sought to determine the effect of a similarly formulated diet in a larger cohort of client-owned dogs over a 12-month feeding period. The hypothesis was that the test diet would slow or prevent echocardiographic left heart enlargement in dogs with mild subclinical DMVD as compared to a control diet over a 12-month feeding period. The study diets were identical to each other in appearance and packaging and were labeled with 1 of 4 letters (A-D) to help maintain double blinding throughout the study. The amount of diet fed was calculated based on the dog's ideal body weight and maintenance energy requirement. Owners were instructed that the study diet should comprise at least 90% of their dog's diet with the remaining 10% comprised of treats and other foods as long as prohibited dietary supplements were not fed.

| MATERIALS AND METHODS
For reasons of other illness, such as dietary indiscretion or transient diarrhea or decreased appetite, study diet holidays of up to 4 consecutive days were permitted as long as the days off the study diet amounted to ≤16 days in total over the duration of the study. Recheck physical examination, echocardiography, serum electrolyte concentrations, renal function tests, and blood pressure were performed 180 and 365 days after randomization. Between these visits, recheck physical examination was performed 90 and 270 days after randomization. At each visit, the quantity of study diet being fed was reevaluated based on body condition, body weight, and discussion with the owner. Study personnel contacted owners by telephone or email every 45 days to ensure compliance with the diet feeding. Adverse events, whether or not thought to be related to the diet, were recorded and classified according to severity and their potential relationship to the diet.
The study's primary outcome measure was the difference in the sum of the absolute percentage change of LAD and LVDd from day 0 to day 365 between groups in the per protocol cohort (Figure 1).
The outcome measure was independent of the dog's body weight or change in weight during the study. Sample size was calculated a priori based on a power of 80%, alpha of 0.05, and a hypothesized clinically relevant difference of ≥8% (SD, 14%) in progression of left heart size between groups (eg, control 8% increase vs. test diet 0% increase). A total of 98 dogs (49 in each group) were needed in the per protocol sample. To this total, 14 dogs were added to account for loss of power in the event the rate of disease progression was not normally distributed. An additional 10 dogs were added to account for withdrawals for a total recruitment target of 122 dogs.
The interobserver and intraobserver variability associated with echocardiography 7,8 tends to decrease study power. The study methods specifically included 3 design characteristics to increase study sensitivity and power. First, at each study timepoint, 2D, M-mode, and color flow echocardiography exclusively was performed by a single board-certified cardiologist (i.e., the site's primary investigator). Second, the study outcome measure used the sum of the LAD and LVDd as a composite measure of global left heart enlargement rather than either measure individually. Third, the primary outcome measure was determined by an echocardiography core laboratory (ECL) with 1 prospectively assigned investigator (BAS). The ECL prospectively established standard processes for image acquisition, transfer, and analysis that each site and investigator followed (see Supplemental Information).
Briefly, LVDd and the LAD and aortic diameter (AoD) were measured from the right parasternal short axis M-mode and 2D views, respectively. The LVDd was measured from the inner edge of the interventri-

| Primary outcome measure
At day 365, dogs receiving the test diet experienced an average increase in the primary outcome measure of 8.0% (95% CI, 2.9%-13.1%) as compared to 8.8% (95% CI, 5.1%-12.5%) in dogs receiving the control diet ( Figure 3). These differences were not significant between groups (P = .79). The rate at which progression of heart enlargement developed was normally distributed (Shapiro-Wilk P = .31). Neither of the individual components of the primary outcome measure significantly differed between groups. Dogs receiving the test diet experienced a 4.4% (95% CI, 1.1%-7.8%) increase in LAD as compared to 5.4% (95% CI, 2.7%-8.1%) in dogs receiving the control diet (P = .65; Figure 4A). Dogs receiving the test diet experienced a 3.6% (95% CI, 0.84%-6.3%) increase in LVDd as compared to 3.4% (95% CI, 1.6%-5.3%) in dogs receiving the control diet (P = .92; Figure 4B). No significant effect of age (P = .21) or breed (P = .1) was found on the primary outcome measure. Results specific to sites in North America (P = .55) or South America (P = .13) were not significant. Analyses of the primary endpoint using LOCF data from day 180 (P = .6) as well as use of the site investigator's echocardiographic measurements in place of the ECL measurements (P = .91) were not significant.

| Secondary outcome measures
The change in the primary outcome measure between 0 and 6 months  In previous studies, 6,10 a diet similar to our test diet was associated with significantly increased circulating MCT, omega-6-to-omega-3 FA ratio, and nitric oxide and carnitine biosynthesis precursors as compared to control diet. Our study protocol did not include metabolomic testing, which might have provided additional insights. In the previous studies, feeding of the dogs was strictly controlled. In our cohort of client-owned dogs, we recognized the impossibility of excluding any and all foods other than the study diet for 365 days, and owners were allowed to provide small amounts of food other than the study diet and to take diet holidays. Although foods or supplements with potential to interfere with the study (e.g., fish oil, antioxidants, cardiac glycosides, other prescription diets) were strictly prohibited, we cannot rule out an effect of other foods.
The previous study 6   In conclusion, a specially formulated diet was not associated with significant changes in LAD and LVIDd when fed to dogs with mild subclinical DMVD for 1 year as compared to control diet. Additional studies are needed to better understand the effect of diet and the metabolome in dogs with DMVD.