J Vet Intern Med 2000;14:513–520

Clinical, Echocardiographic, and Neurohormonal Effects of a Sodium-Restricted Diet in Dogs with Heart Failure John E. Rush, Lisa M. Freeman, Donald J. Brown, Barbara P. Brewer, James N. Ross, Jr., and Peter J. Markwell The use of low-sodium diets in dogs with heart failure is common practice, but randomized, double-blind studies have not been conducted to examine the benefits or problems with this approach. The purpose of this study was to determine the effects of a low-sodium diet on clinical, echocardiographic, and neurohormonal parameters in dogs with heart failure. Dogs with stable chronic heart failure were fed exclusively a low-sodium (LS) and a moderate-sodium (MS) diet for 4 weeks each in a randomized, doubleblind, crossover design. At days 0, 28, and 56, echocardiography and thoracic radiography were performed, and blood was analyzed for electrolytes and neurohormones. Fourteen dogs completed the study (9 with chronic valvular disease and 5 with dilated cardiomyopathy). Electrolyte abnormalities were common during the study, and serum sodium and chloride concentrations decreased significantly on the LS diet. Neurohormones did not change significantly between diet groups. Maximum left atrial (P ⫽ .05) and standard left atrial (P ⫽ .09) size decreased on the LS diet. For dogs with chronic valvular disease, vertebral heart score (P ⫽ .05), left ventricular internal dimension in diastole (P ⫽ .006) and systole (P ⫽ .02), standard left atrial dimension (P ⫽ .03), maximum left atrial dimension (P ⫽ .02), end-diastolic volume index (P ⫽ .02), and end-systolic volume index (P ⫽ .04) decreased significantly on the LS diet compared to the MS diet. Although analysis of these data suggests some benefits of a lowsodium diet, future studies with improved study design are needed to further evaluate the advantages and disadvantages of sodium restriction in dogs with heart failure. Key words: Chronic valvular disease; Congestive heart failure; Dilated cardiomyopathy; Neurohormones; Nutritional management.

S

odium restriction, in conjunction with diuretics and venous vasodilators, is one method to treat abnormally high preload in patients with congestive heart failure (CHF). No randomized, double-blind placebo-controlled trials of sodium restriction in dogs with CHF have been reported. Nonetheless, veterinarians have extrapolated from recommendations made for humans since the 1960s in applying sodium restriction to dogs with CHF. Only recently have the benefits and potential problems associated with sodium restriction been questioned. Despite the lack of research in dogs with CHF, low-sodium diets have been well studied in healthy dogs. An early study in 1964 showed no significant changes in extracellular water, serum sodium concentration, or serum chloride concentration in healthy dogs fed a low-sodium diet.1 That study also showed that healthy dogs were able to maintain sodium and potassium balance on both low- and high-sodium diets.1 Two other studies found that healthy dogs fed a low-sodium diet had no changes in serum sodium and chloride concentrations or extracellular fluid volume compared to those fed a high-sodium diet.2,3 In 1994, a study examined the effects of a low-sodium diet and furosemide in healthy dogs with or without captopril.4 Although withingroup changes in serum electrolyte concentrations were not From the Department of Clinical Sciences, Tufts University School of Veterinary Medicine, North Grafton, MA (Rush, Freeman, Brown, Brewer, Ross); and the Waltham Centre for Pet Nutrition, Melton Mowbray, Leicestershire, UK (Markwell). Previously presented in part at the 1999 ACVIM Forum. Reprint requests: John E. Rush, DVM, MS, Department of Clinical Sciences, Tufts University School of Veterinary Medicine, 200 Westboro Road, North Grafton, MA 01536; e-mail: [email protected]. edu. Submitted December 6, 1999; Revised February 24, April 3, 2000; Accepted May 1, 2000. Copyright 䉷 2000 by the American College of Veterinary Internal Medicine 0891-6640/00/1405-0009/$3.00/0

found in that study, 3 of 6 dogs became hyperkalemic while receiving a low-sodium diet plus furosemide and 2 of 6 became hyperkalemic while receiving a low-sodium diet plus furosemide and captopril.4 The effects of drugs such as furosemide or angiotensin-converting enzyme (ACE) inhibitors alone or potential drug-nutrient interactions can be profound and must be considered. The effects of the lowsodium diet alone were not given in the previously reported study.4 In healthy dogs, low-sodium diets caused increased plasma renin activity (PRA) and plasma aldosterone concentration (PAC) compared to a high-sodium diet, although ACE, atrial natriuretic peptide (ANP), arginine vasopressin (AVP), and endothelin-1 (ET-1) remained unchanged.5,6 However, healthy dogs that received enalapril while eating a low-sodium diet had an exaggerated increase in PRA and a larger decrease in ACE and ANP compared to a dogs eating a high-sodium diet.7 These investigators also found an inverse correlation between PRA and sodium content of the diet.7 The response to feeding a low-sodium diet may be very different in dogs with CHF. Even before clinical signs become apparent in dogs with cardiac disease, the reninangiotensin-aldosterone (RAA) system is activated and excretion of sodium is abnormal.8 Therefore, some authors have recommended institution of a low-sodium diet at the 1st sign of heart disease (ie, in dogs with cardiac murmurs that do not exhibit clinical signs of CHF). Potential activation of the RAA system by sodium restriction raises questions as to whether such activation is beneficial or harmful. A study by Pensinger showed that dogs with cardiac disease but without CHF were able to maintain sodium and potassium balance on both low- and high-sodium diets, similar to healthy dogs.1 However, dogs with CHF retained sodium on the high-sodium diet but did not retain sodium on the low-sodium diet.1 A study in 1994 showed that dogs with CHF fed a low-sodium diet in com-

514

Rush et al

bination with furosemide or furosemide plus captopril had an increase in mean serum creatinine, a decrease in mean serum sodium, and 6 of 10 dogs became hyperkalemic during the 4-week study period.4 More recently, untreated dogs with mild, asymptomatic mitral valve insufficiency had a larger increase in PRA and PAC and a lower ACE activity when changed from a high-sodium diet to a lowsodium diet.9 Sodium intake had no effect on ET-1, ANP, and AVP.9 The purpose of the current study was to determine the clinical, echocardiographic, and neurohormonal effects of a low-sodium diet compared to a moderate-sodium diet in dogs being treated medically for heart failure.

Table 1. Comparison of selected nutrients in low sodium and moderate sodium diets.

Protein Fat Sodium Chloride Potassium Magnesium Calcium Phosphorus Energy (kcal/100 g)

Low Sodium (g/100 kcal)

Moderate Sodium (g/100 kcal)

3.88 7.65 0.04 0.05 0.22 0.01 0.27 0.15 182.9

4.85 7.75 0.07 0.08 0.27 0.01 0.27 0.15 155.0

Materials and Methods Subjects Dogs with stable CHF (modified New York Heart Association functional class II–IV)10 were studied. In this classification class I ⫽ no limitation, physical activity does not cause symptoms; class II ⫽ slight limitation of physical activity, ordinary physical activity results in symptoms; class III ⫽ marked limitation of physical activity, less than ordinary activity leads to symptoms; and class IV ⫽ inability to carry on any activity without symptoms, symptoms present at rest.10 Heart failure was secondary to either dilated cardiomyopathy (DCM) or chronic valvular disease (CVD). Dogs with major concurrent diseases and dogs not expected to live at least 8 weeks were excluded from the study. If medication changes were judged to be necessary at the time of enrollment, enrollment was postponed for 7–10 days so that stabilization on the new medication could be achieved. Medication adjustments were not restricted during the study period if changes were deemed necessary by the attending clinician. Dogs were treated in a standardized fashion, as follows: category I (ACE inhibitor only at 0.5 mg/kg q24h), category II (ACE inhibitor plus furosemide), category III (ACE inhibitor, furosemide, and digoxin), category IV (furosemide, digoxin, and high-dose ACE inhibitor [0.5 mg/kg q12h]), category V (category 4 plus the addition of spironolactone), and category VI (category 5 plus maximal dosage of furosemide). Within each category, furosemide was given at the lowest possible dosage required to control clinical signs. The study was approved by the Tufts University Animal Care and Use Committee, and owners signed a consent form before enrolling their dogs in the study. Owners agreed to feed the study diets exclusively for the duration of the study with no other dog food, table food, or treats.

Estimated diastolic blood pressure was recorded when the flow sound became muffled or altered. Blood was collected by jugular venipuncture (heparinized plasma for tumor necrosis factor [TNF]; serum for biochemistry profile and aldosterone concentration; and ethylenediaminetetraacetic acid plasma for a CBC, ANP, and PRA). Plasma collected for ANP was immediately placed on ice after collection. A freshly voided urine sample was collected for urinalysis. Blood was centrifuged at 4⬚C, separated within 30 minutes, and samples were frozen at ⫺80⬚C until analysis. After baseline analysis, dogs were randomized to begin with either the low-sodium (LS) diet or the moderate-sodium (MS) diet first (Table 1). The LS diet is a commercially available product,a whereas the MS diet was made specifically for the study using the same batch of raw materials as the LS diet. Neither the owners nor the investigators knew which diet was being assigned throughout the study. Owners were instructed to make the change to the new food gradually (over 2–3 days) and to then feed the prescribed diet exclusively for 28 days. Each owner was called at 2-week intervals for progress reports on the dog’s condition and appetite. All dogs were reassessed after eating the initial diet for 28 days. Subjects were presented to the hospital after an overnight fast. At this time, all baseline examinations and measurements were repeated. Dogs then were crossed over to the diet they had not yet received. Owners were instructed to make the dietary change immediately and to offer no other foods while on the study. Each owner was called after 2 weeks for a progress report on the dog’s condition and appetite. After eating the 2nd diet for 28 days, dogs again were reassessed (day 56). All examinations and measurements were repeated as on days 0 and 28.

Study Design Dogs were presented after an overnight fast. After a complete physical examination, standard 2-D and M-mode echocardiograms were performed. Standard left atrial dimension was measured using a right parasternal short-axis view by measuring the left atrial dimension with the M-mode cursor at the level of the aortic root. Maximum left atrial dimension was measured using a right parasternal short-axis view by placing the M-mode cursor just caudal to the aorta at the subjective maximum diameter of the left atrium. The maximum left atrial dimension was then measured at this point from an M-mode echocardiogram. End-systolic volume index (ESVI) and end-diastolic volume index (EDVI) were calculated using the following formulas: ESVI ⫽ (left ventricular internal dimension in systole)3/body surface area, and EDVI ⫽ (left ventricular internal dimension in diastole)3/body surface area. Lateral and dorsoventral thoracic radiographs were taken, and vertebral heart scores were calculated.11 Cardiologists were blinded to diet when performing echocardiograms and when calculating vertebral heart scores. Systolic and estimated diastolic blood pressures were measured using the Doppler technique with the dog resting quietly.

Laboratory Methods The WEHI 164 subclone 13 cytotoxicity bioassay was used to measure TNF-like activity in heparinized plasma, as previously described.12 Results were read from a standard curve using recombinant human TNF tested in the same assay. Plasma n-terminal ANP was measured using a commercial radioimmunoassay.b PACs were measured using a commercial solid-phase radioimmunoassay without extraction previously validated for use in the dog.c,13 PRA was calculated from the difference in angiotensin I concentrations measured in 2 aliquots of plasma, 1 incubated at 37⬚C for 2 hours and the other kept at 4⬚C. Angiotensin I concentration was measured in plasma using a commercial radioimmunoassay.d Unless otherwise noted, all reported values are mean ⫾ standard deviation. The distributions of data were examined graphically. Data that were not normally distributed were subjected to logarithmic transformation. Baseline comparisons between groups employed independent t-tests. Analysis of variance with repeated measures was used to analyze changes in continuous data after the LS and MS diets. Ordinal

jvim 14

509 Mp 515 File # 09em

Sodium-Restricted Diets in Dogs with Heart Failure

515

Table 2. Characteristics of dogs enrolled in the study. No significant differences were found between all dogs enrolled in the study and dogs completing the study. CVD, chronic valvular disease; DCM, dilated cardiomyopathy; N, neutered; IVSd, interventricular septal thickness in diastole; IVSs, interventricular septal thickness in systole; LVIDd, left ventricular internal dimension in diastole; LVIDs, left ventricular internal dimension in systole; LVWd, left ventricular free-wall thickness in diastole; LVWs, left ventricular free-wall thickness in systole.

Age (years) Gender Body weight (kg) Median time since diagnosis (months) Aorta (cm) Left atrium (cm) Maximum left atrium (cm) IVSd (cm) IVSs (cm) LVIDd (cm) LVIDs (cm) LVWd (cm) LVWs (cm) a

All Dogs Enrolled (n ⫽ 18)

Dogs Completing Study (n ⫽ 14)

Dogs With CVD (n ⫽ 9)

Dogs With DCM (n ⫽ 5)

10.2 ⫾ 3.0 14 male (8 N) 4 female (4 N) 22.3 ⫾ 22.2 12.0 (range: 0.5–48.0) 2.1 ⫾ 0.8 3.4 ⫾ 0.8 5.3 ⫾ 1.4 0.8 ⫾ 0.3 1.3 ⫾ 0.3 4.7 ⫾ 1.3 3.0 ⫾ 1.4 0.8 ⫾ 0.3 1.6 ⫾ 1.6

10.8 ⫾ 2.9 12 male (8 N) 2 female (2 N) 21.7 ⫾ 24.1 12.0 (range: 0.5–48.0) 2.1 ⫾ 0.7 3.3 ⫾ 0.8 5.2 ⫾ 1.5 0.8 ⫾ 0.3 1.3 ⫾ 0.3 4.5 ⫾ 1.2 2.8 ⫾ 1.2 0.8 ⫾ 0.3 1.3 ⫾ 0.2

11.8 ⫾ 2.8 8 male (5 N) 1 female (1 N) 8.4 ⫾ 6.6 18.0 (range: 0.5–36.0) 1.7 ⫾ 0.2 2.9 ⫾ 0.6 4.4 ⫾ 1.1 0.7 ⫾ 0.1 1.2 ⫾ 0.1 3.9 ⫾ 0.8 2.1 ⫾ 0.7 0.7 ⫾ 0.1 1.2 ⫾ 0.1

8.7 ⫾ 2.3 4 males (3 N) 1 female (1 N) 42.8 ⫾ 24.3a 12.0 (range: 12.0–48.0) 2.7 ⫾ 0.9a 4.0 ⫾ 0.8a 6.6 ⫾ 1.0a 1.0 ⫾ 0.3 1.4 ⫾ 0.5 5.6 ⫾ 1.0a 4.0 ⫾ 1.1a 1.0 ⫾ 0.3a 1.4 ⫾ 0.3a

P ⬍ .05 between dogs with CVD and DCM.

data were analyzed using chi-square tests. Statistical analysis was performed using a commercial statistical software program.e

Results Eighteen dogs were enrolled in the study (11 with CVD and 7 with DCM). The owner of 1 dog with CVD elected to discontinue the study for unknown reasons within 1 week of enrollment. The owner of 1 dog with DCM did not feed the experimental diet exclusively and elected to discontinue the study on day 35. One year after initiation of the study, 2 dogs refused to eat the experimental diet. The food was analyzed and the fat was found to have precipitated, likely due to storage conditions. A new batch of food would have had different nutritional content, and the study was terminated prematurely. Subject characteristics are shown in Table 2 for all dogs enrolled in the study (n ⫽ 18), dogs completing the study (n ⫽ 14), dogs with CVD that completed the study (n ⫽ 9), and dogs with DCM that completed the study (n ⫽ 5). No differences were found between dogs completing the study and those that were excluded for either owner compliance issues or diet palatability issues. As expected, dogs with DCM were heavier and had larger mean left atrium, aorta, left ventricle, and left ventricular free wall compared to dogs with CVD (Table 2). All remaining data are reported for only the 14 dogs that completed the study. Before enrolling in the study, most dogs ate either a dry commercial dog food only (n ⫽ 6; 2 dogs with CVD, 4 dogs with DCM) or a combination of dry and canned food (n ⫽ 7; 5 dogs with CVD, 2 dogs with DCM). Only 1 dog (CVD) ate canned dog food exclusively. Although owners did not keep complete dietary records before enrolling in the study, the median sodium content in the dogs’ usual diets (not including treats and table food) was 0.06 g/100 kcal (range, 0.02–0.34 g/100 kcal), similar to that of the

MS diet used in the study. Palatability of the study diets was reported by the owners to be excellent for the first 14 dogs and resulted in a high compliance rate by owners. At the time of enrollment, 2 dogs were in medication category I, 5 were in category II, 5 were in category III, 0 were in category IV, 1 was in category V, and 1 was in category VI. At baseline, electrolyte abnormalities included hyponatremia (n ⫽ 4), hypokalemia (n ⫽ 1), and hypomagnesemia (n ⫽ 2). Eleven of 14 dogs had increased baseline concentrations of ANP, whereas 11 of 14 had increased PACs and 10 of 14 had increased PRA concentrations (Fig 1).

Dietary Effects in All Dogs Changes in body weight during the study period were not different between the LS and MS diets. Medication category and furosemide dosage also did not change significantly between diet groups. Serum sodium and chloride concentrations decreased significantly in dogs eating the LS diet compared to the MS diet (⫺4.0 versus ⫹1.4 mEq/L for sodium, P ⫽ .01; ⫺3.1 versus ⫹2.2 mEq/L for chloride, P ⫽ .02; Table 3). A significant sequence effect for sodium (P ⫽ .02) was identified and a carryover effect of the MS diet may have occurred. Serum alanine aminotransferase (Table 3) increased significantly more during feeding of the LS diet than during feeding of the MS diet (⫹12.1 versus ⫺11.6 U/L; P ⫽ .01) with a carryover effect of the LS diet (P ⫽ .02). Other data from the serum biochemistry profile, CBC, and urinalysis were not significantly different between diets. However, 5 dogs became hyponatremic during the 8-week study (all at the end of LS diet) and 4 dogs became hypochloremic during the study (4 at the end of the LS diet and 2 at the end of the MS diet). Only 1 dog was hypokalemic at the start of the study, but at the end of the LS diet period a different dog

516

Rush et al

Fig 1. Concentrations of serum aldosterone (A), plasma renin activity (B), n-terminal atrial natriuretic peptide (C), and tumor necrosis factor (D) at the beginning of the study (start study), the end of the low-sodium diet period (end LS diet), and the end of the moderate-sodium diet period (end MS diet). Dotted horizontal lines on each panel indicate the upper and lower limits of the reference ranges.

was hypokalemic and 1 was hyperkalemic. Two dogs were hyperkalemic at the end of the MS diet period. Six dogs were hypomagnesemic at the end of the LS diet period and 7 dogs were hypomagnesemic at the end of the MS diet period. Although PAC, PRA, and ANP increased more on the LS diet (Fig 1; Table 3), these changes did not reach statistical significance. No electrocardiographic or radiographic differences were

found in dogs while eating the LS versus MS diets. Systolic (⫺5.5 mm Hg versus ⫹1.2 mm Hg, P ⫽ .56) and diastolic (⫺5.5 mm Hg versus ⫹1.4 mm Hg, P ⫽ .35) blood pressures also were not different between LS and MS diets, respectively. Maximum left atrial size decreased in the dogs while eating the LS diet compared to the MS diet (⫺0.47 cm versus ⫹0.18 cm, P ⫽ .05; Fig 2). Standard left atrial size also decreased on the LS diet compared to the MS diet

Table 3. Mean electrolyte and neurohormonal concentrations and alanine aminotransferase (ALT) activities in dogs at baseline and after eating the low-sodium (LS) and moderate-sodium (MS) diets. Baseline Sodium (mEq/L) Chloride (mEq/L) Potassium (mEq/L) Magnesium (mg/dL) Renin (ng/mL/h) Aldosterone (pmol/L) Atrial natriuretic peptide (pg/mL) Tumor necrosis factor (pg/mL) ALT (U/L)

146.8 110.6 4.5 1.7 6.8 652 537 582 64.4

⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾

3.9 5.4 0.4 0.3 4.1 549 512 789 35.1

⌬ LS Diet ⫺4.0 ⫺3.1 ⫹0.3 ⫺0.1 ⫹0.8 ⫹995 ⫹98 ⫹162 ⫹18.1

⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾

4.2 4.3 0.6 0.3 4.0 1,241 373 1,897 17.2

⌬ MS Diet ⫹1.4 ⫹2.2 ⫹0.0 ⫺0.1 ⫺0.7 ⫹216 ⫺165 ⫹316 ⫺11.6

⌬ LS diet, change in values after eating the LS diet; ⌬ MS diet, change in values after eating the MS diet.

⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾

5.3 4.4 0.5 0.2 4.1 1,427 591 1,910 38.4

P Value .01 .02 .31 .78 .37 .15 .30 .87 .01

jvim 14

Sodium-Restricted Diets in Dogs with Heart Failure

509 Mp 517 File # 09em

517

Fig 2. (A) Mean (⫾SD) changes in echocardiographic parameters for dogs (n ⫽ 14) eating a low-sodium diet (closed bars) and a moderatesodium diet (open bars). (B) Mean (⫾SD) changes in echocardiographic parameters for dogs with chronic valvular disease (n ⫽ 9) eating a lowsodium diet (closed bars) and a moderate-sodium diet (open bars). *P ⬍ .05 between low-sodium and moderate-sodium diets; LA, standard left atrial dimension; MaxLA, maximum left atrial dimension; LVIDd/s, left ventricular internal dimension in diastole/systole; IVSd/s, interventricular septal thickness in diastole/systole; LVWd/s, left ventricular free-wall thickness in diastole/systole.

(⫺0.19 cm versus ⫹0.13 cm, P ⫽ .09), but this change did not reach statistical significance (Fig 2). Other echocardiographic measurements were not different between diets (Fig 2) for the 14 dogs as a group.

Dietary Effects in Dogs with CVD Dogs with CVD may respond differently to diet and other treatments than dogs with DCM. Consequently, dogs with CVD were analyzed separately. Too few dogs were enrolled in the DCM group to perform a similar analysis. When dogs with CVD were analyzed separately, serum sodium and chloride concentrations still decreased more on the diet LS than on the MS diet (⫺4.1 versus ⫹0.2 mEq/ L for sodium, P ⫽ .05; ⫺3.9 versus ⫹2.6 mEq/L for chlo-

ride, P ⫽ .05). Other data from a CBC, serum biochemistry profile, and urinalysis did not differ between diets. No significant changes were found for neurohormones. Vertebral heart scores decreased significantly in dogs with CVD eating the LS diet (⫺0.3) compared to the MS diet (⫹0.5, P ⫽ .05). Several echocardiographic parameters decreased significantly in dogs on the LS compared to the MS diet, including right ventricular diameter in diastole (⫺0.12 versus ⫹0.10 cm, P ⫽ .03), left ventricular internal dimension in diastole (⫺0.19 versus ⫹0.29 cm, P ⫽ .006), left ventricular internal dimension in systole (⫺0.04 versus ⫹0.15 cm, P ⫽ .02), standard left atrial dimension (⫺0.17 versus ⫹0.21 cm, P ⫽ .03), and maximum left atrial dimension (⫺0.63 versus ⫹0.38 cm, P ⫽ .02). In addition, EDVI (⫺24.4 versus ⫹35.2 mL/m2, P ⫽ .02) and ESVI

518

Rush et al

(⫺1.5 versus ⫹5.2 mL/m2, P ⫽ .04) decreased significantly in dogs on the LS diet compared to the MS diet. No electrocardiographic differences were found between dogs on the diets.

Discussion Clinically, few changes were noted between the dogs eating the LS and MS diets. Both diets were judged to be extremely palatable, and food intake was not a problem. Neither medication category nor furosemide dosage changed significantly between groups. As in previous studies, electrolyte abnormalities were common, even before beginning the study.14 The high frequency of abnormalities that developed during the current study suggests that changes in dietary composition may exacerbate the electrolyte abnormalities in dogs with heart failure. For example, the LS and MS diets used in the current study seemed to be too low in magnesium to maintain normal serum magnesium concentrations in 50% of dogs. Conversely, many commercial diets designed for dogs in heart failure (including the ones used here) contain high potassium concentrations to compensate for the potential urinary loss of potassium due to furosemide use. Current widespread use of ACE inhibitors, which promote potassium retention, may necessitate re-evaluation of optimal dietary potassium content because both the present study and previous studies have identified hyperkalemia as a frequent occurrence with use of ACE inhibitors.4 Unlike healthy dogs, which maintain serum sodium and chloride concentrations despite low or high dietary concentration, serum sodium and chloride concentrations decreased significantly in dogs eating the LS diet in the present study. Echocardiographic changes supported a beneficial effect of the LS diet, but without mortality data the long-term benefit of these results cannot be determined. The reduction in cardiac size observed with the LS diet was most pronounced in dogs with CVD. However, these cardiac changes were not reflected in changes in neurohormone concentrations. These results differ from those of a previous study of dogs with asymptomatic cardiac disease in which a low-sodium diet caused a larger increase in PRA and PAC than did a high-sodium diet.9 The different results of the 2 studies may be due to the presence of heart failure in the current study population, large variation in neurohormone concentrations, small sample size, and medications such as ACE inhibitors and furosemide. For example, furosemide alone induces significant increases in serum creatinine concentrations, PACs, and PRA in healthy geriatric dogs.15 Although medications may have influenced the neurohormonal concentrations, the study was designed to determine the effects of diet, rather than drugs, on the measured outcomes. Therefore, dogs were randomized to the 2 different diet groups so that drug effects also would be randomized. In addition, dogs were stable on medication at the time of the study so that the differences seen between the 2 diets would be due to diet rather than to drugs. Finally, no difference was found between the LS and MS diets with regard to change in medication category or furosemide dosage. One difficulty in designing a study of this type is select-

Table 4. Mean sodium dose and dosage provided by previous studies and by the current study.a Low-Sodium Diet

Study

Control Diet (Moderate-High Sodium)

mg/day

mg/kg/d

mg/day

mg/kg/d

213 57 225 214

9 2 16 15

1,034 1,856 1,288 1,220

41 64 92 87

Dogs with mild cardiac disease (without heart failure) Pedersen et al9 162 17 918

96

Dogs with heart failure (current study) All dogs 535 24 CVD 260 31 DCM 884 21

42 53 36

Healthy dogs Hamlin et al2 Morris et al3 Pedersen et al5 Koch et al7

940 446 1,546

CVD, chronic valvular disease; DCM, dilated cardiomyopathy. a Based on reported mean body weight and dietary sodium content. Maintenance energy requirements were calculated using the formula: 132(body weight in kg)0.75.16

ing the sodium content of the 2 diets because no guideline exists for a specific level of restriction other than guidelines used in studies of human patients. The National Research Council recommends a minimum sodium requirement for dogs of 11 mg/kg/d, but recommendations for dogs with CHF often are below this amount.16 In 1991, recommendations made for dogs with heart disease were 15–25 mg/ kg/d for mild heart disease, 10–15 mg/kg/d for moderate heart disease, and 4–10 mg/kg/d for severe heart disease.17 More recent recommendations do not specifiy amounts and concentrate rather on the percentage of sodium in the diet (ie, moderately restricted or severely restricted). A similar method was used to select the sodium content of the 2 diets used in the present study. However, this method can mask differences in the total sodium dose each dog receives because of differences in energy content of diets and differences in food intake of individual dogs. When one compares the sodium content of the diets used in the current study to those used in previous studies on the basis of the mean sodium dosage, a number of differences become apparent (Table 4). The sodium dosage provided by the low-sodium diets in previous studies ranged from 2 to 17 mg/kg/d, and most dry commercial low-sodium dog foods provide approximately 10–12 mg/kg/d. Moderate- or high-sodium diets used as control diets in previous studies provided between 41 and 96 mg/kg/d, and most commercial premium dry dog foods provide approximately 40 mg/kg/ d. The median sodium content in the prestudy diets of dogs in the current study (not including treats and table food) was 52 mg/kg/d (range, 9–201 mg/kg/d). The diets used in the current study provided 24 mg/kg/d (LS diet) and 42 mg/kg/d (MS diet) for all 14 dogs. For dogs with DCM, an even smaller difference occurred between the 2 diets because of differences in food intake on a kcal/kg basis. Most previous studies have used control diets providing up to 32 times the sodium content of the LS diet, whereas the MS

jvim 14

Sodium-Restricted Diets in Dogs with Heart Failure

diet in the current study provided less than 2 times the sodium content of the LS diet. This small difference may have blunted diet-related differences in clinical, echocardiographic, and neurohormonal parameters. Another issue not addressed in this study is the role of chloride, which also may be important in the clinical and neurohumoral responses in heart failure. In hypertension, for example, dietary chloride also must be restricted to achieve the full benefit of a sodium-restricted diet.18 In the current study, beneficial effects of the LS diet on cardiac size were much more pronounced in dogs with CVD than in dogs with DCM. Although some clinicians treat all dogs with CHF similarly, regardless of underlying cause, fundamental differences in these 2 diseases may require tailored medical and nutritional therapy. Alternatively, differences may have been due to the fact that dogs with DCM may have experienced a much smaller difference in sodium intake while on the LS and MS diets than did the dogs with CVD, because of differences in caloric intake on a kcal/kg basis (Table 4). This effect could have limited detectable differences between the 2 groups. A number of limitations may have affected the current study. One factor is the small number of subjects because of early termination of the study. Sample size calculations showed that a minimum of 24 dogs would have been needed to demonstrate differences in echocardiographic and neurohormonal variables between the 2 diets. Although some differences in echocardiographic data were found, statistical power was insufficient to detect a difference in neurohormonal concentrations. Another limitation is the fact that the sodium dosages provided by the LS and MS diets were not markedly different, as previously discussed. It would be desirable for future studies to design diets on a dosage basis so that dogs in each group would receive a consistent daily dosage of sodium (and chloride). In addition, a larger difference in sodium dosage between the 2 groups would be recommended. Another limitation of this study was the use of clinical patients. Patients received a variety of medications at different dosages and had different home environments, which, given the small sample size, may have introduced considerable variability. Future studies ideally would use a larger sample size with a single cause of CHF to minimize variability. Finally, results of this study may not be applicable to the general population of dogs with CHF. Dogs enrolled in this study were a select population because of the requirement for exclusive feeding of the study diets for 8 weeks. This requirement made recruitment difficult because many owners were not willing to completely discontinue treats and table food. This issue raises the question of whether dogs for which a low-sodium diet is recommended actually receive a low-sodium diet, given the other foods typically included in the total diet. Exclusive feeding of a low-sodium diet may not be a realistic expectation in terms of owner compliance. Prescribing sodium intake on a dosage basis allows the clinician to factor in other foods so that the total recommended dosage can still be achieved. Even if future studies confirm the benefits of a low-sodium diet in dogs with CHF, a more realistic approach may be needed to recommend a commercial dog food that fits with an owner’s feeding practices.

509 Mp 519 File # 09em

519

Footnotes WALTHAM Veterinary Diet Canine Low Sodium, Effem, Inc, Bolton, Ontario, Canada b Phoenix Pharmaceuticals, Mountain View, CA c Coat-a-Count Aldosterone, Diagnostic Products Corporation, Los Angeles, CA d Angiotensin I, Biotecx Radioimmunoassay Kit, Biotecx Laboratories, Inc, Houston, TX e SYSTAT statistical software, Chicago, IL a

Acknowledgments We are grateful for the technical assistance of Denise Heublein and Susan Lombardini. This study was supported by a grant from the Waltham Centre for Pet Nutrition.

References 1. Pensinger R. Dietary control of sodium intake in spontaneous congestive heart failure in dogs. Vet Med 1964;59:752–784. 2. Hamlin R, Smith R, Smith C, Powers T. Effects of a controlled electrolyte diet, low in sodium on healthy dogs. Vet Med 1964;59: 748–751. 3. Morris M, Patton R, Teeter S. Low sodium diet in heart disease: How low is low? Vet Med 1976;71:1225–1227. 4. Roudebush P, Allen T, Kuehn N, et al. The effect of combined therapy with captopril, furosemide, and a sodium-restricted diet on serum electrolyte concentrations and renal function in normal dogs and dogs with congestive heart failure. J Vet Intern Med 1994;8:337–342. 5. Pedersen H, Koch J, Jensen A, et al. Some effects of a low sodium diet high in potassium on the renin-angiotensin system and plasma electrolyte concentrations in normal dogs. Acta Vet Scand 1994;35:133–140. 6. Pedersen H, Koch J, Bie P, et al. Effects of a low sodium diet with a high potassium content on plasma endothelin-1, atrial natriuretic peptide and arginine vasopressin in normal dogs. J Vet Med 1994;41: 713–716. 7. Koch J, Pedersen H, Jensen A, et al. Short term effects of acute inhibition of the angiotensin-converting enzyme on the renin-angiotensin system and plasma atrial natriuretic peptide in healthy dogs fed a low-sodium diet versus a normal-sodium diet. J Vet Med 1994;41: 121–127. 8. Barger A, Ross R, Price H. Reduced sodium excretion in dogs with mild valvular lesions of the heart and in dogs with congestive failure. Am J Physiol 1955;180:249–260. 9. Pedersen H. Effects of mild mitral valve insufficiency, sodium intake, and place of blood sampling on the renin-angiotensin system in dogs. Acta Vet Scand 1996;37:109–118. 10. Keene BW, Rush JE. Therapy of heart failure. In: Ettinger SJ, Feldman EC, eds. Textbook of Veterinary Internal Medicine, 4th ed. Philadelphia, PA: WB Saunders; 1995:867–892. 11. Buchanan J, Bucheler J. Vertebral scale system to measure canine heart size in radiographs. J Am Vet Med Assoc 1995;206:194– 199. 12. Espevik T, Nissen-Meyer J. A highly sensitive cell line, WEHI 164, close 13, for measuring cytotoxic factor/tumor necrosis factor from human monocytes. J Immunol Methods 1986;95:99–105. 13. Willard M, Refsal K, Thacker E. Evaluation of plasma aldosterone concentrations before and after ACTH administration in clinically normal dogs and in dogs with various diseases. Am J Vet Res 1987; 48:1713–1718. 14. Cobb M, Michell A. Plasma electrolyte concentrations in dogs receiving diuretic therapy for cardiac failure. J Small Anim Pract 1992; 33:526–529.

520

Rush et al

15. Roudebush P, Allen TA. Effect of dietary sodium and furosemide on hematologic, biochemical and endocrine parameters in normal geriatric dogs. J Vet Intern Med 1996;10:171 (abstract). 16. National Research Council. Nutrient Requirements of Dogs. Washington, DC: National Academy Press; 1985:3, 44. 17. Morris ML, Ettinger SJ. Dietary modification in cardiac disease.

In: Ettinger SJ, ed. Textbook of Veterinary Internal Medicine. Philadelphia, PA: WB Saunders; 1991:233–238. 18. Boegehold M, Kotchen T. Relative contributions of dietary Na⫹ and Cl⫺ to salt-sensitive hypertension. Hypertension 1989;14: 579–583.