A high-sodium diet is associated with acute decompensated heart failure in ambulatory heart failure patients: a prospective follow-up study 1 3

Transcription

1 See corresponding editorial on page 229. A high-sodium diet is associated with acute decompensated heart failure in ambulatory heart failure patients: a prospective follow-up study 1 3 JoAnne Arcand, Joan Ivanov, Alexa Sasson, Vanessa Floras, Abdul Al-Hesayen, Eduardo R Azevedo, Susanna Mak, Johane P Allard, and Gary E Newton ABSTRACT Background: A low-sodium diet is an accepted treatment of patients with heart failure (HF), although minimal evidence exists on the appropriate amount of sodium intake for this population. Certain HF guidelines have liberalized dietary sodium recommendations, which actually exceed guidelines for healthy adults. Objectives: We tested the hypothesis that high sodium intake is related to acute decompensated HF (ADHF) in ambulatory HF patients. Secondary outcomes included all-cause hospitalization and mortality. Design: We prospectively enrolled medically stable, ambulatory patients with systolic HF (n = 123; mean 6 SD age: y) from 2 outpatient HF clinics from 2003 to Baseline estimates of dietary sodium and other nutrient intakes were obtained from two 3-d food records. Results: The median follow-up time was 3.0 y. Mean (6SD) sodium intakes were , , and g Na/d in the lower, middle, and upper tertiles, respectively. Cumulative ADHF event rates at 3 y were %, %, and % in the low, middle, and upper tertiles, respectively (log-rank P = 0.001). For ADHF, the upper tertile was associated with an adjusted hazard ratio of 2.55 (95% CI: 1.61, 4.04; P, 0.001). Time-to-event probabilities were significant for mortality (log-rank P = 0.022) but not for all-cause hospitalization (log-rank P = 0.224). The high-sodium tertile was associated with an adjusted hazard ratio of 1.39 (95% CI: 1.06, 1.83; P = 0.018) for all-cause hospitalization and 3.54 (95% CI: 1.46, 8.62; P = 0.005) for mortality. Conclusions: To our knowledge, this study provides the first prospective evidence that ambulatory HF patients who consume higher amounts of sodium are at greater risk of an ADHF event. These data provide support for more stringent sodium intake guidelines than those currently recommended for HF patients. Am J Clin Nutr 2011;93: INTRODUCTION Heart failure, an exceedingly common cardiac condition, is associated with high rates of morbidity and mortality including frequent hospital admissions (1, 2). Sodium restriction has remained a core therapy for managing sodium and fluid retention in patients with acute and chronic heart failure (3 6). However, despite the theoretical basis for restricting sodium intake, the role of dietary sodium in heart failure management has not been fully established in the current era of highly effective medical therapy. To our knowledge, there is no empirical evidence that demonstrates beneficial or adverse outcomes associated with sodium intake in compensated, appropriately medicated heart failure patients, which is a group that makes up the majority of the heart failure population. Recent data have highlighted an alternative hypothesis that sodium restriction may be harmful in heart failure patients. Two clinical trials of sodium restriction showed that a low-sodium diet in combination with high-dose diuretics and fluid restriction promoted increases in hospital readmission and death (7, 8). Mechanistic investigations have also shown an adverse neurohumoral activation after short-term sodium restriction, which further challenged the role of sodium restriction in ambulatory heart failure patients (9 11). Practice guidelines for heart failure patients also demonstrate reduced enthusiasm for dietary sodium restriction. Limiting sodium intake to 2 3 g Na/d is typically recommended, based on expert consensus, for patients with symptomatic heart failure on optimal medical therapy including diuretics (3, 5, 6). However, the current heart failure guidelines of the American College of Cardiology/American Heart Association suggest that as much as 4 g Na/d may be appropriate for a similar patient group (stage C heart failure) (4). This sodium intake guideline is above the recommended tolerable upper intake of 2.3 g Na/d for the healthy adult population (12). We tested the hypothesis that high sodium intake is associated with adverse outcomes, including acute decompensated heart failure (ADHF), and secondary outcomes of all-cause hospitalization and mortality in a cohort of stable, ambulatory, compensated heart failure patients. 1 From the Department of Nutritional Sciences, University of Toronto, Toronto, Canada (JA and JPA); the Division of Cardiology, Department of Medicine, Mount Sinai Hospital and University Health Network, Toronto, Canada (JA, JI, AS, VF, SM, ERA, and GEN); the Institute for Clinical Evaluative Sciences, Toronto, Canada (JI); the Heart and Vascular Program, Department of Medicine, St Michael s Hospital, Toronto, Canada (AA-H); and the Division of Gastroenterology, Department of Medicine, University Health Network, Toronto, Canada (JPA). 2 Supported by a grant-in-aid from the Heart and Stroke Foundation of Ontario (NA-5897). 3 Address reprint requests and correspondence to GE Newton, Mount Sinai Hospital, 600 University Avenue, Suite 1543, Toronto, ON M5G 1X5, Canada. gnewton@mtsinai.on.ca. Received July 22, Accepted for publication October 25, First published online November 17, 2010; doi: /ajcn Am J Clin Nutr 2011;93: Printed in USA. Ó 2011 American Society for Nutrition

2 DIETARY SODIUM AND HEART FAILURE 333 SUBJECTS AND METHODS Study participants Patients included in this analysis were enrolled in a study that assessed dietary intake in heart failure patients. Comprehensive dietary information from this cohort has been published elsewhere (13). A total of 146 patients were approached to participate in the study. Data were available from 123 heart failure patients in the final analysis because 11 patients declined participation, and 12 patients were excluded for failure to submit 2 sets of food records. The sample included eligible ambulatory heart failure patients who were consecutively enrolled between 2003 and 2007 from multidisciplinary heart failure programs in 2 tertiary care hospitals. Eligible patients were y of age, had a left ventricular ejection fraction,35%, were stable without any hospitalizations or emergency room visits in the 3 mo before study entry, and were taking optimal medical therapy including an angiotensin-converting enzyme inhibitor, angiotensin receptor blocker, or a b-blocker. Patients were excluded if they had significant renal dysfunction defined by a serum creatinine concentration.140 mmol/l, hyponatremia defined by a serum sodium concentration,130 mmol/l, or cardiac cachexia defined as unintentional weight loss.10% over 6 mo. No patients were institutionalized (ie, in a nursing home), and all patients consumed a self-selected diet. This study was approved by the Research Ethics Board at Mount Sinai Hospital (Toronto, Canada) and St Michael s Hospital (Toronto, Canada), and all patients gave written informed consent to participate in the study. Assessment of sodium intake Dietary intakes were assessed by two 3-d food records with one at study entry and the other after 6 12 wk. This approach was applied to capture variations in food selection that resulted from seasonal and grocery-cycle factors. Mean intakes reported by these records were used to estimate habitual intakes of sodium and other nutrients. Patients were instructed to record all food and beverages consumed and were blinded to the fact that sodium was one of the primary nutrients under investigation. Patients were required to weigh or measure their food with a scale or with standard household measures, which were provided to study participants. Patients recorded if salt was added at the table or during cooking. If the amount of salt could not be measured by a scale or volume measures, patients were asked to record the number of shakes added to the food so that sodium could be estimated. Recorded days were selected by study investigators and included at least one weekend day because food intakes often differ between weekday and weekend days. All food records were reviewed by a dietitian to clarify food-item descriptions and to identify any missing food items. Food records were analyzed by trained coders, who were blinded to patient identities, with a nutrient software program (ESHA Food Processor SQL v.10.1; ESHA Research, Salem, OR). Food records entered for analysis were checked twice for accuracy by independent coders and by the study dietitian. A subgroup of patients also completed two 24-h urine collections for the verification of reported sodium intake, which was previously reported (13). However, multipleday food records at 2 time points were considered a feasible and ideal approach to capture the high day-to-day variability in sodium consumption (14). Outcomes and follow-up The primary outcome for this study was ADHF. Secondary outcomes included all-cause hospitalization and death or transplantation. Outcomes were assessed from the time of enrollment to the completion date of 31 October Events were identified through medical chart reviews and follow-up telephone interviews with patients and their families. ADHF is a condition that often occurs in patients with preexisting heart failure and is defined as an exacerbation of dyspnea, edema, or fatigue that requires urgent medical treatment (15). The ADHF outcome in this analysis includes those patients hospitalized with a primary diagnosis of ADHF and patients who visited the emergency department for management of ADHF. Hospitalization and emergency department visits were confirmed by discharge summaries that were reviewed for classification of the primary diagnosis. The outcome of mortality included patients who died and patients who received a heart transplant during the follow-up period. The confirmation of death and transplant was obtained by chart review or by contacting family members. Study outcomes were adjudicated by 2 cardiologists (GEN and AA-H) blinded to dietary intake. Statistical analyses All statistical analyses were performed with SAS version 9.1 (SAS Institute Inc, Cary, NC). With the use of PROC RANK procedure (SAS version 9.1; SAS Institute Inc), patients were divided into tertiles based on mean sodium intakes. This approach was planned prospectively, based on our previously published data (13) to split our sample into 3 groups that constituted low, moderate, and high sodium intakes. Continuous variables were described by using descriptive statistics including means, medians, and SDs. Frequencies were used for categorical variables. One-factor analysis of variance was used for univariate comparisons of continuous variables. When the F ratio from the analysis of variance was significant, Scheffe s post hoc test was used to specify pairwise differences. Pearson s chi-square test was used to determine differences among tertiles as they related to categorical variables. The follow-up period was calculated as the time from the completion of the food records to the date of final data collection and follow-up for clinical events. For hypothesis testing, we truncated the follow-up period to 3 y, which was chosen based on a small number of patients followed beyond this time point. The cumulative event probability for time-to-event outcomes was calculated by using the Kaplan-Meier method. Cox regression analysis was used to identify hazard ratios and 95% CIs. Cox regression analysis was first used to determine the independent predictors of the primary outcome. Clinically relevant covariates or those with a significant univariate hazard ratio, as defined by P, 0.35, were included in the multivariate model. Cox regression analysis was also used to determine risk associated with a high sodium intake. To test the hypothesis that a high sodium intake was associated with increased risk of ADHF and secondary outcomes, the upper tertile was compared with the lowest and middle tertiles. Covariates in the multivariate Cox model were selected based on clinical relevance or because a given covariate changed the effect of the high-sodium tertile when included in the model. The covariates included age, sex, caloric intake, left-ventricle ejection fraction, body mass index

3 334 ARCAND ET AL (in kg/m 2 ), furosemide use, and b-blocker use. Because our sample size was small, we developed 2 models of one partially adjusted model and one fully adjusted model that included all covariates. The proportional hazards assumption was confirmed by visual examination of the log (minus log) curves. Significant P values were defined at an a level,0.05. RESULTS Follow-up data were available for all patients. No hospitalizations occurred during the 3-mo period between dietary intake assessments. Mean sodium intakes were g Na/d (range: g Na/d), g Na/d (range: g Na/d), and g Na/d (range: g Na/d) in the low, middle, and upper tertiles, respectively. The mean follow-up time was 2.3 y (median: 3.0 y; range: y). Patients in the lowest sodium tertile had a longer follow-up time ( y; P, 0.05) compared with that of patients in the highest sodium tertile ( y) but not compared with that of patients in the middle sodium tertile ( y). Patients in this study had a mean age of y, a left ventricle ejection fraction of %, and a body mass index of Characteristics of patients in each sodium intake tertile are reported in Table 1. Compared with patients in the lower and middle tertiles, patients in the upper tertile had a significantly higher sodium intake, were more likely to be men, and had higher intakes of calories, fluid, and macronutrients. Distribution of clinical characteristics and comorbidities were similar between tertiles. Medication use was also similar among tertiles; however, differences in frequencies of b-blocker use were observed. The total number of months in which patients participated in the multidisciplinary heart failure program were similar between the tertiles (P = NS; data not shown). In addition, the number of clinic visits per patient in the year before enrollment in the study were similar between tertiles (P = NS; data not shown). Cumulative ADHF events for the lower, middle, and upper tertiles were 5 6 3%, 5 6 3%, % after 1 y and %, %, and % after 3 y (log-rank P = 0.001; Figure 1A). The hazard ratio adjusted for age, sex, caloric intake, and left-ventricular ejection fraction was 2.36 (95% CI: 1.54, 3.60; P, 0.001), and the hazard ratio was 2.55 (95% CI: 1.61, 4.04; P, 0.001) when the model was further adjusted for additional covariates including body mass index, b-blocker use, and furosemide use (Table 2). High sodium intake (2.8 g Na/d) was the only independent predictor of the primary endpoint of ADHF (hazard ratio: 1.66; 95% CI: 1.23, 2.24). At 3 y, cumulative events for all-cause hospitalization were %, %, and % for patients in the lower, middle, and upper tertiles, respectively, which was not significantly different among tertiles (log-rank P = 0.224; Figure 1B). TABLE 1 Patient characteristics g Na/d (n = 41) Na/d (n = 41) 2.8 Na/d (n = 41) P Men (%) Age (y) BMI (kg/m 2 ) NYHA class I and II (%) NYHA class III and IV (%) LVEF (%) Systolic BP (mm Hg) Diastolic BP (mm Hg) Heart rate (beats/min) Serum sodium (mmol/l) Serum creatinine (mmol/l) Coronary artery disease (%) Diabetes (%) Hypertension (%) Current smoker (%) Furosemide use (%) Furosemide dose (mg/d) Spironolactone use (%) b-blocker use (%) ACEI/ARB use (%) Dietary intake Sodium (g) ,5,0.001 Energy (kcal) ,5,0.001 Carbohydrate (g) ,5,0.001 Protein (g) ,0.001 Fat (g) ,5,0.001 Fluid (L) NYHA, New York Heart Association; LVEF, left-ventricle ejection fraction; BP, blood pressure; ACEI/ARB, angiotensin-converting enzyme inhibitor/ angiotensin receptor blocker. One-factor ANOVA with Scheffe s post hoc analysis was used to conduct between-group comparisons. 2 Mean 6 SD (all such values). 3 Middle tertile ( g Na/d) compared with lower tertile (1.9 g Na/d), P, Upper tertile (2.8 g Na/d) compared with lower tertile (1.9 g Na/d), P, Upper tertile (2.8 g Na/d) compared with middle tertile ( g Na/d), P, 0.05.

4 DIETARY SODIUM AND HEART FAILURE 335 TABLE 2 Outcomes for heart failure patients with a high sodium intake g Na/d (n = 123) P FIGURE 1. Kaplan-Meier plots for sodium intakes and clinical outcomes in heart failure patients. A: Acute decompensated heart failure (ADHF). B: All-cause hospitalization. C: Mortality. The mortality variable included patients who received a heart transplant. Mean (6SD) sodium intakes were g Na/d in the lower tertile (red line), gNa/dinthemiddle tertile (black line), and g Na/d in the upper tertile (blue line). Risk of all-cause hospitalization in the highest sodium tertile was only significant after adjustment for covariates in both the partially and fully adjusted multivariable models (Table 2). The Acute decompensated heart failure Univariate model 1.66 (1.23, 2.24),0.001 Multivariate model (1.54, 3.60),0.001 Multivariate model (1.61, 4.04),0.001 All-cause hospitalization Univariate model 1.20 (0.97, 1.47) Multivariate model (1.04, 1.79) Multivariate model (1.06, 1.83) Mortality 2 Univariate model 1.72 (1.08, 2.73) Multivariate model (1.31, 4.47) Multivariate model (1.46, 8.62) All values are hazard ratios (95% CIs) for the upper tertile compared with the lowest and middle tertiles. Multivariate model 1 included the covariates age, sex, energy intake, and left-ventricle ejection fraction. Multivariate model 2 included model 1 covariates plus the covariates b-blockers, furosemide, and BMI. 2 Included patients who had a transplant during the follow-up period. 3-y cumulative mortality was 8 6 5% in the lower tertile and % in the highest tertile with no events observed in the middle tertile (log-rank P = 0.022; Figure 1C). One patient in the lowest tertile and 5 patients in the highest tertile died of cardiovascular causes, which included conditions such as worsening heart failure, myocardial infarction, and stroke. One patient each in the lowest and highest tertiles died of noncardiovascular causes (ie, cancer, accidental). One patient in the lowest tertile had a heart transplant. Adjusted and unadjusted hazard ratios for mortality were significant (Table 2). DISCUSSION To our knowledge, this is the first study to show any relation between dietary sodium intake and subsequent clinical outcomes in stable compensated heart failure patients. Specifically, our study showed that heart failure patients who consumed a highsodium diet (2.8 g Na/d), compared with patients who consumed lower amounts of dietary sodium, had a higher incidence of early ADHF. Furthermore, we showed that patients who consumed a high-sodium diet had a 2.5-fold increased risk of ADHF and an elevated risk of all-cause hospitalization and mortality when adjusted for covariates. By demonstrating increased risk associated with high sodium intakes, this analysis challenges recent data that concluded that lower-sodium diets are associated with adverse outcomes in heart failure (7, 8) and suggests that current heart failure guidelines should be more aggressive in recommending lower sodium intake. These findings are particularly novel and relevant because they apply to a broad group of ambulatory heart failure patients with systolic heart failure who are already receiving optimal medical therapy. In the past, the therapeutic importance of sodium restriction in the setting of heart failure was well accepted. The recent controversy associated with dietary sodium restriction in heart failure has risen from contradictory findings reported between retrospective observational studies and by more recent clinical and mechanistic investigations. Observational studies that identified a high-sodium diet as a risk factor for heart failure hospitalization

5 336 ARCAND ET AL assessed dietary sodium retrospectively and did not rely on validated dietary assessment techniques, such as 24-h urines collections or food records to classify sodium intake (16, 17). Disease-management trials have included a low-sodium diet as part of a multidisciplinary approach; however, because these studies tested several concurrent interventions, the independent contribution of a low-sodium diet to outcomes could not be assessed (18). Other small clinical studies that restricted sodium intake as an educational intervention for heart failure patients were underpowered to explore outcomes (19, 20). Finally, mechanistic investigations have shown an adverse neurohumoral activation in heart failure patients after short-term dietary sodium restriction (9 11). More recently, Paterna et al (7) conducted a randomized controlled trial in 252 heart failure patients who had a recent ADHF hospitalization. Patients were randomly assigned to receive a sodium-restricted diet (80 mmol Na/d; 1840 mg Na/d) or a normal sodium diet (120 mmol Na/d; 2760 mg Na/d). Along with sodium restriction, patients were simultaneously prescribed a fluid restriction (1 2 L fluid/d) as well as a high dose of furosemide (250 or 500 mg furosemide twice daily) at study randomization, which remained unadjusted over the study period. Based on these co-interventions, it is difficult to elucidate the sole contribution of sodium intakes to clinical outcomes. Furthermore, concerns related to background medical therapies limited the generalizability to the general heart failure population because only 9% of patients in each study group received b-blocker therapy. Nevertheless, these investigators showed that patients who received the sodium-restricted diet as part of their study treatment regimen had an increased risk of hospitalization and the combined endpoint of hospitalization and mortality after 6 mo. These findings are contrary to the data presented in the current study. Certainly, the concurrent interventions in this study may have contributed to adverse outcomes because loop diuretic therapy is associated with associated with neurohumoral activation and mortality in heart failure (21, 22). This is relevant in the context of the already described neurohormonal activation that occurs when dietary sodium is restricted (9 11). Furthermore, we studied ambulatory patients who were euvolemic with longstanding stable heart failure on optimal medical therapy, which is a group that is highly generalizable to the larger heart failure population. Our study took an observational approach to evaluate clinical outcomes associated with sodium intake. Although causality cannot be proven by using observational data, this is the first study, to our knowledge, to link a high-sodium intake to adverse outcomes in this relatively low-risk group of heart failure patients. A high-sodium diet has been associated with adverse outcomes in healthy populations, including incident hypertension and related outcomes of stroke and heart failure. The Dietary Reference Intake recommendation for sodium for healthy adults (aged y) is an adequate intake of 1.5 g Na/d and a tolerable upper intake of 2.3 g Na/d (12). The average sodium intake of most individuals with a North American diet far exceeds these recommendations (23). Our recent report from the same patient cohort identified that sodium intake was similar in stable heart failure patients compared with in an age-matched control group (13). For patients living with established heart failure, an excessive sodium intake has even greater consequences because abnormalities in renal sodium handling promote sodium and fluid retention (24). Although medical therapies including b-blockers and angiotensin-converting enzyme inhibitors aim to correct renal sodium handling, our data indicated that a lower-sodium diet may be required to assist in maintaining euvolemia and clinical stability over the long term. Based on the average intakes in our tertiles, our data suggested that the sodium intake recommendation for stable heart failure patients should not be dissimilar from the established tolerable upper intake for healthy adults of 2.3 g Na/d. This suggestion is more aggressive than current heart failure treatment guidelines, which recommend 3 4 g Na/d sodium diet (4). Whether heart failure patients would benefit from an even lower amount of sodium intake, similar to the Dietary Reference Intake adequate intake of 1.5 g Na/d, will require further study. This may be especially true for patients in refractory heart failure for which a daily sodium intake 2 g Na/d is recommended (3 5). The sample of heart failure patients studied was relatively small; however, we included a well-characterized group with a rigorous dietary assessment for the classification of sodium intake. Despite the relatively small sample size, this is the first observation, to our knowledge, that relates sodium intake to any outcome for patients with stable chronic heart failure. It could be argued that patients nonadherent to sodium-intake guidelines are also more likely to be nonadherent to medical therapies and other beneficial lifestyle factors. Indeed, an analysis of the CHARM (Candesartan in Heart Failure Assessment of Reduction in Morbidity and Mortality) trial data showed that the adherence to medical therapy was an indicator of better outcomes, regardless of whether the adherence was to the placebo or treatment drug (25). The adherence to medical therapies was not measured in our study based on the complexities of applying proper measurement techniques. However, there are relevant differences in factors that influence an adherence diet and medication. For example, adherence to sodium restriction is dependent on knowledge of the sodium content of foods, access to low sodium foods, desire to follow a low-sodium diet, and related social and lifestyle factors that allow for sodium reduction (26). In contrast, medication adherence is more likely related to the cost of the medication, potential side effects, as well as memory and dosing schedules (27). To our knowledge, there are no studies that have directly compared the relation between diet and medication adherence; therefore, these factors should be considered independently. In conclusion, the findings of this study provide insight into the role of dietary sodium in the management of heart failure patients who were clinically stable and received optimal medical therapy. To our knowledge, this is the first evidence that stable heart failure patients who consumed high amounts of sodium were at greater risk of early ADHF hospitalization compared with patients who consumed lower amounts of dietary sodium. We thank the clinical and administrative staff of the heart function clinics at Mount Sinai Hospital and St Michael s Hospital. We also thank Mavra Ahmed and Chelsea Kaplansky for their assistance with data collection and analysis. The authors responsibilities were as follows GEN: had full access to all study data and took full responsibility for the integrity and accuracy of the data; GEN, JPA, JI, AA-H, SM, ERA, and JA: were responsible for the study design and concept; JA, VF, and AS: were responsible for logistical and administrative details, data acquisition and analyses, and database management; JA, GEN, SM, and ERA: were responsible for data acquisition; GEN and AA-H: adjudicated clinical outcome data; JA, JI, and GEN: conducted the statistical analyses and interpretation of the data; JA: drafted the manuscript;

6 DIETARY SODIUM AND HEART FAILURE 337 and all authors: were involved in the critical revision of the manuscript. None of the authors had a conflict of interest. REFERENCES 1. Ko DT, Tu JV, Masoudi FA, et al. Quality of care and outcomes of older patients with heart failure hospitalized in the United States and Canada. Arch Intern Med 2005;165: Ko DT, Alter DA, Austin PC, et al. Life expectancy after an index hospitalization for patients with heart failure: a population-based study. Am Heart J 2008;155: Arnold JM, Liu P, Demers C, et al. Canadian Cardiovascular Society consensus conference recommendations on heart failure 2006: diagnosis and management. Can J Cardiol 2006;22: Hunt SA. ACC/AHA 2005 guideline update for the diagnosis and management of chronic heart failure in the adult: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2005;46:e Heart Failure Society of America. Nonpharmacologic management and health care maintenance in patients with chronic heart failure. J Card Fail 2006;12:e Dickstein K, Cohen-Solal A, Filippatos G, et al. ESC guidelines for the diagnosis and treatment of acute and chronic heart failure 2008: the Task Force for the diagnosis and treatment of acute and chronic heart failure 2008 of the European Society of Cardiology. Developed in collaboration with the Heart Failure Association of the ESC (HFA) and endorsed by the European Society of Intensive Care Medicine (ESICM). Eur J Heart Fail 2008;10: Paterna S, Gaspare P, Fasullo S, Sarullo FM, Di PP. Normal-sodium diet compared with low-sodium diet in compensated congestive heart failure: is sodium an old enemy or a new friend? Clin Sci (Lond) 2008;114: Paterna S, Parrinello G, Cannizzaro S, et al. Medium term effects of different dosage of diuretic, sodium, and fluid administration on neurohormonal and clinical outcome in patients with recently compensated heart failure. Am J Cardiol 2009;103: Damgaard M, Norsk P, Gustafsson F, et al. Hemodynamic and neuroendocrine responses to changes in sodium intake in compensated heart failure. Am J Physiol Regul Integr Comp Physiol 2006;290:R Alvelos M, Ferreira A, Bettencourt P, et al. The effect of dietary sodium restriction on neurohumoral activity and renal dopaminergic response in patients with heart failure. Eur J Heart Fail 2004;6: Cody RJ, Covit AB, Schaer GL, Laragh JH, Sealey JE, Feldschuh J. Sodium and water balance in chronic congestive heart failure. J Clin Invest 1986;77: Institute of Medicine. Dietary Reference Intakes for water, potassium, sodium, chloride, and sulfate. Washington, DC: The National Academies Press, Arcand J, Floras V, Ahmed M, et al. Nutritional inadequacies in patients with stable heart failure. J Am Diet Assoc 2009;109: Caggiula AW, Wing RR, Nowalk MP, Milas NC, Lee S, Langford H. The measurement of sodium and potassium intake. Am J Clin Nutr 1985;42: Allen LA, O Connor CM. Management of acute decompensated heart failure. CMAJ 2007;176: Fonarow GC, Abraham WT, Albert NM, et al. Factors identified as precipitating hospital admissions for heart failure and clinical outcomes: findings from OPTIMIZE-HF. Arch Intern Med 2008;168: Tsuyuki RT, McKelvie RS, Arnold JM, et al. Acute precipitants of congestive heart failure exacerbations. Arch Intern Med 2001;161: McAlister FA, Lawson FM, Teo KK, Armstrong PW. A systematic review of randomized trials of disease management programs in heart failure. Am J Med 2001;110: Arcand JA, Brazel S, Joliffe C, et al. Education by a dietitian in patients with heart failure results in improved adherence with a sodium-restricted diet: a randomized trial. Am Heart J 2005;150: Dunbar SB, Clark PC, Deaton C, Smith AL, De AK, O Brien MC. Family education and support interventions in heart failure: a pilot study. Nurs Res 2005;54: Eshaghian S, Horwich TB, Fonarow GC. Relation of loop diuretic dose to mortality in advanced heart failure. Am J Cardiol 2006;97: Galve E, Mallol A, Catalan R, et al. Clinical and neurohumoral consequences of diuretic withdrawal in patients with chronic, stabilized heart failure and systolic dysfunction. Eur J Heart Fail 2005;7: Garriguet D. Sodium consumption at all ages. Health Rep 2007;18: Volpe M, Tritto C, DeLuca N, et al. Abnormalities of sodium handling and of cardiovascular adaptations during high salt diet in patients with mild heart failure. Circulation 1993;88: Granger BB, Swedberg K, Ekman I, et al. Adherence to candesartan and placebo and outcomes in chronic heart failure in the CHARM programme: double-blind, randomised, controlled clinical trial. Lancet 2005;366: Bentley B, De Jong MJ, Moser DK, Peden AR. Factors related to nonadherence to low sodium diet recommendations in heart failure patients. Eur J Cardiovasc Nurs 2005;4: Wu JR, Moser DK, Chung ML, Lennie TA. Predictors of medication adherence using a multidimensional adherence model in patients with heart failure. J Card Fail 2008;14: