NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Committee on the Consequences of Sodium Reduction in Populations; Food and Nutrition Board; Board on Population Health and Public Health Practice; Institute of Medicine; Strom BL, Yaktine AL, Oria M, editors. Sodium Intake in Populations: Assessment of Evidence. Washington (DC): National Academies Press (US); 2013 Aug 27.

Cover of Sodium Intake in Populations

Sodium Intake in Populations: Assessment of Evidence.

Show details

4Sodium Intake and Health Outcomes

This chapter reviews and assesses new evidence for associations between dietary sodium intake and outcomes published in the peer-reviewed literature through 2012. The health outcomes reviewed by the committee include cardiovascular disease (CVD), including stroke CVD mortality and all-cause mortality, congestive heart failure (CHF), chronic kidney disease (CKD), diabetes, cancer, and “other” outcomes, such as asthma and depression.

An estimated 76.4 million adults 20 years of age and older in the United States have high blood pressure (Roger et al., 2011). Mean dietary intake of sodium among the general U.S. population averages 3,400 mg daily, while federal nutrition policy guidance, the Dietary Guidelines for Americans 2010 (HHS and USDA, 2010a), recommends sodium intakes of less than 2,300 mg daily for adolescents and adults 14 years of age and older, and 1,500 mg daily for African Americans, individuals 51 years of age and older, and individuals with hypertension, diabetes, or CKD. Evidence underlying this recommendation can be found in a number of sources, including the report Dietary Reference Intakes for Water, Sodium, Chloride, and Sulfate (IOM, 2005), and the Report of the Dietary Guidelines Advisory Committee on the Dietary Guidelines for Americans, 2010 (DGAC) (HHS and USDA, 2010b). Excess dietary sodium has been identified as a potential etiologic risk factor for CVD, based on evidence for a dose-dependent increase in blood pressure in response to increasing sodium intake, as well as evidence from studies published before 2003 of sodium intake and risk of stroke or coronary heart disease (IOM, 2005).

The DGAC report (HHS and USDA, 2010b) included a review of evidence on the impact of dietary patterns low in sodium and/or low in saturated fat and high in potassium (e.g., DASH [Dietary Approaches to Stop Hypertension] and Mediterranean diets) on risk of CVD, stroke, and mortality and concluded that plant-based, lower-sodium dietary patterns had a beneficial impact on CVD risk. However, the dietary patterns included in the evidence review included dietary modifications other than sodium reduction that have been shown to have a cumulative impact on risk of CVD and related diseases. These include increased potassium, reduced intake of saturated and trans fats, and increased intake of dietary fiber. Based on their review of the evidence, the DGAC (HHS and USDA, 2010b) concluded that reduced risk of CVD, stroke, and disease-related mortality was associated with total dietary modification. Nevertheless, decreasing sodium intake has a potential role in reducing risk of CVD, stroke, and mortality, and this was the primary focus of the committee's review.

CARDIOVASCULAR DISEASE, STROKE, AND MORTALITY

As noted in Chapter 3, blood pressure is used as a surrogate indicator for CVD, stroke, and mortality risk, especially among individuals who are already at risk of disease. The committee's review of the strength of new evidence for dietary sodium and its effects on blood pressure concurred with previously established evidence (see Chapter 3). This evidence that high sodium intakes can indirectly mediate risk of adverse health outcomes underpinned the committee's assessment of evidence on associations between sodium intake and direct health outcomes. Taking the evidence for blood pressure effects into consideration as background, the committee focused its review on new evidence on sodium intake and direct health outcomes, particularly evidence from intervention studies where available.

Each outcome is discussed in turn, presenting the available data organized by population group, and within each group organized alphabetically by the last name of the first author. Each study is described by its population, size, and characteristics; study design, purpose, and length; sodium intake measure and method; range of intake, reference intake, and adjustments; outcome measure, confounders, and adjustments; and direction and significance of effect. For each major outcome of CVD, CHF, and CKD, the committee provides a summary table evaluating each study using as criteria the generalizability of the study population to U.S. populations and the appropriateness of the methodology used to support the findings and conclusions.

Finally, a summary of findings and conclusions is given on each major outcome for general populations and for population subgroups of interest as described in the statement of task, specifically those with hypertension or prehypertension, those 51 years of age and older, African Americans, and those with diabetes, CKD, and CHF (see Appendix F for evidence tables).

Studies on General Populations

The committee identified studies related to CVD from its literature search that met the criteria for inclusion described in Appendix F, Table F-1, and, on further examination, were found to be relevant to the committee's task. The cardiovascular health outcomes reviewed were CVD, stroke, stroke mortality or CVD mortality, and all-cause mortality. The committee's summary of the evidence for each cardiovascular health outcome is shown in Tables 4-1 through 4-6.

Cohen et al. (2006)

Population size and characteristics Cohen et al. (2006) obtained data from participants in the National Health and Nutrition Examination Survey (NHANES) II from among the general U.S. population (n=7,154 participants 30-74 years of age). Individuals with self-reported preexisting disease, as well as those who reported being on a low-sodium diet for hypertension, were excluded.

Study design, purpose, and length This secondary analysis of NHANES data was carried out to assess the potential impact of dietary sodium intake on risk of CVD and all-cause mortality over a mean of 13.7 years of follow-up.

Sodium intake measure and method Dietary sodium intake was assessed from a 24-hour dietary recall administered in the NHANES survey. The NHANES II dataset does not include sodium from salt added at the table. In addition, because only 1 day of intake was used, the sodium measure used in this study may not have represented the usual dietary intake of sodium at the individual level.

Range of intake, reference, and adjustments Dietary sodium intake was categorized as <2,300 mg [n=3,443] or ≥2,300 mg [n=3,711]; or by quartiles, corresponding to lowest to highest sodium intakes of <1,645, 1,645-2,359, 2,360-3,345, and ≥3,346 mg per day, using the highest intake quartile as reference; and also as a continuous variable (per 1,000 mg). Sodium intake was adjusted for calories and the highest and lowest 1 percent of the calorie intake range were excluded from the analysis. The correlation between sodium and caloric intake for most age/sex groups is greater than 0.7 (IOM, 2010, pp. 129–130).

Outcome measure, confounders, and adjustments Outcomes measured were CVD mortality, all-cause mortality, coronary heart disease mortality, and cerebrovascular disease mortality. All analyses were adjusted for prior hypertension and systolic blood pressure, which may be in the causal pathway.

Direction and significance of effect For quartiles of sodium intake the lower sodium intake quartiles were not associated with CVD mortality (hazard ratio [HR]=1.31 [confidence interval (CI): 0.90, 1.89] p=0.14), 2 (HR=1.39 [CI: 0.91, 2.11] p=0.11), and 3 (HR=0.89 [CI: 0.64, 1.25] p=0.49). Similarly, no significant associations were found between sodium intake quartiles and all-cause or stroke mortality. However, when analyzed for intakes less than 2,300 mg per day compared to 2,300 mg per day or greater, lower sodium intake was statistically significantly associated with increased risk of all-cause mortality. Models of sodium density, expressed as a sodium-to-calorie ratio, showed a statistically significant inverse association with all-cause mortality (HR=0.89 [CI: 0.79, 1.00] p=0.05). In other words, lower sodium intake was associated with higher all-cause mortality. For dietary sodium intake measured as a continuous variable, a statistically significant inverse relationship was found between sodium intake and CVD mortality whether expressed as sodium per mg (HR=0.89 [CI: 0.80, 0.99] p=0.03) or as sodium per calorie (HR=0.80 [CI: 0.68, 0.94] p=0.008).

Interactions Although data were not shown, the authors reported that they found no evidence of interactions by age, race, or prevalence of diabetes or hypertension.

Cohen et al. (2008)

Population size and characteristics Cohen et al. (2008) analyzed data from participants in the NHANES III survey from the general U.S. population (n=8,699 participants 30 years of age and older). Individuals with self-reported preexisting disease, as well as those who reported being on a low-sodium diet for hypertension, were excluded.

Study design, purpose, and length This secondary analysis of NHANES data was carried out to assess the potential impact of dietary sodium on risk of CVD and all-cause mortality over a mean period of 8.7 years.

Sodium intake measure and method Energy-adjusted dietary sodium intake was estimated from one 24-hour dietary recall. Use of added salt was determined by the responses “does not add,” “adds some,” or “adds a lot.”

Range of intake, reference, and adjustments Sodium intake was categorized into quartiles of <2,060 [n=2,174], 2,060-2,921 [n=2,175], 2,922-4,047 [n=2,175], and 4,048-9,946 mg per day [n=2,175], using the highest intake quartile as reference. Intake analyses were adjusted for added table salt.

Outcome measure, confounders, and adjustments Outcomes included CVD and all-cause mortality and were adjusted for blood pressure and hypertension, which may be in the causal pathway. Other variables examined as potential confounders included sex, age, serum cholesterol, race, treatment for hypertension, blood pressure, smoking, alcohol consumption weight, body mass index (BMI), history of diabetes, education, and added table salt.

Direction and significance of effect When fully adjusted, there was a statistically significant higher risk of CVD mortality (p=0.03) with the lowest vs. the highest quartile of sodium intake that was not present for all-cause mortality (p=0.17). For sodium as a continuous variable and residuals-adjusted sodium, nonsignificant inverse trends were found for both CVD and all-cause mortality. Similar trends were found using sensitivity analyses on the subset of participants (n=5,560) who reported not adding salt during cooking or at the table.

Interactions Although data were not shown, the authors reported that they found no evidence of interactions by age, race, or prevalence of diabetes or hypertension.

Gardener et al. (2012)

Population size and characteristics Gardener et al. (2012) analyzed data from participants in the Northern Manhattan Study (n=2,657) who had no previous diagnosis of stroke, were older than 40 years of age (mean 69 ±10 years) and were ethnically diverse (21 percent white, 24 percent African American, 53 percent Hispanic).

Study design, purpose, and length This multiethnic population-based prospective cohort study examined associations between sodium consumption and risk of stroke and combined vascular events (stroke, myocardial infarction [MI], and vascular death) over a mean of 10 years.

Sodium intake measure and method Dietary sodium intake was estimated using the Block National Cancer Institute food frequency questionnaire (FFQ). Average sodium intake was examined continuously.

Range of intake, reference, and adjustments Sodium intake was calculated from self-reported data and categorized into tertiles of ≤1,500, 1,501-3,999, and ≥4,000 mg per day. The lowest tertile was used as the reference.

Outcome measure, confounders, and adjustments The primary outcome measure was incident stroke of all subtypes; secondary outcomes were confirmed incident of combined vascular event (combined), MI incident, and vascular death. Analytic models were adjusted for demographics, behavioral risk factors, and vascular risk factors (diabetes, hypercholesterolemia, hypertension, and continuous blood pressure measurements).

Direction and significance of effect This study found that sodium intake was positively associated with increased risk of stroke. Using sodium as a continuous variable, stroke risk increased 17 percent for each 500 mg per day higher sodium intake (HR=1.17 [CI: 1.07, 1.27]). However, the authors noted that the relationship did not appear linear. Participants who consumed more than 4,000 mg sodium daily had a 2.5-fold increase in risk of total stroke compared to those who consumed less than 1,500 mg per day (HR=2.50 [CI:1.23, 5.07]). This difference persisted after adjustment for vascular risk factors. Those who consumed more than 1,500 but less than 4,000 mg of sodium daily had an approximate 30 percent increased risk, though this was not statistically significant. Each 500 mg per day higher sodium intake was associated with a 16 percent greater risk of ischemic stroke. Among those who consumed more than 4,000 mg compared to those who consumed less than 1,500 mg per day (reference intake), the risk was 2.4-fold greater. Consumption of more than 4,000 mg per day also was associated with an increased risk of combined vascular events, while the results were less consistent for lower levels of sodium consumption and cardiovascular events.

Interactions Although data were not reported, the authors found no evidence of interactions by age, race, or prevalence of diabetes or hypertension.

Larsson et al. (2008)

Population size and characteristics Larsson et al. (2008) analyzed data on 26,556 male smokers in Finland, 50-69 years of age, who participated in the Alpha-Tocopherol, Beta-Carotene Cancer Prevention primary prevention trial (ATBC Cancer Prevention Study Group, 1994). Participants had smoked at least five cigarettes daily at baseline. From the original cohort, men who had evidence of disease, for example, history of cancer, and those receiving anticoagulant therapy or taking excess vitamin supplements were excluded from the analysis, although those with a history of diabetes or coronary heart disease were included.

Study design, purpose, and length This large prospective study followed a male cohort to assess relationships between magnesium, calcium, potassium, and sodium intake and risk of stroke for a mean of 13.6 years.

Sodium intake measure and method Energy-adjusted dietary sodium intake was estimated from a self-administered, validated 276-item FFQ that included items commonly consumed in Finland. Use of cooking salt was included in the questionnaire, whereas salt added at the table was not captured.

Range of intake, reference, and adjustments Median sodium intake estimates were divided into quintiles with means of 3,909, 4,438, 4,810, 5,212, and 5,848 mg per day, for Q1 through Q5, respectively, using the lowest quintile as the reference. Thus, the average sodium intake in the United States would be within the lowest quintile of this study.

Outcome measure, confounders, and adjustments Outcome measure was first-ever stroke occurring between the date of randomization (between 1985 and 1988) and the first occurrence of stroke by the end of the study period (December 31, 2004). Strokes were classified as cerebral infarction, intracerebral hemorrhage, subarachnoid hemorrhage, or unspecified stroke, and identified from the Finnish National Hospital Discharge Register and the National Register of Causes of Death. Multivariate analysis adjusted for age, supplementation group, number of cigarettes smoked daily, BMI, systolic and diastolic blood pressure, serum total cholesterol and high-density cholesterol, as well as history of diabetes and coronary heart disease, leisure-time physical activity, and intake of alcohol and total energy.

Direction and significance of effect The analyses found no significant association between dietary sodium intake and risk of any stroke subtype (p for trend for multivariate relative risk [RR]=0.99, 0.06, and 0.55 for cerebral infarction, intracerebral hemorrhage, and subarachnoid hemorrhage, respectively).

Nagata et al. (2004)

Population size and characteristics Nagata et al. (2004) analyzed a cohort of Japanese adults 35 years of age and older using data collected in the Takayama population-based study. The study included 13,355 men and 15,724 women.

Study design, purpose, and length This population-based cohort study examined associations between dietary sodium intake and stroke mortality risk over a period of 7 years.

Sodium intake measure and method Dietary sodium intake was estimated from a 169-item semiquantitative FFQ. The instrument included questions on use of table salt and salty condiments. Use of cooking salt was not included in the questionnaire.

Range of intake, reference, and adjustments Sodium intake levels were represented by tertile for men (means 4,070, 5,209, and 6,613 mg per day) and women (means 3,799, 4,801, and 5,930 mg per day), using the lowest tertile as the reference. Sodium intake was energy-adjusted when used in analyses. Thus, the average sodium intake in the United States would be within the lowest tertile of this study.

Outcome measure, confounders, and adjustments The outcome measure was stroke mortality (data obtained from National Vital Statistics for Japan). Stroke types included subarachnoid hemorrhage, intracerebral hemorrhage, ischemic stroke, and stroke of undetermined type. Adjustment variables included age, education level, marital status, BMI, smoking, alcohol use, and history of diabetes or hypertension.

Direction and significance of effect Over the 7-year follow-up period, 43 subarachnoid hemorrhages, 59 intracerebral hemorrhage deaths, and 137 ischemic stroke deaths occurred. Among men, a 2.3-fold increased risk of stroke mortality (significantly positive for intracerebral hemorrhage and ischemic stroke death) was associated with the highest tertile of sodium intake (mean intake of 7,194 mg per day) after adjustment for other dietary variables (HR=2.33 [CI: 1.23, 4.45] p for trend=0.009). The authors also reported a borderline significant trend between high sodium intake (mean intake of 6,478 mg per day) and total stroke and ischemic stroke death among women (total stroke HR=1.70 [CI: 0.96, 3.02] p for trend=0.07); ischemic stroke (HR=2.10 [CI: 0.96, 4.62] p for trend=0.05).

Stolarz-Skrzypek et al. (2011)

Population size and characteristics Stolarz-Skrzypek et al. (2011) obtained data from two population-based prospective cohort studies (the Flemish Study on Environment, Genes, and Health Outcomes and the European Project on Genes in Hypertension) to examine a cohort of 2,856 participants recruited from a random sample of households in several European countries. The cohort included men and women 20-39, 40-59, and 60 or more years of age.

Study design, purpose, and length This study prospectively examined associations between sodium and changes in blood pressure and risk of CVD mortality and all-cause mortality over a median of 7.9 years.

Sodium intake measure and method Timed 24-hour urine samples were collected 1 week following blood pressure measurements and analyzed for sodium and potassium.

Range of intake, reference, and adjustments Twenty-four-hour sodium excretion values were categorized into tertiles of low (50–126 mmol [1,150-2,898 mg] for women; 50-158 mmol [1,150-3,634 mg] for men); medium (127–177 mmol [2,921-4,071 mg] for women; 159-221 mmol [3,657-5,083 mg] for men); and high (178–400 mmol [4,094-9,200 mg] for women; 222-400 mmol [5,106-9,200 mg] for men).

Outcome measure, confounders, and adjustments Risk of CVD events was assessed for each quartile against risk of CVD events for the whole study population. Cardiovascular outcomes, fatal and nonfatal stroke, fatal and nonfatal MI, fatal and nonfatal left ventricular heart failure, aortic aneurysm, cor pulmonale, and pulmonary or arterial embolism were validated by physicians against medical records. Covariables in the regression analyses were study population, sex, age, blood pressure level, BMI, alcohol intake, use of antihypertensive drugs, urinary potassium excretion, education, smoking status, total cholesterol, and diabetes.

Direction and significance of effect Overall, the authors found that lower sodium intake was associated with higher risk of CVD mortality. In the low, medium, and high tertiles of sodium excretion the CVD mortalities were 50 (4.1% [CI: 3.5%, 4.7%]); 24 (1.9% [CI: 1.5%, 2.3%]); and 10 (0.8% [CI: 0.5%, 1.1%]) events, respectively, after adjustment for risk factors, including baseline hypertension and blood pressure level. The risk of CVD mortality was statistically significantly higher in the low versus the high tertile (HR=1.56 [CI: 1.02, 2.36] p=0.04) with a significant trend over tertiles (p=0.02). All-cause mortality showed a trend in risk of CVD mortality for low and medium tertiles, although it was not statistically significant (HR=1.14 [CI: 0.87, 1.50] and 64 (HR=0.94 [CI: 0.75-1.18]). Likewise, there was no significant effect on total CVD incidence. Analysis of supplemental data from this study demonstrated that individuals assigned to the low-sodium tertiles were older, less educated, had greater comorbidities, lower urine volume, and lower serum creatinine than those in higher-sodium tertiles.

Takachi et al. (2010)

Population size and characteristics Takachi et al. (2010) examined data from the Japan Public Health Center-based Prospective Study, conducted in two cohorts. Cohort I and II participants were 40-59 and 40-69 years of age, respectively. Those with a history of cancer or coronary heart disease were excluded, leaving a final study population of 77,500 (35,730 men and 41,770 women).

Study design, purpose, and length The objective of this prospective study was to assess associations between sodium and salted food consumption and risk of either cancer or CVD. Participants were followed from the beginning in 1990 (cohort I) or 1993 (cohort II) until December 31, 2004.

Sodium intake measure and method Dietary sodium intake data were determined from a 138-item FFQ that included cooking salt, soy sauce, table salt, and other salty condiments.

Range of intake, reference, and adjustments Energy-adjusted sodium intake per day was categorized by quintile: medians were 3,084, 4,005, 4,709, 5,503, and 6,844 mg per day for Q1 through Q5, respectively. Thus, the average sodium intake in the United States would be close to the lowest quintile of this study.

Outcome measure, confounders, and adjustments Cardiovascular outcomes included diagnosis of MI and diagnosis of stroke confirmed by computed tomography (CT) scan and/or magnetic resonance imaging (MRI) from medical records. Adjustment variables for analysis were sex and age, with additional adjustment for BMI, smoking status, alcohol consumption, physical activity, and quintiles of energy, potassium, and calcium.

Direction and significance of effect Adjusted multivariate analysis found a significant positive association between sodium consumption at the highest compared to the lowest quintile and risk of stroke (HR=1.21 [CI: 1.01, 1.43] p for trend=0.03) and between use of cooking and table salt and risk of stroke (HR=1.21 [CI: 1.02, 1.44] p for trend=0.05), but not between use of dried salted fish and risk of stroke. Increased intake of dried and salted fish was associated with lower risk of MI. The risk of the composite CVD endpoint was elevated in the highest quintile of sodium (HR=1.19 [CI: 1.01, 1.40] p for trend=0.06). The results also showed correlation with other variables, such as dried and salted fish, although the impact of those variables on the outcomes is unknown.

Umesawa et al. (2008)

Population size and characteristics Umesawa et al. (2008) examined data from participants in the Japan Collaborative Cohort Study for Evaluation of Cancer Risks (23,119 men and 35,611 women) 40-79 years of age.

Study design, purpose, and length This large prospective study examined associations between dietary sodium intake and mortality from stroke; stroke related to subarachnoid hemorrhage, intraparenchymal hemorrhage, or ischemic stroke; coronary heart disease; and total CVD over a mean follow-up period of 12.7 years.

Sodium intake measure and method Energy-adjusted dietary sodium intake was estimated from responses to a 35-item FFQ and the results were calibrated against a previous validation study that included items from the FFQ and four 3-day dietary records.

Range of intake, reference, and adjustments Dietary sodium intake was categorized into quintiles from lowest to highest with Q1=101 ±30 (2,323 ±690), Q2=146 ±11 (3,358 ±253), Q3=182 ±11 (4,186 ±253), Q4=220 ±12 (5,060 ±276), and Q5=272 ±36 (6,256 ±828) mmol (mg) per day. The calibrated sodium intake based on a validation study was twice as high in all five quintiles in which median values were assigned for each quintile and the significance of the variables was tested.

Outcome measure, confounders, and adjustments Outcome measures were stroke (including ischemic stroke), CVD, or CHD mortality derived from data obtained from death certificates for targeted populations in each study locale. Adjustments were made for CVD risk factors, including hypertension, and for potassium intake.

Direction and significance of effect The authors found an association between greater dietary sodium intake and greater mortality from total stroke, ischemic stroke, and total CVD. Multivariable hazard ratios were strongest with the highest compared to the lowest quintiles of sodium intake: HR=1.55 (CI: 1.21, 2.00) for total stroke, HR=2.04 (CI: 1.41, 2.94) for ischemic stroke, and HR=1.42 (CI: 1.20, 1.69) for total CVD mortality. The positive associations found between dietary sodium intake and mortality risk of total and ischemic stroke and CVD were independent of potassium intake and body weight. No significant association was found between sodium intake and risk of CHD.

Yang et al. (2011)

Population size and characteristics Yang et al. (2011) analyzed data from 12,267 adults 20 years of age and older from the NHANES III. NHANES participants were linked to mortality data from the National Death Index through December 2006.

Study design, purpose, and length This secondary analysis of NHANES III data was carried out to examine associations between sodium intake, potassium intake, the sodium-to-potassium ratio, and risk of CVD mortality and all-cause mortality over an average of 14.8 years.

Sodium intake measure and method Estimates of dietary sodium intake were derived from the NHANES 24-hour dietary recall. Within-person variability was calculated using 7 percent of individuals and this variability was used to adjust the sodium intake.

TABLE 4-1. Weaknesses and Strengths of Population Studies and Methods of Studies on Cardiovascular Health Outcomes: CVD Outcomes in General Populations.

TABLE 4-1

Weaknesses and Strengths of Population Studies and Methods of Studies on Cardiovascular Health Outcomes: CVD Outcomes in General Populations.

Range of intake, reference, and adjustments Sodium intake levels, categorized into quartiles from lowest to highest are Q1=2,176; Q2=3,040; Q3=3,864; and Q4=5,135 mg per day, using the lowest intake as reference. Nutrient disease associations were estimated as continuous variables for all-cause and CVD mortality. Because relationships between estimated usual intakes and all-cause and CVD mortality were approximately linear, the percentile distributions of estimated usual intakes were calculated as the middle value of each quartile. Usual intake estimates were derived using a method adjusting for within-person variation developed by the National Cancer Institute (Tooze et al., 2006). Usual sodium (and potassium) intake estimates for CVD mortality, all-cause mortality, and ischemic heart disease (IHD) mortality were reported.

Outcome measure, confounders, and adjustments Outcome measures were CVD (including IHD) mortality and all-cause mortality. Unlike Cohen et al. (2008), the investigators elected not to adjust for measured blood pressure or antihypertensive treatment, arguing that these factors may be in the causal pathway linking dietary sodium to health outcomes.

Direction and significance of effect After multivariable adjustment, higher usual sodium intake was found to be directly associated with all-cause mortality (HR=1.20 [CI: 1.03, 1.41] per 1,000 mg per day), but not with CVD mortality or IHD mortality. However, the finding that correction for regression dilution increased the effect on all-cause mortality, but not on CVD mortality, is inconsistent with the theoretical causal pathway.

Interactions Although data were not reported, the authors found no evidence of interactions by age, race, or prevalence of diabetes or hypertension.

Studies in Populations 51 Years of Age and Older

Geleijnse et al. (2007)

Population size and characteristics Geleijnse et al. (2007) examined a subset of data from the older cohort of the Rotterdam prospective cohort study that included 1,448 randomly selected participants 55 years of age and older living in the Netherlands.

Study design, purpose, and length This case-cohort design examined the relationships between sodium and potassium intake and incidence of MI and stroke and of CVD mortality and all-cause mortality. Participants were followed for a median of 5.5 years.

TABLE 4-2. Weaknesses and Strengths of Population Study and Methods of Studies on Cardiovascular Health Outcomes: CVD Outcomes in Populations 51 Years of Age and Older.

TABLE 4-2

Weaknesses and Strengths of Population Study and Methods of Studies on Cardiovascular Health Outcomes: CVD Outcomes in Populations 51 Years of Age and Older.

Sodium intake measure and method Urinary sodium, potassium, and creatinine excretion were estimated from a single overnight urine sample collected at home.

Range of intake, reference, and adjustments In a sub-cohort free of CVD and hypertension, 24-hour urinary sodium was divided into quartiles for analysis of all-cause mortality. Quartile levels were 66, 105, and 151 mmol (1,518, 2,415, and 3,473 mg) per day. Adjustments were made for total energy, alcohol intake, saturated fat intake, and 24-hour urinary potassium.

Outcome measure, confounders, and adjustments Outcome measures were CVD mortality and all-cause mortality. Mean baseline blood pressure for the cohort was 140 (standard deviation [SD]=22) mmHg systolic and 74 (SD=11) mmHg diastolic; 37 percent had a diagnosis of hypertension. Adjustments were made for age, sex, 24-hour urinary creatinine, BMI, smoking status, diabetes, use of diuretics, and education.

Direction and significance of effect This study found no significant difference between urinary sodium level and risk of CVD mortality or all-cause mortality. Using the lower quartile as the reference, the study found an inverse association with CVD mortality that was borderline statistically significant (RR=0.77 [CI: 0.60, 1.01] per 1 SD). After excluding participants with a history of CVD or hypertension, the difference was attenuated and nonsignificant. In an examination across quartiles of urinary sodium level, using the lowest quartile as reference, all-cause mortality in disease-free participants relative risk by quartile were 0.80 (CI: 0.43,1.49), 0.66 (CI: 0.34, 1.27), and 0.98 (CI: 0.54, 1.78), respectively.

Studies with Additional Analysis of Data by Age

Five of the nine reported studies in the general population listed above also analyzed the data on health outcomes by age and found no interaction (Cohen et al., 2006, 2008; Cook et al., 2007; Gardener et al., 2012; Yang et al., 2011).

Studies on Populations with Chronic Kidney Disease

Dong et al. (2010)

Population size and characteristics Dong et al. (2010) collected data on 305 incident peritoneal dialysis patients (with diabetes or with preexisting CVD) in Japan. The cohort included 129 men and 176 women with a mean age of 59.4 years.

Study design, purpose, and length The purpose of this retrospective cohort study was to determine whether dietary sodium intake was correlated with CVD mortality and all-cause mortality over a period of 31.4 (±13.7) months.

Sodium intake measure and method Dietary sodium intake was estimated by a dietitian from 3-day diet records that also asked about use of added salt.

Range of intake, reference, and adjustments Sodium intake was categorized by tertile: mean 1,410 (±0.17), 1,810 (±0.11), and 2,470 (±0.54) mg per day for low, middle, and high tertiles, respectively (sodium removal, dialysate + urine=2,200 [±1,210], 2,780 [±1,090], and 3,030 [±1,100] mg per day for low, middle, and high tertiles, respectively). Statistical comparisons were made among intake tertiles. Sodium intake during the first 3 months was represented as the baseline sodium intake. Daily total protein and daily energy intakes were normalized for standard body weight. All of the measurements during the study were averaged.

Outcome measure, confounders, and adjustments Outcome measures were all-cause and CVD mortality. Covariates for baseline sodium intake included age, sex, BMI, history of diabetes, CVD, and biochemical measures.

Direction and significance of effect The authors found that the lowest sodium intake was associated with increased mortality risk. Mean baseline systolic blood pressure was similar across tertiles of sodium intake. Compared to the highest sodium intake tertile, those in the lowest tertile were at 55 percent lower risk of all-cause mortality (p=0.02), and 67 percent lower risk of CVD mortality (p=0.07). Similarly, when analyzed as a continuous variable, average sodium intake was correlated with overall mortality (HR=0.44 [CI: 0.20, 0.95] p=0.04), and CVD mortality (HR=0.11 [CI: 0.03, 0.48] p=0.003), suggesting that average low sodium intake was a predictor of CVD mortality.

Heerspink et al. (2012)

Population size and characteristics Heerspink et al. (2012) examined participants in the Reduction of Endpoints in NIDDM [non-insulin- dependent diabetes mellitus] with the Angiotensin II Antagonist Losartan study (250 centers in 28 countries in Asia, Europe, and the Americas) and the Irbesartan Diabetic Nephropathy Trial (209 centers in the Americas, Australia, Europe, and Israel). Participants were 30-70 years of age, with type 2 diabetes and overt proteinuria.

Study design, purpose, and length This prospective cohort study evaluated whether the effect of randomization to treatment with angiotensin receptor blockers (ARBs) with kidney disease progression and CVD was modified by dietary sodium intake over a period of 30 months.

Sodium intake measure and method Dietary sodium intake was assessed using multiple (average of five collections per participant) 24-hour urinary excretion samples collected throughout the study. The 24-hour urine sodium measure was indexed to creatinine in an attempt to control for quality of collections.

Range of intake, reference, and adjustments The 24-hour sodium-to-creatinine ratio tertiles were <121, 121-153, and >153 mmol/g (<2,783, 2,783-3,519, and >3,519 mg/g, respectively). Sensitivity analyses were conducted using 24-hour urine sodium without indexing to urine creatinine.

Outcome measure, confounders, and adjustments The outcomes were CKD progression, defined as doubling of serum creatinine or incident end-stage renal disease (ESRD), and CVD, defined as a composite of CVD death, MI, stroke, hospitalization for CHF, or revascularization procedure.

Direction and significance of effect Results from this study suggest that ARBs were more effective at decreasing CKD progression and CVD when sodium intake was in the lowest tertile [<121 mmol per day (<2,783 mg per day)].

TABLE 4-3. Weaknesses and Strengths of Population Studies and Methods of Studies on Cardiovascular Health Outcomes: CVD Outcomes in CKD Populations.

TABLE 4-3

Weaknesses and Strengths of Population Studies and Methods of Studies on Cardiovascular Health Outcomes: CVD Outcomes in CKD Populations.

Studies on Populations with Cardiovascular Disease

Costa et al. (2012)

Population size and characteristics Costa et al. (2012) followed a subset of 372 participants in the Brasilia Heart Study. This study included only individuals who had experienced a recent (within 24 hours) MI and who sought medical care, not individuals from the general population.

Study design, purpose, and length This prospective cohort study was carried out to examine the influence of high vs. low sodium intake on inflammatory-oxidative response, cardiac remodeling, and total mortality after MI, for up to 4 years. Cardiac MRI was used to identify cardiac changes and areas of infarction. Study participants received lifestyle counseling upon discharge, which included diet counseling.

Sodium intake measure and method Dietary sodium intake was estimated from a 62-item validated FFQ.

Range of intake, reference, and adjustments Energy-adjusted sodium intake was categorized as either high (≥1,200 mg per day) or low (<1,200 mg per day). A validation study of 24-hour urine excretions in 21 patients indicated that values were consistently 1,500 mg ±500 mg greater than the estimated daily sodium intake from the questionnaire. Sodium intake was assessed for associations with alcohol intake, smoking, income, and education level.

Outcome measure, confounders, and adjustments The primary outcome was total mortality in the first 30 days following MI. Secondary outcomes were total mortality during 4-year follow-up, composite endpoint of fatal or nonfatal MI, and unstable angina. Outcomes were recorded from 48 hours following onset of symptoms of MI. Associations with outcomes were tested for potential confounders, namely blood pressure changes during hospitalization, use of antihypertensive medication, and presence of pulmonary congestion.

Direction and significance of effect Significant correlations were found between sodium intake and percentage of fat and calories in daily intake. In the first 30 days following MI, total mortality was statistically significantly higher in the high- compared to the low-sodium groups (p=0.04), and the association remained similar after excluding non-CVD-related mortality (p=0.02). Multivariate analyses found a 2.9-fold risk of mortality with high sodium intake. Overall, for the first 30 days and up to 4 years afterward, total mortality was significantly associated with high sodium intake (p<0.05).

Kono et al. (2011)

Population size and characteristics Kono et al. (2011) examined a Japanese cohort of 78 men and 24 women, with a mean age of 64 years. All participants had a recent (within 2 weeks) hospitalization for acute ischemic stroke and thus the study did not evaluate participants from the general population. Both initial and later vascular events were confirmed by a neurologist using clinical data, including CT scan or MRI. Participants were divided into groups with large vessel disease (LVD) or small vessel disease (SVD).

Study design, purpose, and length This prospective cohort study examined the association of timed urine sodium excretion with recurrence rates and risk of stroke and other vascular events, including MI, angina pectoris, and peripheral artery disease over a period of 3 years.

Sodium intake measure and method For 3 consecutive days, participants collected urine for 8 hours overnight using a self-monitoring device.

Range of intake, reference, and adjustments Mean daily sodium intake was calculated for each participant, divided into four groups, and analyzed according to median value of physical activity and sodium intake. For analysis, sodium intake was expressed as greater or less than 3,201 mg per 24 hours.

Outcome measure, confounders, and adjustments Outcome measures were cumulative recurrence rates of stroke. Adjustments were made for age and medications. Deaths without stroke recurrence were excluded from the analysis.

Direction and significance of effect Cumulative risk analysis found that a salt intake of greater than the median of 10,700 mg per day (4,280 mg of sodium) was associated with higher stroke recurrence rate (HR=2.43 [CI: 1.04, 5.68] p=0.04). Univariate analysis of lifestyle management also found that poor lifestyle, defined by both high salt intake (≥10,700 mg [4,000 mg sodium] per day) and low physical activity (<5,800 steps per day), was significantly associated with stroke recurrence (HR=1.71 [CI: 1.11, 2.62] p=0.013).

O'Donnell et al. (2011)

Population size and characteristics O'Donnell et al. (2011) reanalyzed data from two randomized controlled drug trials, the ONgoing Telmisartan Alone and in combination with Ramipril Global Endpoint Trial (ONTARGET) and the Telmisartan Randomized AssessmeNt Study in aCe iNtolerant subjects with cardiovascular Disease (TRANSCEND). Both trials included participants with established CVD. The combined study sample had a total of 28,880 participants, 55 years of age and older, recruited from 40 countries. A subset of the study cohort (n=2,625) provided a 2-year follow-up and final urinary measurement.

Study design, purpose, and length This study was a prospective observational analysis of two cohorts followed for a median of 56 months (4.7 years).

Sodium intake measure and method Sodium intake was estimated from spot urine sodium measurements from a single morning fasting urine sample collected before the run-in period of the trials. Estimates were extrapolated to 24-hour estimated sodium excretion using the Kawasaki formula (see Chapter 2 for further discussion). This formula was developed in a healthy Asian population with higher dietary sodium intake than the usual intake in the United States. However, O'Donnell et al. (2011) validated the sodium measure in 105 individuals in the Prospective Urban Rural Epidemiology study, and reported correlations of estimated and measured 24-hour urine sodium of 0.55.

Range of intake, reference, and adjustments Sodium excretion was categorized into seven ranges: <2,000, 2,000-2,990, 3,000-3,990, 4,000-5,990, 6,000-6,990, 7,000-8,000, and >8,000 mg per day. The reference sodium excretion was 4,000-5,990 mg per day.

Outcome measure, confounders, and adjustments Risk estimates were made for each outcome (CVD-related mortality, MI, stroke, and hospitalization for CHF) and for a composite of all outcomes. Models for analysis were adjusted for age, sex, race/ethnicity, prior history of stroke or MI, creatinine, BMI, comorbid vascular risk factors (hypertension, diabetes, atrial fibrillation, smoking, low-density lipoprotein, high-density lipoprotein, drug treatment, fruit and vegetable consumption, physical activity, baseline blood pressure, change in systolic blood pressure, and urinary potassium).

TABLE 4-4. Weaknesses and Strengths of Population Studies and Methods of Studies on Cardiovascular Health Outcomes: CVD Outcomes in Populations with Preexisting CVD.

TABLE 4-4

Weaknesses and Strengths of Population Studies and Methods of Studies on Cardiovascular Health Outcomes: CVD Outcomes in Populations with Preexisting CVD.

Direction and significance of effect For the composite outcome, multivariate analysis found a U-shaped relationship between 24-hour urine sodium and the composite outcome of CVD death, MI, stroke, and hospitalization for CHF. Compared with the reference sodium excretion (4,000-5,990 mg per day), excretion of 7,000-8,000 (HR=1.15 [CI: 1.00, 1.32]) and >8,000 mg per day (HR=1.49 [CI: 1.28, 1.75]), and excretion of 2,000-2,990 (HR=1.16 [CI: 1.04, 1.28]) and <2,000 mg per day (HR=1.21 [CI: 1.04, 1.43]) were associated with increased risk of composite CVD and mortality. Increased risk of CVD-related mortality was associated with excretion of 7,000-8,000 mg per day (HR=1.53 [CI: 1.26, 1.86]), excretion of >8,000 mg per day (HR=1.66 [CI: 1.31, 2.10]), excretion of 2,000-2,990 mg per day (HR=1.19 [CI: 1.02, 1.39]), and excretion of <2,000 mg per day (HR=1.37 [CI: 1.09, 1.73]) compared to the reference excretion. Compared with the reference excretion, excretion of 6,000-6,990 mg per day (HR=1.21 [CI: 1.03, 1.43]) was associated with increased risk of MI, while excretion of >8,000 mg per day was associated with increased risk of stroke (HR=1.48 [CI: 1.09, 2.01]) and hospitalization for CHF (HR=1.51 [CI: 1.12, 2.05]). Excretion of 2,000-2,990 mg per day was associated with increased risk of hospitalization for CHF (HR=1.23 [CI: 1.01, 1.49]). Subgroup analyses found no effect of covariates on the associations between sodium excretion and nonfatal CVD events.

Interactions The authors reported HRs to be similar when data were adjusted for blood pressure. Unpublished data sent to the committee indicate no evidence of interactions by prevalence of diabetes.

Studies on Populations with Prehypertension

Cook et al. (2007)

Population size and characteristics Cook et al. (2007) obtained follow-up data from a subset of 2,415 prehypertensive participants included in a previous randomized comparison of 744 and 2,382 participants from two randomized controlled trials (RCTs), the Trials of Hypertension Prevention (TOHP) I and II, respectively, with mortality follow-up on all participants (3,126 total participants and 2,415 responders). (The original studies were conducted to evaluate the effect of dietary sodium on blood pressure, rather than on health outcomes—see Chapter 3.) Despite the original randomized trial design, those remaining in the follow-up of the intervention group were older than those in the follow-up of the control group: 43.4 (±6.6) vs. 42.6 (±6.5) years (p=0.074) in TOHP I and 43.9 (±6.2) vs. 43.3 (±6.1) years (p=0.015) in TOHP II.

Study design, purpose, and length This prospective cohort study was an observational follow-up conducted about 10 years after the end of TOHP I and 5 years after the end of TOHP II. The purpose of the study was to examine the effects of reduced sodium intake on CVD events 10-15 years following the original trial.

Sodium intake measure and method Sodium intake was assessed by repeated measures of 24-hour urine collections over an 18-month period. Additional information was collected on self-reported sodium intake on a final follow-up questionnaire and included participants' preference for salty and low-sodium foods, usual use of low-sodium products, whether they read food labels for sodium, or whether they tracked daily sodium intake to assess long-term patterns of sodium use.

Range of intake, reference, and adjustments Baseline sodium in the intervention group in TOHP I was 154.5 (±59.9) mmol (3,556 [±1,378] mg)/24 hours and in TOHP II was 182.9 (±78.4) mmol (4,207 [±1,803] mg)/24 hours, with similar levels in the control groups. Over the course of the study, the intervention group experienced a sodium reduction of 55.2 mmol (1,270 mg)/24 hours in TOHP I and 42.5 mmol (978 mg)/24 hours in TOHP II, while the control group experienced a reduction of 11.3 mmol (260 mg)/24 hours in TOHP I and 9.8 mmol (225 mg)/24 hours in TOHP II.

Outcome measure, confounders, and adjustments The primary outcome measure was CVD, which included MI, stroke, coronary artery bypass graft, percutaneous transluminal coronary angioplasty, or death with a cardiovascular cause. Adjustments included clinic, age, race, sex, and weight loss intervention assignment. Adjustments in additional analyses included baseline weight and sodium excretion.

Direction and significance of effect In unadjusted analysis, the low-sodium intervention had somewhat lower CVD event risk compared to the control group (with a nonsignificant p=0.21 in analyses stratified by study). However, in models adjusted for trial, clinic, age, race, and sex, the intervention group had 25 percent lower risk of nonfatal CVD compared to control groups (RR=0.75 [CI: 0.57, 0.99] p=0.04), an association that was of borderline statistical significance. Further adjustment for baseline sodium excretion and body weight found a 30 percent lower risk (RR=0.70 [CI: 0.53, 0.94] p=0.02). Analyses for total mortality, which was the only outcome on which information was available in all participants, found a nonsignificant 20 percent lower mortality in the sodium reduction group compared to the usual sodium intake group (RR=0.80 [CI: 0.51, 1.26] p=0.34).

Interactions Although data were not reported, the authors found no evidence of interactions by age or race.

Cook et al. (2009)

Population size and characteristics Cook et al. (2009) also evaluated a slightly smaller number (2,275) of the prehypertensive participants who were in the control arms of TOHP I and II (37 of whom participated in both trials).

Study design, purpose, and length This prospective cohort study examined relationships between sodium and potassium intake and sodium-to-potassium ratio with CVD events among participants from TOHP I and TOHP II over 10-15 years of follow-up.

Sodium intake measure and method Sodium and potassium intake were determined from repeated measures of 24-hour urinary sodium excretion collected at baseline plus scheduled intervals in the TOHP I (five to seven collections over 18 months) and TOHP II (three to five collections over 3-4 years) intervention trials, with a mean of 4.8 measures for all participants.

Range of intake, reference, and adjustments Sodium was determined as mean urinary excretion levels. Mean baseline sodium excretion for all participants (TOHP I and II) was 176.1 mmol (4,050 mg)/24 hours for men and 138.3 mmol (3,105 mg)/24 hours for women. The overall median urinary sodium excretion collected over 18 months for all participants was 158 mmol (3,634 mg)/24 hours (interquartile range, 127-194 mmol [2,921-4,462 mg]/24 hours). For men, the median excretion was 171 mmol (3,933 mg)/24 hours, and for women, it was 134 mmol (3,082 mg)/24 hours.

Outcome measure, confounders, and adjustments Outcome measures (CVD events) and adjustments were as described for Cook et al. (2007).

Direction and significance of effect Analyses for a linear effect of sodium on CVD outcomes found a nonsignificant 25 percent increase risk of CVD associated with a 100 mmol (2,300 mg)/24-hour higher sodium excretion (RR=1.25 [CI: 0.91, 1.72] p=0.18). This association remained after adjustment for baseline and changes in blood pressure and medication use, which suggests that any effect on outcomes may be independent of effects of blood pressure. After adjustment for potassium excretion, the association of sodium and CVD was strengthened and rendered borderline statistically significant. Each 100 mmol (2,300 mg)/24-hour higher sodium excretion was associated with a 42 percent higher risk of CVD events (RR=1.42 [CI: 0.99, 2.04] p=0.05). Across sex-specific quartiles of sodium excretion, using lowest quartile as reference, no significant trend was detected for CVD risk (from lowest to highest, RR=1.00, 0.99 [CI: 0.62, 1.58], 1.16 [CI: 0.73, 1.84], and 1.20 [CI: 0.73, 1.97] p for trend=0.38).

Additional Studies with Analysis of Statistical Interactions by Blood Pressure

Several other studies discussed in this chapter analyzed data on health outcomes by blood pressure and found no statistical interactions (Cohen et al., 2006, 2008; Gardener et al., 2012; O'Donnell et al., 2011; Yang et al., 2011).

Studies on Populations with Diabetes

Ekinci et al. (2011)

Population size and characteristics Ekinci et al. (2011) examined participants with type 2 diabetes (n=638; 56 percent male), who were attending a diabetes clinic in Australia. The mean age of all participants was 64 years and mean duration of diabetes was 11 years. Forty-seven percent of the participants were obese. During follow-up, all patients continued their standard medical care, including antihypertensive, lipid lowering, and anti-diabetic medications. Patients with type 1 diabetes or diabetes secondary to medication or pancreatitis were excluded from the study.

Study design, purpose, and length This prospective cohort study was carried out to examine associations between dietary sodium intake and all-cause and CVD mortality in patients with type 2 diabetes.

Sodium intake measure and method Sodium intake was estimated from 24-hour urine collection. Patients were given general dietary advice as part of their routine care by a dietitian, but follow-up urinalysis for sodium was not performed.

Range of intake, reference, and adjustments Participants were stratified into tertiles: <3,450, 3,540-4,784, and ≥4,785 mg sodium per day. Those in the lowest sodium tertile were older, less likely to be on medication, and had lower estimated glomerular filtration rate (eGFR).

Outcome measure, confounders, and adjustments The primary outcome was death from any cause. Cardiovascular mortality also was included as an outcome. Adjustment was made for age, sex, duration of diabetes, atrial fibrillation, the presence and severity of CKD in the analyses.

Direction and significance of effect The hazard ratios for all-cause mortality and CVD mortality for each 100 mmol per day (2,300 mg per day) higher sodium intake were 0.72 ([CI: 0.55, 0.94] p=0.017) and 0.65 ([CI: 0.44, 0.95] p=0.026), respectively.

Tikellis et al. (2013)

Population size and characteristics Tikellis et al. (2013)1 analyzed data from the Finnish Diabetic Nephropathy (FinnDiane) study. This study included participants with type 1 diabetes but excluded those with preexisting CVD as well as those with ESRD. The cohort included 2,648 adults 18 years of age and older, with a mean age of 38 years.

TABLE 4-5. Weaknesses and Strengths of Population Studies and Methods of Studies on Cardiovascular Health Outcomes. CVD Outcomes in Populations with Hypertension and Prehypertension.

TABLE 4-5

Weaknesses and Strengths of Population Studies and Methods of Studies on Cardiovascular Health Outcomes. CVD Outcomes in Populations with Hypertension and Prehypertension.

Study design, purpose, and length This prospective cohort study examined associations between sodium intake and CVD and mortality outcomes in adults with type 1 diabetes for a median of 10 years.

Sodium intake measure and method Sodium intake was estimated from one 24-hour urinary sodium collected at baseline.

Range of intake, reference, and adjustments Urinary sodium excretion was categorized into quartiles. The second and third quartiles were collapsed for analysis so that the categories used for analysis were <102 (<2,346), 102-187 (2,346-4,301), or >187 (>4,301) mmol (mg) per day. Adjustments were made for parameters associated with daily urinary sodium excretion.

Outcome measure, confounders, and adjustments Outcome measures were CVD mortality or all-cause mortality. CVD analyses were performed using spline analysis with a single knot at 102 mmol (2,346 mg) per day for CVD incidence and at 141 mmol (3,243 mg) per day for mortality. Adjustments were made for age, sex, glycemic control, presence and severity of CKD, and total cholesterol and triglycerides.

Direction and significance of effect Adjusted multivariate regression analysis found urinary sodium excretion was associated with incident CVD, with increased risk at both the highest and lowest urine sodium excretion levels. When analyzed as independent outcomes, no significant associations were found between urinary sodium excretion and new CVD or stroke after adjustment for other risk factors.

TABLE 4-6. Weaknesses and Strengths of Population Studies and Methods of Studies on Cardiovascular Health Outcomes: CVD Outcomes in Populations with Diabetes.

TABLE 4-6

Weaknesses and Strengths of Population Studies and Methods of Studies on Cardiovascular Health Outcomes: CVD Outcomes in Populations with Diabetes.

Additional Studies with Analysis of Data by Diabetes Prevalence

Two other studies discussed in this chapter analyzed the data on health outcomes by diabetes prevalence and found no interaction (Cohen et al., 2006; O'Donnell et al., 2011).

Studies in Populations with Congestive Heart Failure

The committee also reviewed evidence on the association of sodium intake with CHF. Its summary of the evidence for CHF is shown in Tables 4-7 and 4-8.

TABLE 4-7. Weaknesses and Strengths of Population Studies and Methods of Studies on Cardiovascular Health Outcomes. Observational Trials: CVD Outcomes in Populations with CHF.

TABLE 4-7

Weaknesses and Strengths of Population Studies and Methods of Studies on Cardiovascular Health Outcomes. Observational Trials: CVD Outcomes in Populations with CHF.

TABLE 4-8. Weaknesses and Strengths of Population Studies and Methods of Studies on Cardiovascular Health Outcomes. Randomized Control Trials: CVD Outcomes in Populations with CHF.

TABLE 4-8

Weaknesses and Strengths of Population Studies and Methods of Studies on Cardiovascular Health Outcomes. Randomized Control Trials: CVD Outcomes in Populations with CHF.

Arcand et al. (2011)

Population size and characteristics Arcand et al. (2011) examined data from 123 New York Heart Association (NYHA)2 Class I/II and class III/IV medically stable CHF patients enrolled in multidisciplinary CHF programs in two tertiary care hospitals in Canada. Patients were between 18 and 85 years of age, had a left ventricular ejection fraction (LVEF) of <35 percent, were medically stable for at least 3 months, and were on standard medical therapy. Patients were excluded if they had significant renal dysfunction or cardiac cachexia. All patients were consuming a self-selected diet.

Study design, purpose, and length This small prospective cohort study followed participants for 3 years to determine whether a high sodium intake is related to acute decompensated heart failure (ADHF) in ambulatory patients.

Sodium intake measure and method Sodium intake was measured using two 3-day food records: at baseline and after 6-12 weeks. Intake estimates were verified by 24-hour urine analysis in a subset of patients.

Range of intake, reference, and adjustments Calorie-adjusted sodium intakes were pooled as tertiles with cut-points at 1,900 mg per day and 2,700 mg per day.

Outcome measure, confounders, and adjustments The primary outcome measure was ADHF. Secondary outcomes were all-cause hospitalization and death or transplantation. Adjustments were made for age, sex, energy intake, LVEF, beta blockers, furosemide, and BMI.

Direction and significance of effect High sodium intake levels (≥2,800 mg per day) were significantly associated with ADHF (HR=2.55 [CI: 1.61, 4.04] p=0.001), all-cause hospitalization (HR=1.39 [CI: 1.06, 1.83]), and mortality (HR=3.54 [CI: 1.46, 8.62] p=0.005).

Lennie et al. (2011)

Population size and characteristics Lennie et al. (2011) evaluated a cohort of 302 patients with CHF NYHA Class I to IV recruited from six large community hospitals or medical centers in Georgia, Indiana, Kentucky, and Ohio. Patients were eligible if they had a confirmed diagnosis of chronic CHF with reduced LVEF, had been on medication for at least 3 months, and could read and speak English. Those who were referred for transplantation, had a history of acute MI, valvular heart disease, peripartum heart failure, myocarditis, inflammatory disease, ESRD, or coexisting terminal illness were excluded.

Study design, purpose, and length This small prospective cohort study examined differences in cardiac event-free survival between patients with sodium intake either above or below 3,000 mg per day over a period of 12 months.

Sodium intake measure and method Sodium intake was estimated from a single 24-hour urine collection.

Range of intake, reference, and adjustments The analysis was conducted using 3,000 mg urinary sodium (130 mmol) as the only cut-point for dietary salt intake. Patients were divided into two groups using the 3,000-mg cut-point and stratified by NYHA Class I/II vs. Class III/IV. Adjustment was made for BMI.

Outcome measure, confounders, and adjustments The primary outcomes of the study were the composite endpoint of time to first event for emergency department or hospital admission for CHF or other cardiac-related cause, and all-cause mortality. This model adjusted for age, sex, etiology of CHF, BMI, LVEF, and total comorbidity score.

Direction and significance of effect Results for event-free survival at a urinary sodium of ≥3,000 mg per day varied by the severity of patient symptoms. Among participants stratified into NYHA Class I/II, sodium intake greater than 3,000 mg per day was correlated with a lower disease incidence compared to those with a sodium intake less than 3,000 mg per day (HR=0.44 [CI: 0.20, 0.97] p=0.40). Conversely, participants stratified into NYHA Class III/IV and a sodium intake greater than 3,000 mg per day had a higher disease incidence than those with sodium intakes less than 3,000 mg per day (HR=2.54 [CI: 1.10, 5.84] p=0.028).

Parrinello et al. (2009)

Population size and characteristics Parrinello et al. (2009) examined 173 previously hospitalized patients who had a recent event of decompensated CHF in Italy. Characteristics for inclusion were heart failure consistent with the definition of NYHA functional classification of CHF, uncompensated CHF, or Class IV that was unresponsive to treatment. Participants with cerebral vascular disease, dementia, cancer, uncompensated diabetes, and severe hepatic disease were excluded.

Study design, purpose, and length This RCT was designed to evaluate the effects of dietary sodium restriction at two levels on neurohormonal and cytokine activation and on clinical outcomes over a period of 12 months.

Sodium intake measure and method The treatment groups received one of two levels of a prescribed sodium-restricted diet. Both treatment groups included daily intake of 1,000 ml of fluid and 125-250 mg furosemide twice daily. No follow-up sodium intake assessment was performed.

Range of intake, reference, and adjustments Dietary sodium levels were controlled at either 120 or 80 mmol (2,760 or 1,840 mg) for the modestly restricted and restricted diets, respectively. Adherence to the diet was confirmed by 24-hour urine analysis at baseline, 6 months, and 12 months.

Outcome measure, confounders, and adjustments The primary outcome measures were neurohormonal markers, cytokine levels, hospital readmission, and mortality.

Direction and significance of effect During the 12 months of follow-up, participants receiving the restricted sodium diet had a greater number of hospital readmissions (adjusted risk reduction [ARR]=14.2 [CI: 5.65, 22.7] p<0.005) and higher mortality (ARR=14.2 [CI: 5.65, 22.7] p<0.005) compared to those on the modestly restricted diet.

Paterna et al. (2008)

Population size and characteristics Paterna et al. (2008) examined 232 Italian patients 53-83 years of age and classified as NYHA Class II and recently hospitalized for decompensated heart failure. Study participants with cerebral vascular disease, dementia, cancer, uncompensated diabetes, and severe hepatic disease were excluded. Eligible patients were those unresponsive to treatment regimens with high doses of oral furosemide up to 250-500 mg per day and/or combinations of diuretics (thiazide, loop diuretic, and spironolactone), angiotensin-converting enzyme (ACE) inhibitors (captopril; 75-150 mg per day), digitalis, beta blockers and nitrates, and who also had aggressive co-treatment with high doses of diuretics (furosemide [250 mg twice daily] and severe fluid restriction [1,000 ml per day]).

Study design, purpose, and length This RCT evaluated the effects of two different sodium levels on risk of hospital readmission. Study participants were evaluated by two physicians blinded to the protocol every week for the first month, then every 2 weeks for the next 2 months, and then every month for the reminder of the study period (6 months).

Sodium intake measure and method Sodium intake levels were prescribed to participants in treatment groups.

Range of intake, reference, and adjustments Participants in treatment groups were randomized to a diet with low (80 mmol per day [1,840 mg per day]) or normal (120 mmol per day [2,760 mg per day]) sodium for 180 days; adherence to the diet prescribed by dietitians was ascertained every week.

Outcome measure, confounders, and adjustments The outcome measure was hospital readmission after having been hospitalized for CHF and then discharged within the previous 30 days.

Direction and significance of effect The lower sodium intake group experienced a significantly higher number of hospital readmissions compared to the normal sodium intake group (absolute risk reduction=18.69% [CI: 9.29, 28.08] p<0.05) and a higher but not significantly higher mortality compared to the normal sodium intake group (absolute risk reduction=8.07% [CI: 0.71, 15.43]).

Paterna et al. (2009)

Population size and characteristics Paterna et al. (2009) examined 410 (205 intervention and 205 control) recently hospitalized patients with decompensated CHF NYHA Class II-IV who were 55-83 years of age.

Study design, purpose, and length A 2×2×2 factorial double-blinded RCT was used to assess the effect of dietary sodium intake in combination with a diuretic and fluid regimen on risk of mortality.

Sodium intake measure and method Sodium intake ranges for participants were estimated based on food diaries.

Range of intake, reference, and adjustments Sodium intake ranges were categorized into 120 mmol (2,760 mg) or 80 mmol (1,840 mg) per day.

Outcome measure, confounders, and adjustments The primary outcome measures were hospital readmission and mortality. The independent variables were fluid intake limits (1,000 and 2,000 ml per day), sodium intake (120 and 80 mmol per day [2,760 and 1,840 mg per day]), and furosemide treatment (500 and 250 mg per day). Study participants were evaluated every week for the first month and then every 2 weeks for the next 2 months and then every month for the remainder of the study period (6 months).

Direction and significance of effect A significant association was found between the low sodium intake (80 mmol per day [1,840 mg per day]) and hospital readmissions (odds ratio [OR]=2.46 [CI: 1.84, 3.29] p<0.0001). The group with normal sodium diet also had fewer deaths compared to all groups receiving a low-sodium diet combined. The effect of low sodium intake on health outcomes within subgroups by dosage of furosemide and fluid restriction could not be examined due to the small number of participants in the subgroups.

SYNTHESIS OF THE EVIDENCE

Methodological Approach

A number of factors influenced the committee's assessment of the evidence reviewed. These included the variability in methodological approaches used to evaluate relationships between sodium intake and risk of health outcomes, study design, limitations in the quantitative measures of both dietary intake and urinary excretion of sodium, confounder adjustment, and the number of relevant studies available. The committee considered studies that determined sodium intake levels through multiple high-quality 24-hour urine collections to be the best design. In addition to inconsistencies in sodium intake measures, methodological flaws included the possibility of confounding, measurement error (e.g., systematic underestimates of sodium intake in those at highest risk of outcomes), and reverse causation (e.g., individuals with existing underlying disease that leads to low sodium intake and eventually the outcome of interest).

Assessing the impact of sodium intake on health outcomes was further complicated by wide variability in intake ranges among studies. For example, in the studies reviewed, high sodium intake ranged from about 2,700 to more than 10,000 mg per day; the high intake ranges of some studies overlapped with the lower ranges in others. The wide range of typical intakes across various population groups, as well as differences in the methods used to measure dietary sodium among different studies, meant that the committee could not derive a numerical definition for high or low intakes in its findings and conclusions. Rather, it could consider sodium intake levels only within the context of an individual study. Thus, in its findings and conclusions, the committee's description of sodium intake reflects the levels in the ranges described in the evidence reviewed. Likewise, the extreme variability in intake levels among population groups precluded the committee from establishing a “healthy” intake range.

Findings for Cardiovascular Disease, Stroke, and Mortality

The committee reviewed evidence that included a broad range of population groups and methodological approaches to determine relationships between sodium intake and direct measures of disease, specifically CVD, stroke, and mortality, including all-cause mortality. All of the evidence considered was observational, mostly prospective cohort studies that examined associations between sodium intake and risk of adverse health outcomes. The populations studied were disproportionately from outside the United States and many included groups that consumed levels of sodium much higher than 3,400 mg per day, the average amount consumed by U.S. adults. Many studies also focused on populations with hypertension, borderline hypertension, or other preexisting diseases or conditions that put them at risk of developing disease.

General Population

Based on its assessment of the evidence, the committee found, first, the evidence reviewed on specific adverse health outcomes consistently indicates an association in the general population between excessive sodium intakes and increased risk of CVD, particularly for stroke. In particular, data from studies using FFQs generally supported an association between high sodium intake and increased risk of CVD, particularly stroke (Gardener et al., 2012; Nagata et al., 2004; Takachi et al., 2010; Umesawa et al., 2008), although not consistently (Larsson et al., 2008). Several of these studies evaluated populations with sodium intakes much higher than the average U.S. intake of 3,400 mg per day (Nagata et al., 2004; Takachi et al., 2010; Umesawa et al., 2008). Gardener et al. (2012), however, found an effect on stroke with sodium intakes starting at 1,500 mg per day (HR=1.17 per 500 mg increase in sodium). Sodium intake data from FFQs, however, are limited by errors in estimating discretionary sodium intake (salt added in cooking or at the table), which accounts for an estimated 11 percent of sodium intakes (Mattes and Donnelly, 1991).

The committee found, in contrast, that the evidence from the current literature is inconsistent with regard to associations with sodium intakes below 2,300 mg per day, with results ranging from lower, similar, or higher risk of CVD, stroke, or mortality, including all-cause mortality. All of the studies identified have limitations of different types. The evidence in some cases is suggestive, however, of associations between lower sodium intake (below 2,300 mg per day) and potential increased risk of adverse health outcomes, though reverse causation, confounding, and systematic measurement error cannot be ruled out.

For example, studies using data from NHANES III (which is representative of the general U.S. population) used 24-hour recall data to estimate sodium intake, which, as discussed in Chapter 2, could introduce considerable error in the measurement of sodium levels. In addition, such studies showed inconsistent results, depending on the methodological approach. Two studies (Cohen et al., 2006, using NHANES II; Cohen et al., 2008, using NHANES III) found an increased risk of CVD at lower sodium levels, while one study (Yang et al., 2011, also using NHANES III) found a lower risk of all-cause mortality at lower sodium intake levels. These studies, however, are limited by the sodium intake measurement used. In addition, they differ on the corrections made for sodium measurements as well as for calorie intake adjustment. Additionally, Cohen et al. (2006, 2008) did not adjust for within-individual day-to-day variation in intake, whereas that adjustment was made in Yang et al. (2011) As another example, Stolarz-Skryzpek et al. (2011), in a general population, also reported higher CVD outcomes in the lower sodium intake group. However, Stolarz-Skryzpek et al. (2011) were limited by the possibility of unmeasured confounding and undercollection of urine specimens in the lowest sodium tertile.

Population Subgroups

Some studies addressed questions related to associations between sodium intake and health outcomes in population subgroups. The committee evaluated one large (O'Donnell et al., 2011) and six small (Arcand et al., 2011; Dong et al., 2009; Ekinci et al., 2011; Heerspink et al., 2012; Kono et al., 2011; Tikellis et al., 2013) prospective cohort studies in patients with preexisting CHF, stroke, MI, CKD, and diabetes, using various methods of sodium assessment. The outcomes estimates were extremely heterogeneous, with HR values ranging between 0.11 (Dong et al., 2009) and 3.54 (Arcand et al., 2011). Because these populations are typically advised to reduce their sodium intake, reverse causation cannot be ruled out as a factor in the relationship.

In contrast, two related observational studies (Cook et al., 2007, 2009) used three to seven 24-hour urine collections, the best available method, to measure sodium intake levels. Both of these studies were conducted in prehypertensive individuals. Cook et al. (2007), an observational follow-up of the TOHP I and II sodium reduction trials, found a 25 percent reduction in CVD incidence, as well as a nonsignificant 20 percent reduction in total mortality when average levels of sodium intake decreased from approximately 3,600 to 2,300 mg per day in the intervention group in TOHP I and from 4,200 to 3,200 mg per day in TOHP II. Cook et al. (2009), also an observational study that followed participants in the TOHP I and II trials, found a linear increase in the risk of total CVD events with increasing sodium intake levels after adjusting for potassium intake. However, there were relatively few outcomes in the lowest ranges of sodium intake, leading to unstable estimates in those ranges. The committee found that these studies suggest an association between a decrease in CVD event rates and sodium intakes down to 2,300 mg per day, and perhaps below, although based on small numbers.

The committee identified and evaluated three RCTs and two cohort studies that examined associations between sodium intake at low, moderate, and high levels and health outcomes in study participants with CHF at various levels of severity. Although the results from the effects of dietary sodium on outcomes in these participants appear inconsistent, several factors might have contributed to the disparate findings. Three RCTs with a similar sodium reduction protocol from a single site in Italy consistently demonstrate higher adverse events (hospital readmission and mortality) associated with lower-sodium diets. Treatment regimens in the three RCTs (Parrinello et al., 2009; Paterna et al., 2008, 2009) included low rates of beta blocker use and high-dose furosemide diuretic use combined with significant fluid restriction, which does not reflect contemporary U.S. management of patients with CHF.3 However, the committee could not identify weaknesses in the study designs. Further, the uniformity of the results even under very different basic treatments suggests a need for additional trials to be conducted by other investigators with participants under treatments similar to those used in the United States.

The observational studies may suggest a difference in the effect of lower sodium consumption depending on the degree of compensation of the CHF patients (NYHA I/II vs. III/IV), but this was not observed in the Paterna trials (one trial was done in NYHA II and one in decompensated CHF NYHA Class II to IV). Additional difficulty comparing the trials and the observational studies may arise from the differences in CHF patients after hospital discharge (e.g., the Paterna trials) and stable outpatients with heart failure (e.g., the observational studies in CHF clinics in Lennie et al. [2011] and Arcand et al. [2011]). Lastly, adjustment for different potential confounders may have influenced the outcomes observed in the cohort studies, leading to different interpretations. For example, while Arcand et al. (2011) controlled for caloric intake, none of the studies controlled for education, which has been shown to be associated with better health outcomes and lower sodium intake. In contrast to the Italian studies, which show consistency in the results, the results from Arcand et al. (2011) are inconsistent in that the number of hospital readmissions is largest in the middle sodium intake category (66 percent), although it has the lowest mortality (0 percent).

Importantly, CHF is the only health outcome for which RCTs exist. Therefore, although differences in disease management preclude a definitive assessment of the effects of low sodium intake (i.e., 1,840 mg per day) for CHF patients in the United States, the evidence suggests that low sodium intakes may lead to higher risk of adverse events in mid- to late-stage CHF patients with reduced ejection fraction receiving aggressive therapeutic regimens. In addition, a cohort study in a population including individuals with CVD supports the conclusion reached by the Italian studies. That is, low sodium intake levels might result in increased CHF episodes (O'Donnell et al., 2011). However, O'Donnell et al. (2011) used spot urine collection, and the relationship between low sodium intakes and higher CVD risk appeared limited to the outcome of CVD death and CHF, with no apparent relationship between low sodium and risk of MI or stroke.

Nevertheless, three randomized trials in CHF patients (Parrinello et al., 2009; Paterna et al., 2008, 2009) found that a sodium intake of 1,840 (vs. 2,870) mg per day was associated with increased mortality in this population subgroup. These trials, however, were limited to patients with mid- to late-stage CHF and reduced ejection fraction. In addition, patients were receiving aggressive therapeutic regimens that were very different from current standards of care, and thus, the results may not be generalizable.

Finally, the committee found that overall, the paucity of evidence in the general population and population subgroups strongly points to the need for further research to better define relationships between sodium intake and risk of CVD, stroke, and mortality, particularly at the lowest levels of sodium intake within the U.S. population.

STUDIES ON KIDNEY DISEASE

Two studies explored the relationship between sodium intake and CKD (Heerspink et al., 2012; Thomas et al., 2011). Collectively, the committee found these studies lacking in clarity about the risk of kidney disease progression or ESRD associated with sodium intake. Both of the studies reviewed evaluated populations with diabetes who had macroalbuminuria. These two studies show conflicting results about either benefits or risks associated with sodium intake in diabetic patients with macroalbuminuria. Thomas et al. (2011) demonstrated an inverse association of sodium intake with ESRD risk, which suggests that lower sodium intake may increase ESRD risk. On the other hand, the study by Heerspink et al. (2012) suggests that use of renin-angiotensin-aldosterone system (RAAS)-blocking agents may be more beneficial in patients with low sodium intake. RAAS blockade is the first-line therapy for treatment of diabetic nephropathy (Arauz-Pacheo et al., 2003; NKF, 2007) and has consistently been shown to delay progression of diabetic nephropathy into ESRD. Thus, Heerspink et al. (2012) suggest that sodium restriction may be beneficial rather than harmful in preventing kidney disease progression in kidney disease patients with diabetes and macroalbuminuria.

Overall, the studies published since 2003 reviewed by the committee provide inconsistent data about the relationship of sodium intake levels and kidney disease progression in patients with type 2 diabetes and mac-roalbuminuria, and some evidence suggests that low sodium intake may be harmful in this population. The committee found no studies published since 2003 evaluating the risk of ESRD among individuals with kidney disease at baseline, who had nondiabetic forms of CKD, or in individuals with diabetes but without macroalbuminuria. Some studies suggest that lower sodium intake may lead to lower proteinuria, and that proteinuria is a strong risk factor for CKD progression (described in Chapter 3). However, recent large-scale clinical trials demonstrate that decrements in proteinuria are not always associated with slower progression of CKD (Mann et al., 2008; Parving et al., 2012). For these reasons, to reach its conclusions, the committee relied primarily on data evaluating relationships between sodium level and CKD progression and dialysis initiation rather than changes in proteinuria.

STUDIES ON METABOLIC SYNDROME, DIABETES, AND GASTRIC CANCER

The committee identified from its literature search two cross-sectional studies that examined associations between sodium intake and risk of metabolic syndrome (Rodrigues et al., 2009; Teramoto et al., 2011). The committee also identified two prospective cohort studies that examined associations between sodium intake and risk of developing diabetes (Hu et al., 2005; Roy and Janal, 2010) and one study that examined the role of genetic polymorphisms linked to sodium intake in risk of diabetes (Daimon et al., 2008). These studies did not meet the committee's criteria for further evaluation of the strengths and weaknesses of the study and its relevance to the committee's task.

The committee identified seven prospective cohort studies (Murata et al., 2010; Peleteiro et al., 2011; Shikata et al., 2006; Sjodahl et al., 2008; Takachi et al., 2010; Tsugane et al., 2004; van den Brandt et al., 2003) and five case-control studies (Lazarević et al., 2011; Lee et al., 2003; Pelucchi et al., 2009; Strumylaite et al., 2006; Zhang and Zhang, 2011) that examined associations between sodium intake and risk of gastric cancer. These studies had a number of limitations, including that the reported intakes of the populations studied were not relevant to intake levels in the United States. Overall, the prospective cohort studies showed conflicting results for risk of gastric cancer and the case-control studies were potentially biased due to recall bias. Another possible effect modifier is infection with H. pylori. There is no agreement in the published literature, however, about whether the relationship between infection and sodium intake modifies risk of gastric cancer (Lee et al., 2003; Peleteiro et al., 2011). Taken together, the limitations in these studies precluded further evaluation. Although some evidence suggests that high sodium intake may be associated with increased risk of gastric cancer, the committee found that evidence was not definitive for an effect at low intake ranges.

Details about the design, characteristics, and outcomes for each of these studies are tabulated in Appendix F.

ADDITIONAL HEALTH OUTCOMES

In addition to the more commonly studied health outcomes above, the committee identified studies that examined outcomes related to ascites and reflux (Aanen et al., 2006; Gu et al., 2012; Nilsson et al., 2004), pulmonary function, including asthma and pulmonary hyperresponsiveness (Gotshall et al., 2004; Hirayama et al., 2010; Mickleborough et al., 2005; Sausenthaler et al., 2005), genitourinary symptoms, including kidney stone formation and urinary tract symptoms (Eisner et al., 2009; Maserejian et al., 2009; Meschi et al., 2012; Yun et al., 2010), depression (Song, 2009), and quality of life (Ramirez et al., 2004). The studies identified were inconsistent in methodological approach and results and, for the majority of specific outcomes, only one study was found for a given outcome. For example, although three studies addressed the potential association of sodium dietary intake and stone formation (Eisner et al., 2009; Meschi et al., 2012; Yun et al., 2010), the results were inconsistent. Meschi et al. (2012) found a positive correlation between sodium intake and calcium nephrolithiasis in a retrospective study of Italian women and Yun et al. (2010) found that in a cohort of stone formers, those with hypernatriuresis were more likely to develop stones in 3-year follow-up. The role of sodium intake in the risk of stone formation is not clear, and some authors suggest that while an increase in dietary sodium intake might increase urine calcium, it also might increase urine volume and decrease the urinary supersaturation of calcium oxalate (Eisner et al., 2009). In addition, these studies mirror many of the limitations of other studies reviewed in this chapter, including inconsistent and inadequate sodium intake assessment. Although the evidence was insufficient for the committee to draw conclusions about associations between sodium intake and these health outcomes it recognizes that other studies are ongoing and may be useful in the future.

REFERENCES

  • Aanen MC, Bredenoord AJ, Smout AJPM. Effect of dietary sodium chloride on gastro-oesophageal reflux: A randomized controlled trial. Scandinavian Journal of Gastroenterology. 2006;41(10):1141–1146. [PubMed: 16990198]
  • Arauz-Pacheo C, Parrott MA, Raskin P. Treatment of hypertension in adults with diabetes. Diabetes Care. 2003;26(Suppl 1):S80–S82. [PubMed: 12502624]
  • Arcand J, Ivanov J, Sasson A, Floras V, Al-Hesayen A, Azevedo ER, Mak S, Allard JP, Newton GE. A high-sodium diet is associated with acute decompensated heart failure in ambulatory heart failure patients: A prospective follow-up study. American Journal of Clinical Nutrition. 2011;93(2):332–337. [PubMed: 21084647]
  • ATBC Cancer Prevention Study Group. The Alpha-Tocopherol, Beta-Carotene Lung Cancer Prevention Study: Design, methods, participant characteristics, and compliance. Annals of Epidemiology. 1994;4:1–10. [PubMed: 8205268]
  • Cohen HW, Hailpern SM, Fang J, Alderman MH. Sodium intake and mortality in the NHANES II follow-up study. American Journal of Medicine. 2006;119(3):275–e7-275.e14. [PubMed: 16490476]
  • Cohen HW, Hailpern SM, Alderman MH. Sodium intake and mortality follow-up in the Third National Health and Nutrition Examination Survey (NHANES III). Journal of General Internal Medicine. 2008;23(9):1297–1302. [PMC free article: PMC2518033] [PubMed: 18465175]
  • Cook NR, Cutler JA, Obarzanek E, Buring JE, Rexrode KM, Kumanyika SK, Appel LJ, Whelton PK. Long term effects of dietary sodium reduction on cardiovascular disease outcomes: Observational follow-up of the trials of hypertension prevention (TOHP). British Medical Journal. 2007;334(7599):885–888. [PMC free article: PMC1857760] [PubMed: 17449506]
  • Cook NR, Obarzanek E, Cutler JA, Buring JE, Rexrode KM, Kumanyika SK, Appel LJ, Whelton PK. Joint effects of sodium and potassium intake on subsequent cardiovascular disease: The trials of hypertension prevention follow-up study. Archives of Internal Medicine. 2009;169(1):32–40. [PMC free article: PMC2629129] [PubMed: 19139321]
  • Costa APR, de Paula RCS, Carvalho GF, Araújo JP, Andrade JM, de Almeida OLR, de Faria EC, Freitas WM, Coelho OR, Ramires JAF, Quinaglia e Silva JC, Sposito AC. High sodium intake adversely affects oxidative-inflammatory response, cardiac remodelling and mortality after myocardial infarction. Atherosclerosis. 2012;222(1):284–291. [PubMed: 22436606]
  • Daimon M, Sato H, Sasaki S, Toriyama S, Emi M, Muramatsu M, Hunt SC, Hopkins PN, Karasawa S, Wada K, Jimbu Y, Kameda W, Susa S, Oizumi T, Fukao A, Kubota I, Kawata S, Kato T. Salt consumption-dependent association of the GNB3 gene polymorphism with type 2 DM. Biochemical & Biophysical Research Communications. 2008;374(3):576–580. [PubMed: 18656447]
  • Dong J, Li Y, Yang Z, Luo J. Low dietary sodium intake increases the death risk in peritoneal dialysis. Clinical Journal of the American Society of Nephrology. 2010;5(2):240–247. [PMC free article: PMC2827592] [PubMed: 20019116]
  • Eisner BH, Eisenberg ML, Stoller ML. Impact of urine sodium on urine risk factors for calcium oxalate nephrolithiasis. Journal of Urology. 2009;182(5):2330–2333. [PubMed: 19758639]
  • Ekinci EI, Clarke S, Thomas MC, Moran JL, Cheong K, MacIsaac RJ, Jerums G. Dietary salt intake and mortality in patients with type 2 diabetes. Diabetes Care. 2011;34(3):703–709. [PMC free article: PMC3041211] [PubMed: 21289228]
  • Gardener H, Rundek T, Wright CB, Elkind MSV, Sacco RL. Dietary sodium and risk of stroke in the Northern Manhattan Study. Stroke. 2012;43(5):1200–1205. [PMC free article: PMC3347890] [PubMed: 22499576]
  • Geleijnse JM, Witteman JCM, Stijnen T, Kloos MW, Hofman A, Grobbee DE. Sodium and potassium intake and risk of cardiovascular events and all-cause mortality: the Rotterdam Study. European Journal of Epidemiology. 2007;22(11):763–770. [PMC free article: PMC2071962] [PubMed: 17902026]
  • Gotshall RW, Rasmussen JJ, Fedorczak LJ. Effect of one week versus two weeks of dietary NaCl restriction on severity of exercise-induced bronchoconstriction. Journal of Exercise Physiology Online. 2004;7(1):1–7.
  • Gu XB, Yang XJ, Zhu HY, Xu BY. Effect of a diet with unrestricted sodium on ascites in patients with hepatic cirrhosis. Gut and Liver. 2012;6(3):355–361. [PMC free article: PMC3404174] [PubMed: 22844565]
  • Heerspink HJL, Holtkamp FA, Parving HH, Navis GJ, Lewis JB, Ritz E, De Graeff PA, De Zeeuw D. Moderation of dietary sodium potentiates the renal and cardiovascular protective effects of angiotensin receptor blockers. Kidney International. 2012;82(3):330–337. [PubMed: 22437412]
  • HHS and USDA (U.S. Department of Health and Human Services and U.S. Department of Agriculture). Dietary Guidelines for Americans, 2010. Washington, DC: U.S. Government Printing Office; 2010a. [February 4, 2013]. http://www​.cnpp.usda​.gov/Publications/DietaryGuidelines​/2010​/PolicyDoc/qPolicyDoc.pdf.
  • HHS and USDA. Report of the Dietary Guidelines Advisory Committee on the Dietary Guidelines for Americans, 2010, to the Secretary of Agriculture and the Secretary of Health and Human Services. Washington, DC: USDA/ARS; 2010b. [February 1, 2013]. http://www​.cnpp.usda​.gov/Publications/DietaryGuidelines​/2010​/DGAC/Report/2010DGACReport-camera-ready-Jan11-11.pdf.
  • Hirayama F, Lee AH, Oura A, Mori M, Hiramatsu N, Taniguchi H. Dietary intake of six minerals in relation to the risk of chronic obstructive pulmonary disease. Asia Pacific Journal of Clinical Nutrition. 2010;19(4):572–577. [PubMed: 21147720]
  • Hu G, Jousilahti P, Peltonen M, Lindstrom J, Tuomilehto J. Urinary sodium and potassium excretion and the risk of type 2 diabetes: A prospective study in Finland. Diabetologia. 2005;48(8):1477–1483. [PubMed: 15971060]
  • IOM (Institute of Medicine). Dietary reference intakes for water, potassium, sodium, chloride, and sulfate. Washington, DC: The National Academies Press; 2005.
  • IOM. Strategies to reduce sodium intake in the United States. Washington, DC: The National Academies Press; 2010.
  • Kono Y, Yamada S, Kamisaka K, Araki A, Fujioka Y, Yasui K, Hasegawa Y, Koike Y. Recurrence risk after noncardioembolic mild ischemic stroke in a Japanese population. Cerebrovascular Diseases. 2011;31(4):365–372. [PubMed: 21252505]
  • Larsson SC, Virtanen MJ, Mars M, Männistö S, Pietinen P, Albanes D, Virtamo J. Magnesium, calcium, potassium, and sodium intakes and risk of stroke in male smokers. Archives of Internal Medicine. 2008;168(5):459–465. [PubMed: 18332289]
  • Lazarević K, Nagorni A, Bogdanović D, Rančić N, Stošić L, Milutinović S. Dietary micronutrients and gastric cancer: Hospital based study. Central European Journal of Medicine. 2011;6(6):783–787.
  • Lee SA, Kang D, Shim KN, Choe JW, Hong WS, Choi H. Effect of diet and Helicobacter pylori infection to the risk of early gastric cancer. Journal of Epidemiology. 2003;13(3):162–168. [PubMed: 12749604]
  • Lennie TA, Song EK, Wu JR, Chung ML, Dunbar SB, Pressler SJ, Moser DK. Three gram sodium intake is associated with longer event-free survival only in patients with advanced heart failure. Journal of Cardiac Failure. 2011;17(4):325–330. [PMC free article: PMC3073718] [PubMed: 21440871]
  • Mann JF, Schmieder RE, McQueen M, Dyal L, Schumacher H, Pogue J, Wang X, Maggioni A, Budaj A, Chaithiraphan S, Dickstein K, Keltai M, Metsärinne K, Oto A, Parkhomenko A, Piegas LS, Svendsen TL, Teo KK, Yusuf S. Renal outcomes with telmisartan, ramipril, or both, in people at high vascular risk (the ONTARGET study): A multicentre, randomised, double-blind, controlled trial. The Lancet. 2008;372(9638):547–553. [PubMed: 18707986]
  • Maserejian NN, Giovannucci EL, McKinlay JB. Dietary macronutrients, cholesterol, and sodium and lower urinary tract symptoms in men. European Urology. 2009;55(5):1179–1189. [PMC free article: PMC2735823] [PubMed: 18692303]
  • Mattes RD, Donnelly D. Relative contributions of dietary sodium sources. Journal of the American College of Nutrition. 1991;10(4):383–393. [PubMed: 1910064]
  • Meschi T, Nouvenne A, Ticinesi A, Prati B, Guerra A, Allegri F, Pigna F, Soldati L, Vezzoli G, Gambaro G, Lauretani F, Maggio M, Borghi L. Dietary habits in women with recurrent idiopathic calcium nephrolithiasis. Journal of Translational Medicine. 2012;10(1) [PMC free article: PMC3337252] [PubMed: 22453026]
  • Mickleborough TD, Lindley MR, Ray S. Dietary salt, airway inflammation, and diffusion capacity in exercise-induced asthma. Medicine and Science in Sports and Exercise. 2005;37(6):904–914. [PubMed: 15947713]
  • Murata A, Fujino Y, Pham TM, Kubo T, Mizoue T, Tokui N, Matsuda S, Yoshimura T. Prospective cohort study evaluating the relationship between salted food intake and gastrointestinal tract cancer mortality in Japan. Asia Pacific Journal of Clinical Nutrition. 2010;19(4):564–571. [PubMed: 21147719]
  • Nagata C, Takatsuka N, Shimizu N, Shimizu H. Sodium intake and risk of death from stroke in Japanese men and women. Stroke. 2004;35(7):1543–1547. [PubMed: 15143292]
  • Nilsson M, Johnsen R, Ye W, Hveem K, Lagergren J. Lifestyle related risk factors in the aetiology of gastrooesophageal reflux. Gut. 2004;53(12):1730–1735. [PMC free article: PMC1774312] [PubMed: 15542505]
  • NKF (National Kidney Foundation). KDOQI clinical practice guidelines and clinical practice recommendations for diabetes and chronic kidney disease—guidelines 3: management of hypertension in diabetes and chronic kidney disease. 2007. [April 10, 2013]. http://www​.kidney.org​/professionals/kdoqi​/guideline_diabetes/guide3.htm. [PubMed: 17276798]
  • O'Donnell MJ, Yusuf S, Mente A, Gao P, Mann JF, Teo K, McQueen M, Sleight P, Sharma AM, Dans A, Probstfield J, Schmieder RE. Urinary sodium and potassium excretion and risk of cardiovascular events. Journal of the American Medical Association. 2011;306(20):2229–2238. [PubMed: 22110105]
  • Parrinello G, Di Pasquale P, Licata G, Torres D, Giammanco M, Fasullo S, Mezzero M, Paterna S. Long-term effects of dietary sodium intake on cytokines and neurohormonal activation in patients with recently compensated congestive heart failure. Journal of Cardiac Failure. 2009;15(10):864–873. [PubMed: 19944363]
  • Parving HH, Brenner BM, McMurray JJV, De Zeeuw D, Haffner SM, Solomon SD, Chaturvedi N, Persson F, Desai AS, Nicolaides M, Richard A, Xiang Z, Brunel P, Pfeffer MA. Cardiorenal end points in a trial of aliskiren for type 2 diabetes. New England Journal of Medicine. 2012;367(23):2204–2213. [PubMed: 23121378]
  • Paterna S, Gaspare P, Fasullo S, Sarullo FM, Di Pasquale P. Normal-sodium diet compared with low-sodium diet in compensated congestive heart failure: Is sodium an old enemy or a new friend. Clinical Science. 2008;114(3):221–230. [PubMed: 17688420]
  • Paterna S, Parrinello G, Cannizzaro S, Fasullo S, Torres D, Sarullo FM, Di Pasquale P. Medium term effects of different dosage of diuretic, sodium, and fluid administration on neurohormonal and clinical outcome in patients with recently compensated heart failure. American Journal of Cardiology. 2009;103(1):93–102. [PubMed: 19101237]
  • Peleteiro B, Lopes C, Figueiredo C, Lunet N. Salt intake and gastric cancer risk according to Helicobacter pylori infection, smoking, tumour site and histological type. British Journal of Cancer. 2011;104(1):198–207. [PMC free article: PMC3039805] [PubMed: 21081930]
  • Pelucchi C, Tramacere I, Bertuccio P, Tavani A, Negri E, La Vecchia C. Dietary intake of selected micronutrients and gastric cancer risk: An Italian case-control study. Annals of Oncology. 2009;20(1):160–165. [PubMed: 18669867]
  • Ramirez EC, Martinez LC, Tejeda AO, Gonzalez VR, David RN, Lafuente EA. Effects of a nutritional intervention on body composition, clinical status, and quality of life in patients with heart failure. Nutrition. 2004;20(10):890–895. [PubMed: 15474877]
  • Rodrigues SL, Baldo MP, de Sa Cunha R, Andreao RV, Del Carmen Bisi Molina M, Goncalves CP, Dantas EM, Mill JG. Salt excretion in normotensive individuals with metabolic syndrome: A population-based study. Hypertension Research—Clinical & Experimental. 2009;32(10):906–910. [PubMed: 19696779]
  • Roger VL, Go AS, Lloyd-Jones DM, Adams RJ, Berry JD, Brown TM, Carnethon MR, Dai S, de Simone G, Ford ES, Fox CS, Fullerton HJ, Gillespie C, Greenlund KJ, Hailpern SM, Heit JA, Ho PM, Howard VJ, Kissela BM, Kittner SJ, Lackland DT, Lichtman JH, Lisabeth LD, Makuc DM, Marcus GM, Marelli A, Matchar DB, McDermott MM, Meigs JB, Moy CS, Mozaffarian D, Mussolino ME, Nichol G, Paynter NP, Rosamond WD, Sorlie PD, Stafford RS, Turan TN, Turner MB, Wong ND, Wylie-Rosett J. Heart disease and stroke statistics—2011 update: A report from the American Heart Association. Circulation. 2011;123(4):e18–e209. [PMC free article: PMC4418670] [PubMed: 21160056]
  • Roy MS, Janal MN. High caloric and sodium intakes as risk factors for progression of retinopathy in type 1 diabetes mellitus. Archives of Ophthalmology. 2010;128(1):33–39. [PubMed: 20065214]
  • Sausenthaler S, Kompauer I, Brasche S, Linseisen J, Heinrich J. Sodium intake and bronchial hyperresponsiveness in adults. Respiratory Medicine. 2005;99(7):864–870. [PubMed: 15939248]
  • Shikata K, Kiyohara Y, Kubo M, Yonemoto K, Ninomiya T, Shirota T, Tanizaki Y, Doi Y, Tanaka K, Oishi Y, Matsumoto T, Iida M. A prospective study of dietary salt intake and gastric cancer incidence in a defined Japanese population: The Hisayama study. International Journal of Cancer. 2006;119(1):196–201. [PubMed: 16450397]
  • Sjödahl K, Jia C, Vatten L, Nilsen T, Hveem K, Lagergren J. Salt and gastric adenocarcinoma: A population-based cohort study in Norway. Cancer Epidemiology Biomarkers and Prevention. 2008;17(8):1997–2001. [PubMed: 18708389]
  • Song EK. Adherence to the low-sodium diet plays a role in the interaction between depressive symptoms and prognosis in patients with heart failure. Journal of Cardiovascular Nursing. 2009;24(4):299–305. [PMC free article: PMC3043551] [PubMed: 21206352]
  • Stolarz-Skrzypek K, Kuznetsova T, Thijs L, Tikhonoff V, Seidlerová J, Richart T, Jin Y, Olszanecka A, Malyutina S, Casiglia E, Filipovský J, Kawecka-Jaszcz K, Nikitin Y, Staessen JA. Fatal and nonfatal outcomes, incidence of hypertension, and blood pressure changes in relation to urinary sodium excretion. Journal of the American Medical Association. 2011;305(17):1777–1785. [PubMed: 21540421]
  • Strumylaite L, Zickute J, Dudzevicius J, Dregval L. Salt-preserved foods and risk of gastric cancer. Medicina. 2006;42(2):164–170. [PubMed: 16528133]
  • Takachi R, Inoue M, Shimazu T, Sasazuki S, Ishihara J, Sawada N, Yamaji T, Iwasaki M, Iso H, Tsubono Y, Tsugane S. Consumption of sodium and salted foods in relation to cancer and cardiovascular disease: The Japan Public Health Center-based prospective study. American Journal of Clinical Nutrition. 2010;91(2):456–464. [PubMed: 20016010]
  • Teramoto T, Kawamori R, Miyazaki S, Teramukai S. Sodium intake in men and potassium intake in women determine the prevalence of metabolic syndrome in Japanese hypertensive patients: OMEGA Study. Hypertension Research. 2011;34(8):957–962. [PMC free article: PMC3257031] [PubMed: 21654751]
  • Thomas MC, Moran J, Forsblom C, Harjutsalo V, Thorn L, Ahola A, Waden J, Tolonen N, Saraheima M, Gordin D, Groop PH. The association between dietary sodium intake, ESRD, and all-cause mortality in patients with type 1 diabetes. Diabetes Care. 2011;34(4):861–866. [PMC free article: PMC3064042] [PubMed: 21307382]
  • Tikellis C, Pickering RJ, Tsorotes D, Harjutsalo V, Thorn L, Ahola A, Waden J, Tolonen N, Saraheimo M, Gordin D, Forsblom C, Groop PH, Cooper ME, Moran J, Thomas MC. Association of dietary sodium intake with atherogenesis in experimental diabetes and with cardiovascular disease in patients with type 1 diabetes. Clinical Science. 2013;124(10):617–626. [PubMed: 23216128]
  • Tooze JA, Midthune D, Dodd KW, Freedman LS, Krebs-Smith SM, Subar AF, Guenther PM, Carroll RJ, Kipnis V. A new statistical method for estimating the usual intake of episodically consumed foods with application to their distribution. Journal of the American Dietetic Association. 2006;106(10):1575–1587. [PMC free article: PMC2517157] [PubMed: 17000190]
  • Tsugane S, Sasazuki S, Kobayashi M, Sasaki S. Salt and salted food intake and subsequent risk of gastric cancer among middle-aged Japanese men and women. British Journal of Cancer. 2004;90(1):128–134. [PMC free article: PMC2395341] [PubMed: 14710219]
  • Umesawa M, Iso H, Date C, Yamamoto A, Toyoshima H, Watanabe Y, Kikuchi S, Koizumi A, Kondo T, Inaba Y, Tanabe N, Tamakoshi A. Relations between dietary sodium and potassium intakes and mortality from cardiovascular disease: The Japan Collaborative Cohort study for evaluation of cancer risks. American Journal of Clinical Nutrition. 2008;88(1):195–202. [PubMed: 18614741]
  • van den Brandt PA, Botterweck AAM, Goldbohm RA. Salt intake, cured meat consumption, refrigerator use and stomach cancer incidence: A prospective cohort study (Netherlands). Cancer Causes and Control. 2003;14(5):427–438. [PubMed: 12946037]
  • Yang Q, Liu T, Kuklina EV, Flanders WD, Hong Y, Gillespie C, Chang MH, Gwinn M, Dowling N, Khoury MJ, Hu FB. Sodium and potassium intake and mortality among US adults: Prospective data from the Third National Health and Nutrition Examination Survey. Archives of Internal Medicine. 2011;171(13):1183–1191. [PubMed: 21747015]
  • Yun SJ, Ha YS, Kim WT, Kim YJ, Lee SC, Kim WJ. Sodium restriction as initial conservative treatment for urinary stone disease. Journal of Urology. 2010;184(4):1372–1376. [PubMed: 20723924]
  • Zhang Z, Zhang X. Salt taste preference, sodium intake and gastric cancer in China. Asian Pacific Journal of Cancer Prevention. 2011;12(5):1207–1210. [PubMed: 21875268]

Footnotes

1

Published online, December 2012.

2

The NYHA Functional Classification system is used to classify heart failure based on severity of symptoms and how the person feels during physical activity. Class I patients have cardiac disease but no limitation of physical activity. Class II patients experience slight limitations in physical activity (e.g., fatigue and palpitation). Patients classified as Class III experience limitation in less than ordinary physical activity. Class IV patients are unable to participate in physical activity without discomfort and may experience heart failure symptoms at rest (AHA, 2011).

3

The Guidelines for Heart Failure patients of the Heart Failure Association of America describes therapies appropriate for the different stages of a patient's health (e.g., decompensated heart failure, reduced ejection fraction, end of life). For patients with reduced ejection fraction (≥40 percent), the guideline recommends ACE inhibitors, ARBs (when ACE inhibitors are not tolerated by patients or in post-MI or chronic heart failure patients), beta blockers, and aldosterone antagonists (unless creatinine is >2.5 mg/dL or serum potassium is >5 mmol/L). In addition, diuretic therapy is recommended in certain patients to restore normal volume. For example, the recommended initial daily dose of furosemide is 20-40 mg once or twice per day for a maximum total daily dose of 600 mg. The guideline states that to minimize fluid retention, sodium intake should be limited to 2,000-3,000 mg per day and fluid should be restricted to less than 2,000 ml.

Copyright 2013 by the National Academy of Sciences. All rights reserved.
Bookshelf ID: NBK201520

Views

  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (1004K)

Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to PubMed

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...