Despite efforts over the past several decades to reduce sodium intake in the United States, adults still consume an average of 3,400 milligrams (mg) of sodium every day. A number of scientific bodies and professional health organizations, including the American Heart Association, the American Medical Association, and the American Public Health Association, support reducing dietary sodium intake, and the 2010 Dietary Guidelines for Americans includes as a goal to “reduce daily sodium intake to less than 2,300 milligrams (mg) and further reduce intake to 1,500 mg among persons who are 51 years of age and older and those of any age who are African American or have hypertension, diabetes, or chronic kidney disease.”
A substantial body of evidence supports these efforts to reduce sodium intake. This evidence links excessive dietary sodium to high blood pressure, a surrogate marker for cardiovascular disease (CVD), stroke, and cardiac-related mortality. However, concerns have been raised that a low sodium intake may adversely affect certain risk factors, including blood lipids and insulin resistance, and thus potentially increase risk of heart disease and stroke. In fact, several recent reports have challenged sodium reduction in the population as a strategy to reduce this risk.
1 This summary does not contain references. Citations to support statements made herein are given in the body of the report.
COMMITTEE TASK AND APPROACH
Against the backdrop of questions about sodium reduction in the population, the Centers for Disease Control and Prevention of the U.S. Department of Health and Human Services (HHS) asked the Institute of Medicine (IOM) to convene an expert committee to examine the designs, methodologies, and conclusions of this emerging evidence, as well as other reports published since the 2005 Dietary Reference Intake (DRI) report, Dietary Reference Intakes for Water, Sodium, Chloride, and Sulfate. Specifically, the committee was asked to review and assess the benefits and adverse outcomes (if any) of reducing sodium intake in the population, particularly in the range of 1,500 to 2,300 mg per day, with an emphasis on relevant subgroups. These subgroups include individuals with hypertension and prehypertension, those 51 years of age and older, African Americans, and those with diabetes, chronic kidney disease, and congestive heart failure (CHF). The committee also was asked to comment on the implications for population-based strategies to reduce sodium intake and to identify data and methods gaps and suggest ways to address them.
In approaching its task, the committee first conducted a broad search of the published literature through 2012 to identify relevant scientific publications on sodium and direct health outcomes. The committee was unable to identify studies published prior to 2003 that provided data on how the frequency of direct health outcomes was associated with changes in dietary sodium. Information also was gathered from a public workshop. The committee then developed a strategy to qualitatively assess each study identified as relevant from the search. Although an in-depth review was not conducted, the committee also considered evidence published since 2003 on associations between sodium intake and intermediate markers, particularly blood pressure. This additional evidence on the effect of sodium on blood pressure supported the findings and conclusions in the DRI report, the technical report from the 2010 Dietary Guidelines Advisory Committee (DGAC), and the recently released 2012 report from the World Health Organization, Guideline: Sodium Intake for Adults and Children.
Although blood pressure is a widely accepted surrogate marker, the scientific community continues to debate the use of other intermediate markers and biomarkers generally. Further, and in keeping with the systematic evidence review in the DGAC report, the effects of lowering sodium intake on blood pressure, as with other biomarkers, cannot always be disentangled from the effects of total dietary modification. For example, the committee’s review revealed that in a number of studies, the effects of dietary sodium on CVD outcomes sometimes persisted even after controlling for blood pressure. This suggests that associations between dietary sodium and risk of CVD may be mediated through interaction with other dietary factors (e.g.,
the effects of other electrolytes), or through pathways in addition to blood pressure. The committee sought to synthesize these potential associations. The committee’s approach to assessing the evidence focused on new data about the health effects of sodium intake on measures of health outcomes, rather than on effects mediated through an intermediate marker, namely blood pressure (see Figure D-1 in Appendix D).
The committee’s assessment of the evidence reviewed was guided by a number of factors. These included the study design, the quantitative measures of dietary sodium intake and confounder adjustment, as well as the number and consistency of relevant studies available. Assessing the impact of sodium intake on health outcomes was complicated by variability in the types and quality of measures used in observational studies, so that measures could not be reliably calibrated across studies. These measures also were difficult to assess in comparison to sodium intake in clinical trials. It was the consensus of the committee that the lack of consistency among studies in the methods used for defining sodium intakes at both high and low ends of the range of typical intakes among various population groups precluded deriving a numerical definition for high and low intakes in its findings and conclusions. Likewise, the extreme variability in intake levels between and among population groups precluded the committee from establishing a “healthy” intake range. The committee could consider sodium intake levels only within the context of each individual study.
The evidence for an effect of sodium intake on health outcomes reviewed by the committee included a broad range of population groups and methodological approaches. All of the evidence on the health outcomes related to CVD, stroke, and mortality was observational, mostly prospective cohort studies, whereas the evidence on health outcomes related to heart failure included randomized clinical trials (RCTs). Although the committee considered using a meta-analysis to assess the evidence, this approach was deemed inappropriate for this review because of the marked heterogeneity among the reviewed studies, particularly with respect to variations in measuring sodium intake and adjusting for confounders. For the same reason, the committee did not use a rating system to evaluate individual studies. Instead, studies were reviewed and assessed on an individual basis, and the committee considered the evidence on associations between sodium intake and health outcomes in its totality.
In evaluating each study, the committee considered RCTs a higher-quality study design for determining the effect of sodium on health outcomes than were observational studies. Well-executed cohort studies were considered more important for suggesting associations between dietary intake and health outcomes than were case-control studies because of the potential for bias in dietary assessment of sodium intake with the case-control study design. Finally, cross-sectional studies were included as an
indication of a potential association or to support (or not support) results from other studies. The committee used two major criteria to assess the quality of the evidence for all study designs: (1) the method to estimate sodium intake and the quality of its implementation, and (2) confounder adjustment. Other criteria included the approach used to change sodium intake, the instrument used to estimate sodium intake, the length of the intervention or follow-up of participants, interactions with other factors, and generalizability to the general population or population subgroups.
FINDINGS AND CONCLUSIONS
Recognizing the limitations of the available evidence, the committee found no consistent evidence to support an association between sodium intake and either a beneficial or adverse effect on most direct health outcomes other than some CVD outcomes (including stroke and CVD mortality) and all-cause mortality. Some evidence suggested that decreasing sodium intake could possibly reduce the risk of gastric cancer. However, the evidence was too limited to conclude the converse—that higher sodium intake could possibly increase the risk of gastric cancer. Interpreting these findings was particularly challenging because most studies were conducted outside the United States in populations consuming much higher levels of sodium than those consumed in this country. Thus, the committee focused its findings and conclusions on evidence for associations between sodium intake and risk of CVD-related events and mortality.
Findings and Conclusions for Cardiovascular Disease, Stroke, and Mortality
General U.S. Population
Finding 1: The committee found that the results from studies linking dietary sodium intake with direct health outcomes were highly variable in methodological quality, particularly in assessing sodium intake. The range of limitations included over- or underreporting of intakes or incomplete collection of urine samples. In addition, variability in data collection methodologies limited the committee’s ability to compare results across studies.
Conclusion 1: Although the reviewed evidence on associations between sodium intake and direct health outcomes has methodological flaws and limitations, the committee concluded that, when considered collectively, it indicates a positive relationship between higher levels of sodium intake and risk of CVD. This evidence is consistent with existing evidence on blood pressure as a surrogate indicator of CVD risk.
Finding 2: The committee found that the evidence from studies on direct health outcomes was insufficient and inconsistent regarding an association between sodium intake below 2,300 mg per day and benefit or risk of CVD outcomes (including stroke and CVD mortality) or all-cause mortality in the general U.S. population.
Conclusion 2: The committee determined that evidence from studies on direct health outcomes is inconsistent and insufficient to conclude that lowering sodium intakes below 2,300 mg per day either increases or decreases risk of CVD outcomes (including stroke and CVD mortality) or all-cause mortality in the general U.S. population.
Finding 1: The committee found that the evidence from multiple RCTs that were conducted by a single investigative team indicated that low sodium intake (e.g., down to 1,840 mg per day) may lead to greater risk of adverse events in CHF patients with reduced ejection fraction and who are receiving certain aggressive therapeutic regimens. This association also is supported by one observational study in which low sodium intake levels in patients with CVD and diabetes were associated with higher risk of CHF events.
Conclusion 1: The committee concluded that the available 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 and who are receiving aggressive therapeutic regimens. Because these therapeutic regimens were very different than current standards of care in the United States, the results may not be generalizable. Similar studies in other settings and using regimens more closely resembling those in standard U.S. clinical practice are needed.
Finding 2: The committee found that data among prehypertensive participants from two related studies provided some evidence suggesting a continued benefit of lowering sodium intake in these patients down to 2,300 mg per day (and lower, although based on small numbers in the lower range). In contrast, the committee found no evidence for benefit and some evidence suggesting risk of adverse health outcomes associated with sodium intake levels in ranges approximating 1,500 to 2,300 mg per day in other disease-specific population subgroups, specifically those with diabetes, chronic kidney disease (CKD), or preexisting CVD. In addition to inconsistencies in sodium intake measures, methodological flaws included the possibility of confounding and reverse causality. No relevant evidence was found on health outcomes for other population subgroups considered (i.e., persons
51 years of age and older and African Americans). In studies that explored interactions, race, age, or prevalence of hypertension or diabetes did not change the effect of sodium on health outcomes.
Conclusion 2: The committee concluded that, with the exception of the CHF patients described above, the current body of evidence addressing the association between low sodium intake and health outcomes in the population subgroups considered is limited. The evidence available is inconsistent and limited in its approaches to measuring sodium intake. The evidence also is limited by small numbers of health outcomes and the methodological constraints of observational study designs, including the potential for reverse causality and confounding.
The committee further concluded that, while the current literature provides some evidence for adverse health effects of low sodium intake among individuals with diabetes, CKD, or preexisting CVD, the evidence on both the benefit and harm is not strong enough to indicate that these subgroups should be treated differently from the general U.S. population. Thus, the committee concluded that the evidence on direct health outcomes does not support recommendations to lower sodium intake within these subgroups to, or even below, 1,500 mg per day.
Implications for Population-Based Strategies to Gradually Reduce Sodium Intake in the U.S. Population
As noted in Chapter 1, recommendations of the Panel on Dietary Reference Intakes for Electrolytes and Water of an Adequate Intake for sodium of 1,500 mg per day for all individuals 9 years of age up to 51 years of age was set as an amount necessary to achieve an overall diet that provides an adequate intake of other nutrients and also covers sodium sweat losses. A Tolerable Upper Intake Level for sodium was set at 2,300 mg per day based on evidence showing associations between high sodium intakes and risk of high blood pressure and consequent risk of CVD, stroke, and mortality.
Given this background, overall, the committee found that the available evidence on associations between sodium intake and direct health outcomes is consistent with population-based efforts to lower excessive dietary sodium intakes, but it is not consistent with recommendations that encourage lowering of dietary sodium in the general population to 1,500 mg per day. Further, as noted in the 2010 DGAC report, population subgroups, including those with diabetes, CKD, or preexisting CVD, individuals with hypertension, prehypertension, persons 51 years of age and older, and African Americans represent, in aggregate, a majority of the general U.S. population. Thus, when considered in light of the current state of the
evidence on associations between sodium intake and direct health outcomes for these subgroups, except when data specifically indicate they are different, there is not sufficient evidence to support treating them differently from the general U.S. population.
The committee was not asked to draw conclusions about a specific target range of dietary sodium for the general population or for population subgroups. However, the committee notes that there are important factors it considered that preclude such a conclusion. For example, one factor that is often discussed in the context of other health-related questions is the challenge of defining specific intake levels when the variables of interest are continuous. That is an especially difficult issue in the present circumstances, where the target intake level could theoretically differ for different, large population subgroups.
Other methodological factors that preclude making conclusions about a specific target range for sodium relate to the variability in approaches and study designs in the literature reviewed. Most importantly, quantitative methods for measuring dietary sodium intake have limitations and there are impedients to calibrating those measures across different methodological approaches and study designs. Methodological problems in assessing sodium intake make this particularly challenging.
FUTURE RESEARCH TO ADDRESS GAPS IN DATA AND METHODOLOGY
The committee identified a number of data and methods gaps in studies on sodium intake and risk of adverse health outcomes among population groups. Further research in the areas highlighted below would strengthen the evidence base on the association between lower (1,500 to 2,300 mg) levels of sodium and health outcomes in the general population and population subgroups:
1. Standardized methodological approaches to measure sodium intake in population groups. Specific examples include standardizing the use of 24-hour urine collections and validating sodium intake estimates with data on urine volume, urine creatinine, and body weight;
2. Approaches using dietary sodium intake levels corresponding to levels in current guidelines (i.e., 1,500 to 2,300 mg per day) when examining associations between sodium intake and health outcomes;
3. Analyses examining the effects on health outcomes of dietary sodium in combination with other electrolytes, particularly potassium;
4. Methods that account for confounding factors in dietary studies, including the influence of reported total daily caloric intake on observational associations between sodium and health outcomes, and methods that clarify attributes of individuals with apparently low sodium intake or excretion; and
5. Analyses of interactions with antihypertensive medication and blood pressure in studies examining associations between sodium intake and health outcomes.
In addition, the committee identified a need for RCT research, and observational and mechanistic studies, particularly in population subgroups. Examples of such clinical trials include those to examine
1. effects of a range of sodium levels on risk of cardiovascular events, stroke, and mortality among
a. patients in controlled environments, where randomized trials may be more feasible, such as the elderly in chronic care facilities and other institutionalized individuals; and
b. individuals as part of natural experiments, such as those in other countries where policies affecting sodium consumption are in effect;
2. effects of low-sodium diets on adverse events among CHF patients receiving therapeutic treatment modalities typically used in the United States; and 3.
3. potential beneficial or adverse outcomes of a range of sodium intakes among African Americans, adults 51-70 years of age, 70 years of age and older, and other population subgroups; RCTs may be particularly important within higher-risk patient populations, where reverse causation is a potential limitation of observational studies.
The committee also identified a need for studies to collect and reanalyze
1. data from existing clinical trials that were designed to evaluate sodium and health; and 2.
2. data during extended follow-up periods after completion of a clinical trial to identify health outcomes, such as mortality, that could manifest later in life and after longer follow-up periods. Such trials would not be simple to conduct, however, and careful feasibility assessment is needed first.
In addition to RCT research, mechanisitic studies are needed to examine potential physiologic changes associated with lowering sodium intake and adverse health outcomes. Finally, additional observational research is needed to examine associations between sodium intake and cancer, especially gastric cancer, in the U.S. population, as well as associations between sodium intake and caloric intake in both short-term and longitudinal studies.