9

Sodium: Dietary Reference Intakes for Toxicity

The Tolerable Upper Intake Level (UL) specifies the highest average daily intake level of a nutrient, consumed on a habitual basis, that is likely to pose no risk of adverse health effects for nearly all apparently healthy individuals in a given Dietary Reference Intake (DRI) age, sex, and life-stage group. The potential for adverse health effects increases as intakes increase above the UL. The UL is intended to provide guidance on intake levels that are safe; it is not intended to serve as an intake goal. The Guiding Principles for Developing Dietary Reference Intakes Based on Chronic Disease (Guiding Principles Report) recommended that the UL be retained in the expanded DRI model, but that it should characterize toxicological risk (NASEM, 2017). Although this conceptual revision narrows the scope of the UL, it allows for a more nuanced characterization of the different types of risk that can exist with intake of a nutrient or other food substance. This chapter presents the committee’s review of the evidence on the toxicological effects of excessive sodium intake and its conclusion regarding establishing a sodium UL. For context, the committee’s findings are preceded by a brief summary of the decision made regarding the sodium UL in the Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate (2005 DRI Report) (IOM, 2005).

SODIUM TOLERABLE UPPER INTAKE LEVELS ESTABLISHED IN THE 2005 DRI REPORT

To determine if a UL could be established, the 2005 DRI Report assessed evidence on the relationship between sodium intake and the fol-

lowing indicators: blood pressure; stroke; coronary heart disease; left ventricular mass; calcium excretion, bone mineral density, and kidney stones; pulmonary function; and gastric cancer. Evidence for a relationship between sodium intake and blood pressure, which was described as “direct and progressive” (IOM, 2005, p. 378), ultimately served as the basis for the UL in the 2005 DRI Report. The lowest-observed-adverse-effect level (LOAEL) was informed by three multidose sodium trials (Johnson et al., 2001; MacGregor et al., 1989; Sacks et al., 2001),¹ and it was determined to be the next lowest sodium intake level above the Adequate Intake (AI). A no-observed-adverse-effect level could not be identified owing to the continuous relationship between sodium intake and blood pressure. No uncertainty factor was applied to the LOAEL. A sodium UL of 2,300 mg/d (100 mmol/d) was set for all adult DRI age, sex, and life-stage groups; the UL for children and adolescents 1–18 years of age was extrapolated from the adult UL, based on median energy intake from the 1994–1996 Continuing Survey of Food Intakes by Individuals.

REVIEW OF POTENTIAL INDICATORS OF TOXICOLOGICAL ADVERSE EFFECTS OF EXCESSIVE SODIUM INTAKE

The expanded DRI model shifts consideration of evidence on the relationship between intake and chronic disease indicators to DRIs based on chronic disease. As such, the approach to establish the sodium UL in this report differs from the approach taken in the 2005 DRI Report. For instance, evidence on the relationships between sodium intake and blood pressure, stroke, coronary heart disease, left ventricular mass, bone-related indicators, and kidney disease was reviewed in the 2005 DRI Report as potentially informing the UL, but it is now considered in the context of establishing sodium Chronic Disease Risk Reduction Intakes (CDRRs; see Chapter 10). Similarly, evidence on relationships between sodium intake and pulmonary function and gastric cancer would now be considered as potentially informing the sodium CDRR; however, as described in Appendix D, the available evidence did not support use of these indicators.

For ethical reasons, trials cannot be designed to evaluate whether an intervention will increase the incidence of adverse effects. Consequently, adverse effect data in trials are almost always secondary outcomes. These data, particularly if systematically and carefully reported, can provide use-

___________________

¹Sacks et al. (2001) was included in the Agency for Healthcare Research and Quality systematic review, Sodium and Potassium Intake: Effects on Chronic Disease Outcomes and Risks (AHRQ Systematic Review) (Newberry et al., 2018), but the other two studies were not. The AHRQ Systematic Review excluded Johnson et al. (2001) on the basis of the timing not being of interest and excluded MacGregor et al. (1989) on the basis of study design.

ful information for evaluating the likelihood of adverse effects. However, as secondary outcomes, these trials may not be adequately powered to identify a statistically significant occurrence of an adverse effect. These strengths and limitations need to be taken into account when using data from trials for evaluating the potential for adverse effects.

Guided by the first step of the DRI organizing framework, the committee sought to identify potential indicators of toxicological adverse effects from excessive sodium intake. The sections that follow describe the evidence the committee reviewed to identify indicators that could potentially inform the derivation of the sodium UL, as well as summarize the evidence on the potential indicator identified.

Evidence Reviewed to Identify Potential Toxicological Indicators

The committee conducted a literature scan to identify potential indicators that may be informative for the sodium DRIs (see Appendix D), but it did not reveal any potential indicator of sodium toxicity, separate from consideration of chronic disease–related indicators. Additional exploration of systematic reviews and case reports on toxicity, adverse effects, and poisonings from sodium intake were undertaken in an effort to identify potential toxicological adverse effects. From these efforts, the committee identified a collection of case reports on deaths attributed to high levels of sodium intake. The committee also compiled reported adverse effects of the sodium trials included in the Agency for Healthcare Research and Quality’s systematic review, Sodium and Potassium Intake: Effects on Chronic Disease Outcomes and Risks (AHRQ Systematic Review) (Newberry et al., 2018), and the committee’s supplemental literature searches. The committee notes that the doses used in trials are generally not high enough to cause serious adverse effects, as it would be unethical to randomize participants to such an exposure. The intent of these evidence searches was to identify specific indicators that could potentially inform the sodium UL. The evidence that was compiled is described below.

Case Reports of Death

Several case reports exist in the literature regarding lethal levels of sodium intake, primarily attributable to the ingestion of massive acute doses. A 2017 systematic review summarized evidence on 35 fatalities from acute ingestion of massive doses of salt (Campbell and Train, 2017). Explanations of the massive acute intakes included salt being mistaken for sugar, used as an emetic, used in exorcism rituals, or, in some of the case reports about children, administered by a parent. All cases that documented sodium blood concentrations reported concentrations exceeding 150 mmol/L, indi-

cating hypernatremia. Many, but not all, of the cases included co-ingestion of other potential toxins (e.g., medications for anxiety, depression, schizophrenia) and occurred in individuals with chronic conditions or illnesses (e.g., depression, psychiatric disease, Prader-Willi syndrome).

The estimated level of sodium intake varied across the case reports. The lowest level of sodium intake among the adult cases was estimated to be between 6,800 and 10,200 mg (296 and 444 mmol), ingested as a saline emetic for a suspected antipsychotic medication poisoning in a 48-year-old female (Gresham and Mashru, 1982). In a case of an 83-year-old female with hypertension and dementia, sodium intake was estimated to be between 13,600 and 20,400 mg (592 and 887 mmol) (Engjom and Kildahl-Andersen, 2008). Substantially higher intakes were also reported, including 273,000 mg (11,875 mmol) of sodium consumed by a 34-year-old female (Raya et al., 1992) and less than 400,000 mg (17,399 mmol) of sodium by a 20-year-old female with psychiatric disorders (Ofran et al., 2004). Among children, the lower levels of intake that resulted in death among the identified case reports included an estimated 5,000 mg (219 mmol) of sodium in a 7-month-old female (Martos Sanchez et al., 2000) and less than 7,000 mg (304 mmol) in a 2-year-old with gastrointestinal strictures (Scott and Rotondo, 1947). The estimated doses of sodium intake among children, however, were largely not reported.

Case reports provide evidence that acute ingestion of large doses of sodium, including rapid ingestion of salt in liquid solution, can lead to death. Collectively, the case reports provide information about limits of biological homeostatic controls related to sodium, but they do not necessarily reflect the toxicological effects of habitually elevated intake levels suitable for establishing a sodium UL. In many of the cases, sodium was co-ingested with other potential toxins (e.g., medications), and most poisonings occurred in individuals with coexisting conditions and illnesses. Not all of the case reports provided the dose of sodium ingested. In cases where the level of intake was reported, the amount was often estimated because it was not possible to determine the exact quantity consumed. The absence or imprecision of intake estimates leading to death, coupled with the acute nature of the excessive sodium intake, limits the committee’s ability to use these case reports to inform the sodium UL.

Hypernatremia was also described in the case reports. In general, hypernatremia is associated with symptoms such as nausea, vomiting, headache, and fever but does not always result in death. Hypernatremia was determined not to be an informative indicator of sodium toxicity, as it is typically caused by severe dehydration rather than excessive sodium intake (Adrogué and Madias, 2000; Sterns, 2015). The case reports did not reveal any other potential indicators of sodium toxicity.

Adverse Events Reported in Sodium Trials

The AHRQ Systematic Review did not have a key question regarding adverse events in sodium trials, but it provided a brief summary of commonly reported adverse events in the context of sodium reduction. Building on this work, the committee reviewed descriptions of adverse events reported in trials meeting the inclusion criteria for the AHRQ Systematic Review and the committee’s supplemental literature searches (see Table 9-1).

As outlined in Table 9-1, participants were varied and included healthy normotensive adults, adults with treated and untreated hypertension, pregnant women, and children. Several of the smaller studies did not report marked differences in adverse events between the high- and low-sodium intervention periods (crossover trials) or groups (parallel randomized controlled trials). There was little consistency of the types of adverse events reported and the extent to which they differed between the intervention periods or arms. However, two findings emerged from this review of evidence. First, the crossover studies by Todd et al. (2010, 2012) provide some evidence regarding the level of intake associated with adverse effects. This finding is further considered below. Second, among the reported adverse events, some trials reported reduction of headaches among those in the lower-sodium intervention period or arm. Given this, the committee explored the evidence on headaches as a potential indicator of sodium toxicity in the next section.

Todd et al. (2010) assessed 34 adults with hypertension during three different sodium interventions in a crossover study. Throughout, participants consumed a diet containing 1,380 mg/d (60 mmol/d) sodium. In a random order, participants added to their low-sodium diet 500 mL of tomato juice containing 0, 2,070, or 3,220 mg/d (0, 90, or 140 mmol/d) sodium for a period of 4 weeks each. The investigators reported that seven of the participants withdrew from the highest sodium period because of elevated blood pressure and other symptomology. Only one participant withdrew from the moderate sodium period. In Todd et al. (2012), the tomato juice trial was conducted in 23 normotensive adults. The design was similar to that which was conducted in adults with hypertension, except the highest dose of sodium provided in the tomato juice was 4,370 mg/d (190 mmol/d) rather than 3,220 mg/d (140 mmol/d). Nine of the first 10 participants who completed the highest sodium period reportedly experienced adverse effects, leading the investigators to reduce the amount of sodium in the tomato juice during the highest period to 3,220 mg/d (140 mmol/d). The reduction in sodium content of the tomato juice did not change reported bloating, but it did improve other symptoms experienced (included frequency of cramps

TABLE 9-1 Sodium Trials Included in the AHRQ Systematic Review and the Committee’s Supplemental Literature Search That Provided a Description of Adverse Events

NOTES: Adverse events in the table reflect those reported by the study authors. Omitted from this table are mortality or cardiovascular disease adverse events, unless such occurrences were included in the description of participant withdrawal from the study. Urinary excretion and intake values are presented in mmol. To convert the mmol value to milligrams, multiply the excretion or intake level by 23.0. ACE = angiotensin-converting enzyme; BP = blood pressure; DASH = Dietary Approaches to Stop Hypertension; DBP = diastolic blood pressure; HDL-C = high-density lipoprotein-cholesterol; mm Hg = millimeter mercury; N/A = not applicable; SBP = systolic blood pressure; TOHP = Trials of Hypertension Prevention.

^aFor crossover trials, duration is per dietary period.

^bThis group represents the period or group intended to have the highest level of sodium intake in the study. In several studies, this group reflects usual sodium intake.

^cParticipants were assigned to either the DASH diet (n = 208) or a control diet (n = 204). Within the diet assignment, participants consumed three different levels of sodium.

^dDaily sodium intake was proportionate to total energy intake of each individual participant.

^ePresented as mean urinary sodium excretion for the DASH diet arm/control diet arm, respectively.

^fThis finding was reported in a separate publication (Harsha et al., 2004).

^gUrinary sodium was reported as urinary sodium-to-creatinine ratio. Estimates in the table reflect the estimated amount of dietary sodium consumed. In both studies, participants consumed a low-sodium diet (60 mmol/d) and then received 0, 90, or 140 mmol/d of additional sodium through tomato juice.

^hThreshold for withdrawal in the study was blood pressure > 160/100 mm Hg.

ⁱParticipants consumed a low-sodium diet (60 mmol/d) throughout the study. The highest dose of sodium provided in the tomato juice began at 190 mmol/d, but was reduced to 140 mmol/d due to adverse events.

^jIncluded frequency of cramps upon exercising, joint pain, vomiting, headaches, shortness of breath, and other reported symptoms.

^kPresented as 24-hour urinary sodium excretion during the high–sodium bicarbonate period and high–sodium chloride period, respectively.

^lMedian length of follow-up.

^mPublication included other intervention arms not specific to sodium only, which are not included in this table.

ⁿAs reported in Appel et al., 2001, examined adverse events included excessive weight loss, physical injury from exercise, palpitations, nonischemic chest pain, dizziness, edema, excessive weight gain, or other adverse events.

^oThe publication reported average 48-hour urine to be 204 ± 33 mmol. Value was divided in half to obtain the 24-hour estimate reported in this table.

^pThe publication reported average 48-hour urine to be 321 ± 36 mmol. Value was divided in half to obtain the 24-hour estimate reported in this table.

^qConventional low-salt diet education versus intensive low-salt intervention. Both groups received angiotensin II receptor blocker throughout the trial.

^rBetween-group difference in urinary sodium excretion was reported to be 38 mmol. Baseline urinary sodium excretion was 93 mmol/d in the low-sodium group and 98 mmol/d in the high-sodium group.

^sIdentified after the end of the study. Data were removed from analyses.

^tIntervention was started in the 14th week of pregnancy and stopped at delivery. Duration in this table assumes length of pregnancy is 40 weeks.

^uPublication states that urinary sodium excretion in the low-sodium group was approximately one-third that of the unrestricted group. Values could not be determined from the figure in the publication. Values in this table are based on baseline values of 24-hour sodium excretion in the unrestricted group.

^vNumber of participants in this table reflect the number of children randomized. Only 17 of the 41 randomized to the intervention completed the full program.

^wReflects 32 children randomized to the intervention, including those who dropped out. Baseline sodium excretion was lower among those who actively participated in the program than those in the control group and those who dropped out.

^xValues expressed in the publication were mmol/10 hours, based on an overnight urine collection. Estimates presented in the publication were multiplied by 2.4 to approximate 24-hour excretion.

^yReflects 32 children randomized to the control group. Baseline sodium excretion was higher among those in the control group than those who actively participated in the program.

upon exercising, vomiting, joint pain, headaches, shortness of breath, and other symptoms).

The committee considered whether the two studies by Todd et al. (2010, 2012) could be used to derive a sodium UL, as both studies provide evidence of an intake–response relationship. Factoring into the committee’s decision were the strengths and limitations of using the available data for such a purpose. The two studies are among the few publications included in the AHRQ Systematic Review that evaluated multiple doses of sodium intake, including an elevated intake level. Furthermore, the reported adverse effects were documented in both adults with hypertension and normotensive adults, suggesting implications for the general population. The studies, however, lacked key information about ingestion of the sodium and the adverse events. Neither the publications nor the clinical trial registry for the studies (ANZCTR, 2010) provided information regarding how participants were instructed to consume the tomato juice and how the participants operationalized the protocol. Interpretation of the reported adverse events would likely differ if participants consumed the 500 mL of tomato juice as a single bolus without food, as opposed to consuming portions over the course of the day with food. Furthermore, the derivation of a UL is driven by the identification of an indicator of toxicological adverse effects. The two studies did not sufficiently characterize the reported symptomology (e.g., level of severity, number of participants reporting each symptom, temporal relationship with ingestion of the tomato juice); furthermore, no specific indicator could be identified from either study. Todd et al. (2010, 2012) provide key evidence of adverse effects from consumption of concentrated sources of sodium. However, the uncertainties about the consumption of the sodium interventions and the limited characterization of the adverse effects prevented the committee from using these two studies to establish sodium ULs.

Review of Evidence on a Potential Indicator

Headaches²

In the Trial of Nonpharmacologic Interventions in the Elderly (TONE), 681 participants 60–80 years of age with hypertension (systolic blood pressure less than 145 mm Hg and diastolic blood pressure less than 85 mm Hg while taking one antihypertensive medication) were randomized to a reduced sodium intervention or control (Appel et al., 2001). On average, sodium intake in the intervention group was 920 mg/d (40 mmol/d) lower

___________________

² Evidence presented in this section were gathered through the committee’s supplemental literature search and information-gathering activities (see Appendix E).

compared to the control group,³ and systolic and diastolic blood pressures were 4.3 and 2.0 mm Hg lower, respectively. Headache was less frequently self-reported as an adverse event in the intervention group compared to the control group (35 versus 54 individuals, rates not given, p = .04). In a follow-up study of headache in 975 individuals in TONE,⁴ of which 90 percent had some follow-up, cumulative incidence of headache in the group not randomized to a sodium reduction intervention (i.e., usual care or weight loss intervention) was 14.3 percent compared to 10.5 percent in those randomized to a sodium intervention (i.e., sodium reduction alone, or in combination with weight loss) (p = .012) (Chen et al., 2016). Adjustment for systolic and diastolic blood pressure did not have an appreciable impact on the study results. As compared to the usual care group, the intervention group that was only sodium reduction (i.e., excluding those in the combined sodium reduction and weight loss arm) had reduced hazards of headaches (hazard ratio = 0.61 [95% confidence interval {CI}: 0.39, 0.95]; p = .03). In analyses that considered 24-hour urinary sodium excretion, each 230 mg/d (10 mmol/d) increase in sodium excretion was associated with a 7 percent increase in the hazard for headaches (p ≤ .001); this relationship persisted after adjustment for systolic and diastolic blood pressure. There was evidence of a nonlinear effect, with the greatest effect on headache above 3,449 mg/d (150 mmol/d) sodium.

In the Dietary Approaches to Stop Hypertension (DASH)-Sodium trial, 412 participants who were classified as having prehypertension or stage I hypertension were randomized to one of two parallel diet arms (DASH versus control), and within these diet arms participants consumed three levels of sodium for 30 days each in a crossover feeding study (Sacks et al., 2001). In the control arm, headache occurred in 47 percent of participants during the high-sodium feeding period compared with 39 percent during the low-sodium period (p < .05). In a follow-up analysis of headache in the DASH-Sodium trial (Amer et al., 2014), headache incidence was 47, 41, and 39 percent for the high-, intermediate-, and low-sodium periods of the control diet arm, and 43, 38, and 36 percent for the high-, intermediate-, and low-sodium periods of the DASH diet arm. In models adjusted for age, sex, race, site, systolic blood pressure, body mass index, smoking, and carryover effects, there were no significant differences between the DASH diet and control diet within sodium level. Controlling for these same covariates, there were significant differences between high and low sodium within each

___________________

³ As assessed through 24-hour urinary sodium excretion. Mean baseline 24-hour urinary excretion was 3,311 mg/d (144 mmol/d) among the intervention group, and 3,334 mg/d (145 mmol/d) among the control group.

⁴ The sample size is larger in this report on TONE, as compared to Appel et al. (2001), as it includes participants who were randomized to the weight loss or combined weight loss and sodium reduction interventions.

diet arm (DASH, p = .04; control, p = .05). Although no test for trend was reported, the decreased headache incidence by sodium level across groups is supportive of an intake–response relationship.

Committee’s synthesis of the evidence Headaches were reported to be reduced during the lower-sodium period or in the lower-sodium group in some of the trials included in the AHRQ Systematic Review (see Table 9-1). Headaches occur commonly among the general population for a variety of reasons, many of which are unknown. The available studies generally lacked detailed information about the type, severity, duration, and frequency of headaches. The persistence or transience of the headache response is also not well characterized. Although post-hoc statistical adjustments suggested that the headache effects may be independent of blood pressure effects, more data are needed to understand if and how headache and blood pressure effects are related and the interplay of sodium intake. Headache was prevalent at low sodium intakes and, as such, there is a lack of information that might be used to identify no-effect or minimum effect intakes for sodium-induced headaches, both of which are important for UL development. Thus, while the committee acknowledges that there is evidence of a relationship between sodium intake and headaches, the uncertainties in the evidence preclude using headaches as a critical adverse effect to establish a sodium UL. The committee, however, notes that the same studies used to evaluate headaches are also part of the evidence base used to establish the sodium CDRR. Therefore, this latter DRI value will cover the range of intakes associated with headache.

THE COMMITTEE’S CONCLUSION REGARDING THE TOLERABLE UPPER INTAKE LEVELS FOR SODIUM

Extreme intakes of sodium, especially ingested as a massive acute dose, have been shown to cause severe adverse effects, including death. Two studies provide evidence that a higher concentration of sodium delivered through tomato juice resulted in more adverse effects than when the same volume of tomato juice had a lower concentration of sodium or no sodium at all (Todd et al., 2010, 2012). Additionally, some sodium trials have indicated changes in the occurrence of headaches with changes in sodium intake, but questions remain regarding the nature and severity of the reported headaches. However, without a specific indicator of a toxicological effect of high sodium intake that can be used to establish a quantitative relationship, the committee concluded that a sodium UL cannot be established.

The committee concludes that there is insufficient evidence of sodium toxicity risk within the apparently healthy population to establish a sodium Tolerable Upper Intake Level (UL).

Cautions are in order regarding the possible adverse consequences of excessive sodium consumption, particularly in the concentrated doses. The limitations that exist in the evidence highlight the need for future monitoring and research opportunities (see Chapter 12).

Conclusion in the Context of the Expanded DRI Model

DRIs based on chronic disease allow for the relationship between intake and chronic disease risk to be characterized under a new DRI category, rather than being embedded in the decision-making process for other DRI categories (e.g., AI, UL). ULs previously have been established based on any type of critical adverse effect attributable to excessive intake of the nutrient or other food substance. As per the guidance provided in the Guiding Principles Report, the expanded DRI model now focuses the UL on characterizing toxicological risk attributable to excessive intake and the new DRI category on characterizing the relationship between intake and chronic disease risk.

In the expanded DRI model, there may be scenarios in which chronic disease risk is reduced by increasing intake of a nutrient or other food substance (see Chapter 2, Figure 2-1). Conceptually, a UL could have added value in such a scenario to ensure increases in intakes are not entering a range in which the benefits of reducing chronic disease risk are outweighed by the risk of inducing a toxic response. In the case of sodium, however, the CDRR indicates that risks of cardiovascular disease decrease with reductions in sodium intake (see Chapter 10). Because the sodium CDRR indicates there are benefits from decreasing intakes, the committee views the absence of a sodium UL as less critical. Nonetheless, the absence of a sodium UL does not necessarily mean excessive sodium intakes pose no risks, but rather likely reflects a lack of evidence on adverse toxicological effects.

Conclusion in the Context of the Sodium DRIs Established in the 2005 DRI Report

The committee’s decision not to establish a sodium UL may appear to be a departure from the decisions made in the 2005 DRI Report. However, this apparent change cannot be meaningfully interpreted without considering the recent expansion of the DRI model. In the absence of recommendations or guidance on how to use chronic disease endpoints in the DRI

process, the sodium UL established in the 2005 DRI Report was based on the direct and progressive relationship between sodium intake and blood pressure. Blood pressure was characterized as being a biomarker for “several diseases of substantial public health importance,” including cardiovascular diseases and end-stage renal disease (IOM, 2005, p. 376). With the expansion of the DRI model, evidence that was used to derive the sodium UL in the 2005 DRI Report was considered by this committee in context of deriving a sodium CDRR (see Chapter 10). Thus, the lack of a sodium UL does not reflect a change in the state of the evidence of the risk associated with excessive sodium intake; rather it reflects a change in the model on how risk is characterized in the DRIs.

REFERENCES

IOM (Institute of Medicine). 2005. Dietary Reference Intakes for water, potassium, sodium, chloride, and sulfate. Washington, DC: The National Academies Press.

Johnson, A. G., T. V. Nguyen, and D. Davis. 2001. Blood pressure is linked to salt intake and modulated by the angiotensinogen gene in normotensive and hypertensive elderly subjects. Journal of Hypertension 19(6):1053-1060.

Knuist, M., G. J. Bonsel, H. A. Zondervan, and P. E. Treffers. 1998. Low sodium diet and pregnancy-induced hypertension: A multi-centre randomised controlled trial. British Journal of Obstetrics and Gynaecology 105(4):430-434.

Kwakernaak, A. J., J. A. Krikken, S. H. Binnenmars, F. W. Visser, M. H. Hemmelder, A. J. Woittiez, H. Groen, G. D. Laverman, and G. Navis. 2014. Effects of sodium restriction and hydrochlorothiazide on RAAS blockade efficacy in diabetic nephropathy: A randomised clinical trial. Lancet Diabetes and Endocrinology 2(5):385-395.

MacGregor, G. A., N. D. Markandu, G. A. Sagnella, D. R. Singer, and F. P. Cappuccio. 1989. Double-blind study of three sodium intakes and long-term effects of sodium restriction in essential hypertension. Lancet 2(8674):1244-1247.

Martos Sanchez, I., P. Ros Perez, E. Otheo de Tejada, J. L. Vazquez Martinez, C. Perez-Caballero, and L. Fernandez Pineda. 2000. Fatal hypernatremia due to accidental administration of table salt. Anales Españoles de Pediatría 53(5):495-498.

Meland, E., and A. Aamland. 2009. Salt restriction among hypertensive patients: Modest blood pressure effect and no adverse effects. Scandinavian Journal of Primary Health Care 27(2):97-103.

NASEM (National Academies of Sciences, Engineering, and Medicine). 2017. Guiding principles for developing Dietary Reference Intakes based on chronic disease. Washington, DC: The National Academies Press.

Newberry, S. J., M. Chung, C. A. M. Anderson, C. Chen, Z. Fu, A. Tang, N. Zhao, M. Booth, J. Marks, S. Hollands, A. Motala, J. K. Larkin, R. Shanman, and S. Hempel. 2018. Sodium and potassium intake: Effects on chronic disease outcomes and risks. Rockville, MD: Agency for Healthcare Research and Quality.

Ofran, Y., D. Lavi, D. Opher, T. A. Weiss, and E. Elinav. 2004. Fatal voluntary salt intake resulting in the highest ever documented sodium plasma level in adults (255 mmol L^-1): A disorder linked to female gender and psychiatric disorders. Journal of Internal Medicine 256(6):525-528.

Puska, P., J. M. Iacono, A. Nissinen, H. J. Korhonen, E. Vartianinen, P. Pietinen, R. Dougherty, U. Leino, M. Mutanen, S. Moisio, and J. Huttunen. 1983. Controlled, randomised trial of the effect of dietary fat on blood pressure. Lancet 1(8314-5):1-5.

Raya, A., P. Giner, P. Aranegui, F. Guerrero, and G. Vazquez. 1992. Fatal acute hypernatremia caused by massive intake of salt. Archives of Internal Medicine 152(3):640, 646.

Sacks, F. M., L. P. Svetkey, W. M. Vollmer, L. J. Appel, G. A. Bray, D. Harsha, E. Obarzanek, P. R. Conlin, E. R. Miller, 3rd, D. G. Simons-Morton, N. Karanja, and P. H. Lin. 2001. Effects on blood pressure of reduced dietary sodium and the Dietary Approaches to Stop Hypertension (DASH) diet. DASH-Sodium Collaborative Research Group. New England Journal of Medicine 344(1):3-10.

Schorr, U., A. Distler, and A. M. Sharma. 1996. Effect of sodium chloride- and sodium bicarbonate-rich mineral water on blood pressure and metabolic parameters in elderly normotensive individuals: A randomized double-blind crossover trial. Journal of Hypertension 14(1):131-135.

Sciarrone, S. E., L. J. Beilin, I. L. Rouse, and P. B. Rogers. 1992. A factorial study of salt restriction and a low-fat/high-fibre diet in hypertensive subjects. Journal of Hypertension 10(3):287-298.

Scott, E. P., and C. C. Rotondo. 1947. Salt intoxication; accidental ingestion of a large amount of sodium chloride; report of a case with autopsy of a two-year-old infant. Kentucky Medical Journal 45(4):107-109.

Singer, D. R., N. D. Markandu, A. L. Sugden, M. A. Miller, and G. A. MacGregor. 1991. Sodium restriction in hypertensive patients treated with a converting enzyme inhibitor and a thiazide. Hypertension 17(6 Pt 1):798-803.

Steegers, E. A., H. P. Van Lakwijk, H. W. Jongsma, J. H. Fast, T. De Boo, T. K. Eskes, and P. R. Hein. 1991. (Patho)physiological implications of chronic dietary sodium restriction during pregnancy: A longitudinal prospective randomized study. British Journal of Obstetrics and Gynaecology 98(10):980-987.

Sterns, R. H. 2015. Disorders of plasma sodium—causes, consequences, and correction. New England Journal of Medicine 372(1):55-65.

Todd, A. S., R. J. Macginley, J. B. Schollum, R. J. Johnson, S. M. Williams, W. H. Sutherland, J. I. Mann, and R. J. Walker. 2010. Dietary salt loading impairs arterial vascular reactivity. American Journal of Clinical Nutrition 91(3):557-564.

Todd, A. S., R. J. Macginley, J. B. Schollum, S. M. Williams, W. H. Sutherland, J. I. Mann, and R. J. Walker. 2012. Dietary sodium loading in normotensive healthy volunteers does not increase arterial vascular reactivity or blood pressure. Nephrology (Carlton) 17(3):249-256.

TOHP (Trials of Hypertension Prevention) Collaborative Research Group. 1992. The effects of nonpharmacologic interventions on blood pressure of persons with high normal levels. Results of the Trials of Hypertension Prevention, Phase I. JAMA 267(9):1213-1220.

van Buul, B. J. A., E. A. P. Steegers, G. D. van der Maten, F. M. C. Delemarre, H. W. Jongsma, H. P. Oosterbaan, and P. A. de Jong. 1997. Dietary sodium restriction does not prevent gestational hypertension: A Dutch two-center randomized trial. Hypertension in Pregnancy 16(3):335-346.

Weir, M. R., A. M. Yadao, D. Purkayastha, and A. N. Charney. 2010. Effects of high- and low-sodium diets on ambulatory blood pressure in patients with hypertension receiving aliskiren. Journal of Cardiovascular Pharmacology and Therapeutics 15(4):356-363.

Whelton, P. K., L. J. Appel, M. A. Espeland, W. B. Applegate, W. H. Ettinger, Jr., J. B. Kostis, S. Kumanyika, C. R. Lacy, K. C. Johnson, S. Folmar, and J. A. Cutler. 1998. Sodium reduction and weight loss in the treatment of hypertension in older persons: A randomized controlled Trial of Nonpharmacologic Interventions in the Elderly (TONE). TONE Collaborative Research Group. JAMA 279(11):839-846.

Wing, L. M., L. F. Arnolda, P. J. Harvey, J. Upton, D. Molloy, G. M. Gabb, A. J. Bune, and J. P. Chalmers. 1998. Low-dose diuretic and/or dietary sodium restriction when blood pressure is resistant to ACE inhibitor. Blood Pressure 7(5-6):299-307.