[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
Purchase Options:
[Skip to Content Landing]
Views 8,085
Citations 0
March 7, 2018

Measurements of 24-Hour Urinary Sodium and Potassium ExcretionImportance and Implications

Author Affiliations
  • 1Division of Nephrology-Hypertension, Department of Medicine, University of California, San Diego
  • 2Division of Preventive Medicine, Department of Family Medicine and Public Health, University of California, San Diego
JAMA. Published online March 7, 2018. doi:10.1001/jama.2018.1153

Almost 50 years ago, sodium reduction was identified as a national priority at the 1969 White House Conference on Food Nutrition and Health,1 yet adherence to recommended sodium levels remains an elusive public health goal. Similarly, underconsumption of potassium has been a public health concern for decades. Excessive sodium intake is etiologically related to hypertension and cardiovascular diseases, and these conditions contribute significantly to morbidity, mortality, and health care costs.2,3 Furthermore, estimates of the health and financial benefits of sodium reduction are substantial.4 Physiologically, the actions of sodium and potassium are inter-related and low potassium intake is also etiologically related to high blood pressure.

The current adequate sodium intake for adults is 1500 mg/d and was established based on data that this level of intake should ensure that the overall diet provides adequate amounts of important nutrients and cover sweat losses.5 As recommended by the National Academy of Medicine (formerly the Institute of Medicine), the current tolerable upper intake level for sodium is 2300 mg/d.5 For potassium, the adequate intake for adults is 4700 mg/d, and there is currently no tolerable upper intake level.5

The study by Cogswell et al in this issue of JAMA evaluated 24-hour urine sodium and potassium concentrations in a sample of 837 US adults in the 2014 National Health and Nutrition Examination Survey (NHANES) who completed a 24-hour urine collection.6 The authors report the mean 24-hour urinary sodium excretion was 3608 mg with important differences by sex, body size, hypertension, and diabetes status. The mean for 24-hour urinary potassium excretion was 2155 mg, which differed by sex.

This simple but elegant cross-sectional study is significant for several reasons. First, this is the first NHANES examination to conduct 24-hour urine collections, which represents a unique and important resource for monitoring nutritional status. Second, NHANES now has a broad array of methodologies to estimate sodium intake allowing comparisons of methods and their relationships with health outcomes. Third, these types of surveillance efforts help identify public health problems and justify priorities for public health resources.

Although the relationship of excess sodium intake with hypertension and cardiovascular disease is reasonably well established, whether there is a lower limit of sodium intake below which risk of adverse outcomes are also increased is uncertain. This is an important issue for setting public policy around adequate intakes of sodium, and the answers have differed by the methods used to measure sodium intake in prior studies. When sodium intake was estimated by 24-hour urine collection, prior studies generally have shown linear relationships of higher sodium excretion with cardiovascular events without clear evidence of a lower limit at which risk may also be increased.7-9 Similarly, studies that used food frequency questionnaires, including prior studies from NHANES, generally have shown linear relationships.10 In contrast, several studies have used spot urine specimens and the urine sodium to creatinine ratio to estimate sodium intake. These studies consistently demonstrated J-shaped or U-shaped relationships with cardiovascular disease and mortality, suggesting that risk of adverse events may be increased if dietary sodium intake is insufficient.11-13

The reasons for these disparate findings are uncertain. The study by Cogswell et al, when considered in context with 2 prior publications, provides information on all 3 sodium measurement techniques in NHANES.6,10,14 These data will allow comparisons of associations with cardiovascular disease and other health outcomes across the 3 measurement techniques, all in a sample representative of the noninstitutionalized US population.

The challenges of obtaining 24-hour urine specimens in large-scale studies are well recognized, and the opportunity to understand associations with much larger sample sizes using spot urine specimens is attractive. Several prior studies have evaluated the precision and bias of spot urine sodium–based estimates compared with 24-hour urine sodium excretion in a variety of settings. These studies have reported variable degrees of bias, which have been as disparate as 800 mg or more, and smaller in others.15,16 However, more important than the differences across measurement techniques at the population mean is whether this bias is stable across the range of sodium intakes. Several studies have demonstrated systematic bias comparing spot and 24-hour urine sodium at extremes of sodium intake.15,16 If spot sodium measurements provide a less-accurate surrogate at lower sodium intake levels, then this may help explain why associations with outcomes are J-shaped with spot urine sodium but are often linear using 24-hour urine specimens.

In the study by Cogswell et al, the ability to address these questions in a nationally representative sample and to evaluate relationships within key subgroups is a major step forward. However, the study should be interpreted in light of several caveats. The study sample is relatively small, particularly in subgroups among populations of interest, such as those with hypertension and in racial/ethnic minority groups. An important strength of the study is the availability of repeated 24-hour urine specimens in a subset of participants; however, these data also show high within-individual and between-individual variability. The main analysis focuses on one 24-hour urine specimen, and the high within-individual variability means that the distribution of 24-hour urine sodium and potassium based on one 24-hour urine specimen will increase the number of individuals at the tails of the distribution. Thus, although mean sodium and potassium excretion levels are likely accurate, the data may be misleading if evaluating the percentage of individuals below target levels because these sodium and potassium levels would fall near the lower tail of the distribution, if relying on a single 24-hour urine estimate.

What actions should be taken given the results of the study by Cogswell et al? The findings suggest population sodium intake levels that are in excess of current recommendations and potassium intake levels that may be below current recommendations. Sodium is currently ubiquitous in the US food supply, and the primary sources (approximately 70%) are commercially processed foods.17 Although individuals and clinicians often consider salt added in food preparation or at the table, elimination of these sources may not meaningfully alter sodium intake in the population when sodium remains high in commercially processed foods. Because of this, strategies to reduce sodium intake should focus at the population level first and should include the industries that supply processed foods, beverages, and menu items. Monitoring the success of these types of public health strategies require a benchmark in a population-based sample, such as NHANES, to which future NHANES 24-hour urine collections can be compared. As such, it is important to continue to fund these critical surveillance sources.

What appears at first glance to be a simple cross-sectional study is in actuality a major advance in the ability to estimate dietary sodium and potassium intake and set public policy. The important work by Cogswell et al informs current sodium and potassium intake in the US population, sets a benchmark for comparison of changes in sodium and potassium intake in future years, and provides a critical platform to interrogate the implications of different sodium measurement techniques in terms of bias, precision, and relationships with health-related outcomes.

Back to top
Article Information

Corresponding Author: Joachim H. Ix, MD, MAS, Division of Nephrology-Hypertension, University of California, San Diego, 3350 La Jolla Village Dr, MC 9111H, San Diego, CA 92161 (joeix@ucsd.edu).

Published Online: March 7, 2018. doi:10.1001/jama.2018.1153

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Ix reported serving on the National Academy of Medicine panel “Review of the Dietary Reference Intakes for Sodium and Potassium” and receiving grant support from the National Institute of Diabetes and Digestive and Kidney Diseases; the National Heart, Lung, and Blood Institute; and the American Heart Association. Dr Anderson reported serving on the National Academy of Medicine panel “Review of the Dietary Reference Intakes for Sodium and Potassium” and receiving grant support from the Agency for Healthcare Research and Quality Evidence-based Practice Center, National Institutes of Health, the National Institute of Diabetes and Digestive and Kidney Diseases; the National Heart, Lung, and Blood Institute; and the American Heart Association.

Mayer  J.  White House conference on food, nutrition, and health.  J Am Diet Assoc. 1969;55(6):553-556.PubMedGoogle Scholar
Heidenreich  PA, Trogdon  JG, Khavjou  OA,  et al.  Forecasting the future of cardiovascular disease in the United States.  Circulation. 2011;123(8):933-944.PubMedGoogle ScholarCrossref
Institute of Medicine.  Sodium intake in populations: assessment of evidence. https://www.nap.edu/read/18311/chapter/1. Accessed February 22, 2018.
Bibbins-Domingo  K, Chertow  GM, Coxson  PG,  et al.  Projected effect of dietary salt reductions on future cardiovascular disease.  N Engl J Med. 2010;362(7):590-599.PubMedGoogle ScholarCrossref
Institute of Medicine.  Dietary reference intakes for water, potassium, sodium, chloride, and sulfate. https://www.nap.edu/read/10925/chapter/1. Accessed February 22, 2018.
Cogswell  ME, Loria  CM, Terry  AL,  et al.  Estimated 24-hour urinary sodium and potassium excretion in US adults [published online March 7, 2018].  JAMA. doi:10.1001/jama.2018.1156Google Scholar
Mills  KT, Chen  J, Yang  W,  et al.  Sodium excretion and the risk of cardiovascular disease in patients with chronic kidney disease.  JAMA. 2016;315(20):2200-2210.PubMedGoogle ScholarCrossref
Joosten  MM, Gansevoort  RT, Mukamal  KJ,  et al.  Sodium excretion and risk of developing coronary heart disease.  Circulation. 2014;129(10):1121-1128.PubMedGoogle ScholarCrossref
Cook  NR, Appel  LJ, Whelton  PK.  Lower levels of sodium intake and reduced cardiovascular risk.  Circulation. 2014;129(9):981-989.PubMedGoogle ScholarCrossref
Yang  Q, Liu  T, Kuklina  EV,  et al.  Sodium and potassium intake and mortality among US adults.  Arch Intern Med. 2011;171(13):1183-1191.PubMedGoogle ScholarCrossref
Mente  A, O’Donnell  M, Rangarajan  S,  et al.  Associations of urinary sodium excretion with cardiovascular events in individuals with and without hypertension.  Lancet. 2016;388(10043):465-475.PubMedGoogle ScholarCrossref
O’Donnell  M, Mente  A, Rangarajan  S,  et al.  Urinary sodium and potassium excretion, mortality, and cardiovascular events.  N Engl J Med. 2014;371(7):612-623.PubMedGoogle ScholarCrossref
O’Donnell  MJ, Yusuf  S, Mente  A,  et al.  Urinary sodium and potassium excretion and risk of cardiovascular events.  JAMA. 2011;306(20):2229-2238.PubMedGoogle Scholar
Pfeiffer  CM, Hughes  JP, Cogswell  ME,  et al.  Urine sodium excretion increased slightly among US adults between 1988 and 2010.  J Nutr. 2014;144(5):698-705.PubMedGoogle ScholarCrossref
Mente  A, O’Donnell  MJ, Dagenais  G,  et al.  Validation and comparison of three formulae to estimate sodium and potassium excretion from a single morning fasting urine compared to 24-h measures in 11 countries.  J Hypertens. 2014;32(5):1005-1014.PubMedGoogle ScholarCrossref
Dougher  CE, Rifkin  DE, Anderson  CA,  et al.  Spot urine sodium measurements do not accurately estimate dietary sodium intake in chronic kidney disease.  Am J Clin Nutr. 2016;104(2):298-305.PubMedGoogle ScholarCrossref
Harnack  LJ, Cogswell  ME, Shikany  JM,  et al.  Sources of sodium in US adults from 3 geographic regions.  Circulation. 2017;135(19):1775-1783.PubMedGoogle ScholarCrossref