Statistics from Altmetric.com
Life, as we know it, evolved in water that contained a steady concentration of salts. Consequently, the functioning of living cells is dependent on the provision of an environment with just this right concentration of electrolytes. The emergence of organisms from the sea and onto land was only possible because of the evolution of kidneys, which preserved the environment of the sea within the organisms. Without the kidneys providing volume, electrolyte and acid-base homoeostasis, no heart could beat, no muscle move, no thought be thought.1 Disturbances of this homoeostasis are thus serious conditions that pose grave dangers to the patient.2 3 Treatment is influenced by the correct identification of the underlying problem: hyponatraemia requires a different approach if it is due to an excess of water compared to a deficiency of sodium.4 We can get help in the assessment of these disorders by ‘asking the kidney’, that is, interrogation of kidney function. The kidney ‘speaks’ via the composition of urine. Thus, by determining this composition we can obtain important information about the nature of the underlying problem. However, the determination, as well as the interpretation of these values, has limitations and potential pitfalls.
In this article, we will concentrate on urine sodium and osmolality. We will briefly review the methodology used for measurement, but mostly reflect on the interpretation of results in the clinical setting and the underlying physiology.
How is urine sodium measured?
By far the most common method for urine sodium analysis is electrochemically using an ion-selective electrode (ISE). There are two types of ISE in general laboratory use: the direct and indirect ISE. The indirect ISE method was developed for plasma where the sample is diluted (in order to increase sample volume and reduce protein concentration) before it is presented to the measuring electrode. Direct ISEs were a later development in which the sample is measured without any prior dilution. Indirect ISE methods, when used in serum/plasma, suffer from interferences from high concentrations of lipids and proteins. Since urinary concentrations of lipids and protein are typically low, the measurement of electrolytes using indirect ISE is usually accurate. However, in nephrotic syndrome, urine sodium concentration may be underestimated due to interference from the high protein content. ISEs for sodium are generally accurate and precise (coefficients of variance (CV) <1.5%). However, if there is a very low concentration, it may be necessary to use a more sensitive method. For example, in our laboratory we measure urines with sodium concentration of <5 mmol/l using a flame photometer which can reliably measure sodium concentrations as low as 1 mmol/l. The method is very sensitive but relatively slow and the equipment requires high maintenance.
How is osmolality measured?
Osmolality is defined as the concentration of solutes per kilogram of solvent, that is, plasma or urine, as opposed to osmolarity which is the concentration per litre of solvent. The most common method of determining urine osmolality is by using a freezing point depression osmometer.5 The freezing point of a solvent or solution is the temperature at which an infinite amount of the solid and liquid phases co-exist in equilibrium. When a solute is added to a solution, its freezing point decreases. The change in freezing point is relative to the amount of solute added; for instance, when 1 mol of non-ionic solute is added to 1 kg of water the freezing point decreases by 1.86°C. The change in freezing point is directly proportional to the number of particles in solution. Measurement of the freezing point therefore allows the determination of dissolved solute concentration in water, that is, osmolality. Freezing point depression osmometers are very precise (CV <1%).
Specimen collection (Should I obtain a spot or timed urine specimen?)
In order for urine chemistry results to be meaningful, it is vital that the method of specimen collection, timing and handling is optimised. Spot urine samples are easier to obtain, especially in non-toilet-trained children, and reflect kidney function for a specific period of time (the interval between the previous void and the obtained sample). Since urine sodium and osmolality are typically used to assess kidney function for an acute clinical problem (see below), the spot sample is often preferable. However, for the assessment of sodium intake in a patient on a prescribed low-salt diet, a 24 h urine must be obtained: excretion of electrolytes can vary over the day, depending on intake, such as a sodium load with a meal. A 24 h collection will average out this variability and thus provide a more consistent picture.
Interpretation of results: underlying physiology
What is a normal value for urine sodium or urine osmolality?
There are no normal values for urine electrolyte concentrations! (Principle No 2, see conclusions and box 1 in the online supplementary appendix). In order to preserve homoeostasis, the kidneys mainly need to match output to input; if more water is consumed, more needs to be excreted; if more salt is eaten, more sodium and chloride excreted. Consequently, instead of normal values, there are ‘expected’ values.6 For instance, in a patient with water intoxication, urine osmolality should be low to reflect the water excretion. Conversely, in a dehydrated patient, urine osmolality should be high to reflect water conservation. Thus, urine electrolytes must be interpreted within the clinical context (principle 1; see box 1 in the online supplementary appendix).
Why do we need a concomitant urine creatinine when measuring urine sodium?
Urine concentration can vary by about 20-fold. A urine sodium concentration of, for example, 25 mmol/l can reflect very low sodium excretion in highly concentrated urine and vice versa (see case 1, box 2 in the online supplementary appendix). For proper interpretation of urine sodium (or any other urine electrolyte) an internal reference, typically urine creatinine, is needed so that ratios or fractional excretion can be calculated (see principle 2, box 1 in the online supplementary appendix). The fractional excretion is preferable, as it corrects for the filtered load. For the assessment of tubular integrity, the fractional excretion of sodium (FENa) is a very useful tool. The formula for calculation is FENa (%)=(Naurine/creatinineurine)/(Naplasma/creatinineplasma×100. A FENa of >1% is usually indicative of tubular damage.7 But again, the clinical context is critical: a FENa >1% is absolutely physiological in a child with salt poisoning, but pathological in a child with dehydration.
What does the fractional excretion of sodium assess?
Sodium is the most abundant ion in the extracellular fluid and the key determinant of extracellular volume. Its plasma concentration is tightly regulated around 140 mmol/l. An average 70 kg adult with a glomerular filtration rate of 100 ml/min (or 144 l/day) thus filters 140 mmol/l×144 l≈20 000 mmol of sodium per day. That is the equivalent of ∼1.2 kg of cooking salt! Unless we can manage to eat that much salt per day, it becomes obvious how critically important reabsorption of sodium by the renal tubules is to our survival. Consequently, the amount of sodium excreted in the urine can be used as a measure of tubular function.
What does urine osmolality assess?
The main purpose of urine osmolality determination is to assess the action of antidiuretic hormone (ADH) in the collecting duct and thus renal water excretion. A urine osmolality higher than plasma indicates ADH is acting and water is preserved. Conversely, if it is lower than plasma osmolality, then little or no ADH is acting and water excretion is increased. Again, no normal values exist, only expected: a child with psychogenic polydipsia should have a low urine osmolality and a dehydrated child should have a high one. Because disorders of plasma water excretion are reflected mostly in plasma sodium, the assessment of urine osmolality is especially important in dysnatraemias (see clinical case 2, box 3 in the online supplementary appendix: in this boy, the urine osmolality at presentation was 677 mosm/kg with a concomitant plasma osmolality of 255 mosm/kg, reflecting ADH action).
In children with dysnatraemias, why is it important to measure urine osmolality?
Plasma sodium is measured in mmol/l, thus the concentration can be altered by changes in the numerator (sodium) or denominator (plasma water). When faced with a dysnatraemia, the first question a clinician should ask is: is this an abnormality of sodium or of water? The most important first step is the clinical assessment: is there evidence of volume depletion or excess? Measuring concurrent plasma and urine sodium (plus creatinine) and osmolality is then the next important step and the clinical examination will provide the expected values for the correct interpretation of the results.
Clinical use of urine indices
Let us now consider the utility of urine sodium, creatinine and osmolality for concrete clinical questions. For most of these questions, only very limited clinical evidence is available. However, in the absence of randomised clinical trials, one can resort to physiology to guide the interpretation of urine electrolytes. In fact, the natural laws underpinning physiology provide, theoretically, an even stronger foundation than any statistical associations obtained from clinical trials. The problem, however, is that our understanding of these laws is still evolving. Moreover, the electrolytes must be interpreted within the clinical context (principle 1; see conclusions and box 1 in the online supplementary appendix) and the difficulties in accurately and irrefutably assessing, for instance, the circulating volume in a patient provide the key limitations in the clinical utility of urine electrolytes.
Clinical question no. 1: In children with acutely decreased urine output, can urine sodium and osmolality distinguish between impaired renal perfusion (‘prerenal failure’) and acute kidney injury?
A typical clinical scenario for this question is given in case no. 1 (box 2 in the online supplementary appendix). Quite a few studies have examined the utility of urine electrolytes in this context. Early studies showed that an isolated urine sodium was less reliable than FENa in distinguishing between prerenal and acute kidney injury (AKI, previously referred to as acute tubular necrosis or ATN), providing clinical confirmation for principle no. 2: FENa in prerenal failure was typically <1%, while it was >3% in AKI.8,–,10 Moreover, those with a urine osmolality >500 mosm/kg were more likely to have prerenal failure, whereas it was typically <350 mosm/kg in those with AKI.10 This, of course, is not surprising, as sodium reabsorption and urinary concentration are active tubular processes and thus require intact tubular function. There are also a few studies in children with similar results. Studies in neonates found that a FENa >2.5% was indicative of AKI.11,–,13 Other studies, however, questioned the utility of urine electrolytes. For instance, some studies question the reliability of FENa or failed to identify a correlation between FENa and spontaneous recovery of renal function.14,–,16 However, there is a serious problem with all these studies: there is no gold standard to distinguish between prerenal failure and AKI against which the utility of urinary indices could be assessed. Most studies use the rapid and spontaneous recovery of kidney function as the retrospective criterion to distinguish between the two. But, of course, renal failure can evolve from prerenal failure to AKI, if perfusion is not improved. So, the timing of the urine sample is important.17 Electrolytes will only provide us with a snapshot of renal function at the time, demonstrating the importance of continuous observation of these patients. Another complicating factor is, of course, the use of loop diuretics, such as furosemide, which can increase urinary FENa (see clinical case 1).
Clinical question no. 2: In children with acute nephrotic syndrome, do urinary sodium levels reliably reflect hydration status?
A challenging problem in paediatric nephrology practice is the assessment of intravascular volume status in children with nephrotic syndrome: complications of volume depletion, such as venous or even arterial thrombosis, have been reported, but also those of volume overload, such as pulmonary oedema (see case no. 2, box 3 in the online supplementary appendix). While total body volume is expanded in the nephrotic state, this is mostly through increase in the interstitial volume (oedema). However, the clinically most relevant compartment is the intravascular volume. Urine sodium is commonly used as an indicator of intravascular volume and the handbook Paediatric Nephrology states that “measurement of urine sodium may be helpful (urine sodium <10 mmol/l in severe hypovolaemia)”.18 But what is the evidence for this? There are surprisingly few clinical data and one main problem obviously is that there is no easy and generally accepted method of measuring plasma volume. We often approximate changes in weight with changes in fluid status, but this is not applicable in oedema-forming states. A commonly used experimental approach is the use of radiolabelled albumin, but this is difficult to interpret in a condition where albumin is lost in the urine. One study in children used clinical signs and symptoms of hypovolaemia, such as tachycardia, abdominal pain, oliguria and peripheral vasoconstriction, and found that those with signs and symptoms of hypovolaemia had a lower FENa (mean of 0.3%) than those without (0.9%), but nevertheless FENa was low in both groups.19 A recent study of nephrotic and severely oedematous patients used a FENa of 1% as cut-off to separate those with volume contraction from those with volume excess.20 The former received albumin and furosemide, the latter only furosemide. Outcome in both groups was similar and the authors concluded that FENa was “useful in distinguishing volume contraction from volume excess”. An interesting conclusion, considering that the study design was already based on this presumption. So, in the absence of good clinical evidence, what is the physiological basis for the use of urine sodium to assess volume status in nephrotic syndrome?
There are two opposite explanations for the development of oedema in nephrotic syndrome21:
1) The underfill hypothesis: here, the primary defect is the loss of proteins in the urine, which lowers the oncotic pressure in the vasculature with subsequent extravasation of salt and water in the interstitium. If we believe this hypothesis, then urine sodium should be a good indicator of intravascular volume, as sodium retention would be the consequence of intravascular volume depletion.
2) The overfill hypothesis: here, the primary defect is enhanced sodium reabsorption associated with proteinuria. This results in increased intravascular volume and ‘overspill’ of salt and water in the interstitium. Support for this explanation was recently provided by the finding that the loss of certain proteases in the urine, such as plasmin, activate the epithelial sodium channel in the collecting duct.22 Moreover, a deficiency of other proteases, such as corin, involved in the activation of natriuretic peptides may contribute to sodium retention.23 If we believe this hypothesis, then urine sodium is useless as an indicator of intravascular volume, as the driver for oedema formation is primary sodium retention.
In clinical practice, there are good examples for both explanations: patients with nephrotic syndrome so depleted that they experience thrombosis of major arteries, and patients, like case no. 2, who are clinically well perfused, hypertensive and experience pulmonary oedema. Importantly, in both scenarios, the urine sodium excretion can be very low. Thus, FENa cannot be used as a reliable indicator of the intravascular volume in nephrotic syndrome.
Clinical question no. 3: In children with hyponatraemia, do urine sodium levels distinguish between salt wasting and water retention?
In any case of hyponatraemia, the first decision point for the clinician is whether the hyponatraemia is due to an excess of water or a deficiency in sodium. This is a recurrently debated point among clinicians, especially in patients with a central nervous system lesion: is the hyponatraemia due to cerebral salt wasting (CSW), that is, a deficiency of salt and thus should be treated with salt supplementation? Or is it is due to the syndrome of inappropriate antidiuretic hormone secretion (SIADH), that is, reflects an excess of water and should be treated with fluid restriction and/or loop diuretics? Can urine sodium help distinguish between the two? A common, but mistaken assumption is that in hyponatraemia, the kidneys should avidly retain sodium in order to normalise plasma sodium levels. Consequently, if urine sodium were elevated, this would then reflect kidney dysfunction, such as in CSW. However, this assumption is based on a misunderstanding of physiology: as discussed above, sodium is the key determinant of extracellular volume and the kidneys regulate sodium excretion to maintain the so-called effective intravascular volume: if the kidneys sense decreased perfusion (eg, dehydration, cardiac failure, renal artery stenosis), then sodium reabsorption is increased and if perfusion is increased (hypervolaemia), sodium reabsorption is decreased. In SIADH, the primary problem is inappropriate retention of water, which expands intravascular volume. Consequently, the physiological response of the kidneys is to excrete sodium in order to reduce the effective intravascular volume appropriately. This may appear unexpected, because of the concomitant hyponatraemia, leading to the erroneous diagnosis of CSW. The patient presented in case 3 is a perfect example (see box 4 in the online supplementary appendix). In this case, a mistaken diagnosis of CSW was made based on the ‘high’ urine sodium. Consequently, sodium supplementation was commenced, which initially only resulted in a further increase in urinary sodium. This seemed to confirm the severity of the patient's salt wasting, but in fact reflected only the attempts of the kidneys to maintain volume homoeostasis. However, once the supplementation was increased to extraordinarily high doses (20 mmol/kg/day), the salt intake exceeded the capacity of the kidneys for salt excretion and this resulted in normalisation of plasma sodium, but resulting in a massively increased intravascular volume, reflected in the severe hypertension: the blood pressure in this infant was 145 mm Hg systolic despite treatment with an antihypertensive. This emphasises the need for proper diagnosis before instigation of treatment. And unfortunately, urine electrolytes cannot help distinguish between CSW and SIADH. Indeed, there are several studies demonstrating elevated urine sodium/fractional excretion in both diagnoses.24,–,26 However, all of these studies have one major inherent flaw: there is no gold standard test to distinguish between CSW and SIADH! The biochemical parameters are exactly the same and the distinction is purely on clinical grounds: with evidence of volume depletion it is CSW, whereas with normovolaemia it is SIADH.27 And, as discussed above, the assessment of intravascular volume is difficult. Recent publications suggest that CSW is rare and perhaps non-existent.28 29 This is in keeping with our own personal experience: while we have seen many cases of SIADH, we have yet to encounter a case of CSW.
Clinical question no. 4: In children with hypernatraemia, can urine sodium and osmolality guide diagnosis and treatment?
Again, with an abnormality in plasma sodium, the first question is: is it an abnormality of sodium or of water? An excess of salt is extremely rare and can essentially only occur in patients without free access to water (eg, salt poisoning of an infant, or iatrogenic infusion of inappropriate intravenous fluids) as otherwise the consequent thirst from the salt load would drive fluid intake and thus prevent hypernatraemia.30 In the majority of cases, hypernatraemia will be due to dehydration, that is, water loss, as in case no. 4 (box 5 in the online supplementary appendix).
There are, to our knowledge, no clinical trials investigating the utility of urine indices to guide rehydration fluids. The clinical question is raised here to highlight the importance of considering renal water losses (due to a defect in urine concentration (eg, nephrogenic diabetes insipidus or NDI) as a cause for the hypernatraemic dehydration, as these children need to be treated differently (see below). Thus, in our opinion, in any child with hypernatraemia, consideration should be given to a potential renal concentrating defect. A specific gravity from dipstick may suffice with an obvious history of extra renal water losses or decreased intake, but in those patients with a history of persistent polyuria despite the dehydration, urine indices should be obtained.
While NDI is a rare disease, recognition is important to prevent potentially severe complications, such as mental impairment.31
Topics for further research
The clinical evidence base for the utility of urine sodium and osmolality, for instance in the setting of AKI, is limited because of the lack of a gold standard for the diagnosis of tubular necrosis versus prerenal failure. Recent research into biomarkers of tubular damage, such as neutrophil gelatinase-associated lipoprotein, may provide a frame of reference for the utility of urine electrolytes in AKI.32,–,34
Another exciting area of research is the recent discovery of extra-renal sodium handling with osmotically inactive storage and release of sodium in the interstitium.35 36 These data may require revision of our current understanding of the physiology and thus the interpretation of urinary sodium.
Urine sodium and osmolality are useful tools in the assessment of sodium and volume homoeostasis. Proper interpretation, however, must obey some general principles:
1) For urine sodium, creatinine is needed as an internal reference value, so that fractional excretion can be calculated.
2) There are no normal values for urine sodium and osmolality. Results must be seen in the clinical context, which defines the expected values. Deviations from these expected values can then be used to identify the problem.
3) Proper understanding of the physiological principles of renal salt handling is critical, especially its connection with volume homoeostasis.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Web Only Data - This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Competing interests None.
Provenance and peer review Commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.