There’s a lady in Resus, she’s 46, she’s got a history of mental health problems.  Her husband tells you she’s been gradually more lethargic over the last few days.  He called the ambulance today because he found her on the sofa in the morning mumbling incoherently.  Her observations are okay.  Her ABC’s are okay, she’s got a normal glucose, but when you go to move her arm to cannulate you notice she’s rigid.  Hypertonic all over.  You do what you can of a neuro exam and find she’s got globally increased reflexes.  Her pupils are fine.  VBG is okay acid base wise but her Na is 154.

Now as we said last week.  Sodium is important, and it’s ubiquitous.  We need it for everything, and every single one of our cells uses a lot of energy to maintain a sodium concentration of between 135-145mmol/l.  When things go wrong with sodium homeostasis, things are very wrong indeed.  Mortality rates are higher for hypernatraemia than hyponatraemia ranging from 45-60% for all patient groups, and can be as high as 80% in the elderly.  Thankfully it is less common than hyponatraemia, and we most often see it in patients as they enter their last phase of life, this adds an ethical dimension to treatment that I’m not going to talk about here (maybe another day).

Features of severe hypernatraemia are hyperthermia, delirium, seizures, and coma.  Patients with milder symptoms can sometimes present with delirium or changes in mental status.  Patients might have features of underlying disease processes (such as Conns or Cushings).

Hypernatreamia is usually caused by combined electrolyte and water loss, it’s just that the water loss is in excess of the electrolyte loss, and is coupled with an inability to replace water via the thirst response (people with low GCS, dementia, mental health problems).  The trick is working out where the water is being lost from.

Net water loss from kidneys

Diabetes insipidus can be neurogenic, or nephrogenic.  Neurogenic is usually due to traumatic brain injury, space occupying lesions or infections.  Nephrogenic can be caused by general renal dysfunction, or electrolyte abnormalities such as hypercalcaemia or hypokalaemia, or HHS.

Net water loss from other sources:

The commonest drug cause of hypernatraemia is Lithium, though other drugs such as amphotericin, diuretics and vasopressin analgoues (demeclocycline) can also contribute to or cause it.  Lithium actually inhibits a protein called GSK3 which is part of how renal cells’ respond to vasopressin.  Colchicine, gentamicin, and rifampicin can also cause diabetes insipidus.  Unreplaced loss from the respiratory system, sweating, burns, GI tract (D+V) or any type of fistulae can also be implicated.

Hypernatraemia from sodium gain

Feeding, or increased oral salt intake (this usually needs to be massive, or Iatrogenic).  Sea water ingestion, hyptertonic enemas, or dialysis.  Primary hyperaldosteronism (Conns), or Cushings syndrome can also cause excess sodium re-absorption.



As with hyponatraemia the symptoms are vague and wide ranging.  Treatment depends on the speed of onset with those with a rapid onset (<48 hours) likely to have more severe symptoms.

Assessment of volume status again here is key, because disorders of sodium metabolism are also disorders of WATER.

Hypovolaemic hypernatraemia – patients have signs of hypovolaemia, plus a high sodium!  If you check a urinary sodium and it is low it suggests that the the loss of Na is coming from somewhere other than the kidneys (normally GI tract).  The Na in these cases is usually elevated at 150-170mmol/L.  I think this is the most common class of hypernatraemia.

Euvolaemic hypernatraemia – can be caused by either renal or extra-renal loss of water without loss of Na.  These patients usually have an inability to respond to thirst, or one of the diabetes insipiduses? Inspidies?  Urine osmolality will be lower than plasma osmolality in patients with renal losses of water.  Serum Na in these cases is usually higher >170mmol/L.

Hypernatreamia with hypervolaemia – least common, these patients have normally been given more Na+ than they need (hypertonic solutions either NG, IV) OR they may have conditions which compound this such as renal or liver dysfunction.  People in this category have sky high sodiums >190mmol/L.

Acute Ix strategy – send urine and plasma for electrolytes and osmolality.

Urine osmolality

Lower than plasma or <300

[very dilute]

Normal 400-800 High >800

[super concentrated]

Central or nephrogenic diabetes insipidus


Incomplete Central or nephrogenic diabetes insipidus



Lots of water loss (and your patient has just run out of free H20)


Total Na+ gain


Extra-renal losses, D+V, burns etc

The mainstay of emergency treatment is infusion of the right amount of normal saline to bring the sodium back down.  You want to do this slowly in chronic hypernatreamia (drop the Na by no more than 10mmol/day).  I think this is why Normal Saline is suggested in the emergency phase rather than 5% Dextrose.

Most sources suggest we calculate the water deficit, and replace the lost fluid (after initial resuscitation fluids) over 24-48 hours with oral or 5% dextrose.  The formulae are similar but I encountered 3 different ones in the 3 sources I used (BMJ best evidence, Life in the Fast Lane, Mushin article).  Most of them changed either B or the way you calculate the Na excess.  The one below is from BMJ best evidence I found it the easiest to actually use (for me).

Deficit = A x B x ([Serum Na/140]-1)

A = is their weight

B = % water (0.6 for men, 0.4 for women)


Patients with hypervolaemic hypernatraemia might require that dirtiest of treatments;  fluid AND diuretics, as by expanding their intravascular volume with IV fluid you will downregulate vasopressin excretion further, compounding the problem.  This group of patients might benefit from dialysis to remove the volume and improve the sodium (carefully).

Once we’ve corrected this (and unless your bed-state is really really bad) most patients will be out of your department but further diagnostic tests may be done to confirm the cause.

Tests that may be required:

Plasma Aldosterone:Renin ratio: a high PCA:PRA ratio supports the diagnosis of Conns.  The ratio should be >1000 and a random aldosterone level should be >250pmol/L. (requires patient to be K+ replete, and have all diuretics and antihypertensives stopped for a few weeks beforehand  ).  If you get an equivocal result you might need to get a saline infusion test for hyperaldosteronism (this is an outpatient thing).

Dexamethasone suppression test (Cushings).  Patient takes 1mg dexamethasone at 23:00 and at 09:00 has blood taken for plasma cortisol.  A positive result is <50nmol/L.

Water deprivation test (for euvolaemic hypernatraemias)

When everything is back to normal.  Water restriction starts in the early morning, with baseline vasopressin level, with hourly Na checks.  Once the sodium is >148mmol/L another vaspopressin level should be taken.  At this point DDAVP (vasopressin agonist) should be given.  Patients with nephrogenic DI fail to respond to DDAVP, and their urine osmolality increases by <50% or <150mOsm/kg from baseline pointing to another cause.

CT or MRI head for cranial DI

CT or MRI adrenals for primary aldosteronism.


So just like our hyponatraemia patients, hypernatraemia patients need a serum and urine electrolytes and osmolality.  We need to decide on their volume status.  We need to resuscitate if required with normal saline, or replace slowly with 5% dextrose.  It is possible to calculate their fluid requirements and we should do this too.  In some circumstances we might even need to infuse 5% dextrose while giving diuretics.

The key question to answer is “Where has all the water gone?”


Hypovolaemic Euvolemic Hypervolaemic


Renal Losses

Renal failure


Post-obstructive diuresis


Non Renal Losses (urinary sodium low)




Failure to drink (psychological, behavioral, inability)


Nephrogenic DI

Drugs, AKI, Electrolytes


Neurogenic DI



Failure to drink (psychological, behavioral, inability)


Iatrogenic infusion/feeding of high Na fluid


Co-existing renal or liver dysfunction



Primary Aldosteronism

Resuscitate with Normal Saline


Calculate deficit and replace losses with 5% dextrose over 48 hours.


Aim for <10mmol/day increase in Na

Calculate deficit and replace losses with 5% dextrose over 48 hours.


Aim for <10mmol/day increase in Na

Calculate deficit, replace any losses carefully to avoid worsening overload using 5% dextrose.


Use IV Furosemide


Aim for <10mmol/day increase in Na

PS if you are wondering about that lady, she ended up having neuroleptic malignant syndrome plus a partial neurogenic diabetes insipidus from mass effect from maxillary bone osteomyelitis.  You know, one of those simple diagnoses…



BMJ best practice Hypernatraemia ( Accessed 24/11/2016

Mushin, A, Mount D Diagnosis and Treatment of Hypernatraemia.  Best Practive and Research Clinical Endocrinology and Metabolism 30;2 March 2016 189-203.  [paywall]


An old man is wheeled into resus.  His GCS hovers around 10.  No carers have come with him, you find out from the paramedics that the care home staff were alerted to a thump as he fell out of bed.  He’s moving all 4 limbs.  His obs are okay.  His CT head is normal.  His VBG shows a Na of 119, but is otherwise okay.

This, for me, is a frighteningly common scenario.  It’s also generally managed badly and in variable ways.  Do you give normal saline?  How fast?  What are the risks?  Aren’t you meant to fluid restrict?  What about urine tests, don’t’ we need to send that off for something?

The Oxford Handbook of Clinical Medicine (which I credit with allowing me to pass both my undergraduate degree and membership exam) is annoyingly vague in the single page it devotes to this issue, and concentrates more on fluid management than low or high Na.  Now the management of low sodium is important because patients with hyponatraemia are 7x more likely to die in hospital than those with a normal sodium.

  • A normal serum sodium is between 134 – 146 mmol/l.
  • Osmolarity is the number of osmoles per litre of solution
  • Osmolality is the number of osmoles per kilogram of solvent
  • (2 x (NA +K) + Glucose + Urea )  Normal is 280 – 295 mOsm/kg

Sodium is a highly reactive alkali metal that readily donates it’s free outer electron to any and all comers.  It doesn’t exist in it’s pure form naturally anywhere on the planet.  Many of it’s salts are water soluble and it’s that property and it is it’s positive charge that makes it essential to pretty much every cellular process in biology.

The Na/K ATPase is ubiquitous to all animal cells.  It exchanges 3 Na+ ions from inside the cell with 2 K+  against their respective concentration gradients.  Maintaining this gradient uses up at least 20% of most cells energy (nerve cells use more).  What’s particular fascinating about this protein is it seems to have roles in signaling as well as maintaining homeostatic functions and resting potentials., but before I vanish down a Wikipedia fueled rabbit hole talking about ouabain-binding signaling I’ll get back to the point of this post.

The point is, is that the maintenance of Na/K concentration is so vital to cellular processes at a basic level that every cell is spending at least one fifth of it’s energy maintaining it.  Now every single one of us knows how to put potassium back into cells, but Hypo and Hypernatraemia feel complex, and aren’t intuitive.  Na+ itself is responsible for 86% of plasma osmolality, which is why small fluctuations in the amount available have relatively massive (and vague) clinical manifestations.

Sodium and water homeostasis are linked.  Principally because as Sodium is the main osmole water will follow it.  If you excrete more sodium, you excrete more water.

Key point“ Water follows sodium around like an overly-keen medical student with boundary issues “

Sodium is tightly controlled by brain and kidney via ADH release and perception of thirst (water intake).  Aldosterone acts to re-absorb sodium from the collecting ducts (also sweat and salivary glands and the colon), water follows this via osmosis.  Urine osmolality reflects how much sodium is being secreted or absorbed, and varies between 50 (water) to 1400 (more salty than sea water).

If ECF  (extra cellular fluid) osmolality increases by 1-2% (more osmoles in solution, the more concentrated it is), the posterior pituitary makes vasopressin.  Vasopressin works on the distal nephron to open aquaporin-2 channels.  This lets more water back into the circulation.  The sensation of thirst is also triggered.  Reductions in ECF osmolality reduce vasopressin secretion.

Hypovolaemia stimulates renin secretion from the juxtaglomerular apparatus.  Renin converts angiotensinogen to angiotensin I, which via ACE is converted to angiontensin II which stimulates the adrenal cortex to produce aldosterone (aldosterone can also be released directly if the Na+ in the ECF drops).  Aldosterone causes sodium to re-absorbed in the nephron, gut, and sweat and salivary glands.  If we re-absorb sodium water is meant to follow.

Let’s pause a minute and consider the effect of us fiddling with the renin-angiontensin system on a patient.

If you give people an ACEi, or an ARB you are effectively stopping people being able to re-absorb sodium, if you stop them doing that they can’t effectively scavenge water back into their system, this is why these drugs are BAD if for whatever reason you have a low BP.  They can compound a mild problem to a severe one, as there is no compensatory mechanism.  Diuretics change the urine osmolality and Na concentration.  They make making the appropriate diagnosis harder.

How should we manage these patients?  This depends on the type of hyponatraemia.

Hyponatraemia is a disorder of Sodium AND WATER.

Classify or die.

Hyponatraemia  = Na+ <135
Acute <48 hours Mild <135
Chronic >48 hours Moderate 125-130
Severe <125


Subtle Mild Severe
Gait abnormalities


Decreased concentration

Cognitive deficits

Increased risk of fracture





Cardiorespiratory distress

Abnormal and deep sleep



Severe symptoms are more likely to occur in acute hyponatraemia, as compensatory mechanisms kick in to protect the brain from cerebral oedema (it makes intracellular osmotic molecules like glutamine which keep water in the cells), which is the cause of the severe symptoms.

Investigation Strategy


You need 4 test results to decide what the cause is (and often you don’t need all four).

  • Serum Osmolality (Estimating the osmolality is a good first step, however you need to measured value for the diagnosis)
  • Urine Osmolality (how concentrated the urine is)
  • Serum U+E (what’s the Na?)
  • Urine Na (how well are the kidney’s scavenging up Na)

You also need to decide if the patient is hypo, hyper, or euvolaemic.  Now this is fraught with difficulty and inconsistency (because clinical examination isn’t accurate).

There are three broad classes based on amount of solute in the serum (hypertonic, isotonic and hypotonic).  Hypotonic is the commonest, and it splits into 3.

Hypertonic hyponatremia occurs when another osmole (normally glucose) is exerting more of an affect than it normally does.  This causes water to leave the cells, causing a relative hyponatraemia that corrects if the offending osmole is put back into line.  Think HHS and DKA.  This is the reason why we correct high glucoses slowly, as rapid shifts cause the cerebral oedema that kills.

Isotonic hyponatraemia (MEASURED Serum Osmolality >275mOsmol/kg) is associated with infusion of dextrose or mannitol, or Iv IG.

Hypotonic hyponatraemia  (MEASURED Serum osmolality <275mOsmol/kg) is Divided into three: hypovolaemic, euvovolaemic, hypervolaemic, and is the commonest class we see in the ED.

Hypovolaemic – Loss of water and Na from ECF causes increased vasopressin secretion, decreased water excretion, causing water retention.

Renal causes – diuretics, osmotic diuresis, renal tubular acidosis, salt wasting nephropathy, mineralocorticoid deficiency, ketones.

Non renal (consider if Urinary Na <10) – vomiting, third spacing (eg pancreatitis + bowel obstruction), diarrhea, bowel prep, sweat, bleeding  If the urinary Na is <10  the kidney is working to scavenge Na back into the ECF.

Euvolaemic (Euvolemic patients will have a normal or low urea and an elevated Urinary Na) SIADH is the most common, also think of Psychogenic polydipsia, hypotonic IVF, adrenal insufficiency, hypothyroidism


Total body water and Na are both decreased but there is more water than sodium.  Vasopressin release causes water retention.

Hypotonic Hyponatraemia  (dilute serum)

Rx – isotonic fluids


Rx – fluid restrict


Rx – fluid restrict


Diuretics (thiazides)

Primary adrenal insufficiency

Cerebral salt wasting

Renal Disease

Third spacing



Bowel Obstruction

Syndrome of inappropriate diuresis

Secondary adrenal insufficiency


High Water intake

Lowe solute intake


Renal, Liver, Heart

Nephrotic Syndrome

Drugs that can cause Hyponatraemia (apart from diuretics)

  • Antipsychotics
  • PPI
  • ACEi
  • ARB
  • Anti-eplieptics
  • SSRI
  • MDMA (ecstasy)

The weirdness:

In loads of cases of hyponatraemia you get inappropriate release of vasopressin.  This acts to further decrease serum sodium.

In hypovolaemic patients, the release of vasopressin is in response to their low BP, in an attempt to scavenge water back into the system  to improve BP, but this acts to lower Na concentration further.  This happens in hypervolaemic patients too (as often their intravascular volume is low too).

In the above cases the inappropriate vasopressin release isn’t the primary cause of the hyponatraemia but it compounds the problem.  However you can get inappropriate vasopressin release as the primary cause of the hyponatremia too.  This should be thought of as a diagnosis of exclusion.

SIADH Criteria:  Serum Osmolality <275, Urine Osmol >100, urinary Na > 30 with normal thyroid, renal and pituitary function (and no use of diuretics).
Carcinoma Lung CNS Medications


GI Tract

GU Tract

















Sodium valproate







Cerebral/Renal salt wasting syndrome

This is relatively rare, and was first described in association with brain tumours and bleeds.  It can occur without them however.  The mechanism is thought to involve the release of a natriuretic factor which causes increased renal excretion of sodium.  As this causes volume depletion (water follows sodium).  ADH is released due to hypovolaemia, and the renin-angiotensin system is activated.  Biochemically these patients have exactly the same kind of findings as SIADH.  However they should be hypovolaemic. 


Guidance for the management of hyponatraemia exist.  They are here and here.  And summarized well here.


The guidelines state that hypertonic saline should be given in the first hour of someone’s presentation if they have features of severe hyponatraemia.   It is entirely possible then to have given two boluses of 3% saline before you’re even thinking about getting your urine results back.

Patients with hypervolaemia or SIAD are best managed with fluid restriction.  If there is no improvement with fluid restriction diuretics or oral NaCl should be started.  Patients with hypovolaemia should be given 0.9% NaCl or another balanced solution at 0.5-1ml/kg/hr (which is really quite slow, a litre in 12 hours or more).  We should not be treating asymptomatic patients aggressively because of the risks of treatment.  If you cannot confirm that this is acute hypernatraemia you should assume it is chronic, and only treat it aggressively if there are severe symptoms.

Interesting points on management I’ve been thinking about:

  • Acute hyponatraemia often doesn’t get managed well in the initial assessment by us in the ED. We often don’t order the serum and urine osmolality until prompted to by medics.  We should do this first line as soon as we realise their sodium is low.
  • We also probably generally do not give hypertonic saline often enough.
  • Catheterisation to assess fluid status, and get that urine to the lab is probably not done quickly enough.
  • I’m not sure if we can run people’s urine through our gas analysers to get a quick urinary sodium off (I think it’d probably work), however I’m not aware of a way to estimate Urine osmolality (if you find one please let me know!).

ODS (Osmotic demyelination syndrome)

This used to be called central pontine myelinolysis, occurs if the Na is corrected too rapidly, causing neuronal cells to dessicate as water follows it’s osmotic gradient out into the ECF.  Symptoms are heterogeneous and similar to symptoms of severe hyponatraemia.  The stand out differences are acute neurological deficits.

Severe symptoms of hyponatraemia Symptoms of ODS

Cardiorespiratory distress

Abnormal and deep sleep




Disturbed consciousness

Gait changes

Respiratory depression







ODS can also be caused by severe liver disease, alcoholism, malnutrition, anorexia, and hyperemesis gravidarum too.

What to do if treatment goes wrong.

If you’ve pushed up someones Na too quickly, the first sign you might get is a massive increase in their urine output.  If Vasopressin activity stops then free water clearance increases resulting in a more rapid increase in sodium than you really want.

If you need to quickly lower Na, then you should stop your IV treatment or diuretics, and give 10ml/kg of 5% dextrose over 1 hour.  It might also be appropriate to give desmopressin IV 2microgrammes, but I’d check with someone first.


The management of hyponatraemia hinges upon the severity and speed of onset of the imbalance.  The classification depends on investigations of urine and serum.  There are 3 types of hyponatraemia hyper, iso, and hypotonic.  There are 3 classes of hypotonic hyponatraemia (hyper, euvo, and hypovolaemic).

Acute or Severe hypontraemia should be treated by at least one infusion of 3% NaCl over 30 minutes.

Next time Hypernatraemia!


European Society of Endocrinology guidelines on the management of Hyponatraemia, European Journal of Endocrinology (March 1 2014). 170-G1G47  Here

Hirst et al The adult patient with hyponatraemia British Journal of Anaesthesia:Education 15(5):248-252(2015).  [pay walled]

Williams et al The clinical management of hyponatraemia Postgraduate Medical Journal 2016;92:407-411 [paywalled]