Potassium handling
Almost all potassium hides inside cells and the kidney fine-tunes the rest in one short segment under aldosterone — so aldosterone, acid-base and the ECG are three views of the same number.
What it normally does
About 98% of the body's potassium sits inside cells. That steep gradient across the membrane is what holds nerve and muscle at their negative resting voltage (resting membrane potential).
A small change in blood potassium changes how easily heart muscle fires — which is where the ECG changes of hyperkalaemia come from.
The kidney filters potassium freely and takes nearly all of it back in the proximal tubule and loop of Henle. The amount that actually leaves the body is decided further downstream, in the late distal tubule, connecting tubule and cortical collecting duct, where aldosterone makes the principal cell pull sodium in through a channel (ENaC), leaving the lumen electrically negative, and that negative lumen drags potassium out through its own channel (ROMK). High tubular flow adds a second, flow-activated potassium channel (BK).
Potassium excretion tracks aldosterone and tracks how much sodium and fluid reach that segment — so anything touching either one moves potassium.
Insulin and adrenaline (through the beta-2 receptor) push potassium into cells by driving the sodium-potassium pump (Na+/K+-ATPase) in muscle. Neither removes any potassium from the body.
It explains the emergency treatment for hyperkalaemia, and explains why a measured potassium can lie badly about total body stores.
Hydrogen and potassium trade places across the cell membrane. When the blood turns acidic with a mineral acid (as in a hyperchloraemic acidosis), hydrogen moves into cells and potassium comes out; alkalosis does the reverse. Organic acidoses such as lactic acidosis and ketoacidosis shift potassium far less, because the anion crosses the membrane with the acid — in ketoacidosis the hyperkalaemia owes more to insulin deficiency and hyperosmolality than to the pH itself. Aldosterone drives hydrogen secretion by the intercalated cells of the same segment in which the principal cells secrete potassium.
A hyperchloraemic acidosis inflates the measured potassium and alkalosis deflates it, and aldosterone disorders characteristically carry a matching acid-base picture — although it may be mild or absent when the disorder is mild.
What goes wrong
- Hyperkalaemia and its ECG← from “About 98% of the body's potassium sits inside …”
Potassium cannot get out. Usually the kidney has failed (acute kidney injury or chronic kidney disease), so the filtered load and the sodium and fluid delivery that drive distal secretion both fall — but the remaining nephrons up-regulate secretion, which is why chronic kidney disease alone rarely causes hyperkalaemia until eGFR is very low or urine output drops; at a moderately reduced eGFR, look for the added drug, the potassium load or the acidosis. Extracellular potassium rises, the gradient across the membrane shrinks and the resting voltage drifts up towards threshold. Sodium channels recover from inactivation only at negative voltages, so they sit inactivated and conduction slows and widens. At the same time the higher potassium conductance makes repolarisation abrupt.
Potassium above about 6.5 mmol/L, or any ECG change at all, is an emergency; the usual causes are kidney failure and the ACE inhibitor plus spironolactone plus NSAID combination, with potassium supplements, trimethoprim and tissue breakdown (rhabdomyolysis, tumour lysis, haemolysis) close behind. Before treating a well patient with a flawless ECG, ask whether the sample was haemolysed, clenched-fist drawn or delayed to the lab (pseudohyperkalaemia). End-stage kidney disease is several times more common in Aboriginal and Torres Strait Islander Australians, and much higher again in remote communities, so this is a presentation you will meet.
You would find: The ECG changes broadly in this order: tall narrow peaked T waves, then a long PR with flattened or absent P waves, then a broad QRS, then the QRS merging with the T into a sine wave, then arrest. The order is a teaching sequence, not a promise — a patient can arrive at a broad complex rhythm with no warning peaked T waves at all, so a normal-looking ECG does not exclude hyperkalaemia. Weakness and paraesthesiae are late and unreliable.
- Too little aldosterone effect (Addison disease, type 4 renal tubular acidosis)← from “The kidney filters potassium freely and takes …”
Without aldosterone the principal cell has fewer active sodium channels, so the lumen never goes properly negative and neither potassium nor hydrogen is efficiently secreted. In Addison disease the adrenal cortex is destroyed, autoimmune in most Australian cases. In diabetic kidney disease the same picture comes from low renin and therefore low aldosterone (hyporeninaemic hypoaldosteronism, type 4 RTA); drugs that block the axis reproduce it closely.
Hyperkalaemia plus hyponatraemia plus postural hypotension is Addison until proven otherwise — a short synacthen test settles it, and in a suspected crisis treatment comes before confirmation. Aldosterone deficiency brings acidosis with it because hydrogen and potassium excretion depend on the same electrical gradient in the same segment.
You would find: Addison: high potassium with a low sodium, postural drop in blood pressure, fatigue, weight loss and pigmented skin creases and buccal mucosa. Type 4 RTA: a stubborn mild hyperkalaemia with a normal anion gap acidosis in a diabetic whose creatinine is only modestly raised.
- Too much aldosterone effect — hypokalaemia with alkalosis← from “The kidney filters potassium freely and takes …”
A strongly negative lumen drives both potassium and hydrogen into the urine. Primary hyperaldosteronism (Conn syndrome) does it directly. Loop and thiazide diuretics do it indirectly by dumping extra sodium and flow onto the collecting duct, with volume depletion raising aldosterone on top. Vomiting does it twice over: volume loss raises aldosterone, and the lost gastric acid starts the alkalosis that shifts potassium into cells.
Primary hyperaldosteronism is a common and treatable cause of resistant hypertension, not a rarity — and most patients with it are normokalaemic, so a normal potassium does not exclude it. Interfering drugs (spironolactone above all, and beta-blockers and ACE inhibitors or ARBs to a lesser degree) distort the aldosterone-to-renin ratio, so check how to prepare the patient before ordering it. Magnesium matters too — intracellular magnesium normally plugs ROMK, so a low magnesium unplugs it, wastes potassium and makes the hypokalaemia refractory to potassium alone.
You would find: Muscle weakness, cramps, constipation and polyuria. ECG shows flattened T waves, prominent U waves and a long apparent QT. Hypertension with an unexplained low potassium means screening for Conn with an aldosterone-to-renin ratio.
- The diabetic ketoacidosis potassium trap← from “Hydrogen and potassium trade places across the…”
Insulin deficiency and hyperosmolality drive potassium out of cells, and the osmotic diuresis then flushes it out in the urine for hours or days. Total body potassium is heavily depleted while the plasma level, propped up by the shift, reads normal or high. Give insulin and fluid and the shift reverses within minutes to hours.
Potassium is checked before the first dose of insulin. Below about 3.5 mmol/L, guidelines have potassium replaced before insulin is started — insulin into a hypokalaemic patient risks arrhythmia. The actual replacement follows the local DKA protocol.
You would find: A patient in DKA with a reassuring potassium of 4.0 mmol/L is severely depleted. After insulin starts, the potassium falls fast and needs hourly attention.
What we give, and how it works
Open a drug to see what it binds, what that does to the cell, and what you then see in the patient.
- Binds
- Extracellular calcium acting at the cardiac myocyte membrane, where it alters the voltage dependence of the fast voltage-gated sodium channel (a surface-charge effect, not receptor binding).
- Which does
- Raising extracellular calcium shifts the effective threshold to a less negative voltage, reopening the gap between the depolarised resting potential of hyperkalaemia and threshold, so normal excitability and conduction are restored without touching the potassium itself.
- So you see
- The QRS narrows and the rhythm stabilises within a few minutes. Serum potassium is unchanged.
- And the same mechanism causes
- Extravasation is the mechanism-derived harm: the salts are hypertonic and caustic, so a tissued cannula causes local skin and soft tissue necrosis — markedly worse with calcium chloride, which is why gluconate is preferred when only peripheral access exists. The classic teaching that calcium precipitates a 'stone heart' in digoxin toxicity is a theoretical extrapolation from animal work and has not held up in the human case series that have looked for it; treat it as a reason for care and senior/toxicology input, not an absolute contraindication, and remember that the definitive treatment of digoxin toxicity with hyperkalaemia is digoxin-specific antibody fragments (Fab), not calcium.
- Handling
- Buys roughly 30 to 60 minutes. Repeat dosing may be needed if the ECG deteriorates again — per local protocol.
Catches people out: It does not treat the hyperkalaemia itself. If calcium is given, something that actually moves or removes potassium has to follow. The two salts are not interchangeable millilitre for millilitre — chloride carries roughly three times the elemental calcium of gluconate for the same volume, so a 'give 10 mL of calcium' handover is not a safe instruction. Check your own hospital's protocol before drawing anything up.
- Binds
- The insulin receptor (a receptor tyrosine kinase) and the beta-2 adrenoceptor (Gs-coupled, raising cyclic AMP) on the muscle cell.
- Which does
- Both routes converge on the same pump — they increase the activity and membrane trafficking of the sodium-potassium pump (Na+/K+-ATPase), which drives potassium into the cell.
- So you see
- Insulin drops the potassium by roughly 0.5 to 1 mmol/L, starting within about 15 minutes and lasting several hours. Nebulised salbutamol gives a similar-sized fall but more slowly, over roughly 30 to 90 minutes.
- And the same mechanism causes
- Insulin signalling also recruits GLUT4 transporters to the muscle membrane and drives glucose into the cell, so hypoglycaemia is the predictable consequence — and giving glucose alongside does not abolish it, because the insulin outlasts the glucose bolus. Hypoglycaemia commonly appears an hour or more later, so the blood glucose has to be watched for several hours, not just once. Salbutamol acts on beta-2 receptors in skeletal muscle and on cardiac beta receptors, so tremor and tachycardia are the predictable price, not a coincidence.
- Handling
- Not one millimole has left the body — the potassium is hiding inside cells and will come back out. Dialysis, a gut binder or a working kidney has to do the removing.
Catches people out: Insulin with glucose is the more reliable of the two; a substantial minority of patients — particularly those on dialysis or on a beta-blocker — barely respond to salbutamol at all, so salbutamol is an adjunct and not a substitute. The asthma-strength nebule is not an effective potassium dose. In diabetic ketoacidosis this same mechanism is the trap — the potassium is checked before insulin is started, and guidelines have potassium replaced first when it is below about 3.5 mmol/L. Follow your local protocol for actual doses.
- Binds
- Spironolactone and eplerenone bind the intracellular mineralocorticoid receptor; amiloride plugs the epithelial sodium channel (ENaC) from the tubule lumen.
- Which does
- Fewer working sodium channels means less sodium is pulled out of the lumen, so the lumen stays less negative and the electrical force dragging potassium out through ROMK falls away.
- So you see
- Sodium and water are lost, potassium is held onto, and blood pressure falls — strikingly so in Conn syndrome. As diuretics they are weak; the potassium and blood-pressure effects are the point.
- And the same mechanism causes
- The same mechanism pushed too far is hyperkalaemia, and a mild hyperchloraemic (normal anion gap) metabolic acidosis with it, because hydrogen secretion by the neighbouring intercalated cells is driven in part by that same negative lumen. Spironolactone adds a second, off-target problem: its steroid structure also antagonises the androgen receptor and has progestogenic activity, so men get breast tenderness and gynaecomastia and women get menstrual disturbance. Eplerenone binds the mineralocorticoid receptor far more selectively and largely avoids it.
- Handling
- Anyone on one needs their potassium and creatinine watched, especially after an intercurrent illness dries them out.
Catches people out: Trimethoprim blocks ENaC in much the same way amiloride does, which is why trimethoprim-sulfamethoxazole causes hyperkalaemia. Stack an ACE inhibitor or ARB (less aldosterone), spironolactone and an NSAID (less renal perfusion) and you have manufactured one of the commonest drug-induced hyperkalaemias in general practice. Risk rises steeply as eGFR falls.
- Binds
- The intracellular mineralocorticoid receptor in the collecting duct principal cell.
- Which does
- The activated receptor increases ENaC and Na+/K+-ATPase expression and activity, restoring the negative lumen that drives potassium and hydrogen into the urine.
- So you see
- Sodium and blood volume are retained, postural dizziness settles, and the potassium comes down.
- And the same mechanism causes
- Overshoot gives the mirror image of Addison disease: hypertension, ankle oedema, hypokalaemia and a metabolic alkalosis.
- Handling
- The dose is judged on postural blood pressure, potassium and sodium, not on how the patient says they feel.
Catches people out: Replacement-dose hydrocortisone does not cover the mineralocorticoid job in primary adrenal failure — the potassium and postural blood pressure stay wrong until fludrocortisone is added. (High stress doses of hydrocortisone do have enough mineralocorticoid activity to carry a patient through an acute adrenal crisis, which is why fludrocortisone is not the urgent drug in that setting; it is the maintenance one.) Secondary adrenal insufficiency spares the renin-angiotensin-aldosterone axis, so it does not need fludrocortisone and does not typically cause hyperkalaemia.
Calcium protects the heart, insulin and salbutamol hide the potassium, and potassium only actually leaves the body through a working kidney, the gut (a binder, or diarrhoea) or dialysis — so in a dialysis patient with a wide slow rhythm, treat for hyperkalaemia before the blood result comes back.
Now test whether it stuck
Reading this through is not the same as being able to reconstruct it. Every question in the bank is free, with a full debrief on each option.