Hypokalaemia: Safe Replacement and the Magnesium Link

Why a low number is a cardiac problem

Potassium sets the resting voltage of every excitable cell. The steep potassium gradient across the cell membrane — high inside, low outside — is what fixes the resting membrane potential. Lower the outside K+ and that gradient distorts, changing how easily cells depolarise and, crucially, how they repolarise afterwards. In skeletal muscle the effect is weakness, cramps and, when severe, frank paralysis. In smooth muscle the gut slows and can stop altogether — an ileus, the silent distended abdomen. But the organ that matters most is the heart. Hypokalaemia prolongs and destabilises repolarisation, and the myocardium becomes electrically irritable: ectopic beats, then re-entrant arrhythmias, and at the extreme ventricular tachycardia or fibrillation. This is why a bald number on a lab report is treated as an emergency long before the patient looks like one.

The electrocardiogram gives away the disorder before catastrophe. The classic sequence is flattening of the T waves, the appearance of U waves (a small deflection after the T), ST-segment depression, and a prolonged apparent QT. None of these is subtle once you look for them, and on a patient on digoxin they are a fire alarm — because low potassium dramatically potentiates digoxin toxicity. Digoxin and potassium compete for the same site on the Na+/K+-ATPase pump; when potassium falls, digoxin binds more avidly, and a dose that was perfectly safe yesterday becomes toxic today. The Cardiovascular chapter on cardiac glycosides treats this interaction in full — but the one-line rule is worth carrying everywhere: never let the potassium run low in a patient on digoxin.

Where the potassium goes — the four routes

Every cause of hypokalaemia is a story of loss, shift, or poor intake. The largest cause by far is renal loss, and here pharmacology dominates: loop and thiazide diuretics are the classic culprits, forcing sodium and water out and wasting potassium along the distal nephron. Mineralocorticoid excess does the same through aldosterone — primary hyperaldosteronism (Conn's syndrome), Cushing's syndrome, and, memorably, liquorice, whose active compound blocks the enzyme that normally protects the kidney from cortisol and so mimics aldosterone. The Endocrine section on the adrenal cortex covers Conn's and Cushing's in detail. The second route is gastrointestinal loss: vomiting, diarrhoea, and laxative abuse strip potassium from the gut (vomiting also drives renal loss indirectly, through the alkalosis and volume depletion it creates).

The third route catches people out because the body's total potassium hasn't changed at all — it has simply shifted into cells. Insulin drives potassium inward (which is exactly why insulin-with-glucose is a treatment for the opposite problem, hyperkalaemia); beta-2 agonists such as salbutamol do the same; alkalosis pushes potassium into cells in exchange for hydrogen ions; and the most dangerous version of all is refeeding syndrome, where feeding a starved patient triggers a surge of insulin that sweeps potassium, phosphate and magnesium into cells at once. The Refeeding-syndrome chapter is essential reading precisely because the shift can crash the potassium hours after the first meal or infusion. The fourth and least common route is simply poor intake — rare on its own, because healthy kidneys conserve potassium well, but it compounds every other cause in the malnourished, the alcohol-dependent, and the elderly.

The magnesium rule: why replacement fails

The potassium that won't come up is nearly always hiding a low magnesium. This is the single highest-yield idea in the whole topic. Hypokalaemia and hypomagnesaemia travel together — the same diuretics, the same diarrhoea, the same alcohol dependence that waste one waste the other. But magnesium is not just a fellow traveller; it is mechanistically upstream. Low magnesium releases the brake on the renal outer-medullary potassium (ROMK) channels in the distal nephron, so the kidney leaks potassium it should be holding onto. The consequence at the bedside is brutally practical: if you replace potassium without correcting the magnesium, the kidney simply wastes the potassium you are giving, and the level will not budge. Refractory hypokalaemia — potassium that won't correct despite generous replacement — should make you check and correct the magnesium every single time. Replace both together. The Magnesium chapter develops this and the arrhythmia connection further.

Replacing potassium without causing harm

The route matters as much as the dose. Wherever the patient can swallow and the deficit is mild or moderate, the oral route is preferred — potassium chloride salts, given as tablets or solution, replace the deficit safely and steadily. Chloride is usually the right anion because many hypokalaemic states (especially vomiting) are also chloride-depleted and alkalotic. Oral replacement almost never causes the dangerous overshoot that intravenous potassium can, which is exactly why it is the default. Reserve intravenous potassium for when the deficit is severe, the patient is symptomatic (arrhythmia, marked weakness, paralysis), or the oral route is unavailable — and then treat it with genuine respect.

Here is the cardinal safety rule, and it is absolute: intravenous potassium must be diluted and rate-limited. Concentrated or fast potassium delivered straight to the heart causes fatal arrhythmia and cardiac arrest — this is not a theoretical risk but a recurring cause of death from medication error. Potassium must never be given as a rapid IV bolus or "push." It is diluted in fluid, run at or below a maximum infusion rate, and for higher rates given through a controlled infusion pump with continuous cardiac (ECG) monitoring, ideally into a larger vein. A useful mental anchor: the concentration and the speed are what kill, not the total amount — the same dose that heals over hours can stop the heart in seconds. Peripheral infusions are also kept relatively dilute because concentrated potassium is intensely irritant to veins.