Collecting duct
The last stretch of the nephron and the body's final say on urine: ADH decides how much water you keep, aldosterone decides how much salt you keep and how much potassium you throw away.
What it normally does
Antidiuretic hormone (ADH, also called vasopressin) is the water tap. Released from the posterior pituitary when the blood gets concentrated (and, more powerfully but less precisely, when blood volume falls), it binds V2 receptors on the blood-side membrane of the principal cell and drives water channels (aquaporin-2) into the wall facing the urine. Water then leaves the tubule and returns to the blood. With ADH switched off, almost no aquaporin-2 sits in that wall and the duct is close to waterproof.
This single switch explains diabetes insipidus, SIADH, and the drugs on this page that turn water reabsorption on or off.
Aldosterone is the salt lever. It binds a receptor inside the principal cell and tells it to build more sodium channels (ENaC) in the urine-facing wall and more Na/K pumps on the blood side. Sodium is pulled out of the urine and water follows, so blood volume rises. The negative charge left behind in the tubule then gives potassium the electrical push to leave the cell through ROMK into the urine.
This explains why too much aldosterone gives high blood pressure with low potassium, and why drugs that block aldosterone or the channel it builds tend to raise potassium.
Alongside the principal cells sit intercalated cells. Type A cells pump hydrogen ions into the urine with an H+-ATPase and generate NEW bicarbonate for the blood — note that filtered bicarbonate is reclaimed far upstream in the proximal tubule, not here; type B cells do the reverse and secrete bicarbonate (through pendrin) when the blood is alkaline. This is the last nephron segment where urine acidification is fine-tuned. Aldosterone helps twice over: the lumen-negative voltage from ENaC sodium entry makes H+ secretion easier, and aldosterone stimulates the H+-ATPase directly. But how much acid actually leaves the body is set less by the pump than by how much ammonia is in the urine to buffer the protons — an unbuffered urine bottoms out near pH 4.5 having carried away almost nothing.
This explains the metabolic alkalosis of aldosterone excess, and — once you add that ammonia production depends on a normal potassium — why aldosterone deficiency or blockade gives hyperkalaemia with a mild acidosis (type 4 renal tubular acidosis) and yet an acid urine.
The collecting duct mostly spends the medullary gradient rather than building it. Water leaves only because the loop of Henle has already stacked sodium chloride into the interstitium, reaching roughly 1200 mOsm/kg at the papilla. The one real contribution the duct makes is urea: under ADH the inner medullary collecting duct reabsorbs urea through UT-A1/UT-A3, and that recycled urea supplies a large share of inner medullary osmolality. Aquaporins open the door; the interstitial gradient does the pulling.
This explains why urine osmolality swings only between roughly 50 and 1200 mOsm/kg, and why a washed-out medulla — from loop diuretics, sustained high urine flow, or very low protein intake — leaves a duct that cannot concentrate no matter how much ADH is present.
What goes wrong
- Central diabetes insipidus (arginine vasopressin deficiency, AVP-D)← from “Antidiuretic hormone (ADH, also called vasopre…”
Damage to the pituitary or its stalk — surgery, head injury, tumour, infiltration such as sarcoid or Langerhans cell histiocytosis — stops ADH release. Aquaporin-2 stays parked in vesicles inside the principal cell and never reaches the wall. Water that should have been reclaimed leaves as urine, the blood concentrates, and the thirst centre screams.
Dilute urine plus concentrated blood plus a response to desmopressin equals central; no response equals nephrogenic. A patient with an intact thirst mechanism and free access to water keeps their sodium near normal — it is the patient who cannot drink (unconscious, post-operative, an infant) who becomes dangerously hypernatraemic.
You would find: Three to twenty litres a day of pale urine, waking repeatedly overnight, unquenchable thirst with a preference for cold water. Urine osmolality under 300 mOsm/kg while plasma osmolality is high-normal and serum sodium sits at the top of the range or above it. On formal testing — water deprivation followed by desmopressin — urine osmolality rises by more than 50% once the desmopressin is given, and that rise is what makes it central. Measuring copeptin (a stable fragment of the ADH precursor, co-secreted with ADH) is increasingly replacing the water deprivation test where it is available.
- Nephrogenic diabetes insipidus (arginine vasopressin resistance, AVP-R)← from “Antidiuretic hormone (ADH, also called vasopre…”
The hormone is present but the cell will not listen. Lithium enters the principal cell through ENaC, accumulates, and downregulates aquaporin-2 production. Chronic hypercalcaemia and chronic hypokalaemia do the same thing. Rare congenital forms are X-linked mutations of the V2 receptor (AVPR2) or, less often, of aquaporin-2 itself. Either way the duct stays waterproof despite high circulating ADH.
In Australian practice lithium is the common acquired cause. The direction of management is to correct calcium or potassium, review the lithium with the prescriber, and consider amiloride — rather than escalating desmopressin, which the duct cannot respond to anyway.
You would find: Identical polyuria and thirst, but ADH (or copeptin) is high and desmopressin does not concentrate the urine. Ask about lithium early — an impaired concentrating ability appears in a large minority of long-term users, though frank symptomatic DI is less common. Then check calcium and potassium.
- Syndrome of inappropriate ADH (SIADH)← from “Antidiuretic hormone (ADH, also called vasopre…”
ADH is released regardless of how dilute the blood already is — from small cell lung cancer, pneumonia, stroke, SSRIs, carbamazepine, MDMA, or pain and nausea after surgery. Aquaporins stay in the wall, water keeps coming back, and the blood is diluted. Total body sodium is roughly normal; this is a water problem, not a salt problem, which is why giving isotonic saline alone can make it worse.
Euvolaemic hypotonic hyponatraemia with inappropriately concentrated urine. Exclude hypothyroidism and adrenal insufficiency before you call it SIADH — they mimic it closely. Correction is limited to about 8 mmol/L in 24 hours (10 is treated as an absolute ceiling, and only in patients at low risk) because faster correction causes osmotic demyelination.
You would find: Serum sodium below 135 mmol/L with a low plasma osmolality, urine osmolality above 100 mOsm/kg and urine sodium above 30 mmol/L, in a patient who looks euvolaemic and has normal thyroid, adrenal and kidney function and is not on a diuretic. Confusion, falls, headache; seizures if it drops fast.
- Primary aldosteronism (Conn syndrome)← from “Aldosterone is the salt lever. It binds a rece…”
An adrenal adenoma or, more commonly, bilateral hyperplasia makes aldosterone independently of renin. ENaC activity stays high, so sodium and water are retained and blood pressure rises, while the persistently negative lumen drives potassium into the urine and, with the direct stimulation of the H+-ATPase, acid too.
It accounts for roughly 5-10% of hypertension and around 20% of resistant hypertension — the commonest potentially curable cause, and routinely missed.
You would find: Hypertension that is hard to control, sometimes with potassium below 3.5 mmol/L and a metabolic alkalosis — though most patients have a normal potassium, so do not wait for it. Screen with the aldosterone-to-renin ratio, having corrected hypokalaemia first (low potassium suppresses aldosterone and gives a false negative) and knowing that mineralocorticoid antagonists in particular, and to a lesser degree beta blockers, ACE inhibitors, ARBs and diuretics, distort the ratio. A positive ratio is a screen, not a diagnosis — confirmatory testing and then adrenal vein sampling to decide surgery versus medical therapy follow.
- Type 4 renal tubular acidosis (hyporeninaemic hypoaldosteronism)← from “Alongside the principal cells sit intercalated…”
Aldosterone is deficient, blocked, or the tubule cannot respond to it, so ENaC is quiet and the lumen never goes strongly negative. Potassium loses its electrical push into the urine and stays in the blood. Trace the acidosis carefully, because the obvious answer is the wrong one: it is not that the intercalated cells have stopped pumping protons. It is largely the hyperkalaemia itself — potassium entering proximal tubule cells raises their internal pH and shuts down ammoniagenesis, and potassium competes with ammonium at NKCC2 in the thick ascending limb so less ammonium is trapped in the medulla. With little ammonia buffer in the urine, net acid excretion falls even though H+ secretion is still working. Diabetic kidney disease is the classic setting; ACE inhibitors, ARBs, mineralocorticoid antagonists, trimethoprim, heparin, calcineurin inhibitors and NSAIDs all push it further.
High potassium plus a normal anion gap acidosis plus an acid urine plus diabetic kidney disease. Diabetic and end-stage kidney disease fall far more heavily on Aboriginal and Torres Strait Islander people — treated end-stage kidney disease runs at several times the non-Indigenous rate and far higher again in remote communities — so this picture turns up earlier and more often in those patients.
You would find: Mild persistent hyperkalaemia (often 5.5-6.5 mmol/L) with a mild normal anion gap metabolic acidosis in a patient with diabetes and reduced eGFR on a renin-angiotensin blocker. Urine pH is usually below 5.5 — the intercalated cells can still acidify, there is simply too little ammonia to carry acid out — and that is what separates it from classic type 1 distal RTA, where distal proton secretion itself has failed and urine pH stays above 5.5.
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
- V2 vasopressin receptor, Gs-coupled, on the principal cell.
- Which does
- Raises cyclic AMP and activates protein kinase A, which phosphorylates aquaporin-2. The vesicles holding AQP2 fuse with the urine-facing membrane and water pores appear in the wall.
- So you see
- Urine volume drops, urine osmolality climbs (commonly above 600 mOsm/kg), and the constant thirst and nocturia settle.
- And the same mechanism causes
- What it brings back across the wall is solute-free water. If the patient keeps drinking, that water dilutes the blood — serum sodium falls, giving headache, nausea, confusion and, if it falls fast, seizures. This is the same effect that makes the drug work, taken too far, and it is why fluid intake is capped rather than left to thirst.
- Handling
- It is a modified vasopressin (1-deamino-8-D-arginine vasopressin). The D-arginine substitution is what strips out most V1 activity, so it does not squeeze blood vessels or push up blood pressure the way vasopressin does; the deamination slows its breakdown and gives it a much longer duration of action.
Catches people out: Limiting fluid intake matters more than any monitoring here. Serum sodium is typically checked a few days after starting and after any dose change. Risk of hyponatraemia is highest in older patients, in children given it for enuresis who keep drinking at night, and in anyone also on a thiazide, an SSRI or carbamazepine.
- Binds
- V2 vasopressin receptor on the principal cell.
- Which does
- Blocks ADH from binding, so cyclic AMP falls and aquaporin-2 is pulled back off the urine-facing membrane into internal vesicles. The duct becomes waterproof again.
- So you see
- The patient passes large volumes of dilute urine while keeping sodium — a water diuresis (aquaresis) rather than a natriuresis. Serum sodium rises.
- And the same mechanism causes
- The patient is losing free water, so becomes very thirsty and dry-mouthed — and if that water loss outruns you, serum sodium rises too fast and causes osmotic demyelination: dysarthria, dysphagia, quadriparesis and a locked-in state appearing days after the sodium was 'fixed'. In a patient who cannot access water (frail, confused, nil by mouth), the same aquaresis produces hypernatraemia and volume depletion.
- Handling
- Sodium is measured frequently — around 6-hourly at first. The rise is held to about 8 mmol/L in 24 hours (10 is treated as an absolute ceiling, and less in high-risk patients: severe hyponatraemia, malnutrition, alcohol use, hypokalaemia, advanced liver disease). It is not combined with tight fluid restriction in the first 24 hours, since the two together overshoot, and water is left freely available.
Catches people out: Started in hospital with sodium monitoring, never casually. Tolvaptan is hepatotoxic at the higher, long-term doses used in polycystic kidney disease, so liver function is monitored monthly at first and then regularly for as long as it is continued. It is a CYP3A4 substrate, so strong inhibitors such as clarithromycin and ketoconazole markedly raise exposure, and moderate inhibitors such as diltiazem raise it appreciably.
- Binds
- Intracellular mineralocorticoid receptor — it competes with aldosterone for the binding site.
- Which does
- The occupied receptor no longer drives transcription of ENaC and the Na/K pump. Fewer sodium channels appear in the urine-facing wall, so less sodium is reabsorbed and the inside of the tubule stays less negative.
- So you see
- Sodium and water are lost, blood pressure falls, and potassium that used to be dumped into the urine is kept.
- And the same mechanism causes
- Hyperkalaemia. Most regulated potassium secretion happens here and in the connecting tubule, driven largely by the lumen-negative voltage that sodium entry creates — remove that push and potassium climbs. A mild normal anion gap acidosis follows, but follow the actual chain rather than the tempting shortcut: it is chiefly the hyperkalaemia that suppresses proximal ammoniagenesis and starves the urine of buffer, not a failure of the intercalated cells to pump protons — which is why these patients still pass an acid urine. It is type 4 renal tubular acidosis, arrived at by drug.
- Handling
- Potassium and creatinine are checked at baseline and again roughly a week after starting or increasing the dose. Spironolactone is a prodrug with long-acting active metabolites, so hyperkalaemia can build over days and does not resolve the moment the drug stops.
Catches people out: Spironolactone is a blunt instrument at steroid receptors — it also blocks androgen receptors and has progestogenic activity, which is exactly why it causes gynaecomastia, breast tenderness, erectile dysfunction and menstrual irregularity. Eplerenone binds the mineralocorticoid receptor selectively and largely avoids this. Hyperkalaemia risk multiplies when combined with an ACE inhibitor, ARB, NSAID, potassium supplement or trimethoprim, or when eGFR is low — generally avoided when eGFR is below about 30 mL/min/1.73m2 or potassium is already above about 5.0 mmol/L.
- Binds
- The pore of the epithelial sodium channel (ENaC) on the luminal surface of the principal cell.
- Which does
- Physically plugs the channel. Sodium cannot enter — and neither can lithium, which reaches the cell through this same channel. The cell stops accumulating lithium and aquaporin-2 expression recovers.
- So you see
- In lithium DI, urine volume falls and the patient can concentrate better. As a diuretic on its own it is weak, but it holds onto potassium.
- And the same mechanism causes
- Hyperkalaemia. No sodium entry means no lumen-negative voltage, and without that voltage potassium loses much of the electrical driving force that normally pushes it out through ROMK into the urine, so it accumulates in the blood. The same lost voltage, plus the hyperkalaemia's effect on ammonia production, gives the mild acidosis of type 4 RTA.
- Handling
- Trimethoprim blocks the same channel — that is why a course of trimethoprim raises potassium, and why it is risky in a patient already on an ACE inhibitor, ARB or spironolactone. Same mechanism, different drug. Amiloride protects the tubule cell from lithium but does not lower the plasma lithium level, so it addresses the concentrating defect and not lithium toxicity.
Catches people out: Avoid or monitor closely in chronic kidney disease and alongside ACE inhibitors, ARBs, potassium supplements or mineralocorticoid antagonists. Blocking ENaC alters lithium handling, so lithium levels need rechecking — and lithium dosing itself is reviewed with the treating psychiatrist, never changed unilaterally.
Polyuria with dilute urine is not automatically diabetes insipidus, and 'give desmopressin and see' is the wrong first move. First separate a water diuresis from a solute one — check glucose, because uncontrolled diabetes mellitus is far commoner than either DI. Then separate DI from primary polydipsia: in DI the plasma osmolality and sodium are high-normal or high, in primary polydipsia they are low-normal or low, and giving desmopressin to someone who is still drinking freely is precisely how you cause hyponatraemia. Only once true hypotonic polyuria is established does the desmopressin response (as part of a supervised water deprivation test, or replaced by copeptin measurement) tell you which half of the ADH pathway has failed: urine concentrates by more than 50% = central, the pituitary; urine stays dilute = nephrogenic, the duct — ask about lithium, then check calcium and potassium.
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.