- Na >145 = too little water for the salt. It is a water problem, not a salt problem.
- Shock or very dry? Give LR or NS first to fix the volume. Then give water.
- Give water: free-water flushes if the gut works, or D5W IV. NS is not water.
- Speed: chronic or unknown → lower Na by about 10 mEq/L per 24 h. Acute (<48 h) → 1 mEq/L per hour back to normal.
- Check Na every 4–6 h (every 2–4 h if acute or Na >165). Add ongoing losses to the plan.
- Big volume of dilute urine (U-osm <300)? Think diabetes insipidus → give desmopressin.
- Most hospital cases are made by us: NPO, no water ordered, diuretics, lactulose, tube feeds, fever.
Pathophysiology in 60 seconds
⌄
- Two defenses keep Na normal: thirst (drink) and ADH (kidney saves water).
- Hypernatremia happens when one defense fails and the patient cannot drink. The thirsty patient who can reach a cup does not get hypernatremia.
- Who cannot drink? Old, confused, intubated, sedated, NPO, babies.
- High Na pulls water out of cells. The brain shrinks. Veins between brain and skull can tear → bleed.
- After ~48 h the brain makes its own osmolytes and goes back to near-normal size. Now, too much water too fast can cause swelling.
Sodium does not cross the cell membrane easily — water moves. High Na pulls water out of brain cells. After ~48 h the brain adapts. That is why acute hypernatremia can be fixed fast, and chronic hypernatremia is fixed at a steady pace.
Think of the patient as a bathtub of salty water. The salt stays the same. The water level is low. The faucet is thirst and drinking. The drain is the kidney, the gut, the skin and the lungs. Your job: open the faucet (give water) and close the drain (fix the cause).
Symptoms
- Thirst first — but old and confused patients often do not feel it or cannot say it.
- Then lethargy, weakness, irritability, confusion, muscle twitching, hyperreflexia.
- Severe (usually Na >160): seizures, coma, intracranial bleed.
- High Na is a marker of high mortality — most of the risk comes from the illness under it.
Causes
⌄
- Don't drink — no thirst, no access (dementia, intubated, NPO, stroke)
- Drained — water lost from skin (fever, sweat, burns), gut (diarrhea, lactulose, NG output), lungs (fast breathing)
- Diabetes insipidus — central or nephrogenic
- Diuresis — osmotic (glucose, urea, mannitol) or loop diuretic
- Doctor-made — hypertonic saline, sodium bicarbonate, lots of NS, salt tabs
| Volume status | Typical causes |
|---|---|
| Low (most common) | GI loss, skin loss, osmotic diuresis, loop diuretic + no water |
| Normal | DI, insensible loss, no water access |
| High | NS or bicarb load in ICU, hypertonic saline, hypertonic feeds, (rare) hyperaldosteronism |
Lithium · high Ca · low K. Also: post-obstruction, recovering ATN, demeclocycline, amphotericin B, pregnancy (vasopressinase).
Diagnosis
⌄
- Labs: BMP, glucose, BUN (urea), Ca, Mg, serum osm.
- Urine: urine osm, urine Na + K, urine volume (strict I/O). Weight.
- Look at the MAR and the I/O: lactulose, furosemide, tube-feed type, mannitol, hypertonic saline, fever curve.
→ central DI (AVP deficiency): pituitary surgery, head trauma, tumor
→ nephrogenic DI (AVP resistance): lithium, high Ca, low K
Osmotic diuresis check: total urine solute = U-osm × 24-h urine volume. More than ~1000 mOsm/day = solute is pulling the water out.
The patient is already "deprived." Give desmopressin 1–2 mcg IV/SC and measure urine osm in 1–2 h. Big rise = central. No rise = nephrogenic. A copeptin level >21 pmol/L (no water deprivation) also points to nephrogenic DI.
Central DI → arginine vasopressin deficiency (AVP-D). Nephrogenic DI → arginine vasopressin resistance (AVP-R). Same disease, new name — do not get fooled on the boards.
Management
⌄
1. Total body water (TBW) = weight (kg) × factor
| Patient | Factor |
|---|---|
| Young man | 0.6 |
| Young woman / old man | 0.5 |
| Old woman | 0.45 |
2. Free water deficit (FWD) = TBW × (Na ÷ 140 − 1)
3. Effect of 1 L of fluid on Na (Adrogué–Madias): ΔNa = (infusate Na − serum Na) ÷ (TBW + 1)
| Fluid | Na (mEq/L) | Free water per 1 L |
|---|---|---|
| D5W / free water | 0 | 1000 mL |
| 0.2% NaCl | 34 | ~780 mL |
| 0.45% NaCl | 77 | ~500 mL |
| LR | 130 | ~130 mL |
| 0.9% NaCl | 154 | none |
80-year-old woman, 50 kg, Na 162, from a nursing home. Chronic.
- TBW = 50 × 0.45 = 22.5 L
- FWD = 22.5 × (162 ÷ 140 − 1) = ~3.5 L (total, over 2–3 days)
- 1 L D5W lowers Na by (0 − 162) ÷ 23.5 = ~7 mEq/L
- Goal −10 in 24 h → ~1.5 L D5W
- Add insensible loss ~1 L/day (more with fever) + measured losses
- Start D5W ~100 mL/h (≈ 2.5 L/day). Recheck Na in 4–6 h and adjust.
- Free-water flushes 250 mL q6h via tube = 1 L/day. Write the total daily free water in the note.
- D5W 100 mL/h = 120 g glucose/day → check glucose (more glucose in urine = more water loss). Use enteral water when you can.
- Can drink? Water at the bedside, within reach. Ask nursing to offer water every 2 h.
- Hypovolemic + high Na: start LR/NS for volume, then switch to 0.45% NaCl or D5W.
- Hypervolemic (ICU, after NS/bicarb): D5W + loop diuretic. A loop alone makes Na worse (loop urine is hypotonic).
Central DI (AVP-D) - Desmopressin 1–2 mcg IV/SC q8–12h, or 0.1–0.2 mg PO BID–TID. - Give it with Na checks. Too much desmopressin + IV water → hyponatremia. - After pituitary surgery: watch the triple response — DI → SIADH (~day 7) → DI again.
Nephrogenic DI (AVP-R) - Fix the cause: correct Ca and K; stop lithium if possible (ask psychiatry). - Amiloride for lithium-induced DI. - Thiazide + low-salt, low-protein diet ± NSAID to lower urine volume.
DI + intact thirst + free access to water → Na usually stays normal. The danger is when they cannot drink.
Chronic: ~10 mEq/L per 24 h (many use a ceiling of 12). Acute: 1 mEq/L per hour. Neither: never let it stall above 150 for days.
- Brain swelling from fast correction is well described in children. In adults, observational data show it is rare.
- Chauhan 2019 (ICU) and Feigin 2023 (4265 adults with severe hypernatremia): faster correction (>0.5 mEq/L/h) was not linked to harm; slow correction was linked to higher mortality.
- Kidney360 2025 (~38,000 adults): the opposite — slow correction had lower in-hospital mortality, but more discharges to hospice or a facility.
- Bottom line: all retrospective, no RCT. Aim for a steady ~10 per day, check often, and do not leave the patient salty for days.
Doses are typical adult doses. Check local protocols, glucose, and renal function. Free-water and D5W rates are a start — the next Na tells you the real dose.
Real-life ward tips
⌄
- Find the drain. Lactulose with 6 stools/day, high-protein tube feeds (urea diuresis), post-obstructive diuresis, ATN recovery, fever, open wounds, and loop diuretics are the usual hidden drains.
- Na still going up on D5W? Check urine volume and U-osm. Calculate electrolyte-free water clearance: urine vol × [1 − (U-Na + U-K) ÷ serum Na]. Add this volume to your water plan.
- Tube feeds contain water (~70–85% water). Count it, but it is usually not enough alone.
- HHS: as glucose falls, Na rises. Follow the corrected Na. When corrected Na is normal or high, switch to 0.45% NaCl.
- Neuro ICU patient on 3% saline with Na goal 145–155? This is on purpose. Do not "fix" it — talk to the team.
- Discharge to a facility: write it in the discharge orders: "Offer water every 2 h while awake. Recheck BMP in 2–3 days."
- Lithium patient with big urine output and Na 150? Do not stop lithium alone without psychiatry. Check level, Ca, and kidney function.
Pitfalls
⌄
- Using NS as "water" in a euvolemic patient — it has zero free water.
- Ordering the free water once and forgetting ongoing losses → Na does not move.
- Correcting too slowly — Na 158 on day 4 with "D5W at 50" is a common miss.
- D5W at a high rate in a diabetic → hyperglycemia → glycosuria → more water loss.
- Desmopressin without Na checks → overshoot into hyponatremia.
- Missing lactulose as the cause in a cirrhotic patient.
- Loop diuretic alone in hypervolemic hypernatremia.
Board corner
⌄
| Stem | Answer |
|---|---|
| Nursing-home patient, fever, Na 160, U-osm 750 | Insensible/extra-renal loss + no access → free water |
| Polyuria after pituitary surgery, Na 152, U-osm 120 | Central DI (AVP-D) → desmopressin |
| U-osm 110 → 125 after desmopressin, on lithium | Nephrogenic DI (AVP-R) → amiloride |
| ICU, high-protein feeds, BUN 80, polyuria, U-osm 450 | Urea osmotic diuresis |
| Cirrhosis on lactulose, 8 stools/day, Na 155 | Osmotic GI water loss → free water, adjust lactulose |
| Hypotensive, dry, Na 158 | NS/LR for volume first, then free water |
| Fluid with no free water | 0.9% NaCl |
| Na rises 18 mEq/L in 4 h after hypertonic saline error | Acute → fast correction (D5W ± dialysis) |
| Hypercalcemia + polyuria + high Na | Nephrogenic DI from high Ca |
| New name for central DI | Arginine vasopressin deficiency (AVP-D) |
References
⌄
- Adrogué HJ, Madias NE. Hypernatremia. NEJM. 2000;342:1493.
- Sterns RH. Disorders of plasma sodium — causes, consequences, and correction. NEJM. 2015;372:55.
- Chauhan K, et al. Rate of correction of hypernatremia and health outcomes in critically ill patients. CJASN. 2019;14:656.
- Feigin E, et al. Rate of correction and all-cause mortality in patients with severe hypernatremia. JAMA Netw Open. 2023;6:e2335415.
- Relationship between rate of hypernatremia correction and outcomes in hospitalized patients. Kidney360. 2025.
- Fenske W, et al. A copeptin-based approach in the diagnosis of diabetes insipidus. NEJM. 2018;379:428.
- Refardt J, et al. Arginine or hypertonic saline–stimulated copeptin to diagnose AVP deficiency. NEJM. 2023;389:1877.
- Arima H, et al. Changing the name of diabetes insipidus: a position statement of the Working Group for Renaming Diabetes Insipidus. 2022.