The Hidden Science of Salt: Why the Low-Sodium Era Got the Curve Wrong

What salt actually does in the body

Sodium is the principal cation of extracellular fluid. It sets extracellular volume, and therefore blood pressure and organ perfusion. It powers the sodium–potassium pump, which consumes a large share of the body's resting energy budget and maintains the membrane potential every nerve impulse and muscle contraction depends on. It drives nutrient absorption in the gut and reabsorption in the kidney, since glucose and amino acid transporters are sodium-coupled. It is required for stomach acid, for adrenal function, and for the fluid volume that carries oxygen to tissue.

The kidney regulates it with extraordinary precision. Given a high intake, sodium is excreted. Given a low intake, aldosterone and the renin–angiotensin system rise and the kidney reclaims nearly all of it — at a cost, because that same hormonal response raises sympathetic tone, insulin resistance, and cholesterol markers. This is the central point the low-sodium era missed: sodium restriction is not passive subtraction. It is an active hormonal stress response.

The blood pressure claim, examined

The claim that salt raises blood pressure is not false. It is smaller and narrower than the messaging implies.

The Cochrane reviews of sodium reduction, which are the most systematic accounting available, find that a substantial cut in intake produces a meaningful blood pressure change in people who already have hypertension and a very small one in people who do not. Alongside that modest benefit, restriction reliably raises renin, aldosterone, noradrenaline, adrenaline, triglycerides, and total cholesterol — measurable adverse shifts that the blood pressure number alone does not capture.

The larger problem is the missing second step. Lowering a risk marker is not the same as lowering risk. No adequately powered randomised trial has ever shown that reducing dietary sodium in the general population reduces cardiovascular death, and the observational literature increasingly points the other way at the low end.

Why the low end is dangerous

The hazards of too little sodium are concrete, well documented, and rarely discussed with patients.

Hyponatremia — blood sodium below roughly 135 mmol/L — is among the most common electrolyte abnormalities on hospital admission. Its acute presentation is dramatic: confusion, seizure, cerebral oedema. But its chronic, "mild," asymptomatic form is the more relevant one for readers, because that form is independently associated with higher mortality over five years, with unsteadiness and falls, with fractures in older adults, and with attention and gait deficits that resolve when sodium is corrected.

Aggressive dietary restriction contributes. So do the medications that frequently accompany it — thiazide diuretics, SSRIs, carbamazepine, and others — which impair free-water excretion. A person on a thiazide, an SSRI, and a low-salt diet has three independent pushes toward the wrong side of the curve at once.

Subclinical under-salting produces a milder version of the same picture: fatigue, lightheadedness on standing, exercise intolerance, poor heat tolerance, muscle cramps, salt craving, and brain fog. These are frequently attributed to age, deconditioning, or anxiety.

The U-curve

Large prospective cohorts measuring sodium excretion have repeatedly found the lowest all-cause mortality and cardiovascular event rates in the middle of the intake range, roughly 3–5 g sodium per day, with rising risk both below and above. The most cited of these is the PURE study, spanning more than eighteen countries.

That work is genuinely contested. Critics argue that spot-urine estimation formulas systematically distort intake at the extremes, and that reverse causation — sick people eating less — inflates the apparent harm of low intake. Those criticisms are worth taking seriously; they are also, notably, the same methodological objections that would apply to the observational studies used to justify restriction in the first place.

What survives the argument is this: the evidence base is not strong enough to support pushing an entire population below its instinctive intake, and the burden of proof for a universal intervention sits with the people proposing it.

The heart failure reversal

Salt restriction in heart failure was the one recommendation with a clean mechanistic rationale. A weakened heart, a congested circulation, sodium-driven fluid retention — restrict the sodium, reduce the congestion.

The trials did not cooperate. A 2018 systematic review found no evidence that restriction reduced deaths, hospitalisations, or length of stay. SODIUM-HF, the largest randomised trial to date with 806 patients across six countries, found that reducing sodium below 1,500 mg/day did not reduce deaths, cardiovascular hospitalisations, or emergency visits versus usual care. Later reviews and meta-analyses have gone further, reporting a signal toward increased mortality with strict restriction, most pronounced in the sickest patients and when restriction was stacked on top of fluid restriction and diuretics.

Acute decompensation from a very high sodium load in advanced heart failure is real and well documented. Chronic prophylactic restriction is a different intervention, and it does not have the evidence behind it. European guidance has begun to reflect this; American practice largely has not.

Not all salt is the same

Refined table salt is roughly 97–99% sodium chloride, stripped of the trace mineral fraction present in the original brine or deposit, and usually carrying an anti-caking agent — silicon dioxide, sodium ferrocyanide, or in some markets aluminosilicates — plus added iodine and sometimes dextrose as an iodine stabiliser.

Unrefined sea salt and rock salts retain magnesium, potassium, calcium, and trace elements. The honest accounting is that these fractions are small: the magnesium in a daily gram of sea salt is not a meaningful magnesium source. The defensible arguments for unrefined salt are the absence of processing additives, better flavour at lower dose, and a marginally more physiological ion profile — not mineral supplementation.

Two real caveats. First, most unrefined sea salts contain little iodine; if you switch away from iodised salt, iodine must come from somewhere else — seaweed, dairy, eggs, fish, or a supplement. Second, ocean-sourced salts carry measurable microplastic contamination in analytical surveys, with rock salts generally lower. Neither point is a reason to fear salt; both are reasons to choose sourcing deliberately.

The potassium half of the equation

The most useful practical finding in this entire literature is not about sodium at all.

Ancestral diets delivered far more potassium than sodium. Modern diets invert that ratio. In the largest trial to test the rebalancing directly — roughly 21,000 participants at high stroke risk, followed for five years — simply replacing ordinary salt with a 75% sodium chloride / 25% potassium chloride substitute produced 14% fewer strokes, 13% fewer major cardiovascular events, and 12% fewer deaths. Total salt intake was not reduced. Only the ratio changed.

Potassium substitutes are not universally safe. Anyone with chronic kidney disease, on ACE inhibitors, ARBs, or potassium-sparing diuretics needs medical supervision, because hyperkalemia is dangerous. For everyone else, the more robust route is dietary: potatoes, beans, leafy greens, avocado, squash, yoghurt, and fruit.

The other sodium: sodium bicarbonate

Salt is not the only sodium compound with a physiological job. Sodium bicarbonate (NaHCO₃) — ordinary baking soda — is the body's principal extracellular buffer, and it is the one sodium salt whose usefulness has almost nothing to do with blood pressure and almost everything to do with acid.

Think of bicarbonate as a biological fire extinguisher. When tissue or blood turns more acidic — from a hard sprint, a heavy protein-and-grain diet, or declining kidney function — bicarbonate absorbs the excess hydrogen ions and restores the narrow pH range life runs on. That single mechanism explains four separate uses.

Digestive relief. Bicarbonate neutralises stomach acid within seconds, which is why it has been a heartburn remedy for a century. Roughly ¼ to ½ teaspoon in half a glass of water, and no more than three or four doses in twenty-four hours.

Athletic edge. By raising buffering capacity outside the muscle cell, bicarbonate delays the pH drop that produces the burn, extending high-intensity efforts lasting about one to seven minutes. The studied dose is 0.2–0.3 g per kg of body weight, taken 60–90 minutes before the event.

Kidney protection. In chronic kidney disease, low serum bicarbonate is both a marker and a driver of decline. Correcting it slows the loss of kidney function — one of the few cheap interventions in nephrology with that claim behind it.

Immune calming. Newer work suggests oral bicarbonate signals through the spleen to shift macrophages toward an anti-inflammatory state, lowering systemic inflammation without broad immunosuppression.

The catch is that bicarbonate is a sodium salt, and a concentrated one. Everything the U-curve says about sodium having a ceiling as well as a floor applies here with less margin: in hypertension, heart failure, or advanced kidney disease, bicarbonate is a supervised intervention, not a kitchen habit.