There are few things in nutrition more suspicious than a substance which human beings have deliberately sought out, traded over enormous distances, taxed, fought over and added to their food for thousands of years suddenly being declared a poison. This does not prove that salt is harmless. Human beings have also spent thousands of years drinking alcohol. It does suggest that a claim as sweeping as “salt is bad for you” deserves rather more examination than the slogan normally receives.

This is the central strength of A Midwestern Doctor’s long essay, “The Hidden Secrets of Salt”, published in The Forgotten Side of Medicine. The original article may be read here: A Midwestern Doctor, “The Hidden Secrets of Salt”. Its argument is that modern medicine has spent roughly half a century turning dietary salt into a public-health villain on evidence much weaker than most patients have been led to suppose. Low sodium intake, the author argues, can activate compensatory hormonal systems, cause fatigue and orthostatic symptoms, contribute to hyponatraemia and perhaps increase cardiovascular mortality. At the same time, aggressive blood-pressure reduction may expose some patients, particularly elderly ones, to falls, impaired perfusion and reduced quality of life.

There is more truth in this than a conventional account might admit. Indeed, the article is at its strongest when it argues against universal rules and in favour of physiological balance. Its discussion of U-shaped risk curves is particularly important. Yet it repeatedly damages its own case by moving too easily from reasonable criticism to conspiratorial explanation, and from suggestive physiology to claims that are much less securely established. Most seriously, it sometimes treats three different things — low dietary sodium, low serum sodium and deliberately reduced blood pressure — as though they formed one continuous phenomenon. They do not.

The result is an article that I would recommend reading, though not obeying indiscriminately. It is a valuable corrective to medical simplification. It should not become another simplification in the opposite direction.

The Heresy Is Not Entirely Heretical

This is not the first time I have written about A Midwestern Doctor’s attack on modern blood-pressure management. In an earlier article I concluded that the supposed “Great Blood Pressure Scam” contained several legitimate criticisms but failed when it tried to turn them into an argument that hypertension itself was largely a medical invention. Good measurement matters; white-coat hypertension exists; frail old people can certainly be overtreated. None of this abolishes the very substantial randomised evidence that lowering genuinely high blood pressure reduces cardiovascular events.

The salt article is better. It is less dependent on denying established relationships and more concerned with asking whether a dose-response relationship may change direction at its extremes. This is an entirely respectable question.

Take blood pressure. There is good evidence that lowering an elevated pressure reduces cardiovascular risk. The current American guideline regards an average pressure of 130–139/80–89 as stage-one hypertension and now recommends medication even for some lower-risk patients if several months of lifestyle modification fail. It also sets a general treatment goal below 130/80, while encouraging systolic pressures below 120 in selected higher-risk patients. Yet the guideline itself acknowledges the increased risks of hypotension, syncope, electrolyte abnormalities and kidney injury with intensive treatment, and specifically allows individualisation where frailty, side-effects or limited life expectancy make aggressive treatment unattractive

That is different from saying that modern medicine simply drives everyone towards 120 regardless of circumstance. But it also means that A Midwestern Doctor is not tilting entirely at windmills. The direction of American guidance really has changed. People who would once have been regarded as having an acceptable pressure can now be classified as hypertensive and, under some circumstances, medicated. Whether this represents progress or overreach cannot be settled by displaying the changing thresholds on a chart. Definitions change because evidence changes, but definitions can also create markets. Both facts can be true at once.

Salt Really Does Affect Blood Pressure

One claim in the article needs immediate qualification. The author repeatedly suggests that the connection between salt and blood pressure is scarcely real. The evidence does not support that.

A major Cochrane review of modest salt reduction found an average fall of about 5.4/2.8 mmHg in people with hypertension and 2.4/1.0 mmHg in those with normal blood pressure after reducing salt intake by about 4.4 grams per day. Another Cochrane analysis, using a somewhat different trial selection and a larger sodium reduction, found only about 1 mmHg of systolic benefit in normotensive white participants but nearly 6 mmHg in hypertensive ones. It also found the compensatory increases in renin, aldosterone and catecholamines that A Midwestern Doctor discusses. So the interesting question is not whether sodium affects blood pressure. It plainly does. The question is how much this matters in different people.

A five- or six-point systolic reduction in someone persistently measuring 165 is not trivial. A one-point change in someone measuring 118 probably is. The public-health slogan — less salt means lower blood pressure means better health — compresses these very different situations into one rule.

The article is therefore wrong to make the salt–pressure relationship sound almost fictitious, but right to insist that salt sensitivity varies. Age, kidney function, starting blood pressure, genetic background and potassium intake all influence the response. There is no reason to suppose that the same sodium target must therefore be optimal for everybody.

The Most Important Point: The U-Curve

Here the article becomes much more interesting.

A Midwestern Doctor reproduces several U-shaped or J-shaped curves. His wider point is that physiological variables often have an optimum range rather than an optimum direction. If very high blood pressure is harmful, it does not follow that zero blood pressure would be ideal. If very high glucose is harmful, it does not follow that the lowest attainable glucose must be healthiest.

This sounds obvious when stated so baldly. Medicine nevertheless has a recurring temptation to turn risk relationships into arrows: cholesterol down, pressure down, sodium down. Once an arrow has been established, institutional thinking can become reluctant to ask where it should stop.

Large international observational studies have indeed reported a J-shaped association between sodium excretion and cardiovascular events. In one cohort of more than 100,000 people, the lowest event rate occurred among those estimated to consume roughly 3–5 grams of sodium per day, with greater risk at both lower and higher levels. Higher potassium intake appeared to reduce some of the risk associated with higher sodium.

But one must be exceedingly careful here. Three to five grams of sodium is not three to five grams of salt. Sodium chloride is only about 39 per cent sodium. Three grams of sodium corresponds to roughly 7.6 grams of salt; five grams corresponds to about 12.7 grams. Confusing these units can make dietary discussions almost meaningless.

There is also a fierce methodological argument about the apparent J-curve. Many large studies estimate daily intake from one spot urine sample rather than collecting every drop of urine for twenty-four hours on repeated days. This can produce substantial measurement error. Reverse causation is another problem: people who are ill, frail or already under medical supervision may deliberately consume less sodium. Their higher mortality may therefore help cause their low sodium intake rather than result from it. Against the J-curve literature stand other analyses attributing substantial cardiovascular mortality to sodium intake above about 2 grams daily.

I therefore would not look at the observational curve and announce that 4 grams of sodium is the new scientifically proven optimum. That would simply replace one dogma with another. But neither would I dismiss the curve. It is enough to make me sceptical of very aggressive population-wide sodium restriction, particularly when applied to healthy people without hypertension. This, I think, is where A Midwestern Doctor makes his most useful contribution.

Hyponatraemia Is Real — But the Article Blurs Two Different Things

The weakest major section concerns hyponatraemia.

The author is absolutely correct that low serum sodium can be dangerous. Severe hyponatraemia can produce confusion, seizures, coma and death. Mild chronic hyponatraemia in older people is associated with falls, fractures and poor outcomes. Diuretics, particularly thiazides, can cause it, as can SSRIs and various diseases.

But serum sodium is not a straightforward measure of how much salt somebody eats.

Hyponatraemia is fundamentally a disorder of water relative to sodium. A person can have low serum sodium while having a normal total amount of sodium, or even an excess, because the body contains still more water. Heart failure itself can produce precisely this form of dilutional hyponatraemia. SIADH, renal disease, cirrhosis and excessive water intake are other causes. This distinction matters. If hospital patients with serum sodium of 130 die more frequently than patients at 140, we cannot infer that people should eat more table salt. Their low sodium may instead be a marker of severe heart failure, kidney disease, medication effects or impaired water excretion. Giving some such patients salt indiscriminately would be unwise.

There are circumstances in which inadequate solute or sodium intake contributes. Thiazide-associated hyponatraemia is especially important in older people, and modern reviews describe interactions between sodium depletion, excessive fluid consumption and impaired water clearance. But this is a much narrower and more clinically useful claim than treating population studies of serum sodium as proof that public advice to reduce salt is killing people. The article would be stronger if it maintained this distinction throughout.

Heart Failure: Here the Orthodoxy Really Has Shifted

The discussion of heart failure is perhaps the most surprising section, because recent evidence has moved towards A Midwestern Doctor rather more than many readers might expect. The old theory was straightforward. Heart failure causes fluid retention. Sodium promotes water retention. Therefore patients with heart failure should severely restrict sodium.

That logic is physiologically plausible. The clinical trials have been disappointing. The 2022 SODIUM-HF trial enrolled 806 patients and found that reducing dietary sodium to below 100 mmol per day did not significantly reduce death, cardiovascular hospitalisation or cardiovascular emergency visits. A systematic review of randomised trials likewise found no convincing reduction in mortality or hospitalisation.

More strikingly, a 2026 meta-analysis of sixteen randomised trials involving 2,260 patients reported higher all-cause mortality among sodium-restricted groups, with a risk ratio of 1.50, and substantially higher cardiac mortality. This is the study quoted in the article, and the quotation is substantially accurate.

I would nevertheless resist declaring the matter closed. Heart-failure dietary trials are difficult. The interventions differ, fluid restriction and diuretics sometimes accompany sodium restriction, adherence is imperfect and several earlier studies were small. A 2024 review consequently concluded not that salt restriction kills heart-failure patients, but that severe restriction has failed to demonstrate the expected reduction in clinical events.

European guidance has already become more cautious. A 2024 Heart Failure Association consensus statement describes a normal sodium intake as approximately 1.5–4 grams per day, permits more liberal intake in stable patients and warns that extreme restriction below roughly 0.5–1 gram daily may be harmful.

This is the sort of development that justifies scepticism about medical absolutes. The sensible conclusion is not “heart-failure patients should eat lots of salt.” It is that a treatment which sounded obvious in theory did not perform nearly so well when subjected to trials. That should encourage humility.

Potassium May Be Half the Story

One of the best parts of the article is its emphasis on potassium. Modern diets often combine large quantities of sodium from processed food with relatively little potassium because fruit, vegetables, pulses and other potassium-rich foods have been displaced. A person can therefore reduce cardiovascular risk not merely by thinking about sodium in isolation but by altering the sodium-to-potassium balance.

The evidence for potassium-enriched salt substitutes is unusually impressive. The large Salt Substitute and Stroke Study used a mixture containing 75 per cent sodium chloride and 25 per cent potassium chloride and found significant reductions in stroke, major cardiovascular events and death. This principle is now sufficiently well supported that the 2025 American blood-pressure guideline explicitly recommends potassium-based salt substitutes for suitable adults, while warning that they can be dangerous in people with impaired potassium excretion, particularly some patients with kidney disease or those taking certain medicines.

This points towards a more intelligent principle than “salt bad”. A diet dominated by manufactured bread, takeaway meals, cured meats and packaged snacks may contain enormous quantities of sodium while providing little potassium. A diet containing ordinary salted home-cooked food alongside potatoes, vegetables, fruit, dairy and legumes may produce a very different electrolyte environment. The salt shaker is not necessarily where the problem lives.

Natural Salt and Refined Salt: Attractive, But Poorly Proven

Here A Midwestern Doctor moves from provocative evidence into speculation. He suggests that natural salts may be healthier because they retain minerals, while refined table salt contains anti-caking agents and undergoes chemical processing. Perhaps there are questions worth examining here. But the evidence offered does not justify treating Himalayan, Celtic or sea salt as metabolically different substances in any major way. They remain overwhelmingly sodium chloride. The trace magnesium, calcium, potassium and iron in artisanal salts exist, but generally in quantities too small to make a meaningful nutritional contribution. Harvard notes that the trace minerals in sea salt are not present in amounts capable of replacing minerals from ordinary food, while the Mayo Clinic concludes that sea salt and table salt have essentially the same nutritional value.

There is also one advantage of refined salt that alternative-health discussions overlook: iodisation. Iodised salt was introduced because iodine deficiency caused goitre and impaired thyroid function. Many fashionable sea and Himalayan salts contain very little iodine unless deliberately fortified.

By all means choose Maldon flakes because they taste better. I certainly would. But turning a culinary preference into a medical doctrine requires evidence.

Zeta Potential: Where I Become Much More Sceptical

The article’s account of zeta potential is the point at which I part company most decisively with its author. Zeta potential is a real physical concept. Suspended particles have electrical properties affecting whether they remain dispersed or aggregate. Red-cell aggregation and blood rheology are also genuine fields of study. From this, however, the essay advances to a much larger explanatory theory. Poor zeta potential is invoked to explain congestion, arrhythmia, circulatory disease and supposed vaccine-related “microstrokes”. Intravenous saline is then interpreted partly as restoration of optimal electrical dispersion.

This is an enormous leap. It may eventually turn out that clinically manipulating erythrocyte surface charge has useful therapeutic applications. Science is full of mechanisms dismissed before they were properly understood. But a plausible piece of biophysics is not the same as a demonstrated cause of disease, still less a validated treatment system.

The claim that aluminium in vaccines “frequently trigger[s] microstrokes” through zeta-potential disruption is particularly serious and requires correspondingly strong clinical evidence. None is supplied in the article at a level that would justify the certainty of the statement.

This problem recurred in the earlier blood-pressure essay. A Midwestern Doctor begins with legitimate observations about circulation and aggregation, then allows one mechanistic idea to become a theory of almost everything. In medicine, grand unified theories should make us nervous. Human physiology has a habit of humiliating them.

Nor Is Every Medical Mistake a Pharmaceutical Plot

I have sympathy with the article’s suspicion of incentives. Pharmaceutical manufacturers make money by selling medicines. Guideline panels can develop intellectual inertia. Doctors are human beings and may find it easier to prescribe another tablet than to spend forty minutes investigating sleep, diet or orthostatic symptoms.

One need not be a Marxist to notice material incentives. But incentives are not proof of conspiracy. The history of changing blood-pressure thresholds cannot simply be explained as pharmaceutical companies progressively redefining healthy people as sick. Antihypertensive medicines are mostly old and generic. Many cost pennies. There is comparatively little monopoly profit in persuading somebody to take generic amlodipine rather than allowing his pressure to remain at 155.

More importantly, major randomised analyses do show fewer strokes, heart attacks and episodes of heart failure when genuinely elevated blood pressure is reduced. The argument should therefore be about who should be treated, how aggressively and with what tolerance for side-effects rather than whether hypertension was fabricated to sell pills.

My earlier conclusion still seems sound: modern medicine sometimes behaves like a factory, but a factory is not the same thing as a conspiracy.

What the Graphs Show — and What They Do Not

The illustrations accompanying the article are rhetorically effective precisely because several of them look devastating. The serum-sodium graph shows mortality rising at both low and high concentrations. That supports a U-shaped association. It does not prove that changing dietary salt will move people safely along the curve. The frailty and blood-pressure graph is more interesting. Among the very old, particularly frail people, lower systolic pressures can associate with higher mortality. This is a useful warning against taking evidence from healthy sixty-year-olds and imposing the same targets on an 88-year-old who becomes dizzy every time she stands up. Yet observational geriatric data contain formidable reverse-causation problems: terminal illness, cardiac failure and general frailty themselves lower blood pressure. The historical guideline table also tells a genuine story. Treatment thresholds have fallen substantially. But a chronology is not an argument. Earlier doctors may have tolerated dangerous hypertension; modern ones may occasionally overtreat. Both propositions can be true. Indeed, this is the recurring theme of the entire controversy. The existence of one error does not prove its opposite.

So How Much Salt Should We Eat?

I am reluctant to answer with a number, because the whole point of this review is that numbers detached from context are dangerous. A young healthy person who exercises heavily in summer and sweats profusely does not have the same sodium requirements as a sedentary elderly patient with severe hypertension. Someone with POTS may have been instructed to increase sodium. Someone with kidney disease or decompensated heart failure may require quite different advice. A patient taking thiazides who develops dizziness deserves electrolyte testing, not an argument from dietary ideology.

For the ordinary healthy person, I think the evidence supports something less dramatic. Do not make an ascetic religion out of avoiding salt. Do not make a countercultural religion out of eating it either. Eat mainly recognisable food. Obtain plenty of potassium from ordinary foods. Salt food sufficiently to make it pleasant, while remembering that industrially manufactured foods can supply astonishing quantities before the salt cellar has even appeared.

And above all, distinguish sodium intake, serum sodium concentration and blood pressure. They interact, but they are not interchangeable measurements of the same thing.

Nothing in Excess

There is something almost Greek about the conclusion, though perhaps not in the way dietary campaigners would prefer.

The inscription at Delphi said μηδὲν ἄγαν — nothing in excess. It is an undramatic principle. It produces no crusade, sells no miracle supplement and allows neither the salt abolitionist nor the Himalayan-salt evangelist complete satisfaction. Yet physiology repeatedly seems to vindicate it. Too much sodium can raise blood pressure and cause trouble in susceptible people. Too little may activate compensatory systems and may be particularly undesirable under some clinical conditions. Excessive blood pressure damages arteries and organs. Excessive reduction can leave an elderly person dizzy, weak and injured on the bathroom floor. The difficulty is discovering where moderation lies for a particular person.

A Midwestern Doctor deserves credit for asking that question when public-health messaging too often assumes it has already been answered. His discussion of severe salt restriction in heart failure is surprisingly well supported by recent evidence. His attention to potassium is useful, and his suspicion of universal treatment targets is reasonable.

Where he goes wrong is in becoming too certain of his counter-narrative. Hyponatraemia is not simply dietary salt deficiency. Natural salt has not been shown to possess mysterious health advantages over ordinary sodium chloride. Zeta potential does not presently provide a demonstrated master explanation of circulatory disease. And the fact that medical institutions possess incentives does not mean that every unwanted guideline was invented for profit.

There is a conservative lesson here. Conservatism in medicine should mean a preference for accumulated experience, incremental change and an unwillingness to disturb a functioning system without good reason. That principle applies equally to established medicine and its critics. If medicine says that everybody must eat less salt, ask for the evidence. If alternative medicine says that everybody should eat more, ask again. And if either side tells you that one number determines health, reach for the salt cellar — though perhaps only metaphorically.

Reading List

A Midwestern Doctor. “The Hidden Secrets of Salt.” The Forgotten Side of Medicine, 2026.
Original article

Graudal, N. A., T. Hubeck-Graudal and G. Jurgens. “Effects of Low Sodium Diet versus High Sodium Diet on Blood Pressure, Renin, Aldosterone, Catecholamines, Cholesterol, and Triglyceride.” Cochrane Database of Systematic Reviews (2020). This is especially useful because its interpretation of sodium reduction is noticeably less enthusiastic than much public-health literature.

He, F. J., J. Li and G. A. MacGregor. “Effect of Longer-Term Modest Salt Reduction on Blood Pressure.” Cochrane Database of Systematic Reviews. An important counterweight, finding meaningful blood-pressure reductions, particularly among hypertensive patients.

O’Donnell, Martin et al. “Urinary Sodium and Potassium Excretion, Mortality, and Cardiovascular Events.” New England Journal of Medicine 371 (2014): 612–623. One of the major studies underlying the argument for a J-shaped sodium relationship.

Mente, Andrew et al. “Joint Association of Urinary Sodium and Potassium Excretion with Cardiovascular Events and Mortality.” BMJ 364 (2019): l772. Particularly valuable for its treatment of potassium and the apparent 3–5 g sodium range.

Mozaffarian, Dariush et al. “Global Sodium Consumption and Death from Cardiovascular Causes.” New England Journal of Medicine 371 (2014): 624–634. The important opposing interpretation, estimating substantial cardiovascular mortality from excess sodium.

Ezekowitz, Justin A. et al. “Reduction of Dietary Sodium to Less Than 100 mmol in Heart Failure (SODIUM-HF).” The Lancet 399 (2022): 1391–1400. The largest modern randomised test of sodium restriction in heart failure.

Mullens, Wilfried et al. “Dietary Sodium and Fluid Intake in Heart Failure: A Clinical Consensus Statement of the Heart Failure Association of the ESC.” European Journal of Heart Failure 26 (2024): 730–741. A useful example of how mainstream guidance has become less dogmatic.

“The Effect of Sodium Restricted Diet on the Prognosis of Heart Failure Patients: A Systematic Review and Meta-analysis.” Frontiers in Cardiovascular Medicine (2026). Sixteen randomised trials and 2,260 patients; its reported mortality findings deserve serious attention, though also further replication and scrutiny.

American Heart Association/American College of Cardiology. 2025 Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults. Essential reading if one wishes to criticise present American treatment thresholds accurately rather than in caricature.

Lewis, James L. III. “Hyponatremia.” Merck Manual Professional Edition, updated 2025. A useful corrective to the common confusion between dietary sodium restriction and low serum sodium.