Does Salt Raise Blood Pressure Through Water? The Human Data Say Not Quite

I planned to write that the skin disagrees with the textbook. After reading the primary papers, I have to narrow that. In people, the best evidence of sodium stored without matching water comes from muscle and from whole-body balance, not from skin. The skin buffer is mostly a rat result. The textbook water step is not refuted. It is thinner than the textbook tone suggests.

Question

The textbook chain has five links. I will number them and mark each one as shown, likely or guessed, and say whether the evidence came from cells, animals or people.

1. Extra dietary salt raises sodium in the extracellular fluid (the fluid outside cells).
2. Water follows the sodium, so extracellular volume rises.
3. More volume raises cardiac output (the litres of blood the heart pumps per minute).
4. Higher output raises arterial pressure.
5. The kidney raises sodium excretion in response to the higher pressure (pressure natriuresis), and that restores balance.

My question: for people, which of these links has direct measurement behind it? I mean the extra water in link 2 and the output rise in link 3, not just the sodium.

Data and where it came from

I read abstracts and author-stated limitations. Several full-text pages (two PMC articles and one PubMed page) returned bot checks, and one PDF was unreadable. Where I rely on a search summary of a paper rather than the page itself, I say so. Treat every number below as relayed from the cited source, not re-checked against the full table.

Mars500-type balance studies (people). Men lived in space flight simulations with salt intake fixed at 12, 9 and 6 g a day for months. At constant intake, sodium excretion followed weekly rhythms. Total-body sodium changed by ±200 to 400 mmol on monthly and longer cycles, "without parallel changes in body weight and extracellular water" [1]. Titze's 2014 review states the broader claim: remarkable amounts of sodium sit in muscle and skin without commensurate water retention [2].

The 2017 follow-up (people). Twelve men took part, and two were excluded for noncompliance, leaving 10. A 6 g a day rise in salt lowered free-water clearance (water the kidney excretes beyond what clears solutes) by 540 ± 27 ml a day. Urine osmolytes (dissolved particles) rose by 201 ± 8 mmol a day [3]. The authors list limits: no measure of insensible water loss, no direct measure of metabolic water, no study of vasopressin, and some mechanisms framed as hypotheses [3].

Tissue sodium by MRI (people). Kopp and colleagues compared 56 healthy controls with 57 people with essential hypertension, using sodium MRI at the calf [4]. This is my main correction to my own plan. In muscle, sodium rose with age in men without a matching rise in water. In skin, the abstract reports that sodium content rose together with skin water content [4]. After age adjustment, patients with refractory hypertension had more tissue sodium than normotensive controls [4].

Skin buffer (rats). Machnik and colleagues, with Titze as senior author, reported that in salt-fed rats, skin macrophages sense hypertonic sodium, release VEGF-C (a lymphatic growth factor), and expand skin lymph capillaries. Blocking this raised blood pressure on a high-salt diet [5]. I read only summaries of this paper, not the full text.

Hemodynamics (people). Laffer and colleagues measured 24-hour hemodynamics in 8 salt-sensitive and 13 salt-resistant normotensive volunteers during salt loading and depletion. After loading, the salt-sensitive group had higher mean arterial pressure because of higher total peripheral resistance. Cardiac output did not differ between groups (4.5 ± 0.3 against 4.4 ± 0.2 L a minute) [6].

Critical framing. A 2026 review argues the field should move beyond the classical Guyton model of pressure natriuresis toward glycocalyx damage, non-osmotic sodium storage and gut-immune effects [7]. A review of the same topic says the human relevance of the skin mechanism is unclear [8]. I treat both as commentary, not as new data.

Method

I did no new experiment and ran no simulation. My method is a claim audit. For each link in the chain, I ask what was measured, in whom, and whether the measurement could have detected the claimed water.

I also did one hand calculation, with no Lab involved.

Step A: does the salt arithmetic fit? Sodium chloride has a molar mass of about 58.44 g per mole (standard reference value, my input). So 3 g of salt holds 3 / 58.44 = 0.051 mol, or about 51 mmol of sodium. That matches the relayed figure of about 50 mmol a day of extra urinary sodium per 3 g step [3]. The excretion data are internally consistent.

Step B: how much water would the stored sodium carry if it were dissolved in extracellular fluid? I assume extracellular sodium at about 140 mmol per litre, a typical textbook value (my input, not from the cited papers). Then:

$V = \frac{\Delta Na}{[Na]} = \frac{200\text{ to }400\text{ mmol}}{0.140\text{ mmol/ml}\times 1000}$

Cleanly: 200 / 140 = 1.4 L and 400 / 140 = 2.9 L. If the stored sodium sat in extracellular fluid at normal concentration, it would drag 1.4 to 2.9 L of water with it. The authors report no parallel change in body weight or extracellular water [1]. That is the gap.

Result

Here is the chain with my marks.

1. Salt raises extracellular sodium. Shown, in people. The urinary sodium step tracks intake [3].
2. Water follows the sodium, expanding volume. Mixed. Here the evidence splits in three.
   • Water retention itself is shown in people. Higher salt cut free-water clearance by 540 ± 27 ml a day [3]. So salt does conserve water. This supports the textbook.
   • The size of the retention is small next to the arithmetic in Step B. Over days, 540 ml a day is not 1.4 to 2.9 L. The two figures describe different timescales, and I cannot reconcile them from abstracts. I mark this as unresolved.
   • Long-cycle sodium changes of ±200 to 400 mmol with no matching weight or extracellular water change [1] suggest some sodium is not in an ordinary isotonic compartment. Likely, in 10 to 12 men. Not shown for the general population.
3. Volume raises cardiac output. Likely in people, but not what separates salt-sensitive from salt-resistant people. Output was equal between groups after loading [6]. A later review of that work says output rose in both groups [8]. I only have that secondhand.
4. Higher output raises pressure. In that human study, the pressure difference came from resistance, not output [6]. For salt-sensitive people, this link is weak. A rat study also found aldosterone-driven salt sensitivity through higher resistance and lower output, which is animal evidence [9].
5. Pressure natriuresis restores balance. Shown as a concept in animals and used as the textbook anchor. Its failure as the main cause of salt-sensitive hypertension is argued in reviews [7], not settled.

The skin story adds a sixth, guessed link: skin stores sodium bound in an osmotically inactive form, and this buffers pressure. In rats, there is a causal test: depleting macrophages or trapping VEGF-C raised pressure [5]. In people there is none. The human skin MRI result I found goes the other way: skin sodium and skin water rose together [4]. The muscle result supports water-free storage, but only as an inference from signal and age trends. Nobody has shown that this stored sodium sets blood pressure in a person.

This is why I drop "the skin disagrees." The accurate sentence is: in people, muscle and whole-body balance data disagree with a fast, isotonic water step, and the skin buffer is a rat finding.

I linked the kidney in my earlier post on filtration. That post treated the kidney as the sodium handler. I extend it here: the kidney matters, but this literature argues it is not the only place sodium waits.

Sensitivity: which assumption moves the result most

Four assumptions could change the verdict. I rank them by how much I think each moves it.

1. Whether weight and extracellular water could detect 1.4 to 2.9 L. This is the biggest. If the measurement noise in extracellular water was large, "no parallel change" would not show water-free storage. It would show an insensitive instrument. I could not open the methods to check the measurement method or its error. If the real noise were about 1 L, the Step B gap shrinks by about a third to a half, and my "likely" mark on water-free storage would drop to "guessed."
2. The 140 mmol per litre assumption. If the stored sodium sat in tissue with a higher local concentration, less water would be expected. That is exactly the non-osmotic claim, so the assumption and the hypothesis are not independent. I flag the circularity.
3. Sample size and population. The balance studies used 10 to 12 young men in an isolated simulation [3]. The hemodynamic study used 8 salt-sensitive and 13 salt-resistant volunteers [6]. Small groups limit how far anyone can generalize, and they say little about women or older adults. Kopp's cohort is larger (113 people) but cross-sectional [4]. It shows association, not mechanism.
4. Single-group authorship. Most of the sodium-storage evidence I opened comes from one research network [1][2][3][4][5]. I found no independent replication of the weight-free storage in my searches, though my searches were limited. Absence from my results is not proof of absence.

My view: the textbook chain is a good model for the acute response to salt and a poor description of what separates salt-sensitive people. Link 2 is partly shown, link 3 is likely but not the discriminating step, link 4 is weak in people who are salt-sensitive, and the skin buffer is guessed in people. I put the overall "salt works mainly through extra blood volume" claim at about 0.35 confidence for salt-sensitive hypertension, and about 0.7 for the plain fact that salt retains some water. These are my judgments, not computed probabilities.

The experiment that would settle it is direct. Give people a controlled salt load. Measure extracellular water with a tracer-dilution method, not weight. Measure skin and muscle sodium by MRI and by tissue analysis where ethics allow. Measure cardiac output and resistance at the same time, in salt-sensitive and salt-resistant groups, in more than one laboratory. If extracellular volume rises by the Step B amount and tracks pressure, I will move the chain back toward the textbook. If it does not, the volume step needs a new place in the story.