Exercise-associated hyponatremia

Why can a marathon runner who drinks steadily at every station collapse from having too much water rather than too little?

A runner takes a cup at every station, exactly as instructed, and finishes confused and unsteady. The medical tent's first instinct for decades was dehydration, and the obvious treatment was more fluid. In the worst cases that treatment killed people.

The strange part is not that overdrinking is possible. It is that the body has an apparatus for shedding surplus water — kidneys that can pass litres a day, and a thirst that quietly stops asking. How does a runner end up dangerously waterlogged while both are working?

Start by being precise about what has gone wrong. The measured abnormality is not the volume of water in the body but the concentration of sodium in the blood, normally somewhere in the range of 135 to 145 millimoles per litre. Below 135 is hyponatremia, and the serious neurological trouble generally appears well below that.

A concentration is a ratio, so it can fall two ways: less sodium on top, or more water underneath. Which is it?

Take the sodium side first, since it is the intuitive one — runners sweat, and sweat is salty. But how salty? Sweat sodium varies a great deal between people and with heat acclimatisation, and typical figures sit well under the concentration in blood. That comparison settles a lot: sweat is dilute relative to plasma, so it removes proportionally more water than sodium, and sweating on its own therefore drives plasma sodium up, not down. Salt loss can deepen the problem once it exists, but cannot start it.

So the water side must be doing the work. Now the harder question: the kidney can excrete a large volume of dilute urine when plasma is over-diluted — several hundred millilitres an hour comfortably, and by some accounts considerably more, though I would not lean on a precise figure. Drinking a cup every couple of kilometres should not defeat that. Why does it?

Because the kidney's willingness to dump water is under hormonal control, and that control is not obeying osmolality alone. Vasopressin — the hormone instructing the kidney to hold water back — is released not only when plasma is concentrated but in response to strenuous exertion, pain, nausea, and reduced circulating volume, all ordinary features of the closing hours of a marathon. So exactly when the runner most needs to shed surplus water, the signal to retain it is switched on for reasons unrelated to concentration. Certain anti-inflammatory drugs, common among runners, push the same way.

That gives us two conditions, and both are necessary. A water surplus must be taken in, and the escape route for that surplus must be blocked. A runner with either alone is fine.

Now the consequence, which explains why this is lethal. Cell membranes let water pass, so diluting the fluid outside cells drives water into them. Most tissues tolerate a little swelling. The brain, enclosed in a rigid skull, does not — and that is the mechanism behind the confusion, the seizures, and the deaths.

Which finally explains the inversion in treatment. Fluid is not the remedy but the cause, and the correct emergency response is concentrated salt solution, which pulls water back out of the brain.

Three corrections, since the popular version of this is almost entirely wrong.

The first concerns salt tablets. If sodium loss were the driver, supplementing salt would prevent the condition. It largely does not, because sweat is dilute and the runner in trouble carries a water excess that salt does not remove. Taking salt while continuing to overdrink treats the wrong term of the ratio.

The second concerns weight. Losing some body mass over a long race is the normal outcome of sweating. The runner who gains weight has taken in more than they lost — and weight gain, not weight loss, is what tracks with low sodium. The study of Boston Marathon finishers published in 2005 is the usual reference point, reporting hyponatremia in a meaningful minority of those tested and its association with weight gain; I state that from recall and would not quote its exact percentage without checking.

That gives the account its falsification test, and it is a clean one. If exercise-associated hyponatremia were driven by sodium loss, the affected runners should be those who lost the most weight, and salt supplementation without a change in drinking should prevent it. Finding either would refute the dilution account outright. The observed pattern runs the other way, which is why the guidance changed.

And the third concerns the advice itself. Runners were told for years to drink ahead of thirst, on the theory that thirst lags behind need. This is where the piece in this collection on the urge to breathe is a contrast rather than a repetition: there the body monitors a proxy — carbon dioxide rather than oxygen — and the proxy turns out to be well chosen. Here the regulator directly measures the thing that matters, plasma concentration, and the problem is that athletes were instructed to override it. Drinking to thirst is now the mainstream recommendation, though how much the older advice was shaped by commercial interest remains a contested and ill-tempered corner of sports science.

One limit on the account as given: it explains dilution and retention well, but the fate of the body's exchangeable sodium during long exercise — how much is temporarily taken out of circulation into bone and other stores — is still argued, and that argument matters for the details of treatment rather than for the basic picture.

• Why thirst turns out to be a better regulator during exercise than a drinking schedule, and where it genuinely does lag.
• Why correcting low sodium too quickly carries its own serious neurological risk, so that the pace of treatment matters as much as its direction.

Every term the collection defines is gathered in the glossary.