Hyponatraemia at the finish line
A 24-year-old woman runs a marathon on a hot, humid day (31 °C, 70% humidity). She weighs 70.0 kg before the start and 71.8 kg at the finish after 5 hours, having drunk about 6 L of water and eaten very little. She took ibuprofen for knee pain at km 15 and felt nauseated from km 30.
In the medical tent she is confused and vomits. Plasma: Na⁺ 124 mmol/L, osmolality 258 mOsm/kg, glucose 5.2 mmol/L, urea normal. A urine sample taken at the finish: osmolality 62 mOsm/kg, Na⁺ 18 mmol/L. A second sample taken 40 minutes later, after two episodes of vomiting: osmolality 480 mOsm/kg. She is not clinically dehydrated - mucous membranes moist, no tachycardia.
Race records show that 3 of 1,400 finishers were treated for the same problem. The organisers are considering three options: weighing runners before and after the race, advising runners to "drink to thirst", or providing salt tablets at aid stations.
1. Calculate what her plasma Na⁺ would be from water retention alone, and show that this does not explain 124 mmol/L. What is the missing factor?
Read the worked answer
Water retention alone. Total body water ≈ 0.6 × 70 kg = 42 L; starting [Na⁺] taken as 140 mmol/L. Total exchangeable Na⁺ ≈ 42 × 140 = 5,880 mmol. Retaining the 1.8 L represented by the weight gain, with no sodium: 5,880 / 43.8 = 134 mmol/L.
So dilution explains a fall of about 6 mmol/L, not 16.
The missing factor: sodium loss. A plasma of 124 mmol/L in 43.8 L of body water requires 124 × 43.8 = 5,431 mmol - about 450 mmol Na⁺ lost. At a sweat sodium of 40–70 mmol/L, 6 L of sweat removes 240–420 mmol; the remainder comes from unreplaced dietary sodium and obligate urinary loss.
Conclusion: this is combined dilution and sodium depletion, not water alone. Two consequences follow: water restriction alone is an incomplete explanation of the pathophysiology, and the 1.8 kg gain is the objective evidence of positive water balance. The history of intake is unreliable - runners consistently under-report how much they drink - which is precisely why a scale is better than a questionnaire.
2. Her urine at the finish is 62 mOsm/kg with a plasma of 258. Explain whether this is an appropriate or inappropriate ADH response, why she still became hyponatraemic, and what the 480 mOsm/kg sample 40 minutes later shows.
Read the worked answer
The 62 mOsm/kg sample is appropriate. With plasma hypo-osmolality, osmoreceptors should switch ADH off and the kidney should produce maximally dilute urine (~50–100 mOsm/kg). That is exactly what happened. The textbook "SIADH" explanation does not apply at the finish line, and reporting it as inappropriate ADH secretion would be an error.
Why she is still hyponatraemic. The kidney's maximal free-water clearance is about 0.7–1.0 L/h (roughly 15–20 L/day). She drank 6 L in 5 hours = 1.2 L/h, which exceeds that ceiling. Even a perfectly regulated kidney cannot excrete water faster than it arrives, so water is retained and distributed through 42 L of body water. The limit, not the hormone, was the problem.
The 480 mOsm/kg sample. Nausea and vomiting are potent non-osmotic stimuli for ADH release. Once ADH is high, the same fluid intake is retained almost completely and plasma Na⁺ can fall rapidly. This explains why she deteriorated after km 30 (nausea), rather than steadily through the race, and why an antiemetic and stopping fluid intake are treatment rather than comfort measures.
Practical point: a urine osmolality must be interpreted together with the timing of symptoms. One dilute sample does not exclude the syndrome, and one concentrated sample does not prove SIADH.
3. Evaluate the three prevention options. Which is best supported, what are the limitations of each, and what study design would the organisers need before changing policy?
Read the worked answer
Weighing runners. Cheap, objective, and the only field measure of water balance: change in body mass. Gains above roughly 2–3% of starting mass are associated with hyponatraemia. A practical protocol is weighing at registration, printing the mass on the bib, weighing at the finish, and flagging anyone above starting mass who has symptoms. Limitations: it does not measure sodium loss, and a single pair of weigh-ins misses the trajectory during the race.
"Drink to thirst". Best supported by the evidence: ad libitum drinking is regulated by thirst and osmolality, and case series identify scheduled over-drinking - not under-drinking - as the main risk factor. Limitations: thirst is blunted by nausea, NSAIDs and heat, and many runners follow fixed-volume advice from coaches or other runners.
Salt tablets. No good evidence for prevention. A sodium load can raise plasma Na⁺, but it is not a licence to drink more, and tablets commonly cause gastrointestinal upset that itself stimulates ADH.
Study design. The 3-in-1,400 figure is a case series with no denominator for fluid intake; it cannot estimate dose–response. The organisers should run a prospective cohort: pre- and post-race weights plus recorded intake on all finishers, with plasma Na⁺ measured in symptomatic runners, which would let them link weight gain and intake rate to hyponatraemia before mandating a policy.
Treatment note: symptomatic hyponatraemia with confusion is treated with hypertonic (3%) saline, not with more water; correction should not exceed 8–10 mmol/L per 24 h to avoid osmotic demyelination. 0.9% saline is only relatively hypertonic against a plasma of 124 mmol/L and acts slowly.