A Current Opinion paper published this month in Sports Medicine asks a question that fifty years of hydration research still can’t answer: what happens to an athlete who spends a decade training slightly under-hydrated?
We know an enormous amount about acute dehydration. We know what a 2% body-mass deficit does to your 10K split, your core temperature, your perceived effort. What we don’t know — and the authors, Ruben Francisco and Lawrence Armstrong of UConn’s Korey Stringer Institute, say this plainly — is whether habitually running a little dry carries any cost beyond today’s workout. Their words: “Whether this physiological state poses long-term health risks for athletes remains unknown.”
That’s an unsatisfying answer. It’s also the honest one, and it’s worth understanding why the question is suddenly being asked at all.
Most athletes drink less than you’d think
The paper builds on a study of 68 elite athletes who kept seven-day food and fluid diaries. Roughly 58% qualified as “low drinkers” — under 35 mL of fluid per kilogram of body weight per day. For a 70 kg runner, that’s less than about 2.5 litres of total daily fluid, everything included.
Here’s the part that makes it interesting. Those low drinkers had measurably more concentrated urine — specific gravity around 1.024 versus roughly 1.015–1.018 in the higher drinkers. But when researchers measured their actual total body water using isotope dilution and DXA, there was no difference. Not in total body water, not in intracellular or extracellular fluid.
In other words: their bodies were fully compensating. Which raises the obvious question — compensating with what?
The hormone doing the quiet work
The answer is vasopressin, your antidiuretic hormone, and the evidence here is much older and much stronger than the athlete data.
A well-replicated 2013 study in the British Journal of Nutrition compared free-living adults drinking about 1.2 litres a day against those drinking 2–4 litres. Blood osmolality was identical between groups — standard blood work would have called both normal. But the low drinkers produced 1.0 L of urine daily versus 2.4 L, at more than double the concentration (767 vs 371 mOsm/kg), with significantly higher circulating vasopressin, cortisol and creatinine.
So the low drinkers weren’t dehydrated in any way a blood panel would flag. They were simply holding the line with more hormonal effort, every day, indefinitely.
Whether that sustained effort matters over decades is the open question. In large population cohorts, high copeptin — a stable marker of vasopressin activity — does predict later type 2 diabetes, with roughly a twofold risk across the top quartile in one 4,700-person Swedish study. That’s an association in the general population, not a demonstrated consequence of under-drinking, and definitely not a finding in athletes.
What long races do to kidneys
The more acute story is better documented, and it surprises people.
When researchers tested 22 runners after the 2015 Hartford Marathon, 82% met the creatinine criteria for stage 1 or 2 acute kidney injury, and 73% showed evidence of tubular injury on urine microscopy. Multi-stage ultramarathon studies report AKI incidence anywhere from 41% to 85% depending on the event and the criteria used.
Before anyone cancels their fall marathon: these markers resolve. In the Hartford cohort, injury and repair biomarkers peaked at day one and had normalised by day two. A 2026 mini-review in the Journal of Applied Physiology describes post-race kidney stress as “generally transient” and resolving without treatment — while closing with a call for longitudinal studies to find out whether repeated episodes add up to anything. Nobody has run that study.
What we do have is a hint that hydration status modifies the hit. At the 2023 Western States 100, only 8 of 35 runners finished euhydrated — and those who did showed blunted rises in kidney injury biomarkers compared with the rest.
The uncomfortable comparison
The reason researchers take this seriously at all is an occupational analogue: chronic kidney disease of non-traditional origin among heat-exposed outdoor workers in Central America. In one cohort screened to exclude pre-existing diabetes, hypertension and kidney disease, age-standardised prevalence reached 14.1% among Salvadoran sugarcane workers. Among 326 Nicaraguan cane workers with normal kidney function before harvest, 34 developed acute kidney injury during a single season — and a year later, 10 had newly reduced kidney function.
This is not your Saturday long run. These are six-plus-hour days of cane cutting in extreme heat across months-long harvests, often without reliable water access, and researchers still debate how much agrochemicals, NSAIDs and other exposures contribute. The relevance is narrower than headlines suggest: it establishes that repeated dehydration-associated kidney injury can accumulate in humans, at a sufficiently extreme dose. It says nothing about where the threshold sits for a recreational athlete.
One detail from that literature does travel, though. Workers appear less likely to develop kidney injury when sweat losses are replaced with sodium-containing fluids rather than plain water — a reminder that rehydration is a sodium problem as much as a water problem.
Where sugar enters the picture
One controlled human crossover deserves mention, with a caveat attached. Twelve adults exercised four hours in 35°C heat, rehydrating with either water or a high-fructose caffeinated soft drink. Stage 1 acute kidney injury appeared in 75% of the soft-drink trial versus 8% of the water trial, alongside larger rises in uric acid and copeptin.
The caveat matters: the test beverage was a soda, roughly 11% fructose with caffeine. A conventional sports drink is about 6% carbohydrate and typically glucose- or maltodextrin-based. This study does not show that Gatorade or Powerade harms kidneys, and nobody should read it that way.
It’s a reasonable argument for matching your sugar load to your actual workload, though. For sessions where you need carbohydrate, you need carbohydrate — that’s what Skratch Labs and classic sports drinks are for. For hot, long, low-intensity days where the problem is sodium and fluid rather than fuel, sugar-free options like LMNT, Nuun’s zero-sugar tablets, or a sodium-dense pickle brine shot such as Fast Pickle cover the electrolyte side without adding a fructose load you didn’t need.
What the evidence does not say
It would be easy to turn all of this into “drink more water, avoid disease.” The trial data won’t support it.
The CKD WIT trial randomised 631 adults with stage 3 kidney disease to hydration coaching for a year. Urine output rose, copeptin fell — and kidney function declined at exactly the same rate in both groups. A 2024 systematic review in JAMA Network Open found 18 randomised trials of increased water intake in total; eight were entirely negative, and the only replicated benefits were modest weight loss and fewer kidney stones. Armstrong’s own 2024 review of 96 studies rated the evidence strong for kidney stones and hyperglycaemia, and insufficient or conflicting for seven other conditions.
And more is not automatically better. Over-drinking during endurance events causes exercise-associated hyponatremia, which is a genuine emergency and has killed runners. The goal is adequacy, not volume.
The practical read
The actionable takeaway from this research isn’t about race day. It’s about the other 23 hours.
Most athletes plan hydration around sessions and then drift through the rest of the day. If you’re consistently producing small volumes of dark, concentrated urine outside of training, that’s the pattern this literature is describing — and the fix is boring: spread fluid across the day rather than backloading it around workouts.
For reference, population-level adequate intakes for total water are about 3.7 L/day for men and 2.7 L/day for women, roughly 20% of which comes from food. Those are medians for a population, not prescriptions for an individual — your sweat rate, climate and training load move the number substantially. Around sessions, the established ACSM guidance still holds: roughly 5–7 mL/kg about four hours beforehand, drink enough during to keep body-mass loss under 2%, and replace around 1.5 L per kilogram lost afterward, with sodium alongside it.
As for whether any of this protects a 45-year-old marathoner’s kidneys twenty years from now — that study hasn’t been done. The most useful thing the Sports Medicine paper does is say so out loud.