The shelf already has an answer ready: electrolyte tabs, isotonic drinks, and powders loaded with sodium, potassium, and magnesium. The message behind all of them is the same: water alone won't cut it once you start sweating.
The research tells a different story. There is a point where fluid and sodium genuinely matter, but it sits further out than the marketing suggests. In a meta-analysis on cycling, drinking during a hard one-hour effort actually made performance slightly worse.7
That's what this article covers: what electrolytes actually do during training, at what point water stops being enough, how much you should sensibly drink, and why the most famous electrolyte story of all, the calf cramp in the final stretch, has the weakest science behind it.
- Starting a session already dehydrated costs you 2.4 % endurance performance on average.4
- Drinking isn't a blanket win: during a hard one-hour ride it cost 2.5 % performance, while over two hours it added 3.2 %.7
- Overdrinking is the bigger risk. At the 2002 Boston Marathon, 13 % of finishers tested had sodium levels that were too low.12
Contents
What electrolytes actually do in your body during training
Electrolytes are dissolved minerals that carry an electric charge in the body. Sodium is by far the most abundant one in sweat, and it governs how much water your body holds outside its cells. That's why it sits at the centre of every serious hydration discussion, while potassium and magnesium play a smaller role, with their daily requirements covered in detail in Magnesium and Muscle Function. The ACSM position stand accordingly names fluid loss above 2 % of body weight as an orientation point, not a fixed drinking target.1
It's not the amount in your sweat that decides the outcome, it's the ratio. When you sweat, you lose water and salt at the same time, but not in the same ratio as in your blood. Sweat is dilute compared with plasma. So if you sweat heavily and only top up with water, you dilute your blood a little further over time.
How large this effect turns out to be varies hugely from person to person. The American College of Sports Medicine position stand explicitly states that sweat rate and the electrolyte content of sweat vary considerably between individuals, and concludes that no single blanket recommendation makes sense, only individually tailored drinking plans.1
Sawka and colleagues set the goal, on behalf of the American College of Sports Medicine, of avoiding a fluid loss of more than 2 % of body weight during exercise as well as large shifts in electrolyte balance. Because of the wide variation between individuals, they recommend determining your own sweat rate by weighing yourself before and after a session rather than relying on fixed drinking volumes.1
The second building block is thirst. It doesn't respond to fluid loss itself, but to the rising concentration in your blood. That's why thirst works reasonably well as a control signal, as long as you eat enough salt and aren't facing extreme conditions. And that's exactly why it fails in two situations: very long efforts in heat, and when someone drinks to a schedule instead of to thirst.
Sodium is the most abundant electrolyte in sweat and controls fluid volume outside the cells. Because sweat rate and sweat salt content vary so much between individuals, the ACSM position stand recommends individually tailored drinking strategies rather than blanket volume targets, using fluid loss above 2 % of body weight purely as an orientation point.1
Does water suffice, or do you need electrolytes?
For the vast majority of sessions, water is enough. Fluid status only becomes relevant once you start a session already dehydrated: a meta-analysis of 15 controlled trials with 186 subjects found a 2.4 % drop in endurance performance and a 4.4 % drop in oxygen uptake at the lactate threshold, at an average body weight loss of 3.6 %.4
Two details in this chart deserve a second look. First, the losses are small: two to four percent shows up in a race, not in an easy weeknight session. Second, this picture covers a dehydrated start, not dehydrating during the session itself. The trials ran mostly up to one hour, at temperatures between 19 and 40 degrees.4
The muscles respond too, although less consistently. A meta-analysis of 28 studies found an 8.3 % drop in muscular endurance and a 5.5 % drop in maximal strength under dehydration. Anaerobic capacity and jump height, on the other hand, showed no statistically meaningful change.6 If you're worried about your strength session the moment your shirt gets damp, you're overestimating the effect considerably.
Savoie and colleagues analysed 28 studies on muscular performance under dehydration. Muscular endurance (−8.3 %), maximal strength (−5.5 %), and anaerobic power (−5.8 %) dropped measurably, while anaerobic capacity (−3.5 %) and jump height (+0.9 %) did not. Notably, there was no relationship between the degree of dehydration and the degree of performance loss, and trained individuals tended to lose less than untrained ones.6
That puts the famous 2 percent mark into context. It comes from the 2007 position stand and is meant as a safety margin, not a cliff edge beyond which performance collapses.1 A more recent systematic review with meta-analysis has run the numbers on exactly that: perceived exertion rose by 0.21 points on the Borg scale per percent of body weight lost, and the authors conclude the effect is practically negligible before a loss of around 3 %.5
In short: water is enough as long as you're training under about an hour, not in serious heat, and you start well hydrated. That's the rule, not the exception.
A dehydrated start costs 2.4 % endurance performance in a meta-analysis of 186 subjects at an average 3.6 % body weight loss (Deshayes et al. 2020). A second review from the same group shows perceived exertion rises by only 0.21 points per percent of weight lost and only becomes practically meaningful past around 3 %. The 2 % threshold is set conservatively.4,5
How much to drink during training, and when does sodium matter?
This is where the most interesting finding of the whole topic sits. A meta-analysis of nine cycling studies broke the effect of drinking down by exercise duration and found a sign that contradicts sports-drink marketing: during a high-intensity hour, drinking cost 2.5 % performance rather than helping it.7
The top bar is the uncomfortable one. During a one-hour effort at 80 % of maximal oxygen uptake, performance with drinking sat below performance without it, and the confidence interval stays entirely on the negative side. You're probably paying more for the fluid sitting in your stomach during that short, hard effort than you gain from the fluid itself. Past one hour the picture flips, and over two hours the advantage sits at 3.2 %.7 For how to time carbohydrates around long sessions like these, see Carbs Around Training.
For practical purposes, the same paper7 gives concrete rates: 0.15 to 0.20 ml per kilogram of body weight per minute for sessions of one to two hours, and from two hours onward either to thirst or 0.14 to 0.27 ml per kilogram per minute. For a 75 kg athlete, that works out to roughly 700 to 900 ml per hour in the middle of that range. It's worth noting the authors' own wording: both drinking to thirst and drinking to a plan are described as appropriate recommendations.
| Situation | What makes sense |
|---|---|
| Strength training, under 60 minutes | Water to thirst, no electrolyte supplement needed |
| Hard hour, high intensity | start well hydrated, drink cautiously during the session |
| One to two hours, moderate | drink regularly, around 0.15 to 0.20 ml/kg/min |
| Over two hours or serious heat | to thirst or 0.14 to 0.27 ml/kg/min, sodium starts to matter too |
| Multiple sessions per day | actively replace losses between sessions, eat enough salt |
Practical overview. Drinking rates from the meta-analysis by Holland and colleagues, framework and individualisation from the ACSM and NATA position stands.1,3,7
And what about sodium? Honesty is warranted here, because the evidence is thinner than for water. A controlled trial with six cyclists over four hours showed that a high sodium intake genuinely preserves extracellular fluid volume better. It had no effect on cardiovascular drift or temperature regulation, though.9
Sanders and colleagues had six cyclists ride for four hours each at 55 % of maximal oxygen uptake, giving them 3.85 L of a carbohydrate solution with three different sodium concentrations. In the low-sodium condition, extracellular volume shrank by about 1.3 L; in the high-sodium condition it grew by around 0.8 L. The corresponding shifts in plasma volume changed neither cardiovascular drift nor thermoregulation.9
An older, often overlooked trial with eight participants goes even further. Over six hours at 30 degrees, it compared water, a mild saline solution, and no fluid at all. Water and the saline solution showed no meaningful difference, whether in heart rate, body temperature, or blood sodium. Without fluid, though, participants stopped roughly ninety minutes earlier on average, after a 6.4 % weight loss.10
The modern, personalised approach only partly holds up under scrutiny either. In a double-blind crossover trial, nine runners exactly replaced their previously measured sodium losses over five hours at 30 degrees. Blood sodium rose more sharply as a result, but total water balance, heart rate, core temperature, and perceived exertion all stayed unchanged.11
The composition of the drink isn't irrelevant, though. A meta-analysis of 28 studies8 compared how well different drinks preserve plasma volume during exercise and found hypotonic carbohydrate-electrolyte drinks came out ahead: minus 6.3 % versus minus 8.7 % for isotonic drinks. The classic isotonic sports drink performs worse in this comparison than the hypotonic version.
Two caveats apply. The paper uses Bayesian 90 % compatibility limits, so the numbers aren't directly comparable with classic confidence intervals. And what was measured was plasma volume, not performance.8
The benefit of drinking during exercise depends on duration: in a meta-analysis of nine cycling studies, drinking worsened performance by 2.5 % over a high-intensity hour but improved it by 3.2 % over two hours (Holland et al. 2017). For sodium during exercise, the evidence is thinner: controlled trials show shifts in fluid balance without a clear effect on performance or temperature.7,9,11
Can you drink too much?
Yes, and it isn't a theoretical worry. At the 2002 Boston Marathon, 488 finishers were tested: 13 % had a sodium level of 135 mmol per litre or below, and 0.6 % were at a critical 120 mmol per litre or lower.12 The heaviest risk factors weren't heat or drink choice, but a long finishing time and weight gain during the race.
Almond and colleagues examined blood samples from 488 participants of the 2002 Boston Marathon. In the multivariate analysis, the odds of having a low sodium level were higher by a factor of 4.2 (odds ratio, 95 % CI 2.2 to 8.2) with substantial weight gain during the race, and by a factor of 7.4 (95 % CI 2.9 to 23.1) for a finishing time over four hours compared with under 3:30.12
Weight gain during a race means you drank more than you sweated out. That's exactly what dilutes the sodium in your blood. A review of marathon runners puts the frequency of exercise-associated low sodium levels, across two large studies, at 7 to 15 %, symptomatic and asymptomatic cases combined.13
This has shifted the guidance landscape over the past two decades. The National Athletic Trainers' Association's 2017 position statement places both directions side by side as equally important: too little fluid and too much fluid both harm performance and health, and athletes should drink enough, but not excessively.3 The joint position of the Academy of Nutrition and Dietetics, Dietitians of Canada, and the ACSM places fluid intake within the broader framework of sports nutrition accordingly.2
If you finish a long race heavier than you started it, you drank too much. Headaches, nausea, or confusion after a long effort need medical assessment, not more water.
Overdrinking is a real risk: at the 2002 Boston Marathon, 13 % of 488 finishers tested had sodium levels of 135 mmol/L or below, with weight gain during the race (odds ratio 4.2) and long finishing times (odds ratio 7.4) as the strongest risk factors (Almond et al. 2005). Current position stands accordingly warn against overhydration just as much as underhydration.3,12
The cramp myth and what's really going on
No argument sells electrolytes as reliably as the calf cramp. This is exactly where the evidence is thinnest: a review notes the electrolyte and dehydration hypothesis rests essentially on case reports totalling 18 cases and one small case-control study of ten people, while four prospective cohort studies don't support it.15
The competing explanation starts somewhere else entirely. An analysis of 69 papers concludes that exercise-associated cramps more likely arise from disrupted neuromuscular control: excitation from the muscle spindles rises, inhibition from the Golgi tendon organs falls, and the resulting action potentials originate in the spinal cord, not in the fatigued muscle itself.14
To be fair, this doesn't disprove the electrolyte explanation, it's just considerably less well supported than the alternative. If you reach for a salt tablet at every cramp, you're probably treating the wrong problem, since the more promising levers are load management and fatigue management. Frequent or very painful cramps are worth a medical check-up. If you regularly train past two hours, the trade-off between long endurance sessions and strength training is covered in Cardio and Muscle Growth.
For exercise-associated muscle cramps, the electrolyte and dehydration hypothesis is weakly supported: it rests on case reports covering 18 cases and one case-control study of ten people, while four prospective cohort studies argue against it (Schwellnus 2009). An analysis of 69 papers instead points to disrupted neuromuscular control originating in the spinal cord (Giuriato et al. 2018).14,15
The Bottom Line
For sessions under an hour, water is the right answer, and the 2 % threshold is a conservative safety margin, not a cliff edge. Things get serious in the other direction: long efforts in the heat, where sodium starts to matter, and overdrinking, where good intentions can turn into a medical problem. Measure your sweat rate once, and you won't need a rule of thumb again.
FAQ: Isotonic Drinks, Tablets, and Everyday Training
Worth it, yes, but later and less often than advertised. For sessions under an hour they add nothing measurable over water, and in the meta-analysis by Holland and colleagues (2017), drinking during a high-intensity hour was actually slightly detrimental to performance, by 2.5 %. Past roughly an hour of moderate effort, that picture flips. Composition is interesting too: in a meta-analysis of 28 studies, hypotonic carbohydrate-electrolyte drinks, meaning drinks more dilute than blood, held plasma volume better than classic isotonic versions (minus 6.3 % versus minus 8.7 %). There are approved EU health claims stating that carbohydrate-electrolyte solutions enhance the absorption of water during physical exercise and contribute to the maintenance of endurance performance during prolonged endurance exercise, but they only apply to drinks with a defined composition and during sustained endurance exercise.
For a normal 45- to 75-minute strength session, there's no case for them. Sweat losses are moderate, the session is short, and an ordinary meal supplies many times the sodium you lose during it. The performance effect is limited too: in the meta-analysis by Savoie and colleagues, muscular endurance dropped by 8.3 % and maximal strength by 5.5 % under dehydration, while anaerobic capacity and jump height showed no statistically meaningful change. Notably, there was no relationship between the degree of dehydration and the degree of performance loss. If you still feel like you sweat heavily and saltily, the cheaper test is to weigh yourself before and after the session and replace the losses through normal food and drink.
Weigh yourself unclothed immediately before and after a typical session, then add whatever you drank. One kilogram of weight lost corresponds to roughly one litre of fluid. So if you're 0.5 kg lighter and drank 500 ml during the session, your loss was around one litre. This is exactly the approach the ACSM position stand recommends, because sweat rate and salt content vary so much between individuals that blanket drinking volumes aren't much use. The National Athletic Trainers' Association position stand takes the same approach and explicitly advises determining sweat rates under different conditions. Two or three measurements at different temperatures are enough for a useful picture, and after that you'll know for yourself what most guides can only estimate.
Recovery matters just as much after sweating: 24 g protein and 3 g leucine per serving, plant-based and without sweeteners. We deliberately do not sell electrolytes; this article is not selling you any.
Share: