Friday, 9 October 2015

Sweat: sweet is better than salty for marathon runners

Thinking back to some lectures I attended, probably about 20 years ago now, I came across another good reason why marathon runners might want to 'over' heat acclimate (i.e. train in more clothing than is comfortable so that they can produce large amounts of sweat).

Heat acclimation has been long been known to improve sweating rates, produce bigger sweat glands as well as increase plasma volume. The end result of heat acclimation includes an improvement in aerobic performance whether measured by VO2max or direct race performance. So far I have generally been concerned with the importance of the expanded plasma volume and its importance during a marathon. A large plasma volume will tend to keep heart rates relatively low. During a marathon it is also well known that there is fluid loss (sweating and breathing) causing a decline in plasma volume resulting in a rise in heart rate. This rise in heart rate may eventually limit the aerobic capacity of a runner causing them to slow - the full explanation for this is rather complicated since it involves understanding the importance of venous return to the heart. But, the key point to understand here is that training can dramatically alter the rate at which plasma volume declines even if it does NOT alter (or reduce) the sweating rate. That is, a heat acclimated runner and one that is not heat acclimated may appear to sweat sufficiently and at the same rate - losing the same amount of weight over the course of the marathon - but one will show a much larger decline in plasma volume than the other. The non-heat acclimated runner is much more likely to exhibit a typical bonk or Wall-like slow down because of the loss of plasma volume. So, what is going on here and why does heat acclimation that may result in slightly more sweat loss, during a marathon, result in the protection of plasma volume?

Kirby & Convertino (1986) looked at sodium loss (amongst other things) during exercise following a 10-day period of heat acclimitization. They found that heat acclimation reduced sodium loss by ~50% (as well as increasing the amount of sweat produced). The question can thus be refined: How does maintaining sodium levels also help maintain plasma volume? To understand this - and how the sodium gives us access to 'free' internal water - we need to look at where water is in the human body (see Figure 1).
Figure 1. Water distribution in the human. Only about 7% of the water is in the blood (plasma). Two thirds of the water in a typical human is inside cells, the rest (about one quarter of the total) is fluid around those cells.
The plasma (or blood) contains a relatively small amount (7% or about 5.5 L) of the total water inside an average person. Most of the water in the human body is inside cells. During a marathon, if you don't drink (and some people do run marathons without drinking) you might lose about 5 L of water - clearly it cannot all come from the blood otherwise there would be nothing left to circulate around the body. So, where does the water come from? The interstitial space is one possible source. Individuals who suffer from a blood haemorrhage (bleeding) certainly do auto transfuse water from the extracellular fluid compartment to the plasma but that is driven by a drop in blood pressure - this is not what happens, or what you want to happen when running a marathon. To get water out of the cells and into the plasma an osmotic gradient is required, and this is where sodium is critical. If sodium levels are maintained in the plasma whilst its volume decreases (because of a dilute sweat production) then the osmolarity of the plasma rises. This rise in osmolarity drives water from the intracellular space into the blood plasma. Under such conditions 5 L of water flow is not a major problem - it is a loss that can easily be tolerated by most of the cells within the body. Or, to put it another way - if you have got large sweat glands capable of producing a dilute sweat you can effectively distribute the water loss between all of the compartments of your body so the 'hit' to the blood plasma is minimized. If you have poorly trained sweat glands then most of the water loss will be from the plasma and interstitial space with an associated loss of blood pressure and increase in heart rate.

So, training wearing clothes that cause you to sweat a lot induces a hypertrophy (growth or training) of the sweat glands. This growth means that the sweat glands can produce more sweat if necessary - but more importantly they retain sodium. By retaining the sodium the effect of sweating is to make the blood plasma more concentrated which drives water out of the large fluid compartment that you are carrying around. The result is you become gradually LIGHTER which makes you more efficient, your blood plasma volume is better maintained so you are less likely to bonk/hit the wall and you may not need to drink at all during a marathon.

This is just one benefit of training HOT. If you are out running feeling thermally comfortable then you are probably missing one important physiological adaptation to homeostasis.


  1. Thanks Christof! I always indulge in a bit of hyperthermal training before VLM -- as witnessed by Katie & Darren on the busway this year. Very interesting point overall about the heat-trained losing water from cells they can afford to lose it from, rather than the blood they can't, simply because they sweat less sodium and so maintain the molality of the bloody plasma better.
    I imagine that in cool weather, with low sweating rates, one doesn't really lose much overall because of the water released by burning glycogen: partly that which was bound up with the stuff iin the first place (like burning wet petrol), and partly that generated by oxidising all those hydrogen atoms. In fact one may need to wee during long runs in cold weather to lose the surplus water -- yet you'll certainly end up lighter, which shouldn't really be counted as "dedhydration", something I feel a lot of people writing about the weight loss of marathon runners neglect.
    Thanks again!
    Charlie Wartnaby

  2. Yes, I think a well trained lean male might have about 0.5kg of glycogen within his legs - and about 1.5kg of water associated with it. So, a 'hard' run that depletes all of the glycogen might liberate just over 1.8 L of water (MW glucose 180, MW H2O 18 with 6 moles of water produced for each mole of glycogen). It is true that burning fat also liberates some more water...But, getting hold of any of this water still means extracting it from the cells and osmosis is just about the only way. You are right that around a 2kg weight loss during a marathon could be sustained with no loss of circulating plasma volume. A 5kg loss - which seems to be around what the faster runners achieve - could require only a very modest 3 L from the very large intracellular store. 5 L is, interestingly, around what a 60 kg runner might need to evaporate to carry away the metabolic heat generated by running 42km (1kcal per kg per km and the latent heat of evap 540 kcal per kg). This would suggest the elites consume very little water when running.

  3. Christof, whilst Kirby and Convertino (1986) reported acclimation over a 10d period, if I understand correctly you advocate training hot over a much longer period. Have there been studies over a longer period, and are the adaptations significantly greater than those reported over 10d? I very much enjoyed your talks in Cambridge last week on the subject and will try your methods for a marathon this spring. Might it be possible to repeat here the slides from those talks?

  4. Dear Andy, heat acclimation appears to induce a range of adaptations which have different time-courses. The fastest response is the plasma volume expansion which occurs as a result of exercise and is enhanced in warm conditions. Most studies tend to stop at around 3 weeks (e.g. since the effects are clearly visible relatively early. The hypertrophy of sweat glands is also quite fast although it is not clear how long it takes for them to reach their peak size. The effect on plasma volume on red blood cell numbers is, however, going to take longer since it relies on the turn-over of red blood cells. They have a life time of around a month and so a more prolonged period of increased plasma volume will be necessary to make more blood (rather than just plasma). Then there is the effect on stroke volume, heart rate and the remodelling of your circulatory system. Whilst there is an immediate effect of plasma volume expansion on lowering heart rates - which I think is due to increased venous pressure - the remodelling of the myocardium to 'normalize' that increased internal volume will take a considerable amount of time. If you start with a little heart, it is unreasonable to expect that simply stretching it will bring all of the benefits of a big heart. The stretch needs to be associated with the same hypertrophy that occurs in skeletal muscle - except it is a growth of the heart muscle. That takes time - possibly years. The problem I have with this is that there is very little scientific literature on the time course of this process. The remodelling occurs, of course, without the thermal load needing to be increased - but, increasing it 'should' amplify the effect.

  5. Really informative. This explains a lot - here is a non-competitive run I did last August, on one of the hottest days of the year, you can see on the pace graph where I "bonked" around 21km -
    I always tend to struggle with hot weather, so I need to acclimatise more by training hot!