Shownotes

Strength and speed training practices of world-class Norwegian coaches 

00:50 -

• My first study is called Strength and Speed Training Practices for Olympic Endurance Sports: Insights from World-Class Coaches. It is one publication within a larger project by a Norwegian group including Thomas Haugen, Espen Tønnessen and Øyvind Sandbakk.
• Earlier papers from the same project covered how Norwegian coaches in sports such as distance running, road cycling, triathlon, swimming, speed skating and cross-country skiing programme their training and what their principles look like.
• This one is specifically about how they use strength and speed training within the broader context of endurance training.
• The methodology is thorough.
• Twelve coaches across eight Olympic sports first answered questionnaires, then did structured interviews, then the researchers systematically summarised the findings and went through them again with the coaches, and finally everything was verified against the training logs of the coaches' most successful athletes. 
• Between them, these coaches have athletes who have won over 400 international medals at Olympic Games, World Championships and European Championships.
• There are two coaches each for cross-country skiing, biathlon, swimming, triathlon and running, and one each for speed skating, rowing and cycling.
• One coach is involved in both triathlon and swimming, and listeners will know who that is: Arild Tveiten, a friend of the podcast who has been on many times. 
• It gives a structured insight into what these coaches actually do, which is very different from searching the internet and piecing together a partial picture of Norwegian training from random sources. 
• The researchers themselves point out that it is a good way to generate hypotheses for future controlled studies.
• The findings should not be read as what is optimal, only as what is practised by this particular group of very successful coaches, and that then feeds the cycle where research often follows coaching practice rather than the other way around.
• Most of the coaches include strength and speed training to varying extents. The five main reasons they give are work economy, acceleration and finishing capacity (more important in some sports than others), injury prevention, bone health, and stopping the high volume of endurance training from eroding the athletes' maximal speed capacity.

Strength Training

05:10 - 

• In terms of volume, combining both strength and speed training, rowing does the most at 150 to 200 hours per year per athlete.
• Triathlon does the least at under 50 hours per year. Swimming was, to my surprise, the second lowest at 50 to 70 hours.
• I should clarify that the study applies to aerobic events lasting at least six minutes with 85 percent or more aerobic contribution, so in swimming that means the 800 metres and upwards, and I suspect my surprise comes from a bias towards hearing about the strength and speed practices of shorter-distance swimmers.
• Cross-country skiing, running, biathlon, speed skating and road cycling all fell in the 50 to 100 hour range.
• Rowers already spend a lot of time on the erg and in the gym, so using that equipment more is a matter of familiarity and does not require another training venue, whereas athletes used to open water, the bike and the run are in a different culture.
• The researchers also list physiological and morphological factors.
• In swimming, added muscle mass increases hydrodynamic drag, whereas in rowing the boat might sit slightly lower in the water but there is very little downside to adding strength and muscle mass.
• In triathlon the main point is simply the limitation on hours in the day. I interviewed Arild and Olav some time ago and from the very first conversation Arild said they just had not found strength training that useful; they do some, mostly core work, and not much lifting.
• When you are at the upper limit of volume, around 1,300 hours per year for someone like Kristian Blummenfelt, there is a limit to what you can add from an energy perspective, and you may compensate for some of the benefits of strength training through the extreme endurance volume itself, for example in terms of physiological resilience, which comes up again later.
• Running is less exposed to bone density problems than non-weight-bearing sports like cycling, so that component of the rationale is less important for triathletes.
• Yet running was among the sports that prescribed the most strength training, which surprised me a little given the traditional approach to running.
• Runners also periodised it more, doing a lot in the preparatory period and reducing it towards competition, whereas cyclists kept a similar amount year-round because their long race season requires them to maintain it.

Strength Training benefits

09:46 -

• The paper includes coach quotes to illustrate concepts.
• One is that some coaches use strength work specifically for bone health, which in cycling means better general health and a lower chance of fractures in a crash, but which you could argue is even more important in running because of the impact loads on bones and tissues.
• Another quote is that coaches prescribe more strength training for female athletes.
• That matters because this cohort is Olympic-level athletes in their twenties and early thirties.
• In my own coaching, a male athlete in his twenties with a decent baseline of natural strength does not necessarily need much strength training as a triathlete, but that is very different for someone in their forties and different again for a female athlete, and that individualisation came through in the quotes, even though age could not be a factor in such a homogeneous group.

Speed traning

12:24 - 

• Speed training was remarkably similar across sports, roughly 10 hours per year in every modality.
• It looks like what most listeners will recognise: strides in running, something like three to five times 10-second sprints for one to three sets with long rests on the bike, and the equivalent high-velocity efforts in swimming, speed skating or skiing.
• Road cycling placed it within easy endurance days, other sports attached it to interval days, either before or after the main set. 

Strength training findings

14:50 - 

• On strength training, the finding that stood out to both of us is that across most sports, and certainly those most relevant to triathletes, runners and cyclists, the loads were not very high.
• A typical session had four to eight exercises, two to four sets, five to ten reps, and usually a high number of reps in reserve, so nowhere near failure.
• That differs from randomised controlled trials, which typically prescribe around 80 percent of one-rep max, equivalent to two to three reps in reserve. 
• Examples given were deadlifts at roughly 0.5 to one times body weight and squats around body weight, which for a 60 to 65 kilo male distance runner is not a heavy squat.
• Lighter loads may replicate the demands of the sport more closely, but you may already be getting that from normal training, which is why much current advice favours heavier loads to create a different stimulus. 
• It may be better suited to injury prevention than to economy. It depends on the objective, and it is something for future research.

Strength training protocols

18:35 -  

• A study comparing five reps at 80 to 85 percent of one-rep max against eight reps at 60 to 70 percent on running economy would be very interesting, because perhaps you do not need to recruit all of the fastest fibres; even a relatively low load is still far above the force demands of a running stride, a pedal stroke or a swim stroke.

Periodisation

19:35 - 

• Periodisation was limited. For speed, there was perhaps a slight reduction in reps towards competition, especially where the competitive period is short and focused, as in distance running ahead of a European or World Championships, combined with pushing from around 90 per cent to 95 percent or slightly more of maximal velocity.
• Strides were the running example throughout; hill sprints were not mentioned at all.
• For strength, there was a shallow gradient from higher reps and lower weights towards lower reps and higher weights, with loads peaking in the middle to late preparation period and tapering as competition approached. 
• So it semi-mimicked traditional periodisation, but not strongly. The density of the season mattered too: sports with a dense competitive period reduced strength work during it, while sports with more spaced-out events kept it consistent.

Applying in training

21:45 - 

• My typical athlete is a triathlete racing from April or May through to September or October, which is a long season, so my approach falls in the long competitive period bucket rather than the short one.
• I mostly coach trail runners with events spread through the year, so the protocol stays fairly consistent, with the exception of deliberate blocks, for example a focused eccentric strength block before the first race of the season with a big downhill for athletes who have not been able to get into the mountains.
• Applying this to age groupers, I would keep strength training consistent through the year if the athlete does it at all.
• The difference is that many age group athletes are already at an age where, without working against sarcopenia, muscle mass loss accelerates, so I think strength training is more important for them than for a cohort of Olympic athletes chasing medals.
• It still depends on the individual. If your goal is purely to be the best triathlete you can be on a fixed time budget, it might not make sense, but for a lot of athletes a somewhat larger share of total training than this study implies is reasonable.
• I also prescribe slightly heavier loads, although not to an extreme, and I always caveat that if a session leaves you sore enough to affect the swim, bike and run in the following days, scale it down next time.
• It is complementary, not the main thing. For many athletes that means five to ten reps in reserve anyway, but for those who can lift heavier without soreness I am comfortable at two to three reps in reserve.
• On speed I am very much aligned with the study: once or twice a week through the year, strides, hill sprints or hill strides, sprints in the pool, less consistently on the bike.
• For a triathlete this can add up given you can do it in all three disciplines.
• For example four times 80 metres on the track as part of a warm-up, which loosens athletes up and makes the first interval feel less hard, or some 25 and 50 metre sprints before a longer swim set to build a feel for speed.
• The difference is periodising the modality of strength: gym-based in winter, then building strength within the sport in season, using more paddle work in the swim, which suits swimrun athletes who race with paddles, and hill repeats with a loaded pack for athletes going to races like UTMB where the required gear weighs four or five kilos. 
• Do not do a long swimrun without plenty of paddle work. I did one on a whim once without any specific preparation and my shoulders were wrecked.

Determinants of physiological resilience in endurance runners

28:01 -

• Daniel's first study is Determinants of Physiological Resilience Among Recreational to Competitive Endurance Runners, lead author Timo Malinen from Finland, published in July 2026.
• Physiological resilience is what most of us call durability, the ability to resist deterioration of the physiological determinants of endurance performance.
• Most durability work so far has been in cycling, with a little research on running after 90 and 120 minutes, and this study extends the duration.
• The authors argue that understanding the determinants of durability allows coaches to build them into training plans.
• Forty-five trained runners took part, 23 of them women, split into competitive and recreational groups with women spread roughly evenly across both.
• Testing ran over three days. Day one covered body composition, fat oxidation and a maximal anaerobic running test.
• Day two was thresholds and VO2 max.
• Day three was the resilience test, a two and a half hour non-stop protocol in four stages.
• Each stage began with an incremental test to establish the speed at VT1, followed by 30 minutes at a pace set from that speed, repeated four times so the change in VT1 speed could be tracked over the session, and then a VO2 max test and an isometric leg strength test at the end.
• The aim was to quantify the percentage decline in VO2 max, maximal speed, thresholds and running economy from fresh to fatigued.
• Every variable changed significantly, with declines of between 2 and 6 percent. The largest were in speed at VO2 max and in running economy.
• To put a 5 percent loss of economy into context, Nike popularised the 4 percent figure for carbon shoes, but with how good non-carbon shoes are now and given individual variation, a typical athlete probably gets around 3 percent.
• The fatigue impairment here is considerably more than you could expect to gain from putting on a pair of carbon shoes.
• The recreational runners had a larger decline in speed at VT1, which is what you would expect from a less-trained group, but they also had a smaller decline in running economy.
• The authors suggest this may relate to how economy was calculated, including a body weight component, with the faster runners losing more weight, or perhaps the recreational runners simply had less economy to lose.
• The important finding is that no single variable was associated with the declines in resilience. Two things emerged in the discussion.
• Higher aerobic fitness was associated with better resilience of speed at VO2 max and at VT1, which fits the difference between competitive and recreational groups.

Leg strength and durability

35:23 -

• Greater maximal leg strength was the factor most associated with limiting the deterioration of running economy.
• That ties in nicely with the strength study: strength can help maintain durability, but the key is working on both components.
• This is the point I was making about Blummenfelt's 1,300 hours.
• At that volume, adding strength work may not improve resilience much further, but for a typical age group athlete my working hypothesis is that you can get an outsized return from some strength training in terms of protecting running economy over a long effort.
• Aerobic training remains the more important lever, but the strength component looks like a decent return on investment.
• To be clear about the levels involved, the recreational group were still running half marathons under two hours with a VO2 max around 50, and the competitive group averaged 63.
• These are consistently training athletes rather than merely healthy adults, which was necessary to complete a two and a half hour test followed by a VO2 max test.
• The 30-minute blocks were prescribed at 85 percent of VT1 but came out around 90 percent based on self-selected pace. I ran the numbers for myself: with a VT1 of roughly 4:10 per kilometre, 85 percent is about 4:55 and 90 percent around 4:45 to 4:50.
• Still easy for that duration, and although the percentage sounds high it is anchored to VT1 rather than VT2 or VO2 max, so it is a sustainable intensity

Findings of this study

39:05 - 

• Like the first study, this one is mainly valuable for generating hypotheses, because it is correlational rather than causal. It does not show that training leg strength improves economy resilience, only that they are associated. 
• Some of Michele Zanini's studies, which showed significant declines in running economy after two hours including on trail, may have been designed to look at that more directly, though I would not say so with high confidence.
• The broad takeaway is that the fitter you are, the better your resilience, and there is no need to overthink whether to target VO2 max, VT1 or VT2.
• A rising tide lifts all boats. Strength was the finding that stood out, because economy was the variable that fell most. VO2 max only dropped 2.3 percent, so the engine is still running well but its output is not translating as efficiently. In a marathon you are running at a fraction of VO2 max anyway, so economy losses matter more.

Intervals in a fatigued state

42:04 -

• One thing for marathon runners that did not come out of this study is doing some intensity in a fatigued state, as in the cycling durability studies where intervals follow two hours of riding.
• Runners may not need to go that long, but running at or faster than race pace after accumulating fatigue is worth including.
• If you are training for a marathon, a 70.3 or an Ironman you get some of this naturally because race-specific sessions are long.
• The more debatable question is whether to move threshold work into long runs.
• I am sceptical of that for a 10k, but in a marathon context I am experimenting with it myself: my Sunday long run is marathon-adjacent, and my other hard session, usually threshold-adjacent, is my second longest run of the week, so I make it a medium-long run at the same time. 
• Marathon fatigue session: five kilometres at marathon pace, seven times one kilometre at threshold, then another five kilometres at marathon pace.
• It combines both stimuli and creates fatigue without needing three times 10 kilometres at marathon pace twice in a week.

Ten common criticisms of the critical power model revisited

45:10 -

• My second study is 10 Common Criticisms of the Critical Power Model Revisited by Black, Burnley, Skiba, Poole, Vanhatalo and Jones.
• Three of them, Mark Burnley, Phil Skiba and Andy Jones, have been on this podcast before.
• Most of this group came out of the University of Exeter and they have investigated the critical power model more than almost any other group over a long period.
• The paper takes ten of the most common criticisms, from published literature and from conferences and similar academic settings, and answers each with the available evidence.
• We should be aware that this group is inherently pro-critical power, which is fair, but I think it is a very good paper: detailed without being long-winded, and it answers most of the criticisms well.
• For me, critical power is one tool in the toolbox alongside FTP and others, but I do like the model, and I side with the authors that many criticisms come from misunderstanding what the model is for.
• The criticisms fall into three groups: the fundamentals and physiological basis of the model, the claim that it implies you can hold critical power indefinitely at the long end and behaves wrongly at the short end, and practical application.
• On the fundamentals, critical power is not merely a mathematical construct, and it is not another way of estimating maximal lactate steady state.
• It is its own estimate of the maximal metabolic steady state, and the argument is that it is a better one, because MLSS relies solely on lactate.

Critical power model criticism

49:10 -

• The criticism I hear most is that the model cannot be right because the curve is asymptotic: you approach critical power at 25 or 30 minutes and then keep getting closer at one, two, three or four hours, which people read as the model claiming you can hold it forever.
• That is a misreading.
• The model is built from time trials in the severe intensity domain, above the maximal metabolic steady state, usually two to fifteen minutes, and the authors recommend three, seven and twelve minutes, close to the three, six and twelve I used to use.
• It does not hold in the extreme domain below about two minutes, and it does not hold below critical power. 
• Critical power is estimated from above, and it is not the case that one watt below is guaranteed steady state and one watt above is severe.
• With at least three trials rather than two you can generate a confidence interval, or you can do what I do in practice and treat 95 percent of critical power as a level where I am fairly confident the athlete is in a metabolic steady state.
• Even that cannot be held forever, because glycogen depletion, rising core temperature, neuromuscular fatigue and many other factors come in. You cannot walk forever either.
• Understood correctly, the model is valid from around two minutes up to the 25 to 40 minutes most people can hold critical power, and it marks the boundary between the severe and heavy domains with an uncertainty band.
• The authors use a Schrödinger's cat analogy: ride exactly at critical power and some people end up in steady-state physiology while others do not, and studies have shown this.
• If you want steady state, leave a margin below; if you want severe-domain work, leave a margin above. 
• It is a good way to prescribe intervals in that domain, and easier days can be controlled with heart rate or something else.

Critical power and W' are static

54:36 - 

• For some reason, this model makes people treat the numbers as trustworthy regardless of context, which is not the case. 
• Two hours of riding in the heavy domain reduced critical power by 9 percent and W' by 20 percent in one study.
• A single three-minute all-out effort reduced critical power by 7 percent in another. Heat, hypoxia and carbohydrate depletion all affect both numbers.
• Another critique is that critical power is just FTP by a different name.
• In triathletes, I find the two tend to be very similar, because W' sits in a narrow range in that population, whereas road cyclists show much more variability in W' and the numbers can diverge.
• I still prefer testing critical power because it gives athlete profiling: the shape of the power-duration curve and whether an athlete is relatively stronger at shorter or longer durations.
• And even where the averages are similar, the two are not the same, because critical power has repeatedly been shown to be a metabolic steady-state boundary and FTP has not.
• That may matter little in practice, but it matters in my head to know I have a physiological tool and not only a practical one.

Critical power testing

57:51 -

• On testing, the criticism is that it needs multiple trials and that the result depends on the durations chosen: three, six and nine minutes will give different numbers from three, ten and fifteen. 
• This is where the authors concede the most. Their answer is to spread trials within the domain of validity, with the longest not too close to critical power and the shortest not too close to the extreme domain, hence three, seven and twelve.
• This one made me think, because I currently use three, five and twenty minutes.
• My reasons are that so many athletes have historical 20-minute tests to compare against, that it maps to FTP, and that five-minute power has reasonably good formulas for estimating VO2 max. I can see the authors' point, and at minimum changing twenty to fifteen, or perhaps back to twelve as I used to, is something I need to consider.
• The suggestion is to do all of them, three, five, twelve and twenty, in one session and keep the FTP comparison, but the authors recommend separate days because one test affects the next, and from experience with two critical speed tests on the same day I did not get the most out of the second one.
• Three tests on separate days is the most I think is practical, even though I only prescribe these once or twice a year. The dynamic nature of the numbers also means test days need similar conditions.
• The number reflects the conditions on the test day, and on a given training day an athlete's actual critical power may be somewhat lower, so athletes need to make good decisions and account for that.
• Every test, whether FTP or VO2 max, has the same limitation. See the number as a band rather than a fixed value, and understand the intent of the session.
• For a threshold-type session where you want steady state, start with a decent margin and work up towards critical power if it feels sustainable.
• In the severe domain it is easier, because I have never seen a coach prescribe intervals as close as possible to threshold from above. All of these numbers are guidelines that get you approximately in the right place, and athletes can relax about whether a metric reflects physiology perfectly.
• One critique I nearly forgot, number nine, is that other models describe the power-duration curve better, for example the power law model or mixed models, which in theory cover the whole curve down to much lower intensities and longer durations rather than only the severe domain.
• The authors' response is that this is true outside the severe domain, but within it critical power describes the curve best, and it gives the most accurate estimate of the maximal metabolic steady state when approached from above.
• I ran my own numbers on both models and I agree. 
• The power law is useful for a rough estimate of what you might hold for three or four hours from good five, twenty and thirty-minute numbers if you have never done a maximal effort of that length, but I would not use it instead of critical power for estimating the metabolic steady state or for anything in the severe domain.
• The paper as a whole explains the applicability of the model and how to use it well, provided the testing is done correctly.

Polarised versus high-intensity training in moderately trained cyclists 

01:07:30 -

• The more controversial study is The Effectiveness of Polarized Versus High-Intensity Training in Moderately Trained Cyclists, published in February 2026.
• The question is whether it is the high-intensity or the low-intensity part of training that drives performance improvement.
• The authors argue that many studies comparing training intensity distributions do not equalise the amount of time at high intensity, so they are not comparing like with like.
• Their design gave two groups exactly the same high-intensity intervals, with the second group also doing a fixed amount of moderate-intensity training, to isolate whether the moderate volume itself makes a difference.
• They also had a secondary aim of looking at the psychological experience of adding that easier volume, which is a nice addition other studies could copy.
• The cyclists trained for six weeks with tests at baseline, three weeks and six weeks. There were 20 participants, only three of them women, and one dropped out.
• Both groups did three sessions per week of five times four minutes at high intensity. The polarised group added 240 minutes per week at 80 to 90 percent of LT1.
• In the context of regular endurance training I think that intensity is exactly right: it ensures the moderate domain, neither too hard nor too easy, and it produced the typical 80-20 polarised split. 
• My earlier comment about 85 to 90 percent of VT1 being easy was specific to the resilience testing protocol, where durability studies often test at or near VT1.
• The performance test was a time trial before and after.
• Both groups improved by an average of 8.5 percent with no difference between them, and the authors conclude that over this period the moderate-intensity training made no difference and the improvement came entirely from the intervals. 
• One detail on the time trial: it was not a fixed distance but a fixed work trial, an idea I think comes originally from Asker Jeukendrup's group, with the amount of work scaled to each participant's peak work rate from the final step of the ramp test rather than the same kilojoules for everyone.
• I could not find the actual calorie totals confirmed in the tables, and the two groups started from different absolute power because the polarised group was 10 kilos heavier on average, so the absolute work would have differed.  
• The psychological questionnaire before interval sessions showed a slight, non-significant advantage for the polarised group in how they felt their training was going and how they expected to perform.
• The reason this is interesting is that as endurance coaches our instinct is that four extra hours per week for six weeks must surely show up.
• The authors acknowledge the small sample and limited statistical power, that six weeks may be too short, and that the added volume is far less than most athletes on a genuinely polarised programme would do.

Training status of participants

1:17:30 -

• To be clear about the training status, LT1 and LT2 for the HIT group were 126 and 182 watts, and for the polarised group 158 and 221 watts, a 40 watt difference at LT2 alongside the 10 kilo weight difference, so in watts per kilo they are closer, and I believe randomisation was based on maximum aerobic power, 318 versus 357 watts.
• Even so, from experience it is harder to improve someone from 221 watts at LT2 than from 182; those are different rungs on the ladder.
• Time trial power in the HIT group went from 178 to 190 watts, a 12 watt gain, while the polarised group went from 226 to 251, a 25 watt gain.
• That sounds meaningful to me, but with these numbers it is not statistically significant.
• I will be blunt about my bias: I think the result is wrong, and that is why I say it is underpowered. I probably would not say that if I thought it was right.
• There is a link to the roundup with Gabriele Gallo a couple of months ago, where we covered the large meta-analysis and regression by Knut Sindre Mølmen and colleagues.
• Sprint interval training, HIT and endurance training all improve most things, without huge differences, but on very different timelines: the higher the intensity the faster you improve and the sooner you plateau.
• My speculation is that for athletes at this level there is a ceiling on the rate of improvement regardless of how you structure training, and three interval sessions per week are enough, maybe more than enough, to hit that ceiling over six weeks.
• The question is when the plateau arrives, whether after five, six, twelve or eighteen weeks, and that is where aerobic volume comes in: slower improvement, but sustained for far longer.

Training program sequencing

1:22:10 -

• The question is whether you should do six weeks of intervals and then transition to moderate volume, or whether the volume is accumulating benefit during those six weeks anyway.
• There is not much sequencing research, and what exists tends to point the other way, placing higher intensity closer to the race, although those studies are on races of 5k up to an hour at most, and whether it applies to a marathon or an Ironman is far less certain.
• Practically, for athletes who have been exercising but not doing structured training, I would not start with three hard interval sessions and nothing else. I would build a couple of weeks of aerobic training with one interval session, and add from there.
• That said, plenty of studies use pure high-intensity programmes for six to eight weeks and even less-trained participants cope and improve, so at least in cycling and swimming, where the mechanical load is low, we sometimes speak of this with more fear than is warranted.
• Running is different because of the mechanical load. 
• The assumption was that in relatively untrained athletes any stimulus produces improvement, so four extra hours should have produced more, and he is in the same boat as me in doubting the result while finding that it made him think about programming intensity for different athletes, which is a good outcome from a study even when you disagree with it.
• His other takeaway is patience: athletes may not see the return on added volume in six weeks, but need two or three months, and that needs explaining to newer athletes.
• One thought that came to me during the discussion: if these athletes were new to structured training, 12 watts from 178 to 190 after six weeks of three high-intensity sessions is not actually that large.
• What the study needs is a third group doing only the four hours of endurance training, because those 12 watts might be the baseline improvement from that alone.
• The authors reference earlier studies suggesting moderate training alone gives less improvement, though I would want to check them in detail.
• There was also no familiarisation trial, only one time trial before and one after, so some of the gain could come from learning to pace the test.

Takeaway of UTMB and closing thoughts 

1:30:12 - 

• One takeaway is Benjamin Roubiol's training. He won UTMB in 18 hours and 15 minutes, a course record by over an hour, and for the six months before he was running over 200 kilometres and 13,000 metres of climbing per week. UTMB was his goal race, so he raced the 90 kilometres of Mont-Blanc at the end of May without a taper, in a week where he had already run over 100 miles, specifically to train in a fatigued state.
• He had struggled on the last three climbs the previous year and was building durability, on a completely different scale from anything in the studies.
• A lot of that volume is very easy, which also touches on the moderate-intensity study.
• It is interesting for road runners too: the running hours are far beyond what a road runner could conceive of, and I wonder whether adding very easy trail running on top of road volume, rather than replacing it, would allow more running volume with less injury risk. 
• I have had the same hypothesis with cycling, but trail running is far more specific.
• Something else that came through my feed from, Precision Fuel and Hydration, was Ingvild Kaspersen's executed nutrition and hydration plan from her CCC win. 
• Her body mass loss from start to finish was around 2 percent, certainly under 3, and her fluid intake was remarkably low, something like 137 millilitres per hour on average.
• It is one data point, but it feeds a hypothesis I have had for a while: that the longest events, beyond about 10 hours at the elite level, select for athletes with low sweat rates, because it makes the logistics so much easier, especially now that aid station stops are 20 seconds to fill a bottle.
• The CCC day was very wet and cool for summer, which after a season of heat waves and heat adaptation would have meant relatively low heat stress, and the route was changed because of the weather, so that may have shaped her plan. Ingvild's winning time was around 12 hours.