Blog

This article was made possible through the generous support of our voluntary subscribers. If you value coverage like this, please support FasterSkier with a voluntary subscription.

For the last couple of decades, the priority for high-intensity training has shifted from threshold, to VO2max, and now back to threshold again. Nowadays threshold training is increasingly presented as the ideal endurance training intensity. The suggestion is a high aerobic stimulus at a low recovery cost, repeatable and sustainable. VO2max is often portrayed as the opposite: a very high stimulus, high stress, very high fatigue, and something to be used carefully, if at all. Folks, we have seen this story before, in the threshold boom of the mid to late ’90s.

From some perspectives this may be intuitively correct and perhaps logical. If we compare one minute of threshold work to one minute at VO2max intensity, the VO2max work is more demanding. That is a no-brainer. There is more carbohydrate use, higher oxygen demand, higher lactate production, greater perceived exertion and generally a greater physiological disturbance. But is that the way training really works? We build training plans around individual workouts and daily and weekly loads, not per minute of high-intensity training.

For example, we might plan 5 x 4 minutes at VO2max intensity, or 20 minutes of total work. Or it might be 4 x 10 minutes at threshold, or 40 minutes of total work. When we look at the load of the whole session, things become far less clear. Both types of training produce similar adaptations, although the adaptive signal from the higher-intensity work is much stronger, and there is little evidence that a threshold session creates less recovery demand than an appropriately controlled VO2max session.

This is really the point I want to make. We need to stop confusing intensity with training load, and, perhaps more important, we need to stop confusing training load with the recovery timeline. They are related, but they are not the same thing.

As I have written before, threshold is a moving target, and defining it is problematic. There are perhaps a dozen or more methods used to determine threshold, and they can produce very different speeds, heart rates, power outputs and lactate values. Some athletes are working just above LT1 and calling it threshold. Others are working around 4 mmol, and others are using ventilatory threshold, critical power, some percentage of race pace, or simply an intensity that feels controlled and hard. So when people say “threshold training,” I am not sure we are even talking about the same thing. That alone should make us careful about assigning a fixed physiological cost or benefit to threshold training.

The generally accepted training model is usually described like this: easy endurance training provides a relatively small stimulus at a relatively low cost, so we need a lot of it. Threshold gives us a much larger aerobic stimulus, still at a manageable recovery cost. VO2max gives us a very large stimulus, but at a very high recovery cost. Again, that makes intuitive sense, but is it true when we compare realistic sessions? I am not convinced it is. In fact, I am certain it is not.

If we think of training load in simple terms, it is a combination of intensity and duration. Higher intensity produces greater load per minute, while longer duration produces more total load. A 3-hour easy workout, for example, produces a bigger training load than a 75-minute workout with 5 x 4 minutes of VO2max intervals. This is obvious, but somehow we often forget it when discussing threshold and VO2max training.

Take a typical VO2max workout of five by four minutes, 20 minutes of work. Using Firstbeat’s TRIMP model, that might create about 55 to 60 TRIMP. Now take a threshold workout of four by ten minutes, 40 minutes of work. That might generate something in the 95 to 100 TRIMP range. With Firstbeat we can also look at EPOC as an estimate of the disturbance of homeostasis. Again, these are estimates and certainly not perfect measures, but 20 minutes at around 90 percent of VO2max might produce an EPOC of around 175 ml/kg, while 40 minutes at around 85 percent might produce an EPOC of 225 ml/kg. Once again the threshold workout produces a significantly higher stress load.

I don’t want to get hung up on whether those exact Firstbeat numbers are accurate, because that is not really the point. The point is that it is entirely possible for the lower-intensity session to produce the larger total load, and that should not surprise anyone, because it is twice as long.

The same thing happens if we use session RPE. Twenty minutes of VO2max work at an RPE of 9 gives us 180 AU (arbitrary units), while forty minutes of threshold at an RPE of 6 gives us 240 AU. Again, the threshold session creates the larger estimated training load. That does not prove threshold is more stressful or that it takes longer to recover from. TRIMP and session RPE are measures or estimates of training load, not measurements of recovery time.

We have at least three different things being discussed as if they were the same: the intensity of the exercise, the total load of the workout, and how long it takes the athlete to recover. A VO2max interval clearly creates greater physiological disturbance per minute. A longer threshold session may create a greater total training load. Neither one, by itself, tells us which athlete will be better recovered 24 or 48 hours later.

High cost per minute is not the same thing as high cost per session, and high session load is not necessarily the same thing as longer recovery.

What Are We Actually Trying to Adapt?

The other issue is that we spend far too much time talking about training zones and not enough time talking about desired adaptations. What adaptations are we trying to create? That should come first.

VO2max training provides a very strong aerobic stimulus at near-maximal oxidative capacity. It can increase maximal cardiac output, improve oxygen transport, create very high muscular oxygen demand and expose the athlete to speeds and power outputs similar to race demands. Threshold training lets us accumulate a large amount of work at a high aerobic fractional utilization. It can improve sustainable speed or power, oxidative capacity at that intensity, and the ability to maintain a high percentage of maximal aerobic power for a longer period of time.

These adaptations overlap a great deal. I sometimes hear people talk as if threshold develops lactate clearance and the muscles’ ability to sustain work, while VO2max develops the central cardiovascular system. But VO2max training creates a large muscular stimulus as well. ATP turnover is very high, calcium flux is high, metabolic disturbance is high, and the signaling associated with mitochondrial biogenesis is substantial. VO2max training is not simply “central cardiovascular training.” It also provides a significant peripheral training stimulus.

Threshold training obviously produces a similar but slightly smaller mitochondrial and muscular signal for adaptation. But I think its real advantage may be something much simpler. Because the intensity is lower, we can normally do considerably more of it. That allows us to accumulate a large amount of work at a high aerobic fractional utilization, with substantial muscular recruitment and a large amount of sport-specific work.

That may be threshold training’s most important characteristic. Not that it creates some unique physiological adaptation unavailable at other intensities, but that it allows us to accumulate considerably more work while still creating a high oxidative demand.

This can be particularly important for an athlete who already has a high VO2max and whose limitation is the fraction of that VO2max that can be sustained. If fractional utilization is relatively low, increasing sustainable aerobic power may be far more important than trying to push VO2max another few percent. For that athlete, a greater amount of threshold work may make very good sense.

But that is an argument based on the needs of the athlete, not an argument that threshold is inherently a better training intensity or that we recover from it more quickly.

Lactate clearance and resynthesis is another area where I think training discussions have become confused. We often hear that threshold training is necessary because the athlete needs to train to “clear lactate.” I have always had a problem with this explanation, because it is neither accurate nor scientifically valid. Remember, it is lactate we are talking about, not lactic acid, as many refer to it.

Lactate is not a waste product that suddenly appears at threshold and then needs to be removed. It is being produced, transported and oxidized all the time, at all intensities, and it is an important fuel. Endurance training improves the ability to transport and oxidize lactate, but there is no reason to believe this only happens at threshold. High-intensity interval training also improves lactate transport and can move lactate threshold substantially.

VO2max intervals also create very high lactate production and turnover. So if we are looking at the lactate shuttle, why would it only, or even preferentially, adapt when lactate is sitting around 2, 3 or 4 mmol? That doesn’t make much physiological sense to me. Producing higher levels of lactate will stimulate resynthesis.

The same is true for mitochondrial biogenesis, oxidative enzymes and capillarization. We cannot neatly assign these adaptations to one training zone, because there is considerable overlap between endurance, threshold and high-intensity training.

That doesn’t mean all intensities are the same. They clearly aren’t. It means we should be careful about inventing unique adaptations for a training zone simply because we have given that zone a name. Threshold training is not the only method, or even the optimal method, to improve lactate oxidation and resynthesis.

What Is Recovery?

Another problem in this discussion is recovery. People will say that threshold is easier to recover from. Maybe, but recover from what? Autonomic disturbance, glycogen depletion, neuromuscular fatigue, muscle damage, mechanical load, dehydration or psychological fatigue? These are not the same thing, and they do not necessarily recover at the same rate. A 2-hour run and a 2-hour rollerski at the same intensity, for example, carry very different mechanical loads to recover from.

HRV may tell us something about autonomic recovery, which should also reflect other aspects of recovery but does not specifically measure them. Resting heart rate can tell us something, but again not everything. Blood lactate tells us something about what is happening during and immediately after the workout, but remarkably little about whether the athlete is going to train well 24 or 48 hours later. Creatine kinase may tell us something about muscle damage, but there is huge individual variation. So when we talk about recovery cost as if it were one number, I think we are being too simplistic.

This is also why I don’t think we can simply use TRIMP, EPOC, duration, distance or session RPE to say one workout takes longer to recover from than another. They can give us useful information about internal load or physiological disturbance, but they are not recovery measurements. Ultimately, recovery must be considered in terms of which physiological systems were stressed and, more practically, what the athlete can do in the days immediately after a hard workout.

Carbohydrate use is another good example. VO2max exercise clearly has a higher rate of carbohydrate use per minute than threshold exercise. But once again, the session matters more than the minute. Twenty minutes at VO2max intensity may have a very high glycogen cost per minute. Forty, fifty or sixty minutes at threshold has a somewhat lower rate, but it is sustained for two or three times as long. We cannot automatically assume that the shorter VO2max workout creates greater total glycogen depletion.

This becomes particularly important when athletes are doing regular quality sessions. If glycogen restoration is an important part of recovery, a long threshold session may impose a very substantial recovery requirement despite producing lower lactate concentrations and feeling much more controlled. An athlete does not have to be lying on the ground after a workout for that workout to have been expensive to recover from.

That is one of the problems I see with the idea that threshold can be repeated frequently simply because lactate is controlled. Low lactate does not necessarily mean low glycogen use, low muscular load or rapid recovery. If we keep extending the duration of the threshold work, at some point the additional work has a cost.

Which brings us to what I think is probably the most meaningful recovery measure, and one that is surprisingly rarely studied: can the athlete perform the next important training session at the expected level?

An Interesting Study

One of the more interesting comparisons was made by Seiler, Haugen and Kuffel with highly trained runners. On one occasion the athletes did a threshold session of around 30 minutes between the first and second ventilatory thresholds, and on another, six three-minute intervals at around 95 to 96 percent of VO2max. I would point out that 95 to 96 percent of VO2max is already hard, and above what I would normally call a controlled VO2max session.

The differences between the workouts were substantial. Lactate during the threshold session was around 2.7 mmol, while during the VO2max intervals it was around 7.1. Perceived exertion was also much higher during the VO2max workout. By almost any acute measure the VO2max workout looked harder, so we might reasonably expect recovery to take longer. But HRV recovery following the two sessions was remarkably similar.

Now, this study does not fully answer the question. The threshold work was only around 30 minutes, HRV measures one part of recovery, and there are all kinds of differences between athletes. But it does challenge an assumption that I think is important. Greater acute metabolic disturbance does not necessarily mean longer recovery, but the dose does.

It also raises what I think is an even more interesting question. What would have happened if the threshold session had contained 40 or 50 minutes of work rather than 30? That would be much closer to the threshold sessions many trained endurance athletes are actually doing.

Control of VO2max Training

There is another issue here that I think is important. A lot of people seem to think VO2max training means going as hard as possible. It doesn’t, at least not the way I see it and use it.

If I give an athlete 5 x 4 minutes of VO2max work, I am not asking them to race every interval or see how much lactate they can produce. I am trying to get them to accumulate time at a very high oxygen uptake, typically around 90 to 95 percent of maximal heart rate and VO2max.

If the objective is a cardiovascular and oxidative stimulus, we need sufficient intensity to drive oxygen consumption toward a very high fractional utilization of VO2max. For many athletes, something in the range of 90 to 95 percent of VO2max may be enough to accomplish that. Once we have produced the stimulus we are looking for, going harder does not necessarily make the workout better.

This becomes important when we compare threshold and VO2max training. If someone compares a carefully controlled threshold session with a set of VO2max intervals performed as hard as possible, then of course the VO2max workout is going to look very stressful. But that may simply be a badly controlled VO2max workout.

A well-controlled VO2max session and an all-out interval session are not the same workout, just as a controlled threshold session and a threshold session that slowly turns into a race are not the same workout. Anyone who has coached for very long has seen both.

What Is Threshold’s Real Advantage?

In my view, the strongest argument for threshold training is not that it produces some special adaptation, but that athletes can accumulate a lot of high-quality aerobic work. Forty minutes at threshold is a lot of work. Fifty minutes is a lot of work. Sixty minutes is a lot of work. For an elite endurance athlete doing a very large total training volume, that may be exactly what is needed, provided the athlete can recover from the overall workload of the training period.

Threshold gives us a way to accumulate a substantial amount of work at high oxidative fractional utilization, with considerable muscular recruitment and a lot of sport-specific work. For an athlete with a high VO2max who needs to improve fractional utilization or sustainable aerobic power, this can be valuable.

I think this is where some of the current discussion around threshold becomes problematic. Because it is controlled, because lactate is relatively low, and because the athlete is not destroyed afterward, threshold starts to be treated almost like endurance training. It isn’t. It is still hard training, and if you do enough of it, it carries a substantial training load and potentially a substantial recovery cost.

The question is not simply whether forty or fifty minutes at threshold creates more adaptation than twenty minutes of VO2max work. It probably creates more of some kinds of stimulus simply because we are doing so much more work. The more useful question is whether the additional adaptation we get from that extra work is worth the additional load and recovery cost.

Are We Asking the Wrong Question?

The debate is usually framed as threshold versus VO2max. Which is better? I don’t think that is the right question. The first question should be what adaptation this athlete needs. The second, and I think this is more important, is what is the smallest dose capable of producing that adaptation? It isn’t necessarily 40, 50 or 60 minutes. The third is what effect that dose will have on the training the athlete can perform over the next several days.

We tend to assume that if some training is good, more of it must create a greater adaptation. Up to a point that is probably true. But the objective of training is not to produce the largest possible stimulus in every workout. It is to produce enough stimulus to create the adaptation while still being able to absorb the rest of the training.

If an athlete has a relatively low VO2max and that is limiting performance, perhaps 15 or 20 minutes of controlled VO2max training gives us exactly the stimulus needed. If that is enough to stimulate the adaptation, why do 50 minutes of threshold just because threshold is currently popular?

On the other hand, if an athlete already has a very high VO2max but relatively poor fractional utilization, and cannot sustain a high percentage of it, then perhaps a larger volume of threshold work is exactly what is required. In that case the ability to accumulate 40 or 50 minutes at high aerobic flux may be the important stimulus.

The Training Week Matters More Than the Workout

I also think we make a mistake when we look at workouts in isolation. A workout can look great on paper and still be a poor training choice, because the real issue is not simply what happened during those 60 or 90 minutes. It is what that workout allows us to do over the next several days.

Suppose a threshold workout produces 10 percent less adaptive stimulus than a VO2max session but allows the athlete to train substantially better over the following two days. Threshold may clearly have the better stimulus-to-recovery ratio and therefore be the better choice.

But reverse the situation. Suppose 20 minutes of well-controlled VO2max work produces the adaptation we want and the athlete is ready to train normally the next day, while 40 or 60 minutes of threshold creates substantial glycogen depletion and muscular fatigue. In that circumstance, the shorter VO2max workout may have the better stimulus-to-recovery ratio.

We shouldn’t decide the answer beforehand simply because one workout carries the label “threshold” and the other “VO2max.” Threshold is not automatically low cost, and VO2max is not automatically high cost. The dose matters, the athlete matters, the training done before the session matters, what needs to be done after the session matters, and the total training week matters.

Dose, Adaptation and Recovery

I am certainly not arguing against threshold training. Rather, I am arguing that the dose needs to be controlled relative to the individual athlete, and that the time to recover is not what is often suggested. I use threshold with athletes who need it, and it can be effective. But I do think we need to be more careful about how it is being presented. It is not automatically what every athlete needs.

Threshold and VO2max training have different characteristics, but their physiological adaptations overlap substantially. VO2max training provides a very large physiological stimulus in a relatively short period of time and appears particularly useful when maximal aerobic capacity is what we are trying to improve. Threshold allows us to accumulate considerably more work at a high aerobic output and can be extremely effective when sustainable aerobic power or fractional utilization is the limitation.

Both can be valuable, and both can be overdone.

In my experience, especially over the last decade, I have seen far more athletes become flat, stale, overtrained or simply stagnant from doing too much threshold training than from doing properly controlled VO2max work. Usually it comes from doing too big a volume of threshold work, too often. That doesn’t mean threshold is bad. It means threshold must be appropriately prescribed and dosed.

I also think the very large amounts of threshold work we sometimes see promoted are primarily appropriate for highly trained elite athletes who are also doing very large overall training volumes and have spent years developing that volume. I would be much more cautious applying the same model to younger athletes, developing athletes, masters athletes, or anyone doing a much smaller overall training volume.

There is another point that seems to get lost. Athletes racing 5, 10 and 15-kilometer events are not racing those events at threshold. I have decades of data showing that even in races of up to 90 minutes, fractional utilization can be well above threshold. The shorter races are at a very high intensity, in what is defined as the VO2max zone. At some point we need to train that capacity directly. Doing more threshold work because it is supposedly lower cost does not eliminate the need to train at race-relevant intensity.

In the end, I think we need to get away from ranking training zones as good, bad, safe or dangerous, and think more in terms of dose, desired adaptation and recovery. What are we trying to adapt, what dose is needed to produce that adaptation (often less than we think), and what does that dose cost us in terms of the training that follows?

The most valuable workout is not necessarily the one that creates the greatest immediate physiological disturbance, nor is it necessarily the one that accumulates the most minutes around some predetermined lactate value. It is the workout that provides enough stimulus to produce the adaptation we are looking for while preserving the athlete’s ability to absorb the rest of the training program.

In the end, that is the balance we are trying to find.

Help Support Our Coverage!

If you value the thoughtful work and access FasterSkier provides, please consider becoming a Voluntary Subscriber. Your support directly fuels our work to cover the people, places, and moments that make our sport special.

Join the FasterSkier community!