• Boris Clark

Froome vs Clark. How to win the Tour de France

Recently MPS coach Boris Clark set himself the challenge of improving his lactate threshold power over the course of 4 weeks.

In the end Boris improved his lactate threshold 7 watts, from 354w to 361w, over the slightly shorter period of 3.5 weeks, or on average 2w per week.

This is a pretty solid gain, but how does it compare to the worlds best? If we could keep this going when would Boris be in shape to win the Tour de France, and what do the numbers indicate?

To answer these questions we put Boris’ data up against Chris Froome and his test results from his trip to the Glaxosmithkline human performance lab.

Froome is a 4 time Tour de France winner, Giro d’Italia winner, 2 time Vuelta Espana winner, and Olympic medallist.

By comparison Clark has 4 NZ national medals, 3 bronze, 1 silver. Has raced at the professional level in Europe, 4th in NZ for the individual pursuit, worn the KOM jersey at the NZ Cycle Classic and sprint jersey at tour of the tropics, scored a few UCI points, finished 5th at the Tour of New Caledonia including a 2nd and 2x3rd places on stages, days in the U23 and KOM jersey there, and multiple top 10 finishes on the hill top finish stages at the Tour of Southland.

So, far from winning the Tour de France, but no slouch. Now what would it take to get to that Tour winning form. But first, a quick brief about what we are measuring with this test, and what we will measure at some point in the future.

Can barely see the difference! Both shots from 2015!

What are we measuring?

For Froomes test there were 5 key factors measured.

1. Power out put at 1mmol above baseline

This is one way of estimating the aerobic threshold. The first significant rise in blood lactate, in basic terms an indication the body is starting to have to put in a fair effort.

2. Power output at 2mmol/L blood lactate

Another way of measuring the aerobic threshold. Using a fixed concentration like this can make it easier to compare between tests and other athletes, and make future tests easier and shorter if you are only interested in measuring one variable

3. Power output at 4mmol/L blood lactate

Power output at 4mmol/L is not always exactly lactate threshold/anaerobic threshold, but is generally close, and highly correlated with power output over the course of a one hour time trial, and power output at lactate threshold. Calculating the lactate threshold can be tricky as it depends on the shape of the lactate curve and the calculation method used. By using 4mmol/L it creates a much easier to measure marker, which is also easier to compare between tests, between athletes, and again, can make future testing quicker and simpler.

4. VO2max

The maximum amount of oxygen that can be utilised during exercise. More oxygen helps you oxidise more fuel (i.e. fat and carbohydrate), with some of that fuel being the clearing of lactate (which is essentially a broken-down carbohydrate). The lactate threshold will be at a particular percentage of VO2max for every athlete, and this depends largely on the maximal glycolytic rate of the athlete.

5. Gross efficiency

How much of the oxygen being used is being used to create power at the crank. Generally, this is around 20-25%. It is easy to forget that while riding a bike, or doing any exercise for that matter, the task at hand is only a small portion of what is going on in your body. Even before you make any power it takes energy to beat your heart, contract your diaphragm, keep all your normal bodily functions going, move your legs, dissipate heat etc. All this costs energy before even getting any power to the pedal. So if an athlete is exercising with a metabolic rate of 1000w, and they currently have a power output of 250w, their gross efficiency of 25%, I.e. 25% of there ‘effort’ is being turned in to power.

This is not to be confused with net efficiency, or cycling economy, which is the energy needed for cycling, after that of baseline oxygen consumption (I.e. when you aren’t doing anything) is subtracted already. This measure is useful to determine how much energy is ‘being wasted’ and not going into power output while cycling, without being influenced by normal resting oxygen consumption levels.

At the moment we don’t have any recent VO2max or gross efficiency data for Boris. We can test his VO2max alongside a couple of extra factors which we can also compare against Froomes test, but have not as yet done this, so that will be in part two which we will do anywhere from a few weeks to a few months down the track.

What we do have is all the lactate data we can compare with all Froomes values.

Clarks submaximal lactate tests pre and post the 3.5 weeks training

The methods used in each test

There are two main differences between how the two tests were done.

1. Froomes test was done as an increase of 25W every 4 minutes with the lactate sample taken in the last 30 seconds of each stage. While Clarks was a 35W increase every 5 minutes with the sample taken within 45 seconds of the end of each stage.

Lactate can take a while to get from the muscle to the blood where we measure it. For this reason, it is possible Froomes power output for a given lactate value are a little higher than they would be if longer duration stages had been used. Clarks test has longer stages which should allow a little more time for the lactate to get from the muscle to the blood and may lower power output for a given lactate level a fraction compared to the methods used in Froomes test. The sample being taken after the completion of the 5-minute stage is due to Boris self-testing. If anything, this should allow extra time for the lactate to get from the muscle to the blood. As lactate is a carbohydrate, and if we reduce workload to essentially nothing, we will not utilise a substantial amount of it for energy, this small break to take the lactate sample should not cause any significant change in lactate concentration between stages.

2. The lactate samples in Froomes test were taken from the ear lobe, while Clarks were taken from the finger.

Lactate samples from the finger tend to have higher values than those from the earlobe. The main reason for this is likely sweat accumulating on the fingertips, with even a tiny amount affecting the lactate reading a little. Clarks test used an alcohol wipe before using the lancing device, the first blood drop was wiped away with a tissue as this allows any residual sweat to be cleared away, then the sample was taken from the next blood drop. It is also possible to end up with a ‘micro-concentration’ of increased lactate in the fingers if the athlete tenses them during the test, which is not possible with the earlobe. At MPS we believe taking lactate samples from the earlobe is the best way to do a lactate test (and it doesn’t hurt at all compared to the finger which hurts a little), but when self-testing the earlobe is not an option. In short: Due to the lactate samples being taken from the finger it is possible Clarks lactate readings are higher and subsequently power output lower than that of Froomes test. While there is no way to know how much difference this would cause, we do know that the numbers are consistent if they are taken from the same sample site.

Crunching the numbers

Froomes weight for these tests was 71 kg and 9.5% body fat which is 4kg higher than his reported race weight of 67kg (or as low as 66kg). His test was done not long after le Tour de France, so perhaps he partied a bit hard! This was the weight immediately pre-test, so probably includes some food and fluid taken onboard, which is likely different to the other reported weights which are likely from first thing fasted in the morning. Clarks weight for the test was 69.9kg, so quite similar. He also races in the range of 67-68.5kg generally, so the comparison is extremely similar.

So what do these numbers mean?

The aerobic threshold: Froomes baseline value was lower than that of Clark, probably due to Clarks test being taken from the fingertip, and Froomes being taken from the ear lobe. This doesn not pose a problem when comparing Clark to Clark over multiple tests, or Froome tor Froome, but does make it difficult to use this value to compare both riders as at the lower end of things a 1mmol/l difference is quite a large difference.

Instead we are better to look at the power at 2mmol/L. We still have to accept there may be some variation because of the lactate sample site, but it should not cause as much variation since at least the lactate concentration is the same in this comparison.

As we mentioned before, this value is an estimate of when the body first starts to put in a fair amount of effort, or the first point where lactate concentrations rise a little bit. For Froome this was 390w, and Clark 314w. That is a pretty astounding difference. 314w is pretty solid output to be at 2mmol/L concentration of lactate. For those not familiar, 2mmol/L is almost nothing. At 390w Froome is just barely kicking into gear. There is a 76w difference between the two at this lactate level which is huge.

This aerobic threshold is a pretty good indicator of sub-maximal fitness. So a good measure of endurance, how long you will be able to ride at a moderate pace for, what that pace will be, and how fresh you will feel come the end of a long race. So while Froome is getting to the end of the race pretty fresh, Clark is probably starting to feel pretty cooked.

The one saving grace for Clark is that he has a higher anaerobic capacity/lactate production rate than Froome (something we can talk more about in part 2), meaning he can produce more energy through anaerobic means than Froome, and likely has a higher peak lactate. So on a more ‘punchy’ finish of 5-15 minutes Clark would stand more of a chance than these numbers would suggest.

Power at 4mmol/L lactate: Similar to threshold, this is the number all cyclists compare each other against. This is the most important determinant of endurance performance, but as mentioned above with the ‘punchy’ finish, not the only one.

Froome is putting out 430w at 4mmol/L lactate, while Clark is putting out 356w. For Froome this is 6w/kg, which is good, but not Tour de France winning, but when adjusted for his reported race weight of 67kg the power out put is 6.4w/kg, which is right up there with what is currently thought to be the maximum of clean human performance. Clarks w/kg is 5.09, or 5.3w/kg adjusted for race weight like Froome. This is still a high number, but no match for Froome.

What does this mean. Well perhaps the news isn’t as bad for Clark as it first looks from these numbers.

Froome may have a staggering 74w extra power over Clark at 4mmol/L lactate, but due to his extremely low anaerobic capacity/glycolytic rate, it’s likely his peak lactate level is not very high. This means while he has an extremely high threshold, both in absolute terms and as a percentage of VO2max, when he goes to ride above threshold, there is not a whole lot of extra energy available.

This compared to Clark who we know from other testing has a higher anaerobic capacity, his threshold is lower both in absolute terms and as a percentage of VO2max, but he has more energy available when it comes to riding above this level.

There is still a problem for Clark here though. While 4mmol/L or lactate threshold is not a clear cut line in reality, prolonged periods over this power output will result in acidosis in the muscle, fatigue, and lowered power output.

So lets say there is a flat stage, with a 30 minute hill top finish at the end. Froome will still beat Clark based on these physiological variables. Froome may be able to do say 440-450w for this effort, whereas Clark may be able to do 400w on a good day. While a defeat, compared to the best cyclist of a generation this isn’t too bad!

The problem comes when you involve multiple hills or bouts above or close to threshold. Imagine an alpine tour stage with 4 large climbs. The pace these are ridden at for Froome will be below threshold, but it might be around 350w if the pace setting is solid. For Clark he is pretty much at threshold here and will surely have bouts of effort a little above threshold. So while his power output for a single effort, even a long effort at that, is actually respectably close, when you add in multiple efforts, Clarks lower threshold will lead to fatigue, and nothing left to compete with at the end.

So what can we do about this?

An interesting question. If we assumed a linear improvement in threshold power like we achieved the last 3.5 weeks, Clark could improve his threshold/power at 4mmol/L 2w per week, meaning it would take 37 weeks to get up to Froomes level.

So by next year Clark should be targeting the win at the Tour de France?

Not so fast. Chances are these improvements will taper off. What was once an ‘easy’ 2w gain a week becomes 1w a week, and that becomes 1w a month etc. It takes a lot of work to build threshold power, and the gains are made over months and months and years of work.

Furthermore, there are two main ways of improving power at threshold. Increasing the aerobic capacity, or decreasing the anaerobic capacity. We know from previous data that the best way for Clark to improve his threshold power is through lowering anaerobic capacity, and there is quite a lot of room to do this. But this can only go so far before you start compromising other parts of performance (e.g. shorter hard efforts), and then the only option is to increase VO2max, which while perfectly possible, may be harder to do causing gains to reduce over time.

There is one study we have read however in which they managed to achieve linear increases in VO2maxfor the whole duration of the study (around 10 weeks), and we suspect due to the difference in lactate sample site Clark may be a bit closer to Froome than these number initially suggest, plus the effect of the previous training block may take a little while yet to truly adapt to, so perhaps there is more improvement to come simply from more rest, so who knows, we’ll report back next time Clark does a test and see how much closer he’s gotten! We should also have some other data then and can get a bit more technical and in-depth.