Logo Loading

Perfect Pacing

Our Perspective on Human Performance

Perfect Pacing

Perfect Pacing

The club 10 or an odd half-marathon now and again don’t seem to satisfy us endurance athletes anymore, as we seek bigger and bigger challenges the fitter we get. The number of sport science publications have also rocketed in the last few years, on hydration, fuelling and general long distance performance. Ultra-distance events are definitely becoming more popular, ultra-marathons for runners like the Comrades, Cape Town, and Western States. 12hr and 24hr TTs like Le Mans, RAAM and 600km AUDAX for super randonneurs like the Verona-Resia-Verona, and of course enough iron-man events you can shake a stick at! All very epic, but if you are motivated then why not.

I even managed to pull a for fun 205 miler out of the bag back in June this year with a fine bunch of cyclists raising money for the AGSD charity in the footsteps of Tommy Godwin, Mr. Ultra-distance himself. The day was well organised by the AGSD with support and food stops to make it an enjoyable experience along the south coast from Winchester to Eastbourne and almost back. I had no intention of racing a social event, but kept an eye on the time. Enjoying the sweet and savoury rice parcels made sure I stopped for longer than necessary. We didn’t finish before dark, about 13hrs. Not very scientific I know, as taking it ‘relatively’ easy for me was my overall pacing strategy for the first half, the second half didn’t turn out to be as demanding terrain as we had been led to believe! (important- the main reason a thorough reconnaissance or study of the route profile will help tremendously with distribution of effort).  I had time to think about more formal time-trial events and the impending 24hr Le Mans TT which 3 of my clients will face on August 23rd, and how a slightly more personalised scientific approach, combining practicality, preference and feel may help achieve success on the day.

Every cycling event will have a varied distribution of required power-outputs to complete the circuit at a certain speed, and a cyclists’ unique physiology will help them perform better during some sections than others, with weight and frontal surface area crucial factors. Road races spend large durations at very low and very high powers, fluctuating significantly around a low average value as attacks, sprints and breaks on hills manifest themselves in and around drafting and ‘free speed’! These fluctuations are highlighted by an increased requirement for anaerobic efforts which stress this capacity. The disproportionate powers are physiologically more stressful, and form the reason behind Coggans derivation of the Normalised Power concept to account for non-linear lactate kinetics and the associated disproportionate stress on the body. Whereas the anaerobic work contribution to a time-trial will probably be much less, depending on the terrain, as the power distribution should closely match the average value.

The reason for this can be explained on physiological terms of functional performance. Both oxygen kinetics, i.e the rate of achieving maximal oxygen consumption (not measured by a VO2max test) and hence aerobic metabolism AND anaerobic lactate formation form the balance between success at each discipline. Reliance on a large aerobic capacity will avoid the need to stress the anaerobic system to the point that accumulation of metabolites and muscle acidosis associated with fatigue restricts performance under time-trial conditions. Conversely, as stated above, a well-conditioned anaerobic system in road-racing is necessary for brief intense efforts to attack your opponents.

However, as we know that sustainable power-outputs are dependent on finely-tuned anaerobic and aerobic systems working in harmony, clearly the duration of the event will dictate how long an athlete can tolerate exposure to a certain level of anaerobic waste products before they finally ‘blow-up’. In essence this is determined by the lactate threshold of the rider, and choosing the optimal average power to pace such an event based on their ability. We can see this from the graph below (Mr.S) indicating the choice of different powers for a particular individual targeting different time-trials distances.

The lactate concentration profile and heart rate (as a surrogate for oxygen cost) show that, relative to the threshold or MLSS, target lactate concentrations will assist in successful execution of perfect pacing. Although not exact, shorter distances at 40km and below, such as the club ten which can last anywhere between 20-25minutes for a trained recreational cyclist can be endured at power-outputs above MLSS associated with relatively high blood lactate levels and heart rates (VO2 consumption). Timing is crucial so that maximal oxygen consumption isn’t achieved before the finish due to cardiac drift and accumulation of anaerobic waste metabolites.


Longer events should be executed at power-outputs much lower than MLSS with good reason. For example, a 80km TT should be achievable with good pacing between MLSS and baseline lactate, around 250 Watts in the example shown. A longer solo 160km event will rely on the first aerobic turning point and highest power output still at a baseline lactate value, indicated at around 235Watts.

The reason why higher power-outputs cannot be sustained for each event duration is due to the curvi-linear response of lactate with work-rate (power-output). Take for instance a cyclist with the lactate profile shown above (Mr. S), who intends on racing a 40km TT. Their MLSS value is 265W at 3mmol/L blood lactate. If the cyclist decides to ride much harder than his perfect pace, for instance riding at a power 280Watts which is 1mmol/L above MLSS, the pace will not be sustainable for the duration of the event, and the rider would have to reduce their effort to a recovery power output corresponding to one which is 1mmol/L less than MLSS, or 235 Watts (for a duration as long as they rode above the MLSS), so that the lactate and associated waste products can be cleared appropriately. This would mean that the average power for the time above MLSS (280W), at MLSS (265W) and below MLSS (235W) would be less ((280+265+235)/3 = 260), than if the rider maintained ‘perfect pace’ at the MLSS (260W vs 265W, respectively). Even if the rider has an exceptional ability to clear lactate or buffer associated inflammatory ions after a greater effort, this will still not allow the rider to start out harder than their MLSS pace or target perfect pace, as the reconstitution of the anaerobic capacity (i.e. how quickly lactate levels return to a steady state) is proportional to how much lower the recovery power output is relative to the MLSS. Basically this means a cyclists with superior lactate clearance will be able to breach their MLSS more frequently, but not necessarily for longer. Unless recovery was forced on a decent or with a tail-wind. Of course, this concept applies to all distances and durations of TTs. The greater the difference that the target pace is above MLSS, then the further below MLSS is necessary for adequate recovery. Hence slowing a rider down considerably.

The opposite applies to target paces chosen below MLSS. A higher recovery power output maybe acceptable, if the cyclist breaches his target power, depending on how close this is to the baseline lactate equivalent. If group riding is employed, a different dynamic for pacing exists in that more frequent breaches of MLSS (similar to a hilly time trial), of greater magnitude can occur due to the opportunity to recover at much lower intensities (in the pack or on descents) while maintaining adequate speed. This usually results in a lower overall average power, but with an elevated Normalised Power relative to this, representing non-linear stresses of higher powers will be greater. However the NP value is an apparent representation of physiological stress, and does not actually indicate that the cyclist is riding within his ability for the entire duration, especially if the cyclist may face riding solo for extended periods. Gauging pacing in a pack or ‘groupetto’ using average power is still a good way to maintain speeds above those which are achievable alone. Although how closely MLSS can be approached or breached (which may frequently be the case) if adequate recovery can be achieved with the use of free speed, would be the main concern and is perfected in cyclists with greater amounts of experience for their ability and knowledge of the route profile. Of course this requires profound experience and awareness of physical ability throughout the duration of the event to avoid negative pitfalls in executing this strategy, especially on unknown courses.

Regardless, a ‘perfect pacing’ strategy is known to be superior for longer distance steady-state endurance events which rely more on aerobic capacity and less on total anaerobic work above threshold. Although these two energy systems are not mutually exclusive in any type of cycling event, and is difficult to assess the importance of their individual importance as they ultimately work in harmony with each other, fine tuning how the anaerobic complements the aerobic systems for optimum performance. Clearly, knowing the terrain of a course and profile will help plan an optimal average power strategy to minimise time on ascents and against headwinds, and maximise chances for recovery on descents and forced non-pedalling sections, if target pace or MLSS has been breached. However, if the MLSS is breached too often for too long then the cyclist risks depleting the anaerobic work capacity numerous times which will drain liver glycogen and will thus eventually grind to a halt, depending on their optimal fuelling strategy. The dreaded ‘bonk’!

This leads me on to briefly discuss how calorific calculations for strategic pace setting could be factored in to ultra-long distance endurance events (12/24hr and iron-man distance events), where adequate nutrition and hydration are major contributors to success on the day. Some studies have observed that shorter distance more intense events are influenced less by optimal fueling, probably because of the reduced capacity to reconstitute the anaerobic system, before it is totally depleted in conjunction with practicalities and desire for food/ aqueous carbohydrate intake.

To demonstrate this I have used an example lactate profile from a different subject (Mr. D) where I have taken example sub-threshold pacing power-output strategies for a 24hr event, based on a final trained MLSS value of 263 Watts. From these, I have been able to work out calorific requirements for the duration, taking into consideration the proportion of fat/ carbohydrate aerobic oxidation and cycling efficiency. Table 1 below shows the total amount of energy that theoretically should be derived from the total calories expended at each target pace, with the recommended rate of calorie ingestion (Table 2) based on an assumed cycling efficiency 22.5%. As fat should be derived from existing adipose tissue stores, as long as the cyclists metabolism allows access to these (an important trainable marker) then the calories should be replaced endogenously assuming the athlete doesn’t have compromised fat stores, even a 75Kg subject at 5% fat will carry almost 4kg of fat, viscerally (around the organs and muscles) and ectopic (under the skin at typical fat storage sites). Although access to these stores throughout the range of work-rates may not be constant, stressing the aerobic use of glycogen by the muscles. Fat in the diet prior to the event day will bolster levels of circulating triglycerides and availability to reliant muscles exercising at sub-threshold intensities.

More important is making sure liver glycogen levels are not critically depleted, even though endogenous carbohydrate oxidation, is not necessarily repressed by exogenous intake of carbohydrates, but do serve to protect all-important liver glycogen levels and the re-circulation of lactate in the Cori cycle. Hence, maintaining adequate carbohydrate intake at an optimal rate matching the expenditure of calories will maximise exercise duration at these target paces. Basal Metabolic Rate should also be added to the total energy expended. Although sodium/potassium and calcium ion-transfer across cell membranes and tissues are markedly increased compared to rest, and so the figure calculated from a standard method maybe an underestimation,  possibly a minor one in relation to the total calories required for 24hrs.

Lactate_pacing_calorific rates_MrD.png

We can see from table 1 that the nearer we get to the MLSS power-output (263Watts in this example, Mr. D), the greater our requirements for carbohydrate is, as fat oxidation is too slow to meet the demands for energy. Improving these ratios in training should be a primary goal, as overall cycling efficiency will increase. Performing a series of steady-state trials (10 minute duration) at these target pace powers allow us to chose the optimal work-rate based on multiple factors. The lactate values at the end of table 1 confirm that 200Watts is the highest average steady-state aerobic power-output for Mr.D to target his 24hr TT pace. This is concluded from the fact that blood lactate concentration at the start (4’30”) of the 200W 10 minute trial is lower (1.4mmol/L) than both 180 and 190Watts (2.2 and 2.4 mmol/L, respectively) although these lower work-rate values at 4’30” are potentially elevated artificially due to the initial warm-up effect of the assessment on muscle metabolism. The lactate value at 200W continues to decrease after a further five minutes to 1.2mmol/L in a similar trend to the previous target powers. In contrast, the subsequent steady state trial at 210W concurs with a higher value at 1.9mmol/L which has no decreasing trend in concentration, after five minutes this remains the same, 1.9mmol/L.

Therefore, an approximate target pace at 200Watts, representing less aerobic stress on muscle glycogen (60% Fat metabolism/ 40% Carbohydrate metabolism) can be selected which should be maintainable for a prolonged period depending on external factors, such as temperature. Even so, focusing on the target speed required for victory, adjusting perfect pace power outputs for the terrain and the potential of drafting may mean that the target pace powers might fluctuate either side of this. Skewing the average median power below 200W if drafting is extended, or skewing the average median power above 200W if the terrain is hilly or the pace is erratic, at the peril of depleting precious muscle glycogen. Either way it would be possible to compensate for calorific expenditure to support a perfect pacing strategy. We can apply this concept in a similar way to events of different durations.

There area multitude of good reference resources for recommended fuelling and hydration strategies, which I will not go into too much detail here. There seems to be some variation in burn rates of carbohydrates (glycogen) under intense exercise (40 to 60g/hr), and optimal absorption rates for carbohydrate intake of possibly up to 90g/hr depending on the format, aqueous or solid.  Ideally isotonic, aka equi-osmolar (similar particle density as interstitial fluid and blood plasma) drinks are ideal for maintaining hydration and re-hydrating, ideal temperatures around 15C to avoid discomfort and promote intake. Carbohydrate concentrations of between 3-8% w/v are optimal for gastric emptying and absorption, but may not match adequate pace of refuelling at more demanding exercise rates. Hypertonic solutions (as recommended in Table 2) with greater concentrations of carbohydrate approaching 25% w/v may meet these demands more closely, to reduce excess volume intake and be more palatable than gel preparations, especially under hot conditions. However, avoiding gastric distress is important. Replacing sugar (glucose-fructose 50:50) based solutions with complex carbohydrate polymers like malto-dextrin or better still, amylopectin starch (a high molecular weight polymers and low osmolality) such as UCAN SuperStarch which facilitate gastric emptying, intestinal digestion and are completely absorbed and achieve a more stable blood glucose levels for longer to promote fat oxidation and avoids insulin fluctuations. See my previous post on SuperStarch here.

Table 1- Perfect Pace Power-output targets and associated calorific EXPENDITURE (Mr.D)


There seems to be some contention with the use of Fructose, in a general dietary context implicated in a host of health problems whether rightly or wrongly, and in sports science nutrition, with some commercial suppliers suggesting increase carbohydrate intake can be achieved through up-regulated GLUT transporters in the intestine and shuttling of cleaved fragments through MCT proteins to the liver. This may well take advantage of our bodies ability to gorge on this energy-rich fuel. However, fructose bypasses the important steps for ATP production as glucose substrates normally undergo, and enters de novo lipogenesis (DNL) as convenient precursors able to influence the formation and esterification of triglycerides and free fatty acids. Fructose is therefore a potent regulator of glycemia and glucose levels and upsets the balance of glycogenolysis, breakdown of glycogen, insulin levels and storage of glucose as glycogen in the liver in the absence of insulin release. Fructose may also disrupt stomach emptying and gastrointestinal absorption due to its impact on osmolality. But I will save the lengthy critical appraisal of Fructose for a different post.

Research has not significantly differentiated between the format of ingested calories and performance outcomes, mainly focusing on the desirable physical properties necessary for easier ingestion and digestion. Maintaining a continuous ‘sludge’ of low-sugar content complex carbohydrate similar to porridge or rice starch may be beneficial depending on the work-rate. The further up the power-output scale you go, i.e. the closer to MLSS and beyond, the muscles will require rapid glucose uptake as they maximise their contraction rate and glycogen is depleted, so higher content sugar products may have an advantage to preserve these stores, albeit sporadically during more intense efforts to avoid insulin spikes. Following a perfect pace strategy may not be ideal for all cycling events, but will give you a good reference to gauge your effort for sustainable pacing, which maybe adapted accordingly to the prevailing conditions. Using a power-meter in this way will help you integrate your perception of feel, with speed and physical ability under the varied conditions, and should support a successful fueling strategy throughout to avoid any upsets on race day.

Table 2- Perfect Pace Power-output targets and associated calorific CONSUMPTION recommendation (Mr.D)


24hrs can feel like a long time, without sleep! Although the event was raced more like a criterium than a Time-Trial, I captured some good data to infer conclusions from pacing strategies, whether the subjects complied with the proposed power-output strategies or not. Unfortunately due to inaccessible power data files, it is more convenient to display lap times as a surrogate of effort. I was able to quickly measure blood lactate within the first minute of arrival for each stop that all riders made together (hence there were many more stops than shown in the table) so that all the riders could get to work on re-fuelling unhindered.

Table 3- Final event results.


Graphically below, we see that each rider started at a similar pace for the first 2hrs of the race (approx 7 min 6sec per lap on average), although Mr.D was working harder during this time than both Mr. S & T shown by a higher blood lactate at almost 2hrs. Mr.D increased the pace further to 6.91 min/lap on average for the next 5 hrs, whereas Mr.S & T reduced or maintained their pace giving Mr. D the overall lead of all three by a large distance margin after 6.7 hrs. However after 11.4 hours, Mr. S had maintained his average pace very well while Mr.D started to slow considerably, and the lead had swapped.

Even though Mr.T had slowed to a practical target pace (his power meter connection failed), he had started to show signs of marked fatigue. At 14 hours into the race (5am the next day) Mr.S was riding very steadily, much below threshold still, indicated by a low lactate value of 1.8mmol/l. Whereas Mr.D had supplemented with a high dose of caffeine which served to facilitate his lap times, although he had previously been unable to take on board solid food for  5hrs prior to this (due to too fast a start) and was entering a period of hypogylcemia (3.5mmol/L lactate conc). Mr.D recovered over the next 3hrs but increased his effort too much, while Mr.S had maintained consistent pacing to secure his overall distance, finishing on his lowest lactate concentration (1.1mmol/L).

Possible conclusions would be: Secondary to drafting and positioning to find ‘free speed’ (whether steady-state or more erratic road-racing like criteriums) where recovery is very important to maintain effective power-outputs, keeping effort relative to absolute lactate values (Mr. S has a higher absolute threshold power, but lower lactate threshold concentration vs Mr. D) and threshold values, consistent pacing with lower values will facilitate endurance and reduce fatiguability over longer durations of work. Being able to recover at intensities much below your threshold, helps to reconstitute energy for subsequent efforts such as short hills even if the power-to-weight ratio is less. Mr.S will be able to recover much more quickly after the hill if the intensity required to maintain pace around the rest of the circuit falls much lower than his threshold value, when compared with Mr.D. Therefore, on a flatter course such as Le Mans, Mr.S may be more naturally inclined to perform better as his threshold power on the flatter sections is 13 watts higher, not much of a difference you would have thought. The energy saving equivalent to eating three large bananas in 24hrs!



Share Button

No Comments

Post a Comment

Your email address will not be published.

Please fill in the missing value *