Shownotes

Michele’s background

04:29 -

• I am a lecturer in applied sport and exercise science at the Open University in the United Kingdom, and a visiting research fellow at Loughborough University, where I completed my PhD and conducted the majority of my research on durability.
• Since 2022, I have also worked as Head of Strength and Conditioning and physiologist for the Italian Triathlon National Team, a role that included participation in the 2024 Olympic Games.
• I define durability as the capacity to withstand fatigue as exercise continues and accumulates over time. It can be expressed as the degree of performance decline, or the decline in key performance determinants, as prolonged exercise progresses.
• More specifically, durability reflects the ability to maintain physiological stability and preserve physiological parameters that are typically measured in a fresh state, even when the athlete becomes fatigued after a given duration or workload.

Classic model for endurance performance

06:10 -

• In 1991, one of the most influential endurance physiologists proposed a model to estimate endurance performance based on three primary parameters.
• These parameters are maximal oxygen uptake (VO₂max), the fractional utilization of VO₂max at lactate threshold or other submaximal intensities, and economy. 
• In the original model, economy referred specifically to running economy, as the framework was developed for running performance. However, the concept has since been expanded to include economy or efficiency in other endurance sports such as cycling, rowing, cross-country skiing, and similar disciplines.
• When combined, these three parameters allow endurance performance to be estimated with a high degree of accuracy, particularly for events lasting between approximately one hour and three to four hours.
• A well-known practical example of this model is Paula Radcliffe’s marathon world record performance. The work of her physiologist, Andy Jones, demonstrated that her marathon time could be predicted very closely—within a narrow margin—using measurements of VO₂max, running economy, and fractional utilization of oxygen.
• The research paper discussing this model was published in 2006, while Radcliffe’s record-setting performance took place in 2003, and her record stood for approximately 14 to 15 years.
• This example represents an exceptional performance and remains one of the clearest illustrations of how endurance performance can be effectively estimated through physiological determinants.

Research on long runs

09:05 -

• The rationale for this work emerged from a very practical perspective. In endurance training, particularly among marathon and half-marathon runners, long runs are a common component of training programmes and are typically prescribed to enhance the ability to resist fatigue.
• From an applied standpoint, this is effectively the practical expression of durability. Our objective was therefore to investigate whether the inclusion of long runs in training differentiated athletes in terms of durability.
• We approached this question using a cross-sectional design. Two groups of athletes were recruited who differed substantially in their training background but were closely matched for 10-kilometre performance, with both groups averaging a time of approximately 39 minutes.
• The primary distinguishing factor between the groups was exposure to long runs. The long-run group completed at least one run per week lasting longer than 90 minutes, whereas the non–long-run group was restricted to training sessions of no more than 70 minutes.
• Training data were collected retrospectively for the six months preceding the laboratory visit and used to characterize training exposure.
• Based on this information, we designed a protocol to assess durability. The initial visit consisted of a step test and a maximal test to determine physiological parameters in a fresh state.
• This was followed by a prolonged 90-minute run performed at lactate threshold intensity, corresponding to the lower boundary of the heavy intensity domain and closely aligned with the intensity expected during marathon performance.
• Running economy was assessed every 15 minutes, and changes from the start of the run to the 90-minute mark were compared between groups.
• In this context, lactate threshold was defined as the point at which blood lactate begins to accumulate above baseline, identified either by an increase of 0.5 mmol·L⁻¹ or using a log-log method, which yields very similar results. In terms of relative intensity, this corresponded to approximately 80% of VO₂max in our study.
• The results revealed a clear difference between the two groups. When examining changes from a fresh state, the group performing regular long runs showed a deterioration in running economy of approximately 3%, whereas the group not performing long runs exhibited a worsening of around 6%.
• This represents roughly a 50% lower drift in running economy in the long-run group compared to the non–long-run group. There was also a baseline difference in running economy, with the long-run group demonstrating a lower energetic cost of running from the outset.
• This likely contributed, at least in part, to the observed differences, as a better initial economy implies lower energetic demands throughout prolonged exercise.
• In addition, we observed differences in neuromuscular capacity. Maximal strength and power were assessed before and after the prolonged run, and the long-run group showed a markedly smaller decline in neuromuscular function compared with the control group.
• This suggests that regular exposure to long runs was associated not only with improved maintenance of running economy under fatigue but also with a greater preservation of neuromuscular function during prolonged exercise.

Neuromuscular function during exercise

13:30 -

• Because running economy is strongly influenced by neuromuscular factors, we decided to quantify changes in neuromuscular function and examine whether these changes were associated with alterations in running economy during prolonged exercise.
• At a group level, we observed a clear agreement: both running economy and neuromuscular function deteriorated with fatigue, and both variables clearly differentiated the two groups.
• However, when we performed correlation analyses across the entire cohort of 26 athletes, no meaningful relationship emerged between the change in running economy and the change in neuromuscular function.
• This finding highlights how multifactorial running economy is and how challenging it is to explain changes in economy through a single physiological factor.
• While neuromuscular function undoubtedly underpins running economy to some extent, in this study, it did not explain the magnitude of economy drift observed during prolonged exercise.
• This suggests that multiple interacting mechanisms contribute to the deterioration of the economy under fatigue.
• From a practical and conceptual perspective, these findings help explain why classical physiological models of endurance performance work well in some contexts but fail in others. In highly durable athletes, such as world-class performers, running economy may drift very little over the duration of a marathon, potentially on the order of 1% or less.
• This limited deterioration allows traditional models based on VO₂max, fractional utilization, and economy to predict performance with high accuracy.
• In contrast, in athletes whose running economy deteriorates substantially—such as a 6% impairment over 90 minutes—performance during a marathon is likely to deviate markedly from model predictions, as these models do not account for fatigue-induced drift.
• This limitation becomes particularly clear when considering lactate threshold–based predictions. While many individuals, even those with little training, may demonstrate a metabolic steady state at a given intensity, they are unable to sustain that intensity for a marathon duration without specific training aimed at fatigue resistance.
• This inability reflects a lack of durability rather than an absence of physiological thresholds. As performance deteriorates with fatigue, models that ignore durability become progressively less accurate.
• At the extreme end of endurance performance, recent applied case studies further support this concept. In the case of elite Ironman performance, simulations conducted prior to a major competition showed that running economy drifted by less than 2% over the entire event, even after prolonged cycling and a full marathon.
• This illustrates why classical models remain highly effective for elite athletes, whose physiological stability under fatigue is exceptional. However, as durability decreases, model assumptions must be adjusted to account for fatigue-related performance decay.

Goals of the durability research

17:31 -

• I believe this process is already underway in cycling, largely because of the widespread availability of power data. In this context, durability can be assessed through the decay of maximal mean power across different durations, such as 10 seconds, 30 seconds, 5 minutes, or 20 minutes.
• Clear differences are already evident between elite or WorldTour athletes and junior or under-23 athletes, indicating that this area is actively being explored in applied practice.
• From a more mechanistic and physiological perspective, I believe that testing must be highly specific to the demands of the event being studied.
• The case study previously mentioned involving Magnus Ditlev was designed with this specificity in mind, which is why it proved effective. However, particularly in highly performing athletes, it is often necessary to push the athlete very close to their limits in order to detect meaningful changes.
• This creates a challenge: testing protocols must be sufficiently demanding to reveal durability-related effects, while avoiding excessive fatigue that would compromise recovery and subsequent training.
• For example, asking a marathon runner to complete 35 kilometres at marathon pace effectively represents 80–85% of a full marathon and would require a substantial recovery period.
• For this reason, testing protocols must be carefully fine-tuned to maximize the information obtained while minimizing unnecessary physiological cost. In elite athletes, this balance is especially critical. In lower-performing athletes, durability-related drift tends to occur earlier, making assessment somewhat easier.

Main takeaways of this research

19:42 -

• I would highlight a few key takeaways. First and foremost, long runs appear to be essential for developing durability. While we did not explore training prescription in detail here, the findings strongly support the inclusion of long runs when the goal is to improve fatigue resistance.
• Another factor that differentiated the groups, which we did not previously discuss, was overall running volume. The group performing long runs averaged approximately 51 kilometres per week, whereas the non–long-run group averaged around 30 kilometres per week.
• This difference in training volume may also have contributed to the observed differences in durability.
• When we examined correlations, the variable most strongly associated with changes in running economy was the duration of the longest long run.
• The second strongest association was total training duration. However, given the cross-sectional nature of the study, I would be very cautious in interpreting these relationships. Correlations in this context cannot establish causality.
• The logical next step would be a longitudinal training intervention in which one group introduces long runs into their training while the other does not. This is a direction we may pursue in future research.
• It is also important to consider the performance context of the athletes involved. The groups were matched on 10-kilometre performance, which allowed us to recruit athletes who did not routinely perform long runs.
• Many of these athletes were primarily track-focused, training for events such as 800 m, 1500 m, and in some cases 3 km or 5 km. The 10-kilometre distance represented a practical compromise, as it is extremely difficult to find athletes who do not perform long runs while preparing specifically for half-marathon or marathon events.
• This is consistent with the notion that as event duration increases, the absence of long-run training becomes increasingly limiting for performance.

Physiological determinants of performance in long runs

22:43 -

• This work builds on another study that we published recently, which included fourteen well-trained marathon runners. The average marathon performance of the group was approximately 2 hours and 46 minutes.
• To be eligible for inclusion, athletes were required to have completed a marathon in under three hours within the six months preceding the start of the study.
• The experimental design began with a standard physiological assessment to determine running economy, physiological thresholds, and VO₂max in a fresh state. Following this initial assessment, athletes completed a prolonged preload run lasting either 90 or 120 minutes at an intensity within the heavy domain.
• This intensity was defined as a 10% delta between the first and second lactate thresholds, placing it just above the first lactate threshold and therefore very close to marathon race intensity.
• After the preload, the step test and maximal test were repeated in order to assess how key physiological determinants changed with accumulated fatigue.
• We observed that several physiological parameters were already altered after 90 minutes of running, which was consistent with our previous research. More notably, after 120 minutes, these changes became substantially more pronounced.
• For example, VO₂max declined by approximately 3% after 90 minutes, but by around 7% after 120 minutes. This indicates that the rate of decline was almost four times greater between 90 and 120 minutes than during the initial 90 minutes of exercise.
• A similar pattern was observed for speed at lactate threshold, which we used as a proxy for performance within the framework of Joyner’s model, as well as for fractional utilization at lactate threshold. Fractional utilization increased slightly, rising from approximately 80% of VO₂max at baseline to around 84–85%.
• In my interpretation, this increase was driven primarily by the reduction in VO₂max. As maximal aerobic capacity declined, the same absolute intensity represented a higher percentage of the available maximum.
• At the same time, the deterioration in running economy contributed to a reduction in threshold speed, which partially offset the increase in fractional utilization.
• We also calculated fractional utilization at the actual speed of the prolonged trial. Respiratory gases were measured at the start of exercise, after 90 minutes, and after 120 minutes.
• Fractional utilization was then calculated at each time point by relating the measured VO₂ to the cost of running and the contemporaneously measured VO₂max. In a fresh state, this calculation was based on baseline VO₂max and economy, while subsequent values reflected the altered physiological state at 90 and 120 minutes.
• Using this approach, we found that the relative exercise intensity increased markedly, from approximately 79% of VO₂max at the start to around 91% after two hours.
• This represents a substantial shift from the lower boundary of the heavy intensity domain to its upper limit. In some athletes, the exercise intensity clearly entered the severe domain, indicating that they would not have been able to sustain the workload for much longer and would likely have reached exhaustion within 15 to 20 minutes.
• We did not examine the relationship between these changes and the athletes’ habitual training characteristics, such as the duration of their typical long runs.
• The study was not designed for this purpose, and with a sample size of fourteen athletes, any correlational analysis would have been difficult to interpret due to the high level of noise inherent in small samples.
• While this is an interesting question and one that we discussed as a potential direction for future research, it was not possible to address it within this study.
• However, related evidence has recently emerged from a study published in the European Journal of Sport Science. In that work, a 90-minute durability test was used to assess changes in physiological parameters, and these changes were subsequently correlated with marathon performance.
• The authors found that the change in speed at lactate threshold from baseline to 90 minutes was associated with marathon performance. This finding supports the idea that durability is not only a physiological construct but also a meaningful determinant of endurance performance.

Main takeaways of the studies

28:26 -

• I believe the most interesting aspect of this work lies in the time-dependent changes observed in physiological determinants of performance.
• Parameters such as running economy, speed at lactate threshold, VO₂max, and fractional utilization do not all change in the same way as exercise duration increases.
• The running economy appeared to deteriorate in a relatively linear manner, whereas the other variables displayed clearly non-linear behaviour. This has important implications for durability testing.
• If the objective is to assess durability in a meaningful way, the prescribed exercise duration and intensity must be highly specific to the demands of the marathon. Otherwise, the test may fail to capture the physiological stresses that are truly relevant to marathon performance.
• Another key observation is that all of the classical determinants of endurance performance change under fatigue, but they do so through different mechanisms.
• Fractional utilization increased primarily as a consequence of the substantial reduction in VO₂max, while running economy and speed at lactate threshold worsened over time.
• This divergence has potential implications for how durability should be interpreted from a physiological standpoint and suggests that changes in performance under fatigue cannot be explained by a single variable.
• The increase in fractional utilization should not be interpreted as a positive adaptation. In my view, it simply reflects the decline in maximal aerobic capacity. As VO₂max decreases, the same absolute workload represents a higher percentage of the available maximum.
• This apparent increase in fractional utilization, therefore, indicates a reduced ceiling rather than improved efficiency. As a result, the range of intensities that can be sustained within the heavy and severe intensity domains becomes narrower.
• In practical terms, this represents a reduction in the athlete’s physiological buffer rather than a compensatory mechanism that offsets fatigue-induced declines in performance.

Practical applications for coaches and athletes

31:26 -

• I will first address this topic from a durability perspective, as this is a relatively newer concept, although much of what we observe aligns closely with established principles of endurance training.
• As discussed previously, the inclusion of long runs appears to be beneficial. From a practical standpoint, this is almost self-evident, as long runs are a fundamental component of marathon preparation regardless of how durability is defined or measured.
• Although further experimental confirmation is still required, this approach appears to be both logical and effective for enhancing marathon performance.
• Training volume also seems to be positively associated with durability, and this relationship has been observed across different endurance sports. In addition, accumulated training exposure over multiple years appears to influence durability, with several longitudinal and cross-sectional studies in cycling supporting this association.
• There is also a growing body of research specifically investigating training interventions aimed at improving durability. In our own work, we have examined strength training interventions and observed positive effects on durability, both in terms of running economy and time to exhaustion.
• Specifically, ten weeks of heavy strength and plyometric training resulted in a reduced drift in running economy and an improved time to exhaustion. Similar benefits of strength training have already been demonstrated in cycling and cross-country skiing.
• From a cycling perspective, some studies have shown that incorporating high-intensity efforts within long rides can be effective in improving durability, as reflected by improved five-minute time trial performance following prolonged exercise.
• Taken together, these findings suggest that durability can be influenced through targeted training strategies.
• From a practical standpoint, however, the overarching message remains consistent with traditional endurance training principles. Gradually increasing training volume and exposure to long-duration sessions is likely to remain one of the most effective ways to improve marathon performance.

Strength training

33:53 -

• I will keep this brief. In the study, we used a very simple strength-training design consisting of five exercises. This is not necessarily how I would approach training in applied practice, as there is usually more time and greater scope to progress gradually from one exercise to another.
• However, for the purposes of the study, the programme included a half squat, a single-leg leg press, and an isometric calf raise to target maximal strength. In addition, we implemented a progression of vertical and horizontal plyometric exercises.
• The vertical plyometrics consisted of traditional jumps and drop jumps, while the horizontal plyometrics included variations of bounds and hops that were progressively advanced over the intervention period.
• While this structure represents a solid starting point, in practice, I would expand upon it. I would include additional power-oriented work and potentially introduce strength-endurance training toward the end of individual sessions or toward the later stages of a periodization block leading into competition.
• I would also place greater emphasis on the posterior chain by incorporating exercises such as hip thrusts or deadlifts, or by progressing from bilateral movements like the squat to unilateral exercises such as the Bulgarian split squat.
• These progressions are difficult to implement in a research setting, as the time required for athletes to learn new movements can significantly reduce the effective duration of the intervention.
• In this study, the training period lasted ten weeks, whereas in applied practice, I would expect to work with an athlete over six to ten months, or even across multiple years, to develop strength in a more comprehensive and sustainable manner.
• Regarding strength-endurance work, this is an area I am still actively experimenting with. There is currently limited evidence supporting the effectiveness of strength-endurance training for performance enhancement, which may be due to the fact that it is often implemented in isolation for relatively short periods.
• In my view, it is more logical to prioritize maximal strength and power development first, followed by strength-endurance work. My practical approach has been to target muscle groups that tend to fatigue earlier in the specific activity being trained.
• When working with triathletes, this has involved developing strength-endurance capacity in the calf muscles and in upper-body muscles that are critical for swimming propulsion.
• In practice, this has involved very high repetition schemes, ranging from approximately 40 to 100 repetitions, typically performed for one or two sets per session and repeated once or twice per week.
• This type of work often induces muscle soreness during the initial sessions, usually lasting a couple of days, but this response tends to diminish after two or three exposures, as expected.
• I cannot definitively quantify the effectiveness of this approach, as it is extremely difficult to attribute changes in performance to a single training stimulus.
• Nonetheless, from a periodization perspective, this strategy appears logical, which is why we have continued to explore it in applied settings.

Intensity in long runs

38:45 -

• There are several relevant studies from the research group in Norway. In these studies, sprint interval training was incorporated into long cycling sessions, typically lasting three to four hours.
• The general approach involved performing short sprint intervals at regular intervals during the ride, for example, once every hour. While I do not recall the precise details of the protocol, the key finding was a clear positive effect on performance under fatigued conditions when assessed before and after the intervention.
• Across this line of work, the group published three or four studies with very similar experimental designs. The main differences related to the timing within the season, with some studies conducted during the preparatory phase and others during altitude training camps, where the focus was on maintaining physiological determinants.
• Despite these contextual differences, the training interventions were remarkably consistent, and the adaptations occurred over relatively short periods of two to three weeks.
• From a physiological perspective, I believe these adaptations may be linked to increased recruitment of type II muscle fibres early in the prolonged sessions.
• This likely accelerates muscle glycogen depletion and disrupts homeostasis, thereby creating a strong stimulus for adaptation. Over time, these fibres appear to adapt to better tolerate prolonged work under fatigued conditions, which may explain the observed improvements in performance durability.

Potential interventions for improving durability

40:49 -

• From a practical standpoint, long runs and long rides represent the most straightforward and effective intervention to improve durability. They are, in many ways, a low-hanging fruit.
• Beyond this, there are several acute interventions that may also influence durability. Some evidence suggests that carbohydrate intake can have a positive effect, with recent data indicating that consuming 120 grams of carbohydrates per hour resulted in a smaller deterioration in running economy compared with intakes of 60 or 90 grams per hour.
• However, this higher intake was also associated with substantial gastrointestinal discomfort, highlighting a clear trade-off, particularly in running, where the tolerable amount of carbohydrate intake is often limited.
• There is also some emerging evidence regarding caffeine, although the current findings are not yet robust. Nevertheless, it would be reasonable to expect caffeine to exert a positive effect.
• In addition, footwear technology may play a role. Carbon-plated and other plated shoes could theoretically have a beneficial effect on durability. However, the current evidence is mixed.
• Several studies published over the past year have reported either improvements in running economy under fatigue or no effect at all.
• Importantly, none of these studies, to my knowledge, demonstrated a clear improvement in durability itself, defined as a reduction in the decay from a fresh to a fatigued state.
• In cases where shoes improved economy in a fresh condition, this effect was either maintained or diminished under fatigue, but the rate of deterioration itself was not altered.
• Regarding training structure, long runs are often included even in programmes targeting shorter events, such as 10-kilometre races, where they are typically performed at a steady, moderate intensity rather than at high intensity.
• I do believe there may be potential benefits to incorporating intervals within long runs, even for athletes not specifically preparing for a marathon, although this remains speculative due to the lack of direct research. This is a concept I have discussed with Renato Canova, and it represents a relatively advanced training stimulus.
• From a physiological perspective, the rationale would be to induce early fatigue in type II muscle fibres. Incorporating high-intensity efforts early in a long run may accelerate muscle glycogen depletion in both type II and type I fibres, increasing the recruitment of less efficient fibres earlier in the session.
• This could potentially enhance the adaptive stimulus. While some argue that the timing of these efforts is less important as long as the fibres are sufficiently fatigued, my view is that performing them earlier may be more effective.
• This mechanism could help explain why, in our own work, improvements in running economy were observed primarily in a fatigued state rather than in a fresh condition.
• It is possible that adaptations occurred predominantly in type II fibres, which are recruited later during prolonged exercise and therefore only influence economy under conditions of fatigue.

Big open questions about durability

45:21 -

• I believe the two main topics that now require further attention are how durability can be effectively trained, depending on the athlete and the specific sport, and how it can be assessed in a valid and practical way.
• Testing durability involves a wide range of methodological considerations. Decisions must be made regarding whether assessments should be conducted indoors or outdoors, whether environmental conditions such as temperature and humidity should be controlled or replicated, how long the exercise bout should last, and at what intensity it should be performed.
• All of these factors must be aligned with the specific demands of the sport if durability testing is to be meaningful.
• This leads to the broader question of ecological validity versus reliability. Highly controlled indoor testing offers strong reproducibility, but may lack relevance to real-world performance.
• Conversely, outdoor testing is more ecologically valid but introduces greater variability. Hybrid approaches may offer a useful compromise. For example, athletes could perform prolonged exercise outdoors and return to the laboratory at predefined intervals for physiological assessments.
• This approach may be advantageous not only from a scientific perspective but also practically, as it reduces the physical and psychological burden of remaining on a treadmill or stationary trainer for extended periods. It may also improve efficiency from a practitioner’s standpoint by reducing the continuous time required in the laboratory.
• This hybrid model has already been applied successfully in previous case-study work, such as the Magnus Ditlev investigation. In addition, the increasing availability of field data presents further opportunities.
• In cycling, field-based power data have been widely used to model performance, and similar approaches could potentially be extended beyond cycling.
• Attempts have been made in running, although this is more challenging due to the greater number of variables influencing running speed compared with cycling power, which is a particularly strong predictor of performance. Nevertheless, advances in this area remain promising.
• Similar approaches could also be applied in triathlon. If comprehensive field data were available from both Olympic-distance and long-distance triathlon races, there would likely be substantial insights to be gained regarding durability and performance across disciplines and race formats.
• Finally, there is an important terminological consideration. Throughout this discussion, the term durability has been used, but within the scientific literature, some groups refer to this concept as physiological resilience. These terms are currently used somewhat interchangeably.
• In a recent response to a viewpoint published in the Journal of Applied Physiology, we proposed a distinction between the two.
• Specifically, we suggested using physiological resilience to describe the physiological factors underpinning performance, such as VO₂max, running economy, and fractional utilization, while reserving the term durability for performance-based outcomes, including time-trial performance, time to exhaustion, race results, or speed at lactate threshold.
• The rationale for this distinction is that similar changes in performance can arise from different underlying physiological mechanisms. One athlete may experience a greater decline in VO₂max, while another may show a larger deterioration in running economy.
• Differentiating between performance outcomes and their physiological underpinnings may therefore provide clearer guidance for training prescription and long-term performance development.

Things that Michelle changed in his coaching approach

50:48 -

• I am not heavily involved in day-to-day endurance coaching. While I work closely with athletes at the national team level, my role there is focused on physiological support rather than endurance training prescription.
• I do, however, work with some non-professional athletes, and my overall approach has remained fairly consistent over the years. Much of my work revolves around providing structured support and gaining a deeper understanding of how athletes respond during competition.
• This includes detailed race monitoring, which we have applied extensively in triathlon using tools such as running power meters to analyze pacing strategies.
• I believe there is growing potential to expand this approach further, particularly by examining the effectiveness of carbohydrate intake during racing and identifying the most appropriate ingestion rates for individual athletes.
• We have not yet systematically tested durability in this applied context, but I believe it could be particularly valuable, especially when working with younger athletes.
• In practice, we often observe that some developing athletes perform well in sprint formats or mixed team relay events, yet struggle significantly when transitioning to Olympic-distance racing.
• Understanding their durability characteristics could help explain these discrepancies and guide more appropriate long-term development strategies.
• From a strength-training perspective, we have recently begun to place greater emphasis on strength-endurance work. Over the past three to four years, we have also consistently used velocity-based training as a means of providing objective feedback to athletes during resistance training.
• This approach has proven useful for monitoring lift quality and managing fatigue. By setting predefined fatigue thresholds, we can reduce training volume when necessary, which may be beneficial for overall training optimization and load management.
• Importantly, this method allows us to adjust training loads in real time. If an athlete lifts a given load at a slower velocity than expected, we can immediately reduce the weight, sometimes even during the warm-up, rather than rigidly adhering to a prescribed load that may not be appropriate given the athlete’s accumulated fatigue from previous sessions.
• Velocity-based training is grounded in the linear relationship between force and velocity for a given athlete. If lifting velocity is reduced on a particular day, this likely reflects a reduction in maximal capacity, indicating that training loads should be adjusted accordingly.
• For this purpose, we use a linear encoder rather than accelerometer-based systems. Linear encoders operate via a tethered string attached to the barbell or moving segment, allowing direct measurement of displacement and velocity.
• Specifically, we use the Vitruve system, which I find to be very user-friendly. There is a free version suitable for individual use, as well as a team-based version, which we have been using.
• Based on our experience, it is a practical and accessible option for practitioners interested in implementing velocity-based training.

Most interesting trends in the scientific world

55:53 -

• There are several emerging areas of interest, although I am recalling these from memory and may be missing some details. Nutrition is certainly one of the major topics, particularly the role of increasing carbohydrate intake and how much athletes are actually able to tolerate and utilize.
• There is some indication that even intakes as high as 150 grams of carbohydrates per hour may not fully meet the needs of certain athletes with very specific metabolic characteristics. However, this approach could be counterproductive for less well-trained athletes.
• Their overall energy expenditure is substantially lower than that of elite athletes, and while intakes of 90 to 120 grams per hour may be appropriate for elite triathletes or cyclists, this does not mean that amateur athletes would benefit to the same extent.
• In fact, they may experience significantly greater gastrointestinal discomfort, as they neither require nor can tolerate such high intakes.
• Sodium bicarbonate is another intervention with potential relevance. It has well-established positive effects in shorter-duration events, such as middle-distance races, and there may also be benefits extending to longer events, possibly even the marathon.
• At present, the evidence in this area is limited, but I expect further research to emerge in the near future.
• Footwear technology is another important consideration. There is substantial evidence demonstrating a positive effect on performance, but its influence on durability remains less clear.
• This is an area that will likely receive more attention going forward. From a training perspective, I find recent longitudinal studies on world-class athletes particularly interesting.
• Until around ten years ago, there was very little research documenting how elite performance develops over time, but this body of work is now growing. While these studies may not directly inform day-to-day training prescriptions, they provide valuable insight into what it actually takes to become a world-class athlete.
• A consistent theme across these studies is that, beyond having exceptional physiological capacity, elite performance is underpinned by more than a decade of regular, consistent, high-volume training.
• Another area that has gained attention more recently concerns very high training volumes sustained across an entire season. Some studies have examined total energy expenditure over long periods and have proposed the existence of a metabolic ceiling, often described as approximately 2.5 times resting metabolic rate.
• Beyond this level, it has been suggested that athletes may struggle to maintain body mass and that training effectiveness may decline.
• A large study involving elite cyclists provided evidence supporting this concept, while several case studies have suggested that values closer to 2.7 or even 2.8 times resting metabolic rate may be achievable for limited periods.
• These findings indicate that while there may be practical limits to sustainable energy expenditure, there is also considerable individual variability.

Michelle’s favourite research studies

1:00:45 -

• Here are a few key references that come to mind. The first is the well-known case study on Paula Radcliffe published by Jones in 2006.
• This work provides a clear illustration of how endurance performance determinants evolve over time, documenting changes across roughly a decade and demonstrating how physiological factors underpin world-class performance development.
• A second important contribution is a comprehensive review published in 2023, which synthesized available case studies and longitudinal investigations involving elite and world-class athletes.
• This review examined long-term training and physiological development across multiple athletes and provided an integrated overview of how performance evolves at the highest level.
• This paper was published in Sports Medicine and included Øyvind Sandbakk as part of the author team.
• More recently, the same research group published another longitudinal study focusing on cross-country skiers.
• This work followed three world-class and three elite skiers, tracking their training characteristics, physiological testing, and performance development over the course of their careers. In some cases, the observation period spanned approximately 15 years or more.
• The study highlighted clear differences in development patterns between elite and world-class athletes and was published in Sports Medicine Open in 2025, again involving Øyvind Sandbakk and colleagues.

Michele’s future study

1:02:50 -

• I will keep this brief, as it may be more appropriate to expand on these findings once the work is published.
• Together with Luís Muguin, a PhD student at Loughborough University, I had the opportunity to access a very large dataset of athletes spanning a wide performance spectrum, from amateur to international level. The data were obtained through Loughborough Sport, which provides performance and commercial sport science services within the university.
• In total, we analyzed data from approximately 900 athletes across running and cycling. All athletes completed a standard step test protocol, from which running economy or cycling efficiency, VO₂max, and physiological thresholds were assessed.
• We then conducted large-scale correlation and multiple linear regression analyses to examine how physiological determinants relate to performance proxies. These proxies included speed at lactate threshold, power at lactate threshold, and speed or power at lactate turn point.
• The regression models were used to quantify the relative contribution of VO₂max, fractional utilization, and economy to these performance-related thresholds.
• Several collaborators provided valuable feedback throughout the process, and the work is now close to completion. We also examined sex-based differences.
• While clear physiological differences between male and female athletes were present, the relationships between physiological determinants and performance were remarkably consistent across sexes. In other words, although absolute values differed, the relative importance of each determinant for performance appeared similar in male and female athletes.
• While it is still early to discuss the results in detail, the primary finding is clear. A high VO₂max is a key requirement for elite performance. Among all variables examined, VO₂max showed the strongest association with performance, followed by economy.
• Fractional utilization was also statistically significant, but its influence was substantially smaller compared with VO₂max and economy. It remains an important factor, but its contribution is comparatively limited.
• An additional methodological point emerged from the analysis. In running, expressing physiological variables relative to body mass produced stronger models, whereas in cycling, absolute values provided better explanatory power.
• In my view, this reflects the non–weight-bearing nature of cycling, particularly under flat conditions, where absolute power output is more closely linked to performance than body-mass–normalised values.
• These findings align well with applied observations in talent development and performance modelling, and I expect the full study to be available in the coming months, at which point the lead author will be able to discuss the work in greater depth.

PhD opportunities

1:07:06 -

• There are two or three final points that may be worth sharing. First, we currently have a fully funded PhD studentship focused on durability, which is open until the 16th of January.
• The position is available to both UK and international students. Information about the studentship has been shared on LinkedIn and Twitter, and anyone interested is welcome to reach out to me directly if they would like to discuss the opportunity in more detail.
• In addition, we are conducting a survey involving high-performing female athletes and the professionals who work with them.
• The aim is to understand how research is perceived within this population and to identify where athletes and practitioners feel the main gaps currently lie.
• The intention is to use this feedback to help guide future research efforts in a more targeted and relevant way. I can share the link to the survey, and it would be greatly appreciated if it could be distributed more widely.
• Finally, Renato Canova and I have recently completed a book together, which is currently being finalized in Italian and is expected to be published early next year. The intention is to subsequently produce an English version.
• This project has been particularly meaningful, as it provided an opportunity to capture and preserve Renato’s extensive experience, knowledge, and historical perspective on training and athletics. Working with him on this has been invaluable, and I hope the book will soon be accessible to a wider international audience.

One sport-related and one non-sport-related recommendation

1:09:41 -

• I am currently based in the United Kingdom, in Loughborough. However, if I were to offer a recommendation, I would point to the place where I was born and raised. I come from the area around Bergamo, in northern Italy, very close to Milan.
• Visiting the city itself is something I always recommend. The historic upper town is particularly beautiful, and simply walking through it, enjoying the green spaces and the local food, is an experience in itself.
• From a sports and outdoor perspective, I would strongly suggest skiing in the nearby Alps. There are several excellent options within reach, including locations such as Bormio or Livigno, which will host the Olympic Games in the near future, as well as the Dolomites.
• Being in these environments offers a unique sense of freedom that I have rarely experienced elsewhere. While hiking can provide something similar, skiing in the mountains, surrounded by snow under clear skies, is truly exceptional and remains a very special experience for me.

LINKS AND RESOURCES

• Michele’s ResearchGate
• Michele’s past appearances on That Triathlon Show
  • Michele Zanini (part 1) | EP#393
  • Michele Zanini (part 2) | EP#394
• Episodes referenced
  • Including sprint training in cycling – training responses and peformance improvements with Nicki Winfield Almquist, PhD | EP#300
  • Nitrate loading, marathons, and endurance sports science with prof. Andy Jones | EP#187
  • Cycling training methods from the WorldTour with Stephen Barrett, Head Coach of Decathlon AG2R La Mondiale and coach of Felix Gall
• Papers cited
  • The Physiology of the World Record Holder for the Women's Marathon - Jones 2006
  • Long-Term Development of Training Characteristics and Performance-Determining Factors in Elite/International and World-Class Endurance Athletes: A Scoping Review - Staff 2023
  • Long-Term Development of Training Characteristics and Physiological Determinants in World-Leading and Medal-Winning Female Cross-Country Skiers: A Three Decade Comparative Analysis - Staff 2025
  • Aerobic Energy Turnover and Exercise Economy Profile During Race Simulation in a World-Record-Breaking Male Full-Distance Triathlete - Christensen 2024 (the Magnus Ditlev case study, already a modern classic in triathlon science)
  • Durability of Parameters Associated With Endurance Running in Marathoners - Hunter 2025
• Female athlete survey 
• Durability PhD studentship: LinkedIn Post  and University portal