Podcast, Science and Physiology

Jem Arnold | EP#420

 December 11, 2023

By  Bernardo Gonçalves


Jem Arnold - That Triathlon Show

Jem Arnold is a PhD candidate at the University of British Columbia, where his primary research focuses on sport-related vascular conditions. Jem is a physiotherapist, a physiologist, and an avid cyclist and trail runner interested in optimising endurance training.

In this episode you'll learn about:

  • Including training above "threshold" in your program improves VO2max, but may not improve Time Trial performance more than training only up to threshold-intensity (a recent meta-analysis)
  • Implications of these findings for athletes and coaches - what does it mean in practice?
  • How can we interpret group-level research for individual-level training programming and application?
  • Why even big sample sizes and data sets might be of limited value for the individual - the larger the sample size, the greater the confidence in the "group mean", but individual prediction may still be way off
  • What is the variability of heart rate, RPE, lactate, and muscle oxygen saturation from day to day? 

Sponsored by:

Zen8 Swim Trainer Logo

The ZEN8 Indoor Swim Trainer allows you to improve technique, power, and swim training consistency. You can target specific aspects of your stroke, like catch and pull-through, work on core activation and body position, and make sure you stay consistent in your swim training even when you don't have time to go to the pool. Try the Zen8 risk-free for 30 days, and get 20% off your first order on zen8swimtrainer.com/tts.


FORM Smart Swim Goggles give you unprecedented real-time feedback in your swim training through a display on the goggle lens. See every split to stay on pace, track your stroke rate and don't let it drop, use heart rate to become more scientific and precise with your training (through integration with Polar HR monitors) and analyse more in-depth metrics post-swim in the app. You can also use a vast library of workouts or training plans, or build your own guided workouts. Get 15% off the goggles with the code TTS15 on formswim.com/tts.


Jem's background

02:33 -

  • I'm a PhD candidate at the University of British Columbia in Vancouver, Canada. 
  • My focus is primarily on working with endurance athletes dealing with vascular issues related to sports. This includes assisting cyclists with flow limitations in the iliac artery and addressing concerns such as a popliteal artery or compartment syndrome in runners and triathletes.
  • In addition to my background as a physiotherapist, my PhD work delves deeply into physiology. I conduct metabolic testing alongside a nurse, exploring near-infrared spectroscopy and muscle oxygenation areas.
  • My work encompasses both clinical and performance aspects. While I previously coached individuals, I currently provide occasional consulting for other coaches and organisations. 
  • Similarly, with metabolic testing, I engage in data analysis for various centres and projects, allowing me to work from home and immerse myself in the fascinating world of data analysis.

Jem's latest meta-analysis overview

04:43 -

  • I had the privilege of collaborating with Michael Rosenblatt on a project that delved into training optimisation. 
  • Michael, who has previously shared his insights on the show, conducted a series of meta-analyses influenced by the challenges posed by COVID during his PhD. 
  • Our collaboration began in Vancouver, and our mutual passion for training and optimisation led us to explore various aspects of the subject.
  • Our initial focus centred on programming a single training session, but our curiosity naturally expanded to encompass broader perspectives. We found ourselves contemplating the structure of training block interventions and even delving into the complexities of periodisation. 
  • Steven Seiler's involvement added a valuable dimension to our project, leveraging data from his previous studies, particularly the renowned four-by-four, four-by-eight, and four-by-16-minute protocols.
  • The fundamental question we aimed to address was comparing two groups of endurance-trained athletes. One group exclusively undertook training below the threshold, akin to a maximal metabolic steady state or FTP, while the other incorporated high-intensity training above the threshold. 
  • Our primary focus was on assessing the impact of this difference on critical parameters such as VO2 max and time trial performance.
  • Michael Rosenblatt spearheaded the meticulous gathering and analysis of studies, with my role primarily involving brainstorming and shaping the narrative around the data. 

Results of the study

07:30 -

  • The key outcome measures of the study focus on VO2 peak (also referred to as VO2 max) and time trial performance. The addition of high-intensity training to the participants' programs was examined, with VO2 peak being tested using a ramp test. The two groups under scrutiny included all low to medium-intensity training and the others incorporating high-intensity interval training.
  • Surprisingly, the group incorporating high-intensity training demonstrated a more significant improvement in VO2 max compared to the low-intensity group. 
  • This finding aligns with prior research suggesting the benefits of high-intensity training on VO2 max. 
  • However, the unexpected result was that time trial performance did not differ between the two groups. Despite the high-intensity group showing a more significant increase in VO2 max, their time trial performance remained on par with the low-intensity group. 
  • This unexpected outcome challenges the assumption that enhanced VO2 max directly translates to improved time trial performance.

The amount of moderate training in all training interventions

09:51 -

  • The challenge lay in the varied ways studies operationalised thresholds, such as MMSS (maximal metabolic steady state), denoting the maximum sustainable energy production oxidatively. 
  • The distinction between thresholds like CP, FTP, and MLSS added complexity. Determining the lower threshold, whether aerobic or lactate threshold, proved challenging. 
  • We had difficulties in explicitly addressing this issue due to the heterogeneity in how thresholds were defined across studies. 
  • Future work might delve into this aspect, but groups engaging in control training or maintaining standard base training were grouped for the current analysis. 
  • The emphasis was on moderate to heavy domain training, encompassing various zones (e.g., one, two, three, four in zone systems) with a recognition of the limitations in teasing out finer distinctions.

How the studies controlled training

11:37 -

  • Each study's design was well-controlled, ensuring comparability. The matching criteria included time, total work, or training volume. 
  • Despite inherent clinical heterogeneity stemming from differences in sports, interval programming, and intervention lengths (ranging from two to 12 weeks), the meta-analysis surprisingly revealed that these variations didn't significantly impact the main results. 
  • The key finding remained consistent: VO2 max exhibited more significant improvement with high-intensity training, while time trial performance demonstrated improvement without significant dependence on training intensity. 
  • The robustness of these outcomes across diverse study parameters underscored the resilience of the main findings.

Why time trial performance did not improve further with high-intensity training

13:25 -

  • There are methodological limitations in this study.
  • One key consideration was the timeframe of the interventions, ranging from two to 12 weeks. Despite this significant variation, the outcomes remained consistent and encouraging. 
  • However, translating physiological improvements, such as enhanced VO2 max, to tangible performance outcomes might require more time. 
  • The possibility of periodisation effects on training blocks and their sequence emerged as a question mark. 
  • Additionally, there is the issue of a taper period before a target event, highlighting the potential impact on training volume and intensity. The absence of information on participants' retesting timing after the intervention led to speculation about the influence of residual fatigue on performance outcomes. 
  • Despite these speculative aspects, it is essential to consider tapering in athletic training, acknowledging its relevance in gradually reaching peak performance. (the need to account for mental and physical exhaustion during intense training blocks)
  • There are responders and non-responders in training interventions, meaning the existence of athletes who exhibit varying degrees of improvement or even regression. 
  • Some well-known studies addressed non-responders by intensifying training through higher volumes of high-intensity exercise. 
  • The results indicated that these non-responders transformed into responders with improved outcomes. 
  • However, this is a controversial subject. There is a specific study where participants expressed reluctance to continue despite experiencing significant improvements in VO2 max due to the perceived intensity and fatigue buildup. 
  • This raises questions about the sustainability of such high-intensity interventions over the long term.

High-intensity training driving metabolic efficiency down

19:50 -

  • There are differences in the timeline for different physiological markers during training. 
  • First, we start with an early increase in blood plasma volume, followed by heightened stroke volume and haemoglobin mass. 
  • These enhancements collectively contribute to improved oxygen-carrying capacity and delivery to muscles. 
  • Subsequently, mitochondrial content or respiratory capacity improvements come into play, further enhancing aerobic function. 
  • We need to understand the potential tradeoff between capacity and efficiency in acute high-intensity training blocks. The body prioritises energy delivery at the expense of efficiency. 
  • The potential transient damage inflicted during such sessions aims to induce tissue damage for subsequent rebuilding and remodelling to a stronger state. 
  • However, testing performance during this post-training, potentially damaged state might lead to observations of increased oxygen consumption (higher VO2 peak) without a corresponding increase in contractile work, indicating reduced efficiency. 

Practical applications

22:19 -

  • Reflecting on our research findings, particularly concerning athletes and coaches, we acknowledged the paper's conservative approach tailored to a sports science audience. 
  • Methodological limitations, such as small sample sizes common in sports science research, were also hinted at.
  • Shifting to the practical implications for athletes and coaches, I highlighted the intriguing point that time trial performance might not significantly differ whether training is below or above the maximal metabolic steady state, FTP, or critical power. 
  • This aligns with Michael Rosenblatt's meta-analysis on high-intensity interval training, where intensity mattered less than the total duration of work within a session.
  • Building on this, I delved further, extending the perspective to multiple sessions and training weeks. 
  • Contrary to the initial assumption that being above the threshold intensity was crucial, our research suggested that, once above the threshold, the specific intensity might not be as decisive as previously thought. 
  • The key takeaway was the potential tradeoff between intensity and duration, with more prolonged training bouts at a slightly lower intensity appearing more effective for time trial performance.
  • In essence, I proposed that the specificity of training at a relevant workload, potentially around threshold intensity, could be more crucial than whether it was above or below the threshold. 
  • This speculation led to the idea that increasing the duration of training might be more beneficial than focusing solely on intensity. 
  • In Stephen Seiler's study, we compare four by 16, four-by-8, and four-by-four intervals. In this context, the four by 16, likely performed around the threshold or slightly below, was less effective than the four by 8, slightly above the threshold. 
  • Interestingly, the four-by-eight intervals were more effective than the four by 4, which was considerably above the threshold. This observation aligns with the comparison to the Michael Rosenblatt study, where exercises strictly above threshold were discussed, and the four-by-eight intervals proved the most effective.
  • Studies comparing high-intensity interval training with more continuous work often use around 70% of VO2 max, placing it solidly in the tempo range. 
  • These workouts typically last around an hour and 40 minutes. The implication is that such training, even below the threshold, does not seem to have a downside regarding time trial performance. This insight leads to the conclusion that coaches and athletes have flexibility in their training approaches, allowing them to decide based on other factors, such as the nature of the event and the athlete's preferences.
  • It is essential to take a pragmatic approach, moving away from hyper-optimisation around physiological outcomes. 
  • While acknowledging the value of understanding the nuances of optimising physiological outcomes, I suggested that the research indicates that the specifics might not matter as much as initially thought. 
  • The critical takeaway is to focus on the basics, emphasising the importance of consistency and adherence to training. This suggests that doing a thing consistently may be more impactful than overly precise optimisation.

High-intensity work as part of the training periodisation

29:09 -

  • We observed a significant increase in my VO2 max, approximately two or three units of VO2 max in millilitres per kilogram per minute. 
  • While this might seem small, it holds significance, especially considering I am a trained athlete. 
  • Over a four-, six-, or 12-week training block, a two or three-unit increase is considered commendable for athletes with years of training experience.
  • We need to understand the correlation between VO2max and endurance performance by looking at it via the joiner and coil model. 
  • This model traditionally considered three factors:
    • VO2max
    • Fractional threshold (threshold relative to VO2max)
    • gross mechanical efficiency or exercise economy
  • However, I mentioned that current discussions would likely include a fourth aspect—durability or fatigue resistance. 
  • Durability assesses how well I maintain factors like VO2 max, threshold, and efficiency as fatigue sets in during prolonged exercise.
  • An increase in my VO2 max might not necessarily translate to an equivalent increase in my time trial performance. 
  • This is due to the dynamic interplay of the other three factors, which can vary independently. 
  • A temporary loss of mechanical or threshold aspects efficiency might accompany increased capacity (VO2max). 
  • This dynamic interaction among factors contributes to the variability in my time trial performance outcomes.
  • Despite identifying a significant increase in VO2max, the holistic performance outcome, particularly in time trial performance, may not distinguish between groups. 
  • The variability within trained athletes, even within a relatively similar group, contributes to the challenge of confidently attributing improvements to specific factors.

Limitations of this study

33:41 -

  • Considering the limitation of sample size is crucial, especially when dealing with outcome measures like VO2 max and time trial performance that exhibit considerable variability within the population of trained endurance athletes. 
  • The effect size observed in VO2 max, around two to three units in millilitres per kilogram per minute, is meaningful but relatively small. 
  • Detecting such a slight difference requires many subjects to confidently identify a systematic change across the population.
  • Since we sample individuals randomly from the population, hoping for a representative distribution of traits influencing endurance performance and VO2max, the need for a larger sample size becomes apparent. 
  • The challenge lies in having enough athletes to observe a systematic increase in the outcome measure across individuals, ensuring that it represents a fundamental change. 
  • In the case of the discussed paper, detecting the effect of VO2max would ideally require a minimum of about 80 participants per group. At the same time, the most extensive study included only 16, making it logistically impossible to achieve the needed sample size.
  • The importance of meta-analyses comes into play here, as they allow pooling of data from multiple small studies, providing a more confident assessment of whether an effect is present. 
  • However, the challenge persists, and this limitation is highlighted when considering time trial performance.
  •  With even fewer subjects and more significant variability in time trial outcomes than VO2max, distinguishing a natural effect becomes more challenging. 
  • Despite both VO2 max and time trial performance improving by the same percentage, the increased variability in time trial outcomes within the population may mask the actual effect, making it harder to identify in smaller studies.
  • It's crucial to consider what a study can reveal. Questions about the population it focuses on, the identified effect size, and the uniformity of the observed changes among participants are crucial to evaluating study findings. 
  • Understanding whether the effect size is substantial and if all participants exhibit a consistent response is essential. 
  • I previously mentioned the concept of responders and non-responders, emphasising that responses can vary within a population.
  • In cases where the study indicates a random distribution of responses within the population, with some experiencing increases and others decreasing, the significance of the finding becomes nuanced. 
  • This highlights the importance of recognising potential variability in individual responses. 
  • While such findings can be meaningful, especially when addressing specific aspects of training, caution is warranted.
  • Coaches can derive insights even from studies with small sample sizes, incorporating them into their overall knowledge base. 
  • However, it's crucial not to make drastic changes to training approaches based solely on the latest research, mainly if the study involves a different population, such as untrained males versus elite-level females. 
  • The application of study findings should align with the characteristics and needs of the specific population of interest.

Individual outcomes of the same training stimulus

40:02 -

  • I'm not a statistician, so take this with a grain of salt when considering statistical advice from strangers online. 
  • However, I'll share my applied understanding of the concept, explicitly focusing on responders and non-responders and the variability of parameters in a population.
  • When we talk about trained athletes as our population of interest and examine parameters like VO2max and time trial performance, let's assume, for the sake of discussion, that there may be more similarity in VO2max within this population and more variability in time trial performance. 
  • I could be wrong, but it is an illustrative example.
  • To clarify, when I mention variability, I refer to the inherent differences among individuals within the population. 
  • For instance, at a specific point in time, if we measure the VO2max of all athletes in the population, they will indeed have distinct values. 
  • This diversity extends to factors like threshold power (FTP), representing natural variability in the population.
  • In research at the group level, we randomly sample participants from this population, aiming for a spread in VO2max values that mirrors the actual distribution in the population. 
  • Group-level research endeavours to make single observations from multiple individuals to construct a representation of reality.
  • On the other hand, individual-level experimentation, akin to what coaches and athletes do, involves making multiple observations from a single individual over time. 
  • In both cases—group and individual levels—we leverage multiple observations to mitigate random variation and noise, hoping that the averaged or synthesised outcome represents a truthful depiction of the underlying reality we aim to describe.
  • Let's step back and revisit the idea of VO2max and performance outcomes in the meta-analysis. 
  • In this context, we're collecting pre and post-test data from various participants with inherent variations. 
  • The average improvement we observed in VO2max was around two to three units. However, it's crucial to recognise that within this average outcome, individual responses vary significantly—some participants experience substantial increases, while others may even show a decrease. 
  • This concept is captured by the terms "responders" and "non-responders."
  • In group-level research, these individual differences are not a primary concern. The aim is to attribute variations to random, unobserved sources of variability that might affect individual responses differently. 
  • Factors like poor sleep, a pre-test coffee, or an unexpected injury introduce noise that we hope cancels out across the group. 
  • If, despite this variability, we observe a systematic improvement at the group level, it instils confidence that, on average, the intervention has a positive effect for most individuals in the population.
  • As a coach, interpreting group-level outcomes allows me to make probabilistic judgments. I can reasonably expect that applying a specific training intervention to the athletes I work with will, more likely than not, yield positive results on average. However, this doesn't provide insights into how one individual will respond because individual characteristics and unobserved factors aren't accounted for in group-level analyses.
  • I cannot wash out those differences when working with a single athlete. I must consider the athlete's unique characteristics, and without knowledge of the individual's specific response factors, my starting point is the available group-level evidence. 
  • It's crucial to recognise that, as a coach, I have no certainty about the direction in which an individual athlete will change. 
  • We gather information over time through repeated observations, treating each training session as an experiment, and gain a better understanding of that specific athlete.
  • When you initially know nothing about the athlete, you often start with a group-level approximation. 
  • However, it's crucial to recognise that there's no such thing as the "median athlete." Every individual falls within the range of variability seen in the population, and their response can be anywhere within that spectrum. 
  • The key is to use ongoing observations and data collection to refine our understanding of the specific athlete and tailor training interventions accordingly. 
  • This personalised approach acknowledges each individual's unique characteristics and responses, moving away from generalisations based on group-level research.

Examples of these statistical concepts

49:00 -

  • Visualising this concept can be helpful. Imagine a bell curve representing the population distribution of a parameter like VO2 max among trained athletes. 
  • The range is broad, with some athletes at lower values (e.g., 35) and others at higher values (e.g., 75), forming the tails of the curve. The majority cluster in the middle, as depicted by the bell curve.
  • If we take the mean of all these athletes, let's say it's around 55, that becomes our group-level estimate. 
  • However, when we randomly sample a small group of, say, ten athletes from this population, the mean VO2 max in that sample may not precisely reflect the true mean of the entire population because of the inherent variability.
  • As we increase the sample size, our mean value estimate approaches the population's actual mean. Yet, within this variability, individual athletes differ significantly. Some may start with a VO2max of 35, while others start at 75. Moreover, their responses to interventions vary, with some improving by ten units and others decreasing by the same amount. 
  • This diversity exists within the population.
  • There's uncertainty when predicting the VO2 max of the next athlete to be tested, even with a group-level mean estimate. 
  • The next athlete could fall anywhere within that standard curve, with a probability that they might be close to the mean. 
  • However, predicting with certainty is impossible without additional information about the specific athlete.
  • This is where coaching becomes crucial. Despite the inherent uncertainty, coaches possess valuable knowledge about their athletes. 
  • The more information available about an athlete—such as training history, preferences, and responses—the better the ability to make informed predictions and have confidence in expected outcomes based on the tailored training provided. 
  • The individualised understanding of athletes enhances the coach's predictive capabilities beyond what can be achieved through group-level estimates alone.

Using Big Data in the context of sports wearables

52:02 -

  • I share your enthusiasm for gadgets and the valuable insights they can provide individually. 
  • I've delved into pilot testing, laden with lab equipment, measuring blood lactate, incorporating temperature sensors, and more. 
  • The intricacies revealed through repeated measurements over time are genuinely fascinating. 
  • However, there are notable challenges regarding the broader use of devices that aggregate data from millions of users, like Garmin or any other device.
  • While these devices have access to a massive treasure trove of population-level data and employ sophisticated algorithms, they often fall short in predicting individual metrics such as VO2max or interpreting metrics like sleep scores. 
  • A common query arises: with all this data, why does my watch struggle to accurately predict my unique physiological parameters?
  • The limitation lies in the real variability present within the population. 
  • An analogy I find apt is envisioning a cloud of data points on an XY axis, representing user observations. A linear regression line may emerge within this cloud, indicating a positive relationship between specific gadget measurements (X) and actual performance outcomes (Y). 
  • On a group level, this correlation is evident. However, as an individual, it's challenging to determine where one stands within this cloud. The distance from one's observation to the arbitrary mean observation remains uncertain.
  • This limitation becomes apparent when considering the attempt to predict individual outcomes using machine learning algorithms. 
  • For instance, BMI (Body Mass Index), a classic metric with explanatory value at a population level, often falls short in predicting health outcomes for individuals. Individuals of the same height but different body compositions may share the same BMI, making it less meaningful on an individual level.
  • Gadgets encounter a similar challenge when extrapolating population-level parameters to inform individual decision-making. 
  • While population-level outcomes offer valuable insights, predicting individual responses and making decisions based on those predictions introduces variability. 
  • It's essential to recognise this weakness and approach individualised interpretations with caution. 
  • The variability in how individuals respond to these numbers underscores the complexity of predicting and applying data at an individual level.

Final thoughts about the discrepancy between sports science and athletes at an individual level

With reference to: https://www.researchgate.net/publication/372920505_Comparing_the_reliability_of_muscle_oxygen_saturation_with_common_performance_and_physiological_markers_across_cycling_exercise_intensity

57:01 -

  • Understanding the variability within individual athletes is crucial when interpreting training data. 
  • Recently, a study by Dr. Assaf Yogev and I aimed to assess the test-retest reliability of various cycling training parameters. 
  • These parameters included heart rate, oxygen exchange (VO2), blood lactate, perceived exertion (RPE), and muscle oxygen saturation (SMO2).
  • This study focused not on the mean values across all athletes but on the variability within each athlete between two distinct training sessions. 
  • The goal was to provide coaches and athletes with a resource outlining the expected range of variability for different parameters. Here are some insights from the study:
    1. Heart Rate: The study found that, between any two sessions, the difference in heart rate might be around plus or minus five beats per minute (BPM). This indicates a relatively consistent range, allowing for slight fluctuations.
    2. SMO2 (Muscle Oxygen Saturation): In contrast, SMO2 showed more significant variability, with an equivalent value of about plus or minus 10. SMO2 measures the percentage of oxygen saturation within the muscle. A change within this range may not necessarily indicate an actual change in fitness; it could be influenced by factors such as sensor placement or room temperature.
  • These findings help athletes and coaches understand the expected variability in different parameters from one training session to another. 
  • Recognising this variability is crucial when interpreting changes in data and determining whether a change is likely due to training adaptations or simply falls within the normal range of day-to-day variations. 
  • For example, a significant deviation from the expected range in heart rate may suggest a meaningful change in fitness. 
  • In contrast, a change within the expected range may be attributed to various factors and not necessarily indicative of a training effect.
  • The study is a valuable resource, allowing athletes and coaches to gauge the significance of changes in training parameters and make more informed decisions about adjusting training protocols based on individual variability.
  • Understanding the variability in parameters like blood lactate, RPE, and others is crucial for athletes and coaches when interpreting training data. Here are some additional insights from the study:
    1. Blood Lactate: The study found that blood lactate exhibited significant variability at low and high intensities. At low intensity, the day-to-day difference within any subject was about plus or minus 0.5 millimoles per litre. This variability is essential, especially during zone two training or when using lactate thresholds as a reference. At higher intensity, the variability increased, with a difference of approximately 1.3 to 2.4 millimoles per litre. This emphasises the need to account for variability, especially in high-intensity training scenarios.
    2. RPE (Rating of Perceived Exertion): RPE, measured on a scale from 1 to 10, showed a difference across the board of about 0.5. At high intensity, the difference was about one on the scale. While RPE exhibited variability, the magnitude was smaller than blood lactate. This suggests that RPE can be functional for athletes and coaches to gauge perceived effort, with less day-to-day variability compared to some physiological measures.
  • These findings provide valuable information for coaches and athletes regarding understanding the typical variability they might encounter in various training parameters. 
  • For example, suppose an athlete observes a difference in blood lactate within the range of variability. In that case, it may not necessarily indicate an actual physiological change but could be attributed to regular fluctuations.
  • The study's focus on how these differences relate to workload steps further enhances its practical applicability. 
  • Parameters like heart rate, VO2, and RPE demonstrated sensitivity to the smallest workload steps, making them valuable tools for monitoring training progress and adjustments.
  • The study's findings, such as +/- 5 BPM for heart rate and +/- 0.5 for RPE as general estimates of variability provide a starting point for coaches who may be working with athletes they know little about initially.
  • However, as you rightly pointed out, coaches often gain a deeper understanding of individual athletes over time. Through repeated observations and interactions, coaches can tailor their approach based on the specific characteristics of each athlete. This personalised knowledge allows coaches to refine their interpretations of physiological parameters and perceived exertion, narrowing down the range of expected variability for a particular athlete.
  • For example, some athletes may consistently show tight variability in heart rate or RPE, indicating a more predictable response to training stimuli. 
  • Others might exhibit more significant swings in these parameters, suggesting a need for closer monitoring and potentially more individualised training adjustments.
  • This personalised approach is a hallmark of effective coaching. By recognising and accounting for each athlete's unique traits and responses, coaches can enhance the precision and relevance of their training interventions. 
  • Over time, this reduces uncertainty and increases the coach's ability to make well-informed decisions that align with the specific needs and characteristics of individual athletes.


Bernardo Gonçalves

Bernardo is a Portuguese elite cyclist and co-founder of SpeedEdge Performance, a company focused on optimising cycling and triathlon performance. He writes the shownotes for That Triathlon Show, and also produces social media content for each new episode.

{"email":"Email address invalid","url":"Website address invalid","required":"Required field missing"}

Explore our products and services