Repeat The Same Training?

I started keeping a handwritten training log (pictured below) after I started running at 13. Back then there were no apps to keep such records for you. I will record how far I ran, time taken, how I felt, what shoes I used, etc. 

I found it useful so I could look back on what training I did when taking part in a similar event. Say my key interval workout for a 1500m track race that I had won previously was 8x200m with 30 seconds rest. I would replicate those sessions along with deciding if I could handle more. Back then, I was a teenager and was able to run personal bests in all my events every year.

I assume that when I responded well to a training program once, I'd respond similarly or better the next time. However, a recent study suggested otherwise.

A group of researchers had subjects do the same 8 week endurance training program twice, but separated by a period of detraining that had to return to baseline. 42 out of the 53 recruited completed the study.

Each 8 week period had 24 supervised 45 minute interval cycling sessions at moderate to maximal intensity. Identical training instructions were applied such that the same training program was repeatedly prescribed to the same individuals. Exercise intensity was individually prescribed relative to performance and adjusted every second week to mainatin a comparable training stimulus across periods.

Results show very poor reproducibility of adaptation. One athlete increased VO2 max by about 600mL min⁻¹ after the first training period. However, there was no improvement at all after the second training period. Yes you read correctly, zero improvement.

Same athlete, same program, but very different outcome. I asked a few coaches whom I treat if this was similar to the athletes they coach. They told me they see it all the time. Training response varies.

The training may be the same but the variations in external factors such as sleep, diet, work stress, travel and illness contributes to the variability in individual training adaptations.

That's what the researchers concluded too. The substantial within-individual fluctuations were largely consistent between training periods and they suggest that this  within-individual fluctuations does not reliably reflect their capacity to adapt to the same training stimulus in the future.

So low stress plus good training leads to adaptation while high stress plus the same training may lead to very different outcomes.

Take away message? Do not assume that because a program worked once, it will work the same way the next time. Monitor the state of your own life stresses and modify accordingly.

Reference

OddenIV, Hamarsland H, Odden TU et al (2026). Limited Reproducibility Of Individual Physiological Adaptations To Repeated Endurance Exercise Training. J App Phy. DOI: 10.1152/japplphysiol.00154.2026

Can You Trust AI With Nutritional And Athletic Performance Advice?

Picture from Sixminutemile.com

I was reading about how 2024 Olympic road race cycling winner Kristen Faulkner built her own artificial intelligence (AI) model to analyse 9 years of her own performance data to help her uncover what's missing from women's sports science. It allowed her to have data that helps explain not just what happened but why.

Faulkner said she coded to build the AI technology that can learn from data, spot patterns and make decisions. These are skills that we usually associate with human intelligence. 

AI is already in our everyday lives. We get Google Maps directing our commute, Spotify suggesting songs on your playlist and hit ChatGPT with any question we might have.

Many people use AI for everyday health, exercise and medical queries. Are these AI driven chatbots reliable and accurate? Our patients already use AI to self diagnose their pain and injuries. Some studies show chatbots are largely accurate, while others reported frequent errors and even a risk for transmitting inaccurate information.

The following research investigated 5 popular AI driven chatbots to evaluate their responses to everyday health and medical queries across 5 categories: cancer, vacines, stem cells, nutrition and athletic performance. Both open ended and closed ended questions were used.

Gemini, Meta AI, DeepSeek, ChatGPT and Grok were the 5 chatbots used. They were each presented with 50 prompts across the 5 topics mentioned above. The researchers used an adversarial framework to strain models towards misinformation or contraindicated advice.

An adversarial framework refers to a system, process or analytical model structured around opposition, competition or conflict. This is a cybersecurity approach used to test the vulnerabilities of AI systems.

Responses were then independently rated by 2 domain experts as non-problematic, somewhat problematic or highly problematic. Citations were assessed for authenticity and completeness while readability evaluated using the Flesch Reading Ease score (100 point scale with higher scores being easier to read).

Results showed that nearly half of ALL responses (49.6 percent) were problematic, 30 percent somewhat and 19.6 percent highly problematic. Nutrition and athletic performance topics had the weakest performance and Grok generated significantly more highly problematic responses than expected.

Reference quality was poor across all chatbots. The median completeness score was 40 percent. No chatbot came up with a fully accurate reference list. Misleading, unreliable or fabricated citations were common. So please be careful if you use them.

All the 5 chatbots produced responses that were rated "difficult" on the Flesch Reading Ease scale, equivalent to university-level reading. Chatbots answered consistently with confidence regardless of accuracy, while rarely declined to respond (2 refusals to answer across 250 total responses).

The researchers concluded that continued deployment of AI chatbots without public education and regulatory oversight risk amplifying health misinformation. Especially in the field of nutrition and athletic performance. They also suggested that public education, professional training and regulatory oversight to ensure that generative AI support rather than replace professionals.

My suggestion when searching for health information is to treat these AI chatbots with a good amount of skepticisim and to verify information with qualified professionals or peer-reviewed sources. There will be some benefit seeking ideas and initial information from a chatbot, but beyond that you will need a real human expert.

Reference

Tikker NB, Marcon AR, Zenone M et al (2026). Generative Artifical Intelligence-Driven Chatbots And Medical Misinformation: An Accuracy, Referencing And Readability Audit. BMJ Open. 16(4): e112695. DOI: 10.1136/bmjopen-2025-112695.

New Way Of Grading ACL Tears

A group of researchers have introduced a new way of grading anterior cruciate (ACL) tears that are relevant to ACL healing potential. This is based on the results of managing 1080 active individuals and athletes using the Cross Bracing Protocol

Remember the Cross Bracing Protocol (CBP)? These are patients who are managed conservatively without surgery, they are put in a brace to allow ACL healing to occur.The injured knee is kept at 90 degrees of knee flexion to reduce the gap between the ruptured ACL remnants. Research shows that 90 percent people regain continuity of the fibers after 3 months of treatment in the CBP. 

Yes, the torn ACL can heal when both ruptured ends attached to each other while the knee is in a bent position. However, more severe ruptures are the least likely to achieve optimal healing.

Not all ACL's that heal (or regain continuity of fibers) in studies had a normal appearance on MRI. 50 percent of subjects in the KANON trial had a normal appearance ACL, the other 50 percent had a thinner/ elongated continuous ligament (Filbay et al, 2023). 

56 percent of those with a continuous ligament 3 months post CBP treatment had a thick ligament with normal course. The other 44 percent had a thinned/ elongated continuous ligament.

In this currrent study, subjects with a thick ACL with normal course at 3 months had better 12 month outcomes when it came to knee function, quality of life and return to sport.

In addition, the following characteristics may affect the likelihood of achieving an optimal healing outcome. Patients with partial avulsion of ACL tissue from the femoral attachment are more likely to have a suboptimal healing or even no healing compared to those with the femoral origin intact (pictured above).

Displacement of the ACL ligament outside the intercondylar notch is also associated with suboptimal healing when managed with CBP. This may be due to a significant disruption to the synovial sheath that encapsulates the ACL. Disruption of this sheath negatively impacts healing potential.

If the distance of the gap between the 2 ends of the ruptured  ACL is too wide (assessed by measuring the largest distance between the torn ends of the ACL), optimal healing is less likely to occur. Among the first 80 people in this study, 7 out of 40  (18 percent) patients achieved a thick/ taut ligament healing had a gap of ≥7mm on MRI.19 out of 32 (59 percent) had a thinned/ elongated ligament while 7 out of 8 (88%) had no ACL healing at 3 months.

When one or both ends of the ACL stumps appear rounded or retracted, also know as ACL involution, it is usually the early stages of a 'non-heal' and may correspond to poor or no healing. This is observed commonly in patients who present for MRI more than 3 weeks post ACL tear.

Besides the above, meniscal injuries, another ligament injury, articular cartilage damage, osteoarthritic features, bone bruising and joint swelling may also affect the likelihood of achieving optimal healing.

The authors concluded that with further research they may be able to assist people in making an informed treatment decision of the likelihood of individuals achieving optimal healing with nonsurgical management.

Reference

Filbay S, Dowsett M, van Haeringen M et al (2025). A New Way Of Grading Severity Of ACL Rupture On Acute MRI To Consider Potential For Noin- Surgical Healing With Cross Bracing Protocol: ACL Acvute Rupture  Characteristics For Healing (ACL-ARCH) MRI Criteria. J Sci Med Sp. 29: 145-148. https://www.jsams.org/article/S1440-2440(25)00411-6/fulltext

Is Your Evening Workout Costing You Sleep?

Picture from Sitnsleep.com

I read that the Kenyan runners usually ran at least twice a day. The first run when it's still dark between 6 to 630 am to beat the heat. It is usually an easy progressive run though sometimes long runs or fartleks can take place. 

A second session often takes place between 10-11 am. This is usually reserved for track interval sessions. Typically they do 2-3 key interval sessions a week with most of the runs really easy so they can recover. When asked why they do the second run before lunch, the reply is so that they have more time to recover before the next day's sessions. 

Sometimes they even do a lighter recovery run between 4 to 5 pm to add to their weekly mileage.

So I was not surprised when I read that the timing of our workouts can affect our sleep (Akhtar and Eleftheriou, 2026). To avoid confounding factors, sleep and activity data from the same day through a model with fixed effects and terms for activity time, activity-sleep gap and the interaction between these terms. 

Data collected from users sleep looked at how sleep was affected by when they decide to exercise. The focus was on cardiovascular exercises, specifically walking, running (outdoor and treadmill), cycling (outdoor and stationary) and hiking. They studied over 100,000+ workout sessions and found that the later you train, the worse you recovered (pictured above).

When you exercise later in the day, your sleep heart rate is higher, your heart rate variabilty (HRV) is lower and you get fewer REM sleep events. You may sleep quicker with later workouts, but it is not better sleep. It is falling asleep faster while recovering less.

The benefits plateau around the 10 hour mark. The sweet spot is a 11 to 14 hour gap between your workout and bedtime. If you can exercise earlier in the day, you can capture most of the benefits. Perhaps that's the reason the Kenyan runners do their second run at 9 am. 

What about those people who can only exercise at 8pm onwards or later? Should you worry about this?  My take is no. Majority of people should NOT worry. I think it is more important to exercise where it fits in your life. 

Don't worry too much about HRV scores or perfect recovery. You are only working out at 8 pm or later because that's the only time when family and work gives you an opportunity to. Many people do not have the luxury to choose the time to exercise and exercising at any time is better than no exercise.

Our bodies are remarkably adaptable too. I am sure if you exercise at 8 pm or later long enough your body will get better sleep. Not perfect sleep but the detriment is less.

Showing up and exercising is way more important than perfect conditions. Consistency beats optimization every time. That is what will compound.

Reference

Akhtar F and Eleftheriou K (2026, May 24). Your Evening Workout Is Costing You Sleep. Terra API. Https://trryterra.co/research/best-time-toworkout--for-sleep#the-honest-takeaway

Warming Up Body And Mind

Picture from Healthywomen.org

All of us are familar with warm ups. Before you run, exercise and especially before a race. You are trying to increase your core temperature, send more blood flow to the limbs and 'wake' the muscles up. Before a running race you may even try some race pace striding so the actual start of the race does not feel like a shock.

So I was surprised to read that combining mental and physical warm ups can improve running times. Curious? Please read on to find out more.

Researchers in that study investigated whether combining cognitive tasks with a standard physical warm up could improve 1-mile performance in runners. 25 runners (11 male, 14 females) with an average weekly mileage of 20 miles (32 km) and 5km personal best timing of 23:31 min were recruited for the study. Each runner completed 3 separate testing sessions.

First was with only a physical warm up which included a 1200m easy jog, 800m alternating 100m jogs and 100m strides and 3 minutes of active stretching drills. 

Next was the same physical warm up plus a low load cognitive task and finally same physical warm up with a high load cognitive task.

In the cognitive warm up conditions, the runners completed four 3-minute cognitive tasks before and between the physical components.These tasks targeted mental functions like switching between tasks, decision making, inhibiting responses to stimuli and memory. They were doing short bursts of focused mental work designed to activate the brain without exhausting it.

 

After that all the runners ran 4 laps of a 400m track with their watch face covered so that they could only rely on feel rather than constant pace feedback.

Ready for the results? The runners were faster after both cognitive plus physical warm ups compared to only doing the physical warm up. The  low load cognitive task improved 1-mile timing by about 8 seconds (2.3%),  while the high load cognitive task improved performance by about 11 seconds (2.8%).

The runners did not appear to muscle their way to faster timings since their perceived effort was lower after the combined warm ups and their average heart rate during the time 1-mile run was also lower. Their readiness to perform was higher, meaning the runners felt more prepared to run hard.

Stride length and cadence did not change meaningfully suggesting that performance boost was not explained by obvious mechanical changes.

Take home message? No, you do not need to download an app to challenge your cognitve skills before every workout or race. However, this study does suggest that the best warm up should prepare both mind and body.

Before you do your track interval sessions, time trials or short, hard and fast sessions or even up to a 5 km race, it may be useful to add a small cognitive task that perhaps include coordination games, quick reaction tasks, fast feet with visual cues to make you engage attention before you start running hard.

The important thing to note is dose. A few short mental tasks coupled with jogging, drills and race pace striding may help you feel sharper and more ready for your workout. A long, draining task may do the opposite.

This sharpens the nervous system rather than stressing your brain. Keep it brief and engaging and please try it in your training first before trying it on race day.

Reference 

Mortimer, H Dallaway N, Diaz-Garcia J et al (2026). Warming Up The Body And Mind: Bombined Cognitive And Exercise Priming Improves 1-Mile Time Trial Performance In Recreational Runners. Eur J Sp Sci. 26(5): e70163. DOI: 10.1002/ejsc.70163

New Guidelines For Fueling

I was sharing with some my patients this past week how I found it amazing that Sebastian Sawe was able to consume 115 grams of carbs an hour in his sub-2 hour marathon. Previous recommended guidelines were between 60-90 grams an hour.

Lo and behold, a patient shared that in the Nice Ironman last year, Casper Stornes (1st place) and Kristian Blummenfelt (3rd) took 120 grams of carbs while running and up to 180 grams on the bike.

Intrigued by what he sent me, I found a recently published review article by Morton et al (2026) which revisits carbohydrate guidelines for endurance athletes. So are the previous fueling recommendations enough?

When training and/ or racing for prolonged periods, carbohydrate intake will spare liver glycogen while maintaining your blood glucose levels. Most importantly, it helps one to sustain higher intensities nearer the end of your race. When everyone else is going faster, you do not want to run out of fuel and slow down.

Fueling is not to just avoid bonking (or hitting the dreaded wall). It is also to delay the shift from using carbohydrates to fat. That shift is not bad, but when you are trying to go fast without carbohydrates it is going to be very difficult. Note that if you are doing Zone 1 or 2 for many hours (like in an ultra marathon) at low intensities, then this shift may not occur.

When one consumes enough carbohydrates during endurance exercise, long ultra race, Ironman event etc, it delays the "crossover point".  This is when carbs stop being the predominant fuel due to glycogen depletion or low carb availability.

You can see from the diagrams above that if no carbs are consumed (0 grams each hour), this happens around 2 hours. With 45-90 grams/ hour, it delays the crossover point by 30-60 minutes. If 120+ grams/ hour of carbs are consumed, there is no crossover point since carbohydrates remain the predominant fuel source while exercising. At least not in cycling studies as the subjects exercised at 90 percent lactate threshold for 3 hours. It may still occur in longer races.

For runners, the review looked at elite male marathoners with personal bests under 2:30 hrs. They completed a 2 hour treadmill protocol close to marathon race intensity while consuming 60, 90 or 120 grams of carbs each hour. 

Those who ingested 120 grams used more of the ingested carbs leading to higher carbohydrate oxidation. They had a roughly 3 percent improvement in running economy compared to 60 grams per hour. There was also greater carbohydrate contribution to total energy expenditure late in the run. Now you know why Sabastion Sawe never slowed down in his sub-2 hour London marathon. 

There's a catch. Gastrointestional (GI) symptoms were also higher at 120 grams/ hour. Nausea, stomach fullness, and cramping were worse in the highest carb intake. More carbs may offer an advantage, but only if your gut can absorb and tolerate them.

The researchers say that runners should use 90 grams of carbs or more per hour as a realistic target for long, hard events. For trained athletes, the range may extend closer to 120 grams per hour. Fueling should be treated like a trainable skill, provided the gut has been trained to handle it.

Runners will find it more challenging compared to cycling since it's much easier to eat on the bike. All the moving, reduced blood flow to the gut, heat stress may lead to higher rates of GI distress.

This is an exciting review for fueling, but do note that most of the research comes from cyling, male athletes and in controlled lab settings. Female athletes are under-represented as well as real world running/ racing conditions.

My takeaway message? For some of you runners, 90-120 grams/ hour will most certainly help your times. Others may benefit from 60-75 grams/ hour. For other runners, it may be just progressing from under-fueling to consistently fueling and getting a personal best no matter how many grams of carbs that may be.

Reference

Morton JP, Fell JM, Gonzalez JT et al (2026). From Metabolism To Medals: Contemporary Perspectives And Revisiting Carbohydrate Guidelines For Fueling Endurance Athletes During Exercise. J Nutrition. 156(5). DOI: 10.1016/j.tjnut.2026.101442

Sub 2 Hour Marathon

Back in 2017, renowned sports scientist Dr Ross Tucker wrote that a sub 2 hour marathon was unlikely to occur for generations. Then Nike introduced the Breaking 2 project which resulted in Eliud Kipchoge running under 2 hours albeit in a paced setting, not in an official race. 

The 2 hour barrier was finally broken last week at the London marathon when winner Sabastian Sawe and runners-up Yomif Kejelcha achieved the breakthrough, the latter running only his first marathon.

The podium

Sawe beat the previous world record by 65 seconds, finishing in 1:59:30 hrs. He ran an amazing negative split, 60:29 for the first half and the next half in 59:01 min. Kejelcha finished 2nd in 1:59:41 hours. The race was so fast that third place finisher Jacob Kiplimo also broke the previous world record (set in 2023) by 7 seconds. He clocked 2:00:28 hrs.

What lead to the breakthrough? Perhaps due to a "perfect storm" or convergence of many of the following factors. 

Sawe's team said he was running in excess of 200 km a week in the 6 weeks leading up to London, with a maximum of 241 km (150 miles). The volume of his training runs is likely an important factor for him to break 2 hours. Not many elite runners can tolerate that sort of volume, especially when most of it is done at low intensity (or Zone 1). High training volume done at relatively low intensity (Zone 1) is associated in faster marathon performances (Muniz-Pumarez et al, 2024).

Sawe's 5 km splits from 30 km onwards were superb. 30-35 km : 13:54 min, 35-40 km : 13:42 min. His final 2.2 km was 5:51 min. His 24th mile was 4:12 min - the fastest single mile ever run in a marathon. As the finishing line beckoned, he was still accelerating. No way he could have done this if he was not fueling well.

Sawe's nutrition was exceptional. I'm not plugging names here, in fact I've not even tried their products, but Sawe is a Maurten athlete. Elite athletes train their gut take up to 120 grams of carbs while competing without GI distress. The previous recommendation was 60 grams per hour previously. Sawe's reported intake was 115 grams an hour. Maurten's research team was with Sawe in Kenya for 32 days across 6 trips between last year and this April to train his gut to absorb that amount. 

Please note that 120 grams of carbs are NOT recommended for the recreational athlete. That is probably why cycling races and marathons are getting faster since there is no depletion in carbohydrate levels.

Sawe also used sodium bicarbonate (also from Maurten) to buffer his lactic acid build up during the race. He took the sodium bicarbonate early since it peaks in the bloodstream  roughly 60-90 minutes after ingestion so the timing of 2 + hours before the race would put peak buffering capacity at the start.

What about his shoes? Adidas says those were the fastest and lightest super shoes ever made. Actually, Adidas had a great day as 4 of the top 5 men were wearing the same Adizero Adios Pro Evo 3. It weighs 97 grams (for men's size 9), 30 percent lighter than its predecessor with a stack height of 39 mm (below the 40 mm limit by World Athletics). 

The Lightstrike Pro foam used is 50 percent lighter than the previous version, along with the carbon plate. Research suggests that the foam and carbon plate can affect the "spring like" bounce of the body as the foot strikes and leaves the ground. The shoes help to store and release energy and acts like the runner is pushing off a springboard so less energy is needed for the run.

London is considered a relatively fast course (though not as fast as Berlin) and the weather conditions was between 13-17 degrees Celsius, close to ideal.

There you have it, exceptional athlete physiology, high mileage training without injury, efficient biomechanics, super shoes , optimized fueling and favorable conditions. The recipe for success for running marathons.

Reference

Muniz-Pumares D, Hunter B, Meyler S et al (2025). The Training Intensity Distribution Of Marathon Runnersd Across Performance Levels. Sports Med. 55: 1023-1035. DOI: 10.1007/s40279-024-02137-7