© Copyright – 2012 – Athletics Illustrated


Dr. Trent Stellingwerff is a former track and field athlete who competed in the NCAA for division 1 Cornell University, where he was selected as a co-captain. He also competed in the CIS for the University of Guelph, where he twice earned All-Canadian status.

Stellingwerff was an academically decorated student who made the Dean’s list at Cornell University. In 2006 he took a position in Switzerland for the Nestle Research Centre (Powerbar) as a Senior Research Scientist in Sport Nutrition, Energy and Performance. Stellingwerff has also served as the Nutrition and Physiology Consultant for Athletics Canada. He now lives in Victoria, BC and is a Senior Physiologist working with the Canadian Sport Centre Pacific.

Stellingwerff’s latest paper that will be published in October on the website  of the International Journal of Sports Nutrition and Exercise Metabolism (IJSNEM ) is on nutrition, training and physiology of three elite marathoners. This is a case study with athletes that he has worked with for the last three years, they are marathon runners Reid Coolsaet, Dylan Wykes and Rob Watson.

The case study (below the interview) is an accurate description of how the athletes trained (volume and intensity), how many times they practiced fuelling, how many times they undertook fasted or low-carbohydrate training, and how much caffeine they used.

CHO = carbohydrate

Asked whether fasted runs are something that he will continue to do in training Coolsaet said, “I have done fasted runs in the past but not many for the Olympic marathon. I plan on doing 1-2-per-week for future marathons.” Wykes and Watson agreed, “low carb training was something that I have done for a long time now, even before this study and before I was a marathoner. I will keep on doing it,” said Watson.”

CK: What is low CHO-availability training?

TS: It is very important to fully understand what this is. First it is NOT a low-carbohydrate (CHO) diet (such as Atkins).  Instead it is just moving around dietary carbohydrate availability around training, so daily carbohydrate intake is still high. These approaches are also for the more elite athletes, not for beginners who can just learn to run more.  We also do NOT do this during the final prep-phase into the marathon – as at this time we are fuelling with CHO sports drink to adapt the body for race day.  There are two main ways to do low-CHO-availability training for the elites:

1) You have low CHO-availability from outside the muscle.  This is basically waking up in the morning, after not eating for >8hrs, and other than some water, and  bit of coffee, head straight out the door for a run, and or workout.  Here you will start out with high muscle glycogen (so high CHO availability inside the muscle), but if the run extends beyond one hour, limited CHO availability from outside the muscle (since you are not running and taking in CHO sports drinks and the liver glycogen (CHO) is low, due to the fact that you haven’t eating in over 8 hrs).  This overnight fasted training was done two to three times-per-week by elite marathoners, and we slowly extended the amount of time they could go and the quality of the run as they adapted over the weeks (just like increasing the run distance and paces).

2) The second to have low CHO-availability is from inside the muscle with low muscle glycogen training. This is done by waking up, having breakfast and having the athletes do a hard, high-quality workout mid-morning.  The high quality workout would drastically decrease muscle glycogen. But after the workout the recovery foods and beverages over the next three to six hours are not rich in carbohydrate, but focus more on protein and fat intake (healthy fats), with minimal CHO intake. The athletes would then go out four to six hours later and do another run, or low-grade tempo run, with low muscle glycogen. This training is much more difficult and was only done about 10-15% of total low CHO-availability sessions.

CK: Did you find that the athletes lost weight when running on low CHO-availability? I don’t mean loss of fluids through sweat rate. What I am think of is the body, utilizing stored carbs. Is this too anecdotal or are there too many variables to even bother measuring weight loss?

TS: Good question. Since it was a case-study, and we couldn’t control everything (e.g. every day dietary intake), this question is impossible to answer.  In short, all marathon runners throughout their 16 week preps lost body weight, but this was probably due to a multitude of things, such as:  increased training load causing more energy deficits throughout the day, which might have been somewhat impacted by running on low CHO-availability.

CK: Was the purpose of the study of the three marathoners using periodic low availability of CHO during training done to find the results of the train low – race high theory?

TS: Main purpose was to describe and characterize the training and nutrition interventions that I do with the three athletes, as part of the reason as to why they are running so well (beyond great genetics, great training programs etc.)

CK: If this study was exclusively about low CHO-availability, I assume the caffeine consumption would have clouded results with:

  • Placebo effect
  • Potentially cause GI discomfort
  • Have an effect on Rate of Perceived Exertion (RPE). There are studies to indicate that athletes consuming caffeine had performance increases and lower RPE after controlled training bouts (I assume you know well of this)…

TS: Caffeine has been proven to have a ‘real’ physiological performance effect. However, for ANY intervention you give an athlete, if you create “belief” that the intervention might work, you are creating a placebo effect.  I would rather call it “belief effect”, and this is part of being a great sport scientist. From every intervention, I am looking for a  true physiological performance benefit + creating a “belief effect” (or placebo effect) so the athletes are benefit as much as possible.

There is some minor suggestions in other studies, that for some individuals, caffeine might increase the chance for GI upset  – very individual though.

Yes, caffeine lowers the RPE or effort ratings.  Some studies have shown lower pain ratings, and the ability to focus more late in races and training when it gets really tough.

CK: Do you know or assume at least that a long run versus a zone three workout with similar low CHO-availability have differing affects on the results?

TS: I characterized the training into three main zones, so one could see how much the guys trained in Zone one (aerobic) vs. zone two (lactate threshold) vs. zone three (VO2max and high intensity training).  There is some evidence to suggest a polarized approach is ideal, which is somewhat what we found with these guys.

CK: Based on the information gathered from the 16 weeks of observing the three marathoner’s training and nutrition would you, as a coach and physiologist, recommend more consistent focus of their training with low CHO-availability?

TS:  Very good question, that doesn’t have an easy answer!  So anything with a difficult answer will result in a scientist giving a long answer.

First, from an academic research perspective, there are now about seven or eight studies that have examined this as an intervention (training with low CHO-availability), with mixed results.  All studies show that training with low CHO availability does result in different adaptations in the muscle (with muscle biopsies) that would be suggestive of an enhanced performance outcomes (e.g. increase in fat-transporters in the muscle, and aerobic adaptations). However, when it comes to the performance measures, only a few studies have looked at this, with mixed outcomes: a couple of studies showing a benefit, with a few others showing no difference (but not a decrease in performance).

But, the issue with most of these studies is they only look at training for a few weeks at the most as it is so difficult to control in a study setting (so maybe 15 low CHO-availability sessions).  And, of course, athletes train for a much, much longer duration in preparation for a marathon, or over the season.

Secondly, with the athletes I have worked with, this applied sports-nutrition case-study is the evolution of a lot of work with them to get to this point.  I would never start with a rookie marathon runner and throw the sport-science kitchen sink at them! But, instead, over years, slowly integrate and trial and error different interventions and ideas on an individual level to see what work’s best for the individual athlete.
BUT, to go back to your original question, for world-class athletes who are already VERY adapted to training, and who are already saturated with running volume (where doing more running drastically increases risk for overuse injuries, such as these guys who are all over 200km of running per week!)  I do believe these interesting nutrition interventions involving CHO-availability can further “adapt” and drive the muscle up to another level of adaptation.

CK: Like training effects, is the muscle adaptation in training with low CHO-availability very run-muscle specific?

TS: To examine/measure the adaptation the studies require muscle biopsies, and these tend to all come from the quadriceps muscles.  So, I would be hypothesizing on the other muscles.  But, at least for runners, since this is their dominate active muscles, I would suspect that majority of the adaptations would occur in running related muscles.  Conversely,  I would hypothesize for rowers that the adaptation would  be in more muscle groups due to more muscles being used during exercise.

CK: Did you see a consistency between this case study and others around the RPE during prolonged exercise and maximal workouts when supplementing with caffeine?

TS: First off, I use RPE (ratings of perceived exertion) a lot in coaching athletes. I think it is a VERY valuable tool and skill set to develop in athletes – the ability of the athlete to be able to individually ‘read’ his or her body.  I use an adapted exercise TRIMP (training impulse) scale of 1 to 10 (see below).

RPE Rating

RPE Rating





Very, very easy






Somewhat hard






Very Hard




Very intense



For beginners it is difficult for them to really know and “feel” how hard something is in training – but through practice they can get better at it.  RPE is the very best indicator a person can have linked to pace judgement – which is a very important skill for all athletes to possess.  (E.g. a lot of beginners ALWAYS go out way too hard in workouts and races, and die badly – IMHO this is due to an underdeveloped sense of pace, which is linked back to their ability to “read” their body, which is essentially RPE training).

So, going back to your question, there is a huge body of scientific evidence showing that caffeine does lower RPE – it makes exercise feel easier at sub-maximal intensities!  (ever so slightly).  But, eventually, the athlete will still reach RPE’s above 8, but hopefully have just gone faster or farther with the caffeine.

CK: So beginners should spend more time running organically, if you will, and really I am assuming everyone should?

TS: Generally YES. But, it is important to “link” up those organic feelings with actually measured paces so that the athlete develops pace sense (you have to measure something (time and distance) to understand pace).  But, once the athlete gets better at this, and further develops their pacing skill set, they can actually rely less on the watch, or GPS or heart-rate and go on feel.  However, whenever there is a change in environment (e.g. altitude), fitness (coming back from time-off, injury, or sickness) I would recommend athletes use the watch, GPS, heart-rate a bit more to re-calibrate their RPE’s to paces.

CK: Were there indications that perhaps the now so-called outdated practice of carbohydrate depleting before a loading phase is an effective tool?

TS: The most recent data from several labs have clearly shown that CHO depleting is no longer necessary for example, five or six days out from a race, and then going on a low CHO diet for three days, and then a high CHO diet for three days leading into the race is not necessary.  Athletes will feel horrible on the low CHO diet for three days, and the last thing you want is an athlete to feel crappy the week of a race. Instead, just with the normal taper (drop of 40 to 60% of volume), along with a very slight increase of carbohydrate intake will result in the same super-compensation (or small 15 to 20% increase) in muscle glycogen prior to a race.

The one caveat I warn people on CHO loading is that for every gram of glycogen you store in your muscles and liver, you also store nearly three grams of water.  That extra water storage can be good for whole-body water balance, but it will also make many athletes feel “sluggish and heavy” during the initial 10k or so of a race. Don’t panic! You’ll burn off some of that weight, and you’ll have extra glycogen for the last 10k of the marathon when you really need it!

CK: Was there any point in adjusting nutritional intake at efforts well below 75% of V02Max, seeings that 75% is the threshold where any effort above that, the primary fuel source is CHO from working muscles?

TS: We actually didn’t measure VO2max and % of VO2max in these athletes, but I can make some estimates. First, world class marathon runners can actually run at 80 to 90% of VO2max during their marathons, and at this pace, nearly 90 to 95% of the fuel they are using to make ATP is from carbohydrate (either blood glucose or glycogen). Now, this CHO energy can be oxidized both aerobically and anaerobically, and early in the marathon much of the CHO is being oxidized aerobically in the mitochondria. So given the amount of running done at marathon pace or faster, yes – world class endurance athletes need a CHO rich diet.  Several good studies have shown the Kenyans eat about 70-75% CHO in their diets!

CK: What percentage of their weekly volume during the first six weeks is below and above the 75% of V02Max?

TS: In the first six weeks of training, the vast majority of running is below 75% of VO2max (~70% of total volume) – but when they go hard, they go very hard in their sessions (e.g. sometimes 35km + workouts with upwards of 20 to 30km of the workout at marathon pace or faster! – so highly specific to marathon preparation).

CK: You show the volume of kms run for the average of the three athletes at 187 and maximum 231 per week. And you referred to the training programs as periodized. Were there any major differences in the programs?

TS: I am maybe going to be a bit provocative here – I actually don’t believe there is any single one training program that is best. I do believe there are a lot of different ways that athletes can be trained, and be very, very successful, as long as some sound fundamental training philosophies around training volumes, intensities, and loading and recovery sequences, strength training and proactive paramedical are undertaken. The key is the coach blends all these elements in a way that best individualizes the training for a given athlete – working on some weaknesses, but never straying far from feeding the baby the bottle and always being sure an athletes strengths as an athlete are optimized (this builds confidence).  The good coach builds confidence by ensuring the sequencing of loads, workouts, races is appropriate.

Where I sometimes see mistakes, is a coach that gets “stuck” doing what they know best with an athlete over a long-term, which results in this athlete having a program that looks the same all the time.  This eventually causes stagnation, as the body is lazy and will always just no longer adapt to the same, or similar, stimulus.  I like looking at training programs where if I take six random weeks out of the program throughout the year, they look quite different, due to different focuses. In this regard, it is amazing to me how often you see an athlete switch coaches, and get a big spike in performance, due to a very different training stimulus. The ‘new’ coach looks like a wizard, but in fact, they just introduced some new stimulus to a body that was ripe for it. So that said, at the time of the study two of the athletes had the same coach (Dave Scott Thomas), and another athlete had another coach (Richard Lee).  I have seen these athletes training close enough to see some differences, but at a top-level approach, I wouldn’t say they are not that fundamentally different.

Case Study: Nutrition and Training Periodization
in Three Elite Marathon Runners – Stellingwerf-Marathon Case Study (IJSNEM2012)

Trent Stellingwerff

The author is with the Canadian Sports Center-Pacific, Victoria, BC  Canada. Laboratory-based studies demonstrate that fuelling (carbohydrate; CHO) and fluid strategies can enhance training adaptations and race-day performance in endurance athletes. Thus, the aim of this case study was to characterize several periodized training and nutrition approaches leading to individualized race-day fluid and fuelling plans for 3 elite male marathoners. The athletes kept detailed training logs on training volume, pace, and subjective ratings of perceived exertion (RPE) for each training session over 16 wk before race day. Training impulse/load calculations (TRIMP; min × RPE = load [arbitrary units; AU]) and 2 central nutritional techniques were implemented:periodic low-CHO-availability training and individualized CHO- and fluid intake assessments. Athletes averaged ~13 training sessions per week for a total average training volume of 182 km/wk and peak volume of 231 km/wk. Weekly TRIMP peaked at 4,437 AU (Wk 9), with a low of 1,887 AU (Wk 16) and an average of 3,082 ± 646 AU. Of the 606 total training sessions, ~74%, 11%, and 15% were completed at an intensity in Zone 1 (very easy to somewhat hard), Zone 2 (at lactate threshold) and Zone 3 (very hard to maximal), respectively. There were 2.5 ± 2.3 low-CHO-availability training bouts per week. On race day athletes consumed 61 ± 15 g CHO in 604 ± 156 ml/hr (10.1% ± 0.3% CHO solution) in the following format: ~15 g CHO in ~150 ml every ~15 min of racing. Their resultant marathon times were 2:11:23, 2:12:39 (both personal bests), and 2:16:17 (a marathon debut). Taken together, these periodized training and nutrition approaches were successfully applied to elite marathoners in training and competition.

Keywords: TRIMP, carbohydrate and fluid, caffeine, endurance, performance.

Since the legendary run of Pheidippides in the Battle of Marathon (490 B.C.) no event has captured the human imagination quite like the marathon, as evidenced by the long interest in marathon-associated physiology and performance (“Proceedings of the 2006 World Congress,” 2007). However, only a few published studies have examined physiological, anthropometric, nutritional, or training characteristics of truly world-class marathoners (Billat, Demarle, Slawinski, Paiva, & Koralsztein, 2001; Billat et al., 2003; Lucia et al., 2006; Onywera, Kiplamai, Boit, & Pitsiladis, 2004). Traditionally, most elite marathon training expertise is tackled anecdotally through word of mouth and practical experience and not systematically documented through validated methodologies in peer-reviewed scientific journals. Scientific interest in the marathon, and what causes fatigue and ultimately limits performance, continues to be highly relevant as evidenced by the recent viewpoint article by Joyner, Ruiz, and Lucia (2011) titled “The Two-Hour Marathon: Who and When?” and the 38 unique published counter-commentaries (Stellingwerff & Jeukendrup, 2011). What has become clear is that there are a myriad of potential physiological, psychological, environmental, and sociological determinants of marathon performance. One distinctive facet to the marathon is that carbohydrate (CHO) metabolism, primarily muscle glycogen, is the dominate fuel at exercise intensities greater than 75% VO2max and can start to become limiting after ~90 min. So despite the fact that endurance running played a significant role in the evolution of man (Bramble & Lieberman, 2004), given the high exercise intensities of marathon racing dictating a high muscle glycogen use (>80% VO2max; O’Brien, Viguie, Mazzeo, & Brooks, 1993), it could be argued that humans are not adapted to racing marathon distances at high exercise intensities. Thus, perhaps under appreciated is the vital role that fuelling and hydration status plays on marathon success, which is not necessarily a prerequisite for success over shorter distances (<90 min). Contemporary scientific studies have started to examine the interactions that altered nutrition approaches may have on endurance-training adaptations. For example, a periodic lack of CHO availability may further enhance training adaptations (Burke, 2010), the gastrointestinal (GI) tract may also be adapted to handling increased fluid and CHO, and an individualized approach to race-day CHO type and fluid intake may contribute to optimal competitive success (Jeukendrup, 2011). However, there have been no studies published that characterize these training and nutrition approaches in elite marathon runners. Therefore, the primary aim of the current case study was to use field data to characterize several training and nutrition techniques, leading to an individualized race-day fluid and fuel plan in a periodized approach with three elite male marathon runners.

Interventions and Methods

Background and Athletes

Each marathoner (Table 1) undertook an individualized 16-week training plan and competed in different marathons at Week 16 but implemented the same nutrition approaches with the same physiology/nutrition practitioner. Each athlete/coach has read, approved, and provided written permission for this publication, which conforms to the principle approval for case studies in the
International Journal of Sports Nutrition and Exercise Metabolism by the ethics committee of the Australian Institute of Sport.

Training Assessment

All athletes kept detailed training logs, including information on duration and volume (min and km), body weight (BW), subjective training feedback (ratings of perceived exertion; RPE), and weather conditions. All the data collected were exclusively used to characterize training, not to affect training decisions. Each training session’s “load” was calculated by multiplying the training duration (minutes) by intensity (RPE scale of 1–10, easy to maximal) into a single arbitrary unit (AU) called training impulse (TRIMP), as previously validated and described by Foster et al. (Foster, 1998; Foster et al., 2001). Session RPE was further divided into three separate intensity zones: Zone 1 = RPE of 0–4 (very easy to somewhat hard); Zone 2 = RPE of 5–7 (hard); and Zone 3 = RPE of >7 (very hard to maximal). These three TRIMP RPE training-zone breakpoints, corresponding to first ventilatory and second ventilatory threshold, respectively, have previously been validated and shown not to differ from TRIMP analysis using either heart rate or blood lactate data (Seiler & Kjerland, 2006).

Dietary Approaches

There were two main nutrition approaches, which were periodized across three training mesocycles: general preparation phase (Weeks 1–6), specific preparation phase (Weeks 7–12), and competition taper phase (Weeks 13–16), with the target marathon being at the end of  Week 16. Each dietary approach was described in writing and verbally to each athlete and coach and individually implemented and continuously monitored. Low-CHO-Availability Training. Athletes were instructed to periodically undertake and develop their ability to tolerate training with low CHO availability (~1–5 times per week as individually tolerated, with emphasis during the general preparation phase). This was done either by limiting exogenous skeletal-muscle CHO availability (by undertaking morning-fasted training) or by limiting endogenous skeletal-muscle CHO availability (by undertaking reduced-muscle-glycogen training). For periodic overnight-fasted training, athletes were instructed to undertake aerobic training (at or below lactate/ventilatory threshold) first thing on waking, with just prerun water and/or coffee consumption. Conversely, low-glycogen training featured doing two training sessions within a single day, but limiting dietary CHO intake after the first training session so that the second training session was undertaken with reduced muscle glycogen stores. Immediately after this second training

bout, athletes consumed optimal CHO and protein to maximize glycogen and protein synthesis. The few times reduced-glycogen training was implemented, a demanding interval-based training session was undertaken in the morning commencing with higher muscle glycogen to ensure high training quality. This type of training would also cause a large reduction in muscle glycogen, which was maintained with low CHO intakes, before the second daily training session (>4 hr), which was prolonged but aerobic in nature (at or below lactate/ventilatory threshold). These low-CHO-availability training sessions are not to be mistaken for a low-CHO diet. Instead, athletes were instructed to merely alter the timing and availability of CHO in their diet around specific training situations. They were taught the amount and timing of CHO to maximize post-exercise glycogen synthesis, as well as the dietary choices required to limit CHO availability when required. Individualized Targeted CHO Fuelling and Hydration During Training and Competition. Athletes were taught to measure CHO- and fluid-intake rates, measure sweat rate via BW changes, and record GI tolerance during key training bouts (1–3 times per week with emphasis during the specific preparation phase and competition taper phase) in a worksheet (Table 2). They were instructed to adjust fluid intake, aiming for ~2–3% BW losses during training sessions. On runs >2 hr, athletes were encouraged to consume at least ~30–60 g CHO/hr and ~400–600 ml/hr, but then urged to further increase intakes to maximize their individual upper tolerance. During the competition taper phase, athletes were encouraged to consume fluids and CHO in every session  >75 min. Within 48 hr post-race, athletes reported their race-day fluids, carbohydrate, and caffeine consumption (Table 3). Since each marathoner was a recognized elite athlete, each got to prepare his own fuel/fluid bottles, which were placed at elite-athlete-only aid stations to facilitate more accurate reporting. Unfortunately, due to the complexities post-race for these elite marathoners (e.g., interviews, doping controls, etc.), pre- to post-race BW tracking for the competitions resulted in either uncertain data or was not possible.

Observations and Outcomes

Marathoner Characteristics
and Performance Outcomes

The characteristics of each marathoner at the time of his race are outlined in Table 1, with data reported as M ± SD. To date, these athletes have won a combined 12 Canadian Championships and competed in 16 global championships (World or Olympic Games). Average training weight was 67.3 ± 3.3 kg, which decreased 2.2% for racing. Marathon times were 2:11:23 and 2:12:39 for Marathoners 1 and 3, respectively (previous personal bests were 2:16:53 and 2:15:15, respectively). Marathoner 2 debuted at 2:16:17.

Training Data

Training data are based on 606 training sessions over 12.6 ± 2.1 training sessions per week per athlete, featuring average weekly training volumes of 173.6 ± 32.5, 213.3 ± 41.2 and 159.6 ± 27.0 km/week for Marathoners 1, 2, and 3, respectively (Figure 1). Each marathoner had his highest training volume during the specific preparation phase, with highs of 228, 266, and 199 km/week, respectively. Lowest weekly training volumes were on race week (114.5 ± 14.1 km), with ~35% of that week’s running volume coming on race day. Due to the highly subjective nature of RPE TRIMP assessment, it can only be compared within a given athlete (Figure 2; Marathoner 1). This athlete had a weekly high TRIMP of 4,437 AU (Week 9) and low of 1,887 AU (Week 16), with a 16-week average of 3,082 ± 646 AU. For all athletes combined, of the 606 total training sessions, 74.3% ± 3.5%, 11.0% ± 1.7%, and 14.7% ± 3.5% of the sessions were completed in Zone 1 (very easy to somewhat hard), Zone 2 (hard), and Zone 3 (very hard to maximal), respectively (Figure 3).

Periodized Dietary Approaches During Marathon Preparation

Figure 4 provides an overview of the two nutrition techniques, which were periodized across each training phase. On average, there were 2.5 ± 2.3, 2.6 ± 2.3, and 1.3 ± 2.3 low-CHO-availability training bouts per week across the general preparation phase, specific preparation phase, and competition taper phase, respectively. There were 107 low-CHO-availability training sessions for all 3 marathoners, of which 11 were reduced-glycogen training and 96 were morning-fasted training. CHO-fueling practice sessions increased during each phase, with an average of 19.0 ± 6.1 CHO-fuelling sessions throughout the 16 weeks,  with most during the specific preparation phase and the competition taper phase. A wide range of individual sweat rates was recorded by Marathoner 2 during training: ~750–1,000 ml/hr (Table 2). This marathoner practised with a wide range of CHO- and fluid-consumption rates and demonstrated great individual GI tolerance to fluid and CHO intake during running (Table 2).

Race-Day Intakes

On average, athletes consumed 61 ± 15 g CHO in 604 ± 156 ml/hr of racing, resulting in a CHO solution of 10.1% ± 0.3% (Table 3). All marathoners used commercial products featuring CHO blends of maltodextrin (glucose) and fructose in the form of sports drinks and CHO gels. On average, marathoners consumed ~15 g CHO in ~150 ml every ~15 min of racing. Two of the marathon runners used caffeine (Table 3). They took ~60% of their total caffeine dose 1 hr before the start as caffeine pills and then consumed ~40% of their remaining dose throughout
the race via CHO gels containing caffeine, which resulted in 3.7 ± 1.4 mg caffeine per kg BW throughout the race. Variability of caffeine intake between athletes was mostly due to previous individual experience.


The aim of this case study was to characterize several training and nutrition techniques, leading to an individualized
race-day fluid and fuel plan. Although there are many variables to consider in marathon performance, together these unique periodized training and nutrition techniques were successfully implemented with three elite marathoners.

Training Load, Distribution, and Tapering Analysis

There are few publications on training characteristics of elite marathoners, and the athletes in the current study averaged
~13 hr/week or ~182 km/week of running (Figure 1), which is comparable to the volume range reported in elite French/Portuguese (Billat et al., 2001) and Kenyan runners (Billat et al., 2003; Onywera et al., 2004). However, their average volume and peak volume (231 km/week) are substantially higher than what was recorded by elite Spaniard (129 km/week) or Eritrean runners (105 km/week; Lucia et al., 2006). Training intensity also needs to be considered when assessing total training load. This case study contains the first published RPE-based TRIMP scores for an elite marathoner (Figure 2). Most contemporary training programs feature cycles of increasing load, followed by consolidation and recovery weeks, which are captured in the week-to-week TRIMP variability. Throughout the first 12 weeks, there was an average TRIMP of 3,350 ± 561 AU/wk, which is comparable to the ~4,000-AU/week threshold that Foster (1998) identified as what is routinely performed by elite endurance athletes. A polarized pattern of training features ~70–75% of training performed below lactate threshold and >15% well above that intensity (Fiskerstrand & Seiler, 2004).

This polarized training distribution has previously been used by successful (e.g., Olympic and World medallists) rowers, cross-country skiers, and track cyclists (Seiler & Tonnessen, 2009). Despite each marathoner having an individualized training program, with no training-plan interventions, a polarized distribution resulted (Figure 3). The 3-week pre-marathon taper featured a 52% reduction in volume with no appreciable change in training frequency (Figure 1). This taper is congruent with the recommendations from a recent meta-analysis on the effects of tapering on performance, which found the ideal length of taper to be ~2–3 weeks, where training volume was decreased 41–60%, without any modification of training intensity or frequency (Bosquet, Montpetit, Arvisais, & Mujika, 2007). Thus, it appears that all three athletes implemented a scientifically supported approach to training volume, distribution, and tapering.

Altering Carbohydrate Availability During Endurance Training

Conventional nutrition recommendations dictate that endurance athletes endeavour to replenish muscle glycogen stores post-exercise to ensure that subsequent training bouts are conducted in a glycogen-compensated state. However, reports from professional cyclists and East African runners indicate that some of those athletes purposely undertake some training in a glycogen-depleted/reduced and/or fasted/water-only state, in an attempt to force muscle adaptation (personal communication). In support of this, recent scientific hypotheses have suggested that periodically decreasing CHO availability may further enhance endurance-training adaptations and possibly performance (for review, see Burke, 2010). Nevertheless, it is difficult to extrapolate the results of laboratory training studies to the real world, where athletes are looking for only marginal improvements in performance over years of training. Despite the fact that low-CHO-availability training is both physiologically and psychologically challenging, these scientific and anecdotal reports suggest

that maximally adapted athletes may need to periodically undertake low-CHO-availability training to fully exploit endurance-training responses. Accordingly, these interventions were individualized at different rates into each  marathoner’s training program, and each was encouraged to increase the frequency, duration, and quality of these sessions throughout the general preparation phase and specific preparation phase as they better tolerated and adapted. There was a purposeful decline in these training sessions during the competition taper phase (Figure 4). Low-CHO-availability training had a large degree of individual variability according to personal acclimatization/ tolerance, as Marathoner 2 conducted ~35% of his training sessions with low CHO availability, compared with ~10% of training sessions for Marathoners 1 and 3. Marathoner 2 already had a lot of experience in doing fasted and low-glycogen training and thus was much more willing to undertake these types of training sessions.

Furthermore, feedback from the athletes indicated that morning-fasted training was physiologically much easier to integrate than the more strenuous reduced-glycogen training, as evidenced by the fact that only 10% of the low-CHO training was reduced-glycogen training.

Individualizing CHO and Fluid Intake During Training

Most studies demonstrate improved endurance performance when subjects consume CHO drinks compared with water, with CHO delivery and oxidation being a central mechanism (for review, see Jeukendrup, 2011). Accordingly, a recent CHO dose-response study showed a step-wise improvement in endurance performance in combination with increased CHO oxidation, with 60 g CHO/hr resulting in better performance than either 30 or 15 g CHO/hr (Smith et al., 2010). This suggests that the more CHO that can be feasibly consumed, the better the potential endurance-performance benefits. Recent evidence suggests that high intake rates (>90 g/hr) of glucose + fructose produce 20–50% greater exogenous CHO-oxidation rates, potentially decreased negative GI side effects (Jeukendrup, 2011), and an 8% increase in time-trial performance versus isocaloric glucose alone (Currell & Jeukendrup, 2008). Accordingly, marathoners used glucose + fructose mixtures at ~60 g/hr in an attempt to minimize adverse GI issues and to enhance performance. The most recent position stand on fluid intake (Sawka et al., 2007) highlights the need for individualized recommendations according to sweat rates (Table 2). Marathoners experimented with a wide range of CHO- and fluid-consumption rates to ascertain individual GI tolerance in varying training intensities, especially in weather conditions that athletes were likely to face on race day. This allowed for an individual intake profile of tolerance for fluids and CHO to be made for each athlete. A recent study examining CHO- and fluid-intake rates (Pfeiffer et al., 2012) showed that 73% of marathon runners did not reach the relatively low recommendation of 30–60 g CHO/hr during racing (Sawka et al., 2007). Thus, during this CHO-fuelling period (specific preparation phase and competition taper phase; Figure 4), athletes were strongly
encouraged to slowly increase intakes in an attempt to

at least reach the threshold of ~60 g CHO/hr, with even higher intakes if tolerated. However, excessive consumption of fluids and CHO during prolonged exercise can also result in adverse GI problems, as we have previously shown through significant
correlation between endurance athletes who have a history of GI problems and subsequent measured GI problems during competition (Pfeiffer et al., 2012). It has been suggested that the GI tract can adapt and be optimized to large fluid and CHO intakes (Jeukendrup, 2011), as GI comfort was shown to improve in just five training bouts with fluid consumption (Lambert et al., 2008). Furthermore, 1 month of endurance training with 100 g CHO/hr increased steady-state exercise CHO oxidation by 14%, which the authors hypothesized was due to increased intestinal CHO transporters (Cox et al., 2010). Therefore, in the last 4 weeks before our athletes’ target marathons, there was a continual emphasis on CHO fuelling and fluid intake (Figure 4). This process of fluid and CHO individualization, and gut adaptation, resulted in each athlete finding his individual balance between the ergogenic effect of maximal CHO and fluid intake on race day and the potential ergolytic effect of GI distress.

Race-Day Nutrition and Hydration Intakes

A recent large (N = 221) CHO- and fluid-intake field study showed a significant positive correlation between higher CHO intake rates and faster finishing times in Ironman and marathon races (Pfeiffer et al., 2012). Thus, all our marathoners attempted to maximize their fuelling and hydration plan that was individually optimized from the many pre-race nutrition/hydration-monitoring sessions (Table 2) by consuming ~15 g CHO in ~150 ml of fluids every ~15 min during the race, along with a total of ~3 mg caffeine/kg BW (Table 3). This fluid intake of 604 ± 156 ml/hr is very similar to the 550 ± 34 ml/hr recently reported in elite marathon runners (Beis, Wright-Whyte, Fudge, Noakes, & Pitsiladis, 2012). Haile Gebrselassie “only” ran 2:06:35 in his first serious marathon, consuming only water. Conversely, Gebrselassie consumed ~60–70 g CHO/hr (personal communication) in ~900–1,000 ml/hr (Beis et al., 2012) during his 2008 2:03:59 marathon world record. World class marathon performance is influenced by incredible genetics, physiological and psychological aptitude, and dedication to handle enormous training loads in an intelligent program. However, a periodized and individualized marathon nutrition and training approach can certainly facilitate the quest for marathon success.


An enormous thank you goes to marathoners Reid Coolsaet, Rob Watson, and Dylan Wykes and their coaches Dave Scott- Thomas and Richard Lee. Without their receptiveness to executing these interventions, and their open collaboration in allowing their data to be showcased, this case-study publication would never have come to fruition.

References: Stellingwerf-MarathonCaseStudy(IJSNEM2012)

Comments are closed.