Nutrition advice for cyclists tends to fall into two camps: under-complicated Instagram rules ("eat whole foods, stay lean") and over-complicated supplement stacks that apply to less than 1% of the riding population. Neither serves the athlete trying to hold power in the final hour of a five-hour ride or recover well enough to train again the next morning.
Over the past several years, the Roadman Cycling Podcast has interviewed more than 1,400 coaches, sports scientists, and professional riders. A recurring pattern emerges when the topic is nutrition: the experts largely agree. The disagreements are narrow and mostly concern context — when a strategy applies, not whether it works. This article synthesises the consensus.
The framing matters. These are the nutrition principles that sports scientists apply to their own athletes, not principles built for the general population. If you ride more than eight hours a week and take your performance seriously, this is the framework they use.
The consensus: most cyclists under-fuel hard days
Asker Jeukendrup, whose research on carbohydrate oxidation rates has shaped World Tour fuelling practice for two decades, is direct on this point: the majority of trained cyclists consume far less carbohydrate during hard efforts than their physiology requires. The gap is not marginal. Studies consistently show recreational and amateur cyclists averaging 40-60g of carbohydrate per hour during rides where their output demands 90g or more.
The consequences are predictable. Power fades in the final third of long rides. Interval sessions lose quality after the first set. Recovery between sessions slows. Yet when athletes start fuelling correctly, they often attribute the performance improvement to fitness gains they can't explain — not realising the adaptation was nutritional.
Tim Spector's work through ZOE adds a useful layer here: individual responses to foods vary considerably based on gut microbiome composition. But this variability operates within a ceiling. If glycogen is not available during high-intensity work, no amount of microbiome optimisation changes the outcome. The individualisation Spector champions is meaningful for food choices and meal timing around training; it does not alter the fundamental requirement for carbohydrate during hard sessions.
The practical implication is that if you are not tracking what you eat on your bike during efforts above Zone 2, you are almost certainly under-fuelling. Most athletes are surprised how much food a proper fuelling protocol requires.
Carbs per hour: the number every expert lands on
Jeukendrup's research established the 90-120g per hour figure, and it has held up. Here's the mechanism: glucose is absorbed through SGLT1 intestinal transporters, which saturate at roughly 60g per hour. Adding fructose, which uses the separate GLUT5 transporter, allows total oxidation to reach 90g and, in well-trained athletes with practised gut capacity, up to 120g per hour.
The ratio that consistently performs best in the research is approximately 2:1 glucose to fructose. In practice, this is achieved with products combining maltodextrin or glucose polymers with fructose — or, more simply, with a mix of glucose drinks and fruit-based foods. There is nothing exotic about the chemistry.
The 60-minute threshold is critical context. Below 60 minutes of hard effort, glycogen stores are rarely the limiting factor. Above 75-90 minutes, they become one. The fuelling target of 90-120g per hour applies to rides of sustained intensity lasting longer than 60-75 minutes. For shorter, high-intensity sessions, pre-session glycogen availability matters more than in-session fuelling.
Dan Lorang, Head of Performance at Red Bull-Bora-Hansgrohe and coach to multiple World Champions, has spoken on the podcast about the logistical discipline his athletes apply to race nutrition — not just knowing the target number but practising it so consistently in training that consuming 90g per hour becomes automatic. The fuelling plan is rehearsed as deliberately as the race tactics.
Use our fuelling calculator to work out your specific per-hour carbohydrate target based on body weight, duration, and intensity.
The fasted training debate
Fasted training generates more debate than any other nutrition topic in endurance sport, mostly because proponents and critics are often talking about different things. The evidence base is not actually contentious when context is applied correctly.
The case for fasted training rests on metabolic adaptation: training in a low-glycogen state upregulates fat oxidation enzymes, increases mitochondrial density, and may improve fat utilisation at submaximal intensities. These adaptations are real and documented. Prof. Stephen Seiler's polarised training research provides the structural context: if the large majority of your training volume is low-intensity work, fasted sessions within that category can produce meaningful adaptation without compromising the quality sessions that drive performance.
The case against fasted training is equally clear: high-intensity work done in a fasted state produces lower power outputs, increases muscle protein catabolism, and does not produce adaptations superior to fuelled high-intensity work. It is the wrong tool for the wrong job.
The consensus position, which Jeukendrup has stated clearly and which most applied practitioners echo, is this: fasted training is appropriate for low-intensity sessions lasting less than 90 minutes where the specific goal is fat oxidation adaptation. It is not appropriate for threshold intervals, VO2max work, race-simulation efforts, or long rides above Zone 2 intensity. Our full breakdown of the evidence is in the fasted vs fuelled article.
One practical point worth emphasising: many athletes who believe they are doing "low-intensity fasted training" are actually riding at an intensity that is too high for the protocol to be appropriate. Fasted sessions require strict zone discipline.
Race weight: what experts actually recommend
Almost every sports scientist who has been on the podcast arrives at the same position on race weight: it matters, but it is not the first lever to pull, and chasing it with caloric restriction during a training block is counterproductive.
The physiological argument is clear. Caloric restriction during a training block suppresses testosterone and IGF-1, increases cortisol, impairs muscle protein synthesis, and reduces the training adaptations the athlete is trying to achieve. Joe Friel, whose work in The Cyclist's Training Bible has shaped periodisation thinking for decades, is consistent on this: you cannot optimise body composition and training adaptation simultaneously if the method for achieving one undermines the other.
The framework that World Tour practitioners apply: fuel training correctly, support recovery, and allow body composition to stabilise before deliberately manipulating it. For most amateur athletes who are chronically under-fuelling, correcting the fuelling deficit alone results in improved body composition over 8-12 weeks — counterintuitively, eating more during training reduces fat mass because recovery improves, training quality increases, and lean muscle is preserved. Anthony's own 9kg loss-while-eating-more story is a worked example of exactly this sequence.
Race weight, where a specific composition target is appropriate, is addressed in a defined window: typically 8-12 weeks out from an A-race, with a modest caloric reduction applied only to easy training days and never to days with quality sessions. The Impey periodised nutrition framework formalises this — fuel the work, restrict only when the work does not require it.
Targeting race weight before establishing a fuelling competence baseline is the wrong sequence. The sequence that experts apply consistently is: fuel correctly first, demonstrate that you can sustain 90-120g per hour in training, then periodise nutrition around body composition goals if they remain relevant.
The gut training consensus
No expert who has engaged seriously with Jeukendrup's research disputes the core finding: the gut is trainable. The intestinal transporters that absorb carbohydrate during exercise — SGLT1 for glucose, GLUT5 for fructose — upregulate with consistent practice. Athletes who regularly consume 80-100g of carbohydrate per hour during training tolerate that intake in races with significantly fewer gastrointestinal symptoms than athletes who only attempt high carbohydrate intake on race day.
The practical consequence is that gut training is not optional for any athlete targeting 90g per hour or above. If the goal is to race at 90-120g per hour, training at 60g per hour and hoping for the best on race day is a reliable path to GI distress. The gut needs the same progressive overload logic applied to training load.
Tim Spector's microbiome research provides a relevant supplement to this. Gut microbiome composition influences the degree of GI symptoms during high carbohydrate intake, and some of the individual variability athletes experience is microbiome-mediated. But the primary driver of gut carbohydrate tolerance is transporter training, not microbiome baseline. Both matter; transporter training is the one athletes have the most direct control over.
The gut training protocol that applied practitioners recommend: start at 60g per hour on long training rides, increase by 10-15g per hour every 2-3 weeks, and target race-level intake at least 6-8 weeks before the A-race. Race-day nutrition should be rehearsed — the same products, the same timing, the same total intake — not improvised.
Dan Bigham, Head of Engineering at Red Bull-Bora-Hansgrohe and former UCI Hour Record holder, has discussed on the podcast how even at the elite level, riders who have not practised high-carb fuelling in training consistently perform worse than their fitness predicts in long, hard races. The limiting factor is not cardiovascular — it is gut capacity that was never trained.
One nutrition change that compounds
Every sport scientist interviewed on the podcast, when asked what single change would most improve an amateur cyclist's performance, gives a version of the same answer: eat more on your hard days. Not different foods. Not supplements. Not a new timing protocol. Simply more carbohydrate during the sessions that require it.
This is the change that compounds because it is not additive — it is multiplicative. When hard sessions are properly fuelled, power outputs are higher, which creates a stronger training stimulus. Recovery is faster, which means quality sessions can be stacked more closely. Strength work is more effective. Glycogen replenishment between sessions is complete, so the next session starts from a full tank rather than a depleted one.
The downstream effects accumulate over weeks and months in ways that no single supplement can replicate. Athletes in the Roadman Cycling Not Done Yet coaching programme routinely report their biggest performance improvements in the six to eight weeks after correcting chronic under-fuelling — a period where their training load has not changed significantly but their fuelling has.
The practical starting point is simple. For your next ride over 90 minutes and above Zone 2 intensity, set a target of 80g of carbohydrate per hour and eat to that number. Use the fuelling calculator to translate that into grams of specific foods or products. Track the data for four weeks. The pattern in the numbers will be more persuasive than any argument made here.
Got a specific question — your own per-hour target, whether to add fructose to your current mix, how to handle a hot race? Ask Roadman for an answer drawn directly from the conversations with the experts above. And if you want the whole nutrition system built around your training week rather than figured out solo, the application for NDY coaching is where that starts.