Most weight-loss conversations in cycling get the order of operations wrong. They start with calories. They argue about thermodynamics. They land on some version of "eat less, ride more" and call it complete.
Prof Tim Spector spent the first half of his career doing what most cycling weight-loss commentators still do — looking at metabolism in narrow, mechanical terms. Then his research at King's College London on twins, on genetics, on the surprising heritability or non-heritability of weight, led him somewhere else. The biggest variable wasn't the genome. It wasn't even the calories. It was the trillions of organisms living in the gut.
I sat down with him for the 99% Get It Wrong episode of the Roadman Cycling Podcast. The conversation reframes most of what cyclists think about food. This is the breakdown — and why it matters specifically for performance.
What the microbiome actually is
The gut microbiome is the population of bacteria, fungi, and viruses living mostly in the large intestine. The numbers are difficult to grasp. Trillions of organisms. More microbial cells than human cells. More microbial genes than human genes by a factor of around 100.
These organisms aren't passive passengers. They digest fibres the body can't break down on its own, producing short-chain fatty acids — butyrate, propionate, acetate — that affect inflammation, gut lining integrity, and energy metabolism. They synthesise vitamins. They train the immune system. They communicate with the brain through neural and hormonal pathways. They influence hunger and satiety signals, glucose regulation, and the way the body partitions energy between storage and use.
Spector's framing is direct: "We've evolved with them so it's very hard to imagine us without microbes. We evolve from them because life started as microbes. You could say that microbes have created humans so they can live inside them."
The shift in thinking required is from seeing food as fuel and the gut as a passive tube, to seeing food as the input that shapes a complex ecosystem that then shapes everything else. What you eat doesn't just feed you. It feeds — or starves — a population of organisms that influence your body in profound ways.
Why this matters for cycling weight loss
The cycling internet's standard weight-loss advice — calorie counting, fasted rides, aggressive deficits — runs into a problem the microbiome research makes clear. Two cyclists eating the same number of calories can have meaningfully different metabolic outcomes depending on their microbial profile.
The ZOE PREDICT studies, which Spector co-leads, demonstrated this with controlled meal experiments. Identical meals produced very different glucose, insulin, and triglyceride responses across individuals — and across the same individual on different days. The microbiome was a major factor in explaining the variability.
For cyclists, this means the calorie-deficit approach is incomplete. A rider with a damaged microbiome — low diversity, dominated by inflammation-driving species, struggling with poor short-chain fatty acid production — will lose weight more slowly, regain it more easily, and pay a higher metabolic cost than a rider with a healthy microbiome eating the same diet.
The corrective isn't a different diet. It's adding the microbiome layer to the existing diet. Most cyclists are eating reasonably well by macronutrient terms but eating poorly by microbial terms — too few different plants, too many ultra-processed convenience foods, too little fermented food, no overnight fasting window. The fix is structural and slower than aggressive calorie restriction, but it produces results that hold.
This connects to the body composition guide and the body recomposition piece. The principle is consistent: sustainable shifts come from ecosystem changes, not from short-term deficits.
The 30-plants rule
The single piece of practical advice Spector gives most often is also the simplest: 30 different plant foods per week.
Plants include everything in the plant kingdom — vegetables, fruits, herbs, spices, nuts, seeds, whole grains, legumes. The number that matters is the variety, not the volume. A rider eating five servings of broccoli and three apples every day is eating a lot of plants but only two plants. A rider eating smaller amounts across 30 different species is feeding a much wider range of microbial organisms, which translates to a more diverse microbiome.
Why does diversity matter? Different microbial species specialise in different fibres. Inulin from chicory, beta-glucan from oats, pectins from apples, polyphenols from berries — each compound supports a specific subset of organisms. A diet narrow in plant variety leaves whole ecological niches unfilled. The microbiome contracts. Diversity drops. The metabolic effects follow.
The practical execution. Add herbs and spices to meals — each one counts. Use seeds in salads. Mix nuts in your post-ride snack. Try new vegetables you don't usually eat. Aim to recognise that a meal with 5-6 plant ingredients is much better than a meal with one or two.
For amateur cyclists, this is one of the highest-impact dietary changes available. It doesn't require calorie counting. It doesn't require giving up favourite foods. It adds rather than subtracts — and the additions feed the system that determines so much of how the body responds to training.
Ultra-processed foods, the central enemy
If 30 plants is the headline addition, ultra-processed foods are the headline subtraction.
Ultra-processed foods aren't simply "junk food." The category includes products that are widely marketed as healthy: protein bars with extensive ingredient lists, plant-based products with emulsifiers and stabilisers, low-calorie drinks with artificial sweeteners. The defining feature isn't sugar or fat content. It's industrial processing — combinations of ingredients you wouldn't find in a home kitchen, designed for shelf stability and palatability rather than nutritional integrity.
The mechanisms by which these foods affect the gut are increasingly well-mapped. Emulsifiers disrupt the mucus layer of the gut lining. Artificial sweeteners alter microbial composition in measurable, often unfavourable ways. Hyperpalatable combinations of fat, sugar, and salt drive over-consumption beyond what the body's satiety signals can correct. Some compounds drive low-grade chronic inflammation that's hard to identify but easy to feel — fatigue, slow recovery, blunted training response.
Cyclists who fuel hard rides with bars and gels need to think about this carefully. Some ride-specific products are reasonable trade-offs — the demand for portable, fast-absorbing carbohydrate during a hard effort justifies products that wouldn't otherwise belong in a daily diet. But the at-home diet, the daily eating pattern that surrounds the riding, is where the long-term effect is set. A rider whose breakfast is a protein shake of mixed processed ingredients, lunch is a packaged sandwich, snacks are bars, dinner is a frozen meal — that rider is fighting against their own gut, and the training adaptations are blunted accordingly.
The fix is direct: more food cooked from recognisable ingredients. Less food from packages with long ingredient lists. The framing isn't "diet" — it's "rebuild the inputs."
Fermented foods and the diversity boost
The third practical lever Spector points at is fermented foods. Live yoghurt, kefir, sauerkraut, kimchi, miso, kombucha. These foods deliver live microbial cultures directly to the gut, supplementing the populations already present.
The evidence base is strongest for the fermented foods that deliver living organisms. Pasteurised products that have been heat-treated for shelf stability lose much of the benefit. The cultures need to be alive when they reach the gut.
For cyclists trying to add fermented foods without making it a chore, the integration is simpler than it sounds. A morning bowl with live yoghurt, berries, and seeds. Sauerkraut as a side with dinner. Kefir as a base for a smoothie. None of these require dramatic dietary overhaul. They're additions, not replacements.
The window of effect is faster than most dietary changes. Microbiome composition shifts visibly within two weeks of consistent fermented food intake, and meaningful health markers — sleep quality, mood, digestive comfort — often follow within months.
The 12-hour overnight window
The fasting question came up in our conversation. Spector's view is more measured than the cycling internet's enthusiasm for extended fasting.
A 12-hour overnight window — finishing dinner by, say, 8pm and not eating again until 8am — is supported by good evidence as a beneficial daily structure. It gives the gut lining time to repair, allows microbial populations to rebalance, and aligns eating with the body's circadian rhythms. Most people can hit this without any deliberate change to their schedule simply by not snacking late at night.
Longer fasts — 16-hour, 20-hour, full-day fasts that have become popular through intermittent fasting protocols — have weaker evidence for additional benefit beyond the 12-14 hour window. For cyclists training hard, longer fasts often interact badly with recovery. The body needs fuel and protein within a reasonable window after hard training to drive adaptation. Pushing fasting windows out to skip post-ride nutrition is counterproductive.
Spector's practical guidance: respect the overnight window, don't push it to the extreme, and prioritise recovery nutrition after hard sessions. The cycling fasted riding myth post covers the related debate about training fasted, with a similar conclusion — most cyclists are doing this wrong and paying for it in performance.
The performance angle
Most of Spector's research is framed in terms of general health outcomes — weight, mood, longevity. The performance angle is less developed in the literature, but the inferences are clear.
A rider with a healthy gut absorbs nutrients more efficiently. The carbohydrates and proteins consumed in the post-ride window arrive at the muscles in better condition. The inflammation response to hard training is better managed. The hormonal environment that drives adaptation is more favourable. Sleep quality, which affects performance directly, is influenced by microbial production of neurotransmitter precursors.
A rider with a damaged gut fights all these things. The fuel doesn't land. The recovery is slow. Inflammation lingers. Sleep is poor. The training stimulus is the same, but the adaptation is smaller.
The cycling weight-loss conversation is incomplete without this layer. The hidden universal motivator of body composition — what most cyclists actually want — runs through the microbiome whether they're paying attention to it or not.
What to do this week
Three changes, in order of impact:
Audit your plant variety. For one week, write down every distinct plant food you eat. Vegetables, fruits, herbs, nuts, seeds, whole grains. Most cyclists land between 10 and 15. The target is 30. Add gradually — different vegetables in your meals, herbs and spices used deliberately, mixed seeds in your breakfast, nuts as snacks instead of bars.
Cut the ultra-processed foods. Identify the ultra-processed items in your daily routine — the protein bars, packaged sauces, sugary drinks, processed snacks — and replace them. Cooking from recognisable ingredients beats supplementation every time.
Respect the 12-hour overnight window. Finish dinner earlier or start breakfast later. A consistent 12-hour gap, every night. No deliberate effort needed beyond the one structural change.
The full conversation with Prof Spector — including his work on twins, the ZOE studies, and his views on cholesterol and saturated fat — is in the 99% Get It Wrong episode. The other nutrition-pillar conversations cover food and performance from different angles.
If you want a structured approach that integrates training with nutrition guidance — including the food-quality work most cycling programs ignore — Roadman coaching treats the diet as part of the program, not a side conversation. The microbiome layer is part of why.