If you've ever counted calories, downloaded MyFitnessPal, weighed your porridge oats and watched the kilos refuse to move, this episode of the conversation is for you. Prof. Tim Spector has run the trials, co-founded Zoe, and is now the most cited voice in nutrition science globally. His position is that the calorie is the wrong unit — particularly for endurance athletes whose bodies are exquisitely calibrated to defend energy availability.
For cyclists trying to drop body fat without losing power, the implications are uncomfortable and useful in equal measure. Here's the conversation.
The calorie isn't a useful unit
Spector's argument starts here. The calorie was a unit invented to describe heat in a sealed bomb calorimeter. The human body is not a sealed bomb calorimeter. Food gets processed by 100 trillion gut microbes that extract, transform and discard nutrients in ways that depend on the microbe population, the food matrix, the cooking method, the time of day, the rider's stress state, and a dozen other variables.
Two riders eating the same 2,000-calorie meal will absorb meaningfully different amounts of energy from it. The same rider eating the same meal in the morning versus the evening will absorb different amounts. The processed version of the same food and the home-cooked version produce different glycaemic and microbial responses even when their stated calorie counts are identical.
The calorie therefore measures something. It just doesn't measure the thing most amateur cyclists assume it measures. It measures the energy contained in the food. It does not measure the energy your body extracts, stores, or burns from that food. The gap between those two numbers is where most calorie-counting attempts fail.
This is why Spector's twin studies — identical twins followed over a decade where one twin dieted repeatedly and one didn't — found the dieting twin heavier and metabolically worse off after ten years. The body adapted. Hunger climbed. Metabolism slowed. The arithmetic that should have worked didn't, because the arithmetic was never really the mechanism.
This is the same point Anthony has made on the podcast for years and lived in his own body: 9kg lost while eating more food, with power going up not down. The mechanism wasn't a deficit. The mechanism was a complete change in what was on the plate, fuelling that finally matched the demand of the work, and a microbiome that could regulate appetite honestly because it was being fed properly.
Microbes as a chemical factory
The frame Spector keeps coming back to: think of your gut microbes as a 1.5–2kg pharmaceutical factory living in your colon. They eat the fibre and complex carbohydrates you can't digest. In return they produce vitamins, appetite-suppressing peptides, neurotransmitters, immune-regulating chemicals, and short-chain fatty acids that protect the gut lining and reduce systemic inflammation.
The output of this factory depends entirely on what you put in. A diverse microbiome — 1,000+ species — produces a diverse range of beneficial chemicals. A depleted microbiome — 200-300 species, which is increasingly common in Western populations — produces fewer of those chemicals and more of the inflammatory ones associated with metabolic disease.
Spector's number: people with poor gut microbial diversity have roughly twice the risk of nearly all the diseases of the Western world. Cardiovascular disease, type 2 diabetes, autoimmune conditions, depression, obesity. Twice the risk. From the same lifestyle factors. The microbiome is the most actionable amplifier or dampener of risk we have.
For cyclists this matters in three specific ways. Recovery — the inflammation-controlling chemicals microbes produce affect how quickly you bounce back from hard sessions. Mood — many of the precursors to serotonin and dopamine are produced by gut microbes, which is why under-fuelled cyclists with poor diet diversity also tend to be the irritable ones in week six of training. Appetite regulation — the satiety signalling that lets you eat to fullness without overshooting comes substantially from microbial metabolites that disappear when your diet narrows.
The 30-plants-a-week target
This is the practical handle Spector gives. Aim for 30 different plants per week. Vegetables, fruits, legumes, whole grains, nuts, seeds, herbs, spices.
The number is high enough that hitting it requires deliberate choices, low enough to be achievable in a normal kitchen. Variety is what matters, not volume. A bowl of porridge with mixed seeds, blueberries and a banana is six plants in one breakfast. A salad with five different vegetables, two herbs, a sprinkle of pumpkin seeds and a vinaigrette of olive oil and lemon is eight more. Add a stir-fry with onion, garlic, ginger, broccoli, peppers, mushrooms, tofu and spring onions — eight more. You're at 22 by Tuesday lunchtime.
Why variety beats volume: each plant species feeds a slightly different population of microbes. The species that loves fermentable fibres in oats is different from the species that loves the polyphenols in dark berries. Eat the same five plants every day at high volume and you support the same five microbial populations. Eat thirty different plants and you support thirty different microbial populations, which dramatically increases the diversity of chemicals being produced.
Frozen counts. Tinned counts. Dried counts. The barrier to 30 isn't access or cost — it's the habit of grabbing the same five things every shop. Buy two new things each week from the produce section. Within a month you've widened your palette without thinking about it.
If your background is the diet-culture, calorie-counting model, this can feel almost too simple. It's not simple in implementation — most amateurs run their kitchen on autopilot — but it's simple in concept. Eat more different real things.
Ultra-processed food as the actual problem
The thing Spector wants Western diets to confront isn't sugar specifically, or fat specifically, or red meat specifically. It's ultra-processed food. UPF.
Definition: anything with a list of ingredients you wouldn't have in your kitchen. Emulsifiers, gums, artificial sweeteners, preservatives, flavour enhancers, modified starches, refined oils used as the dominant fat. Roughly 60% of the average Western diet is now UPF.
Three problems with UPF that show up in the cycling context. It suppresses microbial diversity directly — the additives that make UPF shelf-stable also kill many of the beneficial gut species. It increases appetite — multiple studies show people eat 500-700 more calories per day on a UPF diet versus a whole-food diet of equivalent macros. It bypasses satiety signals because the food matrix is engineered to be hyper-palatable and easy to over-consume.
The cyclist who's "tracking calories rigorously" but eating mostly UPF is fighting their biology in a way the cyclist eating whole food is not. The first rider's body is producing fewer satiety signals, more inflammation, and demanding more food. The second rider's body is producing the opposite.
The honest read on most amateur cyclist body-composition struggles isn't "they need more discipline" — it's "they need to remove UPF and let satiety do the work." Once UPF goes, calorie tracking usually becomes irrelevant within a few weeks. The riders we work with inside the Not Done Yet community who break long-running weight-loss plateaus almost always do it by changing the quality of what they eat, not the quantity.
The exercise-and-weight-loss point that frustrates cyclists
Spector said something a lot of cyclists won't want to hear, and it's important enough to face directly.
Exercise alone, as an intervention for weight loss, has an average effect of zero in randomised controlled trials. Multiple studies, multiple countries, large samples. People assigned to "more exercise, no diet change" do not, on average, lose more weight than people who don't add exercise.
The mechanism is the same as the calorie-restriction one. The body compensates. Hunger climbs. NEAT (non-exercise activity thermogenesis — fidgeting, posture, small movements) drops. The energy budget defends itself. Some people lose weight from added exercise. Roughly the same number gain weight from it because their hunger response is more aggressive than their training-volume increase.
This doesn't mean exercise is useless. It means exercise alone is not a weight-loss intervention. Exercise is a cardiovascular, metabolic, mental health, and longevity intervention. It's also one of the most reliable predictors of keeping weight off once it's been lost — which is why every long-term weight-loss study finds exercise on the success side.
The implication for cyclists trying to drop body fat: the bike is what your body composition will defend on. The kitchen is where the change actually happens. Anthony's 9kg-while-eating-more story is the case study. The training stayed the same. The food changed. The weight came off. The power went up. That's not magic. That's what happens when you stop fighting the satiety system and start fuelling it instead.
Personalisation is real and it complicates everything
The Zoe research has shown — repeatedly — that individual responses to the same food are dramatically different. Two people eat the same croissant. One has a sharp glucose spike and a steep drop into hunger and fatigue at hour three. The other has a modest, smooth glucose curve and feels fine. Same food, same morning, opposite responses.
The reason is genetic, microbial and contextual. Different microbiomes ferment carbohydrate differently. Different insulin sensitivities produce different responses. Different recent meals influence how this meal is processed.
What this means practically: generic diet advice will work for some riders and fail others. The riders for whom it fails are not weak-willed. Their physiology is different. Continuous glucose monitors, blood-fat testing, and microbiome sequencing are slowly making it possible to personalise — but the simple version most cyclists can do today is to observe their own responses honestly.
Pay attention to which meals leave you energised at hour three and which leave you crashed. Track which fuels in your long rides keep you steady and which crash you. Notice which combinations let you eat to fullness and which leave you hungry an hour later. The data isn't lab-precise. It's directional, and direction is enough.
What to do this week
Three actions that operationalise the conversation.
Count plants for one week. Not calories. Not macros. Plants. Make a list each day. Aim for 30 distinct plants by Sunday. Most riders are at 12-15 the first week. By week four, with deliberate variety, 25-30 is normal.
Identify and remove three ultra-processed items from your weekly shop. Cereal bars, flavoured yoghurts, ready-meals, sauces with eight ingredients you can't pronounce. Replace with the closest whole-food version. Don't try to remove all UPF in one go. Remove three. Hold for two weeks. Add three more.
Stop tracking calories for two weeks if you've been tracking. Replace the tracking with two questions at every meal. Is this real food? Did I eat to fullness? Most cyclists who run this experiment report eating less while feeling more satisfied — because they stop overriding satiety to hit an arbitrary number.
The Spector position isn't that calories are meaningless. It's that they're the wrong primary lens. Quality first. Diversity second. Satiety third. The arithmetic falls out of those three. We laid out the broader framework in the cycling weight-loss breakdown and inside the coaching that runs in the Not Done Yet community.
If your body composition has been stuck despite high training volume, the answer is unlikely to be more training or fewer calories. It's more variety on the plate and less food that comes out of a packet. Run that for eight weeks. Then re-measure. The number will have moved.
A note for cyclists already lean
The piece so far has been written for cyclists trying to drop body fat. A short note for the readers in the opposite category — already lean, possibly too lean, possibly under-fuelled.
Spector's framework still applies but the implications flip. You don't need to remove energy. You need to add diversity and quality without removing volume. The 30-plants-a-week target works regardless of total calories. The UPF-removal works regardless of total calories. You can hit both while eating substantially more than you currently are.
Under-fuelled cyclists — particularly women, masters athletes, and high-volume amateurs — are at meaningful risk of relative energy deficiency in sport (REDs). Symptoms include suppressed hormones, frequent illness, recurring injury, persistent low mood, plateaued performance. The fix is more food, not less. The kind of food matters as much as the volume. Whole-food, fibre-rich, plant-diverse, lower-UPF — eaten to satiety. The body composition will defend itself if the inputs are right.
We covered the under-fuelling risk specifically in the under-fuelled cyclist breakdown — energy deficiency is one of the five most common causes of plateaued performance and the one most amateur men silently dismiss because they think it only applies to women. The Energy Availability calculator is the cleanest way to find out where you actually sit before assuming you're fine.
Whichever side of the spectrum you're on, Spector's principles point at the same place. Eat real food. Eat varied food. Eat enough. Let your body's satiety signals do the regulating once the inputs are honest.