Professor of genetic epidemiology at King’s College London who started the TwinsUK study; author of The Diet Myth, Spoon-Fed, Food for Life and Ferment
Food7 ideas
His advice for healthy adults, not medical advice
Checked 17 Oct 2026. Each line is in our own words, with where it was said and an independent study or guideline next to it. Ties: Co-founded ZOE and is a consultant with share options in it; ZOE sells a gut-health test, a food-scoring app and Daily30, a 30-plant powder.
Mixed evidence · In the American Gut Project (over 10,000 volunteers, self-reported), people eating more than 30 plant types a week had more varied gut microbes than those eating under 10. Observational; 30 was a survey answer, not a tested optimum, and he co-wrote the paper Research
Costs: no extra time; herbs, spices, nuts, seeds, beans and grains count; he suggests a tally on the fridge door
Habit
Three portions of fermented food a day: yoghurt, kefir, kraut, kimchi, cheese, miso
Mixed evidence · One small Stanford trial (18 people built up to about six servings a day): gut microbe diversity rose and 19 of 93 inflammation markers fell, but its main immune outcome did not change. ZOE’s own 9,000-person study had no control group Research
Costs: a few euros a week; yoghurt and kefir at breakfast already make two; or ferment your own
Habit
Eat within a 10-hour window, 14 hours overnight without food
Mixed evidence · Trials found no extra weight loss: eating 12:00 to 20:00 vs three meals (116 adults, 12 weeks), and 8:00 to 16:00 plus calorie cuts vs calorie cuts alone (139 adults, a year). The energy and mood gains he cites come from ZOE’s own app study, without a control group Research
Costs: no money; pick your own hours (he eats from about 10:30 to 21:00); black coffee and tea are fine outside the window
For your blueprint
Cut ultra-processed food one swap at a time; distrust “low fat”, “low cal” or “no added sugar” on the front of a pack
Mixed evidence · An umbrella review (45 meta-analyses, 9.9 million people) links more ultra-processed food to heart deaths, type 2 diabetes and depression, mostly on low-quality observational evidence. A 2-week ward trial: about 500 kcal a day more eaten on the ultra-processed diet Research
Costs: more cooking time; some whole foods cost more
For your blueprint
Plants at the centre of the plate; meat and fish on the side
Guideline · The Dutch Health Council’s 2015 guidelines advise a more plant-based, less animal-based diet. A Lancet meta-analysis: the most fibre (25 to 29 g a day) tracks 15 to 30 % lower mortality, and trials show lower weight, blood pressure and cholesterol Research
Costs: no money; often cheaper than meat
For your blueprint
Pick the most colourful, bitter, strong-tasting kinds: red leaves, purple carrots, herbs and spices
Mixed evidence · Flavonoids, one family of polyphenols, tracked lower mortality up to about 500 mg a day in 56,000 Danes followed for 23 years; observational. Colour and bitterness as a guide to polyphenols is his rule of thumb Research
Costs: no time; sometimes a pricier variety
For your blueprint
Experiment on yourself: change what or when you eat, notice how your body responds, keep what suits you
Their own practice · Blood-sugar responses to the same meal differ widely between people (800 people, Weizmann Institute, 2015). Testing food on yourself has not been studied as such; ZOE’s trial of its own personalised app found small effects Research
Costs: no money; a note on how you feel
How to use this
Don't take all of it. Pick one or two lines that fit your week, try them for two weeks, and keep what works for you. That's what Vea is for: a weekly check-in that shows what each habit does for you.
Not an endorsement: these are our short summaries of what each person has published, with a link to the source. They have no connection to Vea. Ideas to try, not medical advice. On this list and want off? [email protected]