The advice to eat 30 plants a week has spread widely enough to feel like settled fact. It came from a specific place: a citizen science study that asked over 10,000 people what they ate, collected their stool samples, and sequenced the bacteria inside them. The American Gut Project ran that experiment, and the 30 plants a week figure traces directly back to one question on its questionnaire.
Knowing where the number came from helps you understand how much weight to give it, and what the researchers themselves said it actually shows.
What was the American Gut Project?
The American Gut Project launched in November 2012 as a collaboration between the Earth Microbiome Project and the Human Food Project. Participants paid a small fee, received a home-sampling kit, filled out a detailed lifestyle questionnaire, and posted their stool sample back to the lab. By May 2017, it included 15,096 samples from 11,336 human participants, primarily in the United States and United Kingdom.
The first major results appeared in 2018 in the journal mSystems (McDonald et al.). The authors described the resulting dataset as “the largest public reference database of the human gut microbiome,” and it remains open access, actively used by other research teams.
How participants described their diet
The questionnaire covered everything from sleep patterns to antibiotic use, but the diet questions are the part that matters here. Participants answered several types of question.
Diet category — a single self-selected label: omnivore, vegetarian, vegan, and similar.
Food frequency — a five-point scale (Never, Rarely, Occasionally, Regularly, Daily) applied separately to fruit, vegetables, different types of meat, sugary drinks, alcohol, and more.
Specialised diets — checkboxes for dairy-free, gluten-free, grain-free, and others.
One question stood apart: “In an average week, how many different plants do you eat?" Participants chose from five bands:
| Plants per week |
|---|
| Fewer than 5 |
| 6–10 |
| 11–20 |
| 21–30 |
| More than 30 |
The paper’s headline comparison focused on the extremes: people reporting fewer than 10 different plants per week versus those reporting more than 30. The questionnaire included a worked example: one can of soup containing three vegetables counted as three plants. The framing was practical and recognition-based, not botanical, which is part of why debates about whether herbs or mushrooms should count have persisted ever since. The guide to what counts for the 30 plants a week challenge goes into those edge cases in detail.
How the gut testing worked
Each sample was analysed using 16S rRNA gene sequencing. The V4 region of this gene varies enough between bacterial species to act as an identifier: by sequencing it and comparing results against a reference database, researchers could establish which bacterial genera were present in each sample and in what relative proportions.
Two diversity measures were central to the analysis. Alpha diversity describes how varied a single person’s microbiome is — the more genera present and the more evenly spread they are, the higher the score. Beta diversity describes how different two people’s microbiomes are from each other.
The method has inherent limits. 16S sequencing identifies bacteria to genus level at best, it does not identify individual species. The authors were also explicit that their results are associative, not causal, and should be read as hypothesis-generating rather than proof.
What the study concluded
The headline finding was that people who reported eating more than 30 different plants per week had higher gut microbiome diversity than those eating fewer than ten. The Microsetta Initiative, which extended this work internationally, later summarised it as: “people who reported consuming 30+ different kinds of plants per week had a more diverse mix of gut microbes."
Some results were less intuitive. Broad lifestyle factors such as age, BMI, and recent antibiotic use proved stronger predictors of microbiome composition than diet labels. Vegetarians, vegans, and omnivores did not cluster into distinct microbiome types. What appeared to matter more was variety across plant foods, regardless of how someone described their overall diet.
Worth stating plainly: the paper’s most detailed follow-up analyses of high and low plant-eaters used a much smaller extreme-groups subset, with roughly 40 participants per group, rather than the full American Gut cohort. The data is self-reported and cross-sectional. It can only show correlation, but not causation.
Where the research stands today
The AGP dataset has been drawn on extensively since 2018. A 2020 paper in mSystems examined fermented food consumption among participants and its links to specific gut compounds. A 2023 analysis in the Journal of Nutrition used AGP data to associate nutrient-based diet quality scores with gut profiles. The original 2018 paper also reported that some mental-health-related associations were reproducible within matched U.S. and U.K. subsets of the AGP data.
Randomised controlled evidence on whether deliberately eating 30 or more plant varieties per week changes gut health remains limited. The specific threshold is still better understood as a useful heuristic than a clinically established target.
How “30 plants a week” became a target
The number moved from dataset to dinner table through press communications, popular science books, and researchers finding it a practical teaching tool. The AGP’s own summaries drew attention to the contrast between participants eating fewer than ten plants per week and those eating more than 30. Tim Spector, who co-founded the Zoe nutrition project and has written widely about gut health, gave the finding broad reach in the UK and beyond. It became shorthand for a more general principle: dietary variety matters, and a concrete number makes it something you can track.
The 30-plant figure emerged from real data, communicated by researchers who believed the underlying finding was meaningful. Whether 29 plants is significantly different from 31 is a question the current evidence cannot answer. What the data does consistently show is that more variety is associated with a richer microbiome, and that most people eat far less variety than they assume until they start paying attention.
Herbyvore is built on that same idea — tracking plant variety using the same practical counting logic the American Gut Project applied, to make the habit visible and easy to maintain.
Important note: any significant changes to your diet, especially if you have medical conditions or specific health goals, should be discussed with a qualified healthcare professional first.