Fruits & Vegetables Prevent Disease

In the not-so-distant past, nutritionists and consumers alike viewed fruits and vegetables as average foods. Decades of research have changed that completely. We now know they carry compounds capable of preventing a great deal of what fills my clinic — heart disease, stroke, several cancers, type 2 diabetes.

What has changed most since I first wrote this is not whether they help. That was settled. It is how much, and the answer is more encouraging than the one I gave you in 2015.

The dose question, finally answered properly

For years the honest answer to “how many should I eat?” was a shrug and “more.” Two large pooled analyses have since given us something much better.

The first, published in 2017, combined 95 studies covering roughly two million people. It found risk falling steadily as intake rose, with the lowest risk of heart disease, stroke, cancer and death at around 800 g a day — about ten portions.¹ That figure was widely reported, and it discouraged a great many people who could not imagine eating ten portions of anything.

The second, published in Circulation in 2021, pooled two large American cohorts with 26 others — around two million participants and over 30 years of follow-up — and asked the question more carefully. Its finding is the one I now quote in clinic: risk of death fell as intake rose to about five servings a day, and then flattened. Beyond five, more produced no further measurable benefit.²

The study I trust most on this question

General population cohorts have a structural weakness when studying plant-based eating: almost nobody in them eats a genuinely plant-based diet. The “highest intake” group is often just people eating a normal Western diet with a bit more salad. That compresses the range and understates what a real dietary pattern can do.

The Adventist Health Study-2 does not have that problem. It follows roughly 96,000 Seventh-day Adventists in North America — a population containing large numbers of lifelong vegans, lacto-ovo vegetarians, pescovegetarians and non-vegetarians, most of whom neither smoke nor drink. It is as close to a natural experiment in dietary pattern as observational nutrition gets,⁶ and because the confounders that plague this field are largely absent, the diet signal comes through unusually cleanly.

What it found is worth stating as a gradient, because that is what it is.

I put more weight on this than on any single finding about a single food group, and it is why my practical advice has shifted over the years from “eat more vegetables” toward “shift the shape of the plate.” Adding a side salad to an otherwise unchanged diet is not what these participants were doing.

Two honest caveats. Adventists differ from the general population in more than diet — the studies adjust for smoking, alcohol, exercise and weight, but no adjustment is perfect. And pescovegetarians doing best is a real finding that does not fit a simple “more plants is always better” story; I would rather report it than tidy it away.

Not everything green counts

The same analysis found something I had not appreciated a decade ago, and it is worth knowing before you count your day.

Starchy vegetables — potatoes, peas, corn — were not associated with lower mortality. Neither were fruit juices. What drove the benefit were leafy greens, and fruits and vegetables rich in beta-carotene and vitamin C: carrots, citrus, berries, tomatoes, cruciferous vegetables.²

So a portion of chips is not a portion of vegetables, and a glass of orange juice is not an orange. I would rather tell you that plainly than let you count a day that does not count.

Why they protect the heart

Fruits and vegetables contain no cholesterol and very little saturated fat. They are low in energy density and high in fibre and potassium, and both of those act directly on cardiovascular risk — fibre on blood lipids, potassium on blood pressure. That last one is badly underappreciated: virtually the entire US population falls short of the potassium recommendation from food alone.

The effect size is large enough to be worth putting in perspective. The reductions in cardiovascular events associated with generous fruit and vegetable intake in these pooled analyses are in the same range as several things we willingly take a daily tablet for.

Cancer, and how to be honest about it

Here the picture is genuinely weaker than I implied in 2015, and I would rather correct myself than repeat a comfortable claim.

The association with total cancer is real but modest, and considerably smaller than the association with cardiovascular disease.¹ Early case-control studies suggested a dramatic protective effect; larger prospective cohorts, which are less vulnerable to recall bias, have shown a more restrained one. The effect appears strongest for cancers of the digestive tract.

That is still worth having. It is not the “magic pill” framing I used a decade ago, and the framing mattered less than the conclusion, which has not changed: eat them.

Where the picture sharpens is again in the Adventist cohort, and specifically for cancers of the digestive tract. The colorectal cancer finding there — roughly 43% lower risk in pescovegetarians, with vegetarian patterns generally protective⁵ — is considerably larger than anything the general-population analyses show for total cancer. That fits a pattern effect rather than a fruit-and-vegetable-portion effect: what is being compared is not one nutrient but an entire way of eating, including what it displaces.

What about supplements of the same nutrients?

This is the obvious next question, and the answer is unambiguous enough that I gave it an article of its own. Trials of isolated antioxidants — beta-carotene, vitamin E — have not reproduced the benefit of the foods that contain them. Two of them found harm and were stopped early.

Whatever is doing the work in a carrot, it is not one extractable compound. That is the single most instructive finding in modern nutrition, and it argues for the plate rather than the bottle.

What I actually tell patients

• Aim for five: two fruit, three vegetables. Hitting five reliably beats aiming at ten and managing three.
• Make the vegetables non-starchy and make them colourful. Leafy greens, cruciferous vegetables, tomatoes, carrots, peppers.
• Eat the fruit, don't drink it. Juice loses the fibre and keeps the sugar.
• Frozen counts. It is picked ripe and frozen quickly; it is often better than a fresh item that has spent a week in transit, and it removes every excuse about spoilage.
• Think in plates, not portions. The Adventist data suggests the gain comes from shifting the composition of meals toward plants, not from bolting servings onto an unchanged diet. What the vegetables replace matters as much as the vegetables.
• Do not wait to be perfect. The steepest part of the benefit curve is at the bottom — going from one serving to three does more for you than going from six to eight. And in the Adventist gradient, even the semi-vegetarian and pescovegetarian patterns did well; nobody had to be a vegan to benefit.

That last point is the one worth remembering. If you are currently eating almost none, you have more to gain than anyone else reading this.