Nuts and longevity: the snack you've been rationing is quietly protective

For most of my twenties I treated nuts as a sort of dietary landmine. Calorie-dense, easy to overdo, the kind of thing you’re meant to eat “in moderation” — which, in practice, meant a sad little portion of six almonds and a vague sense of restraint. Turns out that was almost exactly backwards. The people who eat the most nuts, within reason, are the ones the longevity data likes best.

This is one of those areas where the evidence is unusually good and unusually boring. There’s no biohacking angle, no protocol to optimise, no gadget. It’s a handful of nuts a day. But the strength of the association — and the fact that it holds up in a randomised trial, not just observational data — makes it worth taking seriously.

The number that started the fuss

The landmark study here is genuinely large. In 2013, Bao and colleagues published an analysis in the New England Journal of Medicine pooling two of the best-run cohorts in nutrition research: 76,464 women from the Nurses’ Health Study and 42,498 men from the Health Professionals Follow-up Study. That’s roughly 119,000 people, followed across more than three million person-years, with their nut consumption reassessed every two to four years rather than guessed at once and forgotten.

The finding was clean and dose-dependent. Compared with people who never ate nuts, mortality risk stepped down the more often you ate them: 7% lower for less than once a week, 11% lower for once a week, 13% lower for two to four times a week, and 20% lower — a hazard ratio of 0.80 — for people eating nuts seven or more times a week. The trend was statistically robust (P<0.001), and the inverse associations showed up specifically for deaths from cancer, heart disease, and respiratory disease.

Now, a sensible person reads that and immediately thinks: sure, but people who eat nuts every day are probably doing nineteen other healthy things. Fair enough — that’s the correct instinct. The researchers adjusted for the obvious confounders: smoking, physical activity, body weight, alcohol, overall diet quality, the lot. The association survived. That doesn’t prove causation on its own, but it means the “nut-eaters are just healthier people” objection doesn’t fully explain it away.

It holds up across the whole literature

One large study is interesting. A meta-analysis that pools twenty of them is more convincing. In 2016, Aune and colleagues published exactly that in BMC Medicine, combining prospective studies covering hundreds of thousands of participants and modelling the effect per 28 grams of nuts per day — which, not coincidentally, is about one small handful.

Per daily 28-gram serving, they found a 21% lower risk of cardiovascular disease, a 29% lower risk of coronary heart disease, a 15% lower risk of total cancer, and a 22% lower risk of all-cause mortality (relative risk 0.78). There were even steeper associations for respiratory-disease and diabetes mortality, though those rest on fewer studies and I’d hold them more loosely.

Two details from that paper are worth pocketing. First, the benefit largely plateaus above roughly 28 grams a day — so this is emphatically not a “more is better” situation. Eating half a kilo of cashews does not buy you proportionally more life; it buys you a stomach ache and a lot of calories. Second, the authors estimated that if the associations are causal, a nut intake below 20 grams a day would be attributable to something like 4.4 million premature deaths a year across the Americas, Europe, Southeast Asia and the Western Pacific. That’s the kind of number that sounds made up, and it’s a modelled estimate rather than a measurement, but it gives you a sense of the scale people in this field think is on the table.

The bit that actually convinced me: a randomised trial

Observational data can only take you so far, and I’m naturally sceptical of nutrition epidemiology — the field has embarrassed itself often enough. What tips nuts from “interesting correlation” into “probably real” is that there’s an actual randomised controlled trial.

PREDIMED, run in Spain and published in its corrected form in the New England Journal of Medicine in 2018, randomised around 7,400 people at high cardiovascular risk into three groups: a Mediterranean diet supplemented with extra-virgin olive oil, the same diet supplemented with 30 grams a day of mixed nuts (walnuts, almonds and hazelnuts), or a control group advised simply to eat less fat. Nobody was told to lose weight or exercise more — the intervention was essentially the food.

The nut-supplemented group had roughly a 28% lower rate of major cardiovascular events — heart attack, stroke, or cardiovascular death — than the control group over the follow-up. That is a proper experiment, with the nuts handed out as the thing being tested, showing a hard clinical outcome move. It’s rare in nutrition, and it’s why nuts sit on firmer ground than most “superfoods” that live entirely on observational hype. The same trial is the backbone of the case for olive oil and longevity, incidentally — the two interventions ran side by side.

Why nuts actually work

The nice thing about nuts is that the mechanisms aren’t mysterious. They’re a dense little package of things we already know are good for you.

Unsaturated fats. The bulk of the fat in nuts is mono- and polyunsaturated — the kind that lowers LDL cholesterol when it replaces saturated fat and refined carbohydrate in the diet. This is the least controversial nutrition finding of the last fifty years, and nuts are one of the most concentrated whole-food sources of those fats. Walnuts go a step further and deliver a decent hit of alpha-linolenic acid, the plant-form omega-3, which is why they get their own mention in the omega-3 and longevity post.

Fibre. A handful of almonds carries three to four grams of fibre, most of it the kind that feeds your gut bacteria and slows glucose absorption. Fibre is quietly one of the strongest predictors of a long life, and most people are badly short of it — I’ve written about why in the dietary fibre post.

Plant protein, magnesium, vitamin E and polyphenols. Nuts are a real source of plant protein, and they’re one of the better dietary sources of magnesium — a mineral a surprising number of people run low on. Add vitamin E, an antioxidant, and a spread of polyphenols, and you’ve got a food that hits several pathways at once rather than one flashy compound.

Endothelial function. Nuts are rich in the amino acid L-arginine, a precursor to nitric oxide, which helps blood vessels relax and dilate. Trials feeding people walnuts in particular have shown improvements in the flexibility of the artery lining. Better endothelial function means lower blood pressure and less of the vascular stiffening that underpins cardiovascular ageing.

Lower inflammation. Regular nut consumption is associated with lower levels of inflammatory markers like CRP. Given how much chronic low-grade inflammation drives ageing, anything that nudges it down is doing quiet, useful work in the background.

None of these is dramatic on its own. Stacked together, in a food you can eat by the handful, they add up to a genuinely well-rounded intervention.

The calorie paradox nobody warned me about

Here’s the part that took me an embarrassingly long time to unlearn. Nuts are calorie-dense — around 160 to 200 calories per 28-gram handful — and the instinct is that eating them daily must, arithmetically, make you gain weight. The trials say otherwise.

Across controlled studies, adding nuts to people’s diets consistently fails to produce the weight gain their calorie count predicts. There seem to be three reasons. They’re extremely satiating, so they blunt appetite and quietly displace other snacking. A meaningful chunk of their fat is trapped in the fibrous cell walls and passes through undigested — you don’t actually absorb every calorie on the label. And there’s some evidence that nut consumption slightly bumps up resting energy expenditure. Whatever the exact mechanism, the practical upshot is that you can eat a daily handful without it showing up on the scale.

The one honest caveat: this assumes a handful, not a bag. Nuts are one of the easiest foods to eat mindlessly — the “grazing straight from the container while cooking dinner” failure mode is real, and I know it well. Portion them, and the paradox works in your favour. Don’t, and it stops being a paradox and starts being a calorie problem.

What to actually eat, and how much

Stripping out the noise, here’s what the evidence supports:

Aim for about 28 grams a day. A small handful, or roughly a quarter-cup. That’s the dose where the mortality curve flattens out, so there’s little point in forcing down more. Daily-ish is the pattern that shows up in the data — a big Sunday binge and nothing the rest of the week isn’t the same thing.

Mix them. Different nuts bring different strengths — walnuts for omega-3, almonds for vitamin E and fibre, Brazil nuts for selenium (though cap those at one or two a day, they’re extraordinarily selenium-dense), pistachios and cashews for variety. A mixed handful covers more bases than mono-nutting your way through a kilo of one type.

Unsalted, raw or dry-roasted. The nuts themselves are the point. Dry-roasting is fine and arguably tastier; what you want to avoid is the added-salt, added-oil, honey-roasted, chocolate-coated versions where the packaging around the nut cancels out part of the benefit. Read the ingredient list — good nuts have an ingredient list of one word.

Peanuts count. They’re technically a legume, but their profile is close enough and the data holds up for them too. Given they’re a fraction of the price of tree nuts, peanuts (and unsweetened peanut butter) are the sensible budget play. No shame in it.

Make it frictionless. The single biggest predictor of whether you’ll actually do this is whether it’s easy. Pre-portioned bags, a jar on the desk, a spoon of peanut butter on toast — the specific vehicle matters less than removing the daily decision. This slots neatly into the broader longevity diet, which is really just a collection of these unglamorous, repeatable defaults.

Common misconceptions

“Nuts are fattening, so I should limit them.” The most persistent myth, and the exact opposite of what the trials show. Rationing yourself to six almonds out of fear is leaving the benefit on the table.

“I’ll take a supplement instead.” There’s no nut pill, and that’s rather the point. The benefit appears to come from the whole matrix — fats, fibre, protein, minerals and polyphenols eaten together, in a chewy, satiating food. Isolating one component and swallowing it in a capsule has repeatedly failed to reproduce whole-food results, and nuts are no exception.

“More must be better.” It isn’t. The curve plateaus around a handful. Beyond that you’re mostly adding calories, not longevity.

“Roasted nuts are ruined.” Dry-roasting doesn’t meaningfully wreck the fats. It’s the salt, sugar and frying oil in the fancy formats that erode the advantage, not heat itself.

What this looks like in real life

I keep a jar of mixed unsalted nuts on my desk and genuinely eat a handful most days — usually mid-afternoon, when the alternative would be wandering to the Glockenbach bakery for something with considerably worse macros. On the weekend it’s often a spoon of peanut butter on rye before a morning walk along the Isar. It is not a discipline. It’s a jar within arm’s reach, which is most of the battle.

The app is quietly useful here in a way I didn’t expect. Sar doesn’t lecture me about nuts specifically, but logging them nudges my protein and fibre numbers up in a way I can see, and watching those two lines trend in the right direction is a much better motivator than any amount of willpower. When my afternoon snack is a handful of walnuts instead of whatever’s nearest, the downstream numbers just look better. Bit nerdy, but that feedback loop is what made it stick.

The honest summary: this is one of the highest-evidence, lowest-effort things in the whole longevity toolkit. Roughly 20% lower mortality in the cohort data is a big number for a change this small — and unlike a lot of what I write about, there’s a randomised trial backing up the cardiovascular side rather than pure correlation. A handful a day. Unsalted. That’s genuinely the whole protocol — which, for once, is quite good actually.