Nuts and Seeds

TL;DR. Energy-dense, fat-dominated, and consistently associated with lower mortality in both cohort studies and one large randomised trial. The counterintuitive finding is that nuts do not cause the weight gain their calorie count predicts: a substantial share of their fat is never absorbed, because it stays locked inside intact cell walls and passes through. A handful a day (roughly 30 g) is the intake associated with benefit. Brazil nuts are the exception that needs a limit: one or two a day is a good selenium supplement, ten a day causes selenosis.

1. What they are

Botanically a mess. True nuts (hazelnut, chestnut, acorn) are hard-shelled dry fruits. Most "nuts" are something else:

Cashews are never sold in the shell, because the shell contains anacardic acid, a relative of urushiol from poison ivy, which causes severe burns. They are steamed or roasted to remove it, which is why "raw" cashews are not truly raw.

2. What is inside them

Per 30 g (a standard handful, roughly 20 to 25 almonds or 14 walnut halves):

Points worth flagging:

• Chestnuts are not like other nuts. They are starchy, low in fat, and about a third of the calories. They were a staple carbohydrate in parts of Europe before the potato.
• Walnuts, flax, chia, and hemp are the plant omega-3 sources, supplying ALA. Conversion to EPA and especially DHA is poor (Chapter 13), so they are not a substitute for oily fish or algal oil.
• Flaxseed must be ground. Whole seeds pass through intact and deliver almost nothing. Grind fresh, or buy ground and refrigerate, because the oil oxidises fast.
• Brazil nuts are the most selenium-dense food known, by an enormous margin, and this is a problem as much as a feature.

3. The calorie paradox

In short: Nuts deliver measurably fewer calories than their nutrition label says, because a substantial fraction of their fat is never absorbed.

Standard calorie values come from Atwater factors, which assume near-complete absorption. Nuts break that assumption. Their fat is inside intact plant cell walls, and chewing does not rupture all of them, so some fat passes through in the stool.

Measured metabolisable energy, from USDA work by David Baer and colleagues:

The effect is larger for whole nuts than for butters and flours, because grinding ruptures the cells. Whole almonds deliver about a third fewer calories than almond butter made from the same almonds.

Together with high satiety and increased thermic effect, this explains the repeated observation that adding nuts to the diet does not produce the weight gain their calorie content predicts. Cohort studies generally find nut eaters have lower body weight and less weight gain over time, and trials adding nuts without other changes rarely find the expected gain.

4. What the evidence shows

Established: nuts are dense sources of unsaturated fat, vitamin E, magnesium, and, in some, selenium, copper, and manganese. Ground flaxseed is an effective source of ALA and lignans.

Strong, and unusually so for a single food:

• PREDIMED, randomised, found the Mediterranean diet supplemented with 30 g a day of mixed nuts reduced major cardiovascular events by roughly 28 percent versus the control diet (Chapter 34). That is a hard-endpoint randomised result with nuts as a specified intervention.
• Cohort meta-analyses consistently associate roughly 28 g a day of nuts with 20 to 30 percent lower cardiovascular mortality and around 20 percent lower all-cause mortality, with benefit levelling off around that intake.
• LDL cholesterol: pooled trial data find nut consumption lowers LDL by roughly 0.2 to 0.3 mmol/L, dose-dependently.

Mixed: walnuts for cognition; nuts for type 2 diabetes prevention (associations are present, trials are limited).

Thin: any claim specific to one nut beyond its named nutrients, and every "activated" or "sprouted" nut claim. Soaking nuts reduces phytate somewhat and does not "activate enzymes" in any meaningful sense.

5. Allergy, which is the serious part

Peanut and tree nut allergy are among the most common causes of fatal food anaphylaxis. Peanut allergy affects roughly 1 to 2 percent of children in Western countries and, unlike milk and egg allergy, is usually lifelong.

The most important thing in this chapter is that the advice reversed. For years, guidance was to delay introducing peanut. The LEAP trial (Du Toit and colleagues, NEJM 2015) randomised infants at high risk of peanut allergy to early introduction or avoidance, and found early introduction reduced peanut allergy at age five by around 80 percent. LEAP-On showed the protection persisted after a year of avoidance.

Guidelines worldwide changed. Current advice in the UK, US, and Australia is to introduce peanut (as smooth peanut butter thinned with milk or water, or peanut puffs, never whole nuts) from around 4 to 6 months alongside other solids, and to keep it in the diet regularly. Infants with severe eczema or existing egg allergy should be assessed first.

Practical notes on established allergy: cross-reactivity between tree nuts is common but not universal; peanut and tree nut allergy are separate but frequently coexist; "may contain" labelling is voluntary and inconsistently applied; and anyone with a diagnosed nut anaphylaxis should carry two adrenaline auto-injectors and know how to use them (Chapter 98).

6. Other cautions

• Brazil nuts and selenium. One Brazil nut contains roughly 70 to 90 µg of selenium against an RDA of 55 µg and an upper limit of 400 µg. A handful can exceed the upper limit several times over. Chronic excess causes selenosis: hair loss, brittle nails, garlic-smelling breath, gastrointestinal upset, and neurological symptoms. One or two a day is a sensible supplement; a daily handful is not. Content varies enormously with the soil the tree grew in, which makes the dose unpredictable.
• Aflatoxin. Peanuts, pistachios, almonds, and Brazil nuts are monitored for aflatoxin, a potent liver carcinogen produced by Aspergillus moulds in warm humid storage (Chapter 92). Regulated tightly in the EU and US; discard any nut that is shrivelled, discoloured, or tastes musty or bitter.
• Choking. Whole nuts are a leading cause of fatal choking in children under about five. Use nut butters and ground nuts instead.
• Oxalate. Almonds and cashews are moderately high; relevant to recurrent stone formers.
• Phytate reduces iron and zinc absorption; soaking and roasting help modestly.
• Rancidity. Nut oils are unsaturated and oxidise. Rancid nuts taste sharp and bitter, and the oxidation products are undesirable. Smell before eating.
• Nutmeg is not a nut and is genuinely toxic in quantity (Chapter 18).
• Bitter almonds contain amygdalin and are not sold as food in most countries.

7. How to buy, store, and eat them

• A handful a day, roughly 30 g, is the intake associated with benefit in the cohorts and used in PREDIMED.
• Unsalted and unroasted, or dry-roasted. Roasting in oil adds fat and salted nuts add substantial sodium; a 50 g bag of salted nuts can carry 500 mg or more.
• Whole beats butter beats flour for calories absorbed and for satiety.
• Grind flaxseed immediately before use; chia does not need grinding.
• Store cold. Nuts and seeds go rancid at room temperature within a few months. The fridge gives 6 to 12 months, the freezer a year or more. Nut butters and nut flours go faster.
• Toast them. Dry-toasting in a pan or oven transforms the flavour of almost any nut or seed and takes three minutes.
• Add them to things you already eat: yoghurt, porridge, salad, roast vegetables. Nuts as a standalone snack are easy to over-eat; nuts as a component of a meal are not.
• Tahini and other seed butters are useful for calcium (unhulled sesame) and for anyone avoiding nuts.
• Chia and flax gel. Both form a mucilaginous gel with water, which makes them useful egg substitutes in baking and effective bulking agents. Do not eat dry chia and then drink: a documented case involved a tablespoon of dry chia followed by water causing oesophageal obstruction. Hydrate it first.

8. The bottom line

• Nuts are the rare single food with a randomised hard-endpoint trial behind them: 30 g a day of mixed nuts within a Mediterranean pattern reduced cardiovascular events by around 28 percent in PREDIMED.
• They deliver 5 to 30 percent fewer calories than the label says, because fat trapped in intact cell walls is never absorbed. Whole nuts beat nut butter on this.
• Walnuts, flax, chia, and hemp are the plant omega-3 sources, and ALA converts poorly to EPA and DHA. Grind the flax.
• One or two Brazil nuts a day is a selenium supplement; a handful is a toxic dose.
• Introduce peanut early in infancy rather than avoiding it: the LEAP trial reduced peanut allergy by around 80 percent, and whole nuts remain a choking hazard until about five.
• Store them in the fridge or freezer. Rancid nuts are common and easily detected by smell.

Sources and notes

Composition figures are from USDA FoodData Central. Nut metabolisable energy measurements are Baer et al.'s USDA studies published in the American Journal of Clinical Nutrition and Journal of Nutrition for almonds, walnuts, cashews, and pistachios. PREDIMED's nut arm is Estruch et al., New England Journal of Medicine, 2018. Nut intake and mortality follows Aune et al., BMC Medicine, 2016, and Bao et al., NEJM, 2013. LDL effects follow Del Gobbo et al.'s dose-response meta-analysis. Brazil nut selenium content and selenosis follow Rayman's reviews and case reports; the variation with soil selenium is documented in Brazilian analyses. ALA to EPA and DHA conversion efficiency follows Burdge and Calder's work. Aflatoxin monitoring follows EU and Codex limits. Peanut allergy prevention is the LEAP trial, Du Toit et al., NEJM, 2015, and LEAP-On. Anacardic acid in cashew shells follows dermatology and food processing references. The chia oesophageal obstruction case is reported in American Journal of Gastroenterology case literature. Heavy metals in plant protein powders follow Clean Label Project and independent testing reports.

Open questions. Why nuts are so consistently associated with lower mortality across cohorts, given how small the quantities are, is not mechanistically settled. The metabolisable energy discount has been measured for only a few nuts and may not generalise.

👉 Next: oils and cooking fats.