Nuts: a strong habit signal, and short trials only in people with heart disease

The most familiar sentence about nuts, that 30 g a day helps keep your heart healthy, was put to the European regulator and refused. The United States took the other road and lets a nut claim stand, with the hedge written into the wording, so the label has to say the research is supportive and not conclusive. In the EU exactly one health claim naming a nut is authorised, that walnuts improve the elasticity of blood vessels. Underneath that split between regulators sits a split in the evidence. People who eat the most nuts do get less heart disease. Pooling 63 prospective studies, the highest intake against the lowest carried a relative risk of 0.81 for coronary heart disease and 0.77 for death from any cause. Those studies watched what people already ate, for years, and people who eat nuts differ from people who do not in many other ways. Where nuts were actually handed out, in five trials of 436 adults who already had cardiovascular disease taking 5 to 85 g a day for 6 to 12 weeks, systolic blood pressure fell by 1.16 mmHg at the point estimate, with a confidence interval running from -5.68 to 3.35. That range includes no change at all, and the certainty the authors put on that systolic figure was moderate. Those 436 people were already being treated for heart disease, and the longest of those trials ran 12 weeks. No trial of nuts in healthy adults sits on this page, so read that flat result as what it is, a short experiment in people who are already ill.

The numbers, per 100 g

What your body absorbs

• Across 21 human and in vitro studies, the energy the body actually gets from nuts was consistently below what Atwater factors on the label predict, because part of the fat stays locked inside cell walls. — 21 studies of almonds, cashews, hazelnuts, pistachios, walnuts and peanuts, adults aged 18 and over plus in vitro digestion models, searched to June 2021 (n = 21 studies) Evidence C source
  Metabolizable energy lower than the Atwater prediction for every nut type investigated
• A review of 23 bomb calorimetry studies in adults found that tree nut intake lowered the share of eaten energy that is digested and metabolised, which is a second line of evidence for the same effect. — 23 human studies in adults aged 18 and over, published between January 1973 and July 2024, measured by bomb calorimetry of faeces and urine (n = 23 studies) Evidence C source
  Lower proportions of digestible and metabolizable energy intake with tree nuts and with high fibre diets

What cooking and storage change

• How the nut is prepared changes the energy it yields, with flour above chopped above whole nuts and nut butter above roasted above raw, and with 22.6 kJ/g for pistachios against 18.5 kJ/g for raw almonds. — Same 21 studies of tree nuts and peanuts in adults and in vitro digestion models (n = 21 studies) Evidence C source
  Metabolizable energy 22.6 kJ/g for pistachios and 18.5 kJ/g for raw almonds, rising with grinding and with roasting

What it does to your health

• In patients who already have atherosclerotic cardiovascular disease, pooled trials of 5 to 85 g of nuts a day for 6 to 12 weeks moved systolic blood pressure by 1.16 mmHg, an amount whose confidence interval crosses zero. — 436 adults in 5 randomised trials of secondary cardiovascular prevention, almonds, pecans, Brazil nuts and mixed nuts, median 30 g a day for 6 to 12 weeks (n = 436) Evidence A source
  Systolic mean difference -1.16 mmHg (95% CI -5.68 to 3.35), diastolic 0.10 mmHg (95% CI -2.30 to 2.51)
• Across 53 trial records in 1690 adults, the blood lipid advantage of polyunsaturated fat was larger when it came from oil than when it came from nuts or from whole diets, at P = 0.03 for LDL cholesterol and P = 0.03 for triglycerides. — 1690 adults in randomised trials comparing diets with at least 10% of energy from polyunsaturated fat against diets with at least 15% from monounsaturated fat (n = 1690) Evidence A source
  Subgroup difference by source, oil stronger than nut or diet interventions, P = 0.03 for LDL cholesterol and for triglycerides
• Pooling 63 prospective cohort studies, the highest nut intake against the lowest carried a relative risk of 0.81 for coronary heart disease and 0.77 for death from any cause. — 63 studies in 33 publications, adults followed prospectively, searched to 24 October 2023 (n = 63 studies) Evidence C source
  RR 0.81 (0.75-0.87) for CHD, 0.86 (0.82-0.91) for CVD, 0.74 (0.70-0.78) for CVD mortality, 0.77 (0.73-0.81) for all-cause mortality
• In three cohorts totalling 17,906 adults with 891 dementia cases over 166,762 person-years, the main contributors to the lower dementia risk of a healthful plant-based diet were higher intake of vegetables, nuts, tea or coffee, and legumes. — 6642 adults in the Health and Retirement Study, 3045 in the Framingham Offspring cohort and 8219 in Whitehall II, aged 45 and over and free of dementia at baseline (n = 17906) Evidence C source
  Pooled HR 0.70 (95% CI 0.53 to 0.92) for the highest against the lowest tertile of the plant-based diet index, an effect of the whole index and not an estimate for nuts on their own
• Adding 42 to 84 g of nuts a day to an energy-restricted diet produced significantly greater weight loss in only 4 of 7 trials, and no trial found the nuts made any outcome worse. That amount is 1.5 to 3 of the 30 g portions the regulators wrote about. — Seven randomised trials reported in eight papers, adults on energy-restricted diets of 4 to 52 weeks with 42 to 84 g a day of almonds, peanuts, pistachios or walnuts (n = 7 trials) Evidence C source
  Greater weight loss with nuts in 4 of 7 trials, limited evidence for body composition or glycaemic control

Safety, limits and interactions

• A 78-year-old man with stable chronic kidney disease reached a serum creatinine of 6.0 mg/dL from a baseline of 2.05, with oxalate crystals on biopsy, after eating 3 to 4 oz of cashew nuts daily for a year. That dose is 85 to 113 g a day, three to four ordinary 30 g portions, every day for twelve months. — One 78-year-old man with pre-existing chronic kidney disease (n = 1) Evidence D source
  Serum creatinine 6.0 mg/dL against a baseline of 2.05 mg/dL, biopsy showing acute tubular injury with extensive calcium oxalate deposition
• In food matrices in the laboratory and not in people, across 41 studies of aflatoxin removal from tree nuts, cold plasma and UV-C cut contamination by 70% to 95%, while costing quality, with up to 29% of the alpha-tocopherol lost in hazelnuts treated chemically. — 41 original studies published 2005 to 2025 on almonds, pistachios, Brazil nuts, hazelnuts and walnuts, food matrices rather than people (n = 41 studies) Evidence D source
  70% to 95% aflatoxin reduction by physical methods, close to 100% by ozonation and organic acids, up to 29% loss of alpha-tocopherol in hazelnuts

What people believe that the data does not support

• The EU allows exactly one health claim naming a nut, which is that walnuts improve the elasticity of blood vessels. — EU register of nutrition and health claims, snapshot of 2026-09-21 Evidence E source
  Status Authorised
• The most familiar nut message of all, that 30 g of nuts a day helps maintain heart health, was put to the EU regulator and refused. — EU register of nutrition and health claims, claim covering peanuts and tree nuts other than brazil, macadamia and cashew Evidence E source
  Status Non-authorised
• The United States took the opposite road to the EU and lets a nut claim stand, but only with the hedge written into the wording, which is that the research is supportive and not conclusive. — US food labels, FDA letter of enforcement discretion on macadamia nuts and coronary heart disease, 2017 Evidence E source
  Qualified claim permitted at 1.5 ounces a day, with the hedging sentence mandatory

The bottom line

Start with what the word covers. Our table holds 89 USDA records under the heading nuts, and they include almond paste, sweetened canned coconut cream and roasted chestnuts. Those behave nothing like a handful of almonds. Read the specific record rather than the category. On calories, the label overstates what you get. Across 21 human and in vitro studies of almonds, cashews, hazelnuts, pistachios, walnuts and peanuts, the energy the body actually takes from nuts came in below what the Atwater factors on the label predict, because part of the fat stays locked inside the cell walls. A review of 23 bomb calorimetry studies in adults found the same thing by a different method. The form matters: flour yields more than chopped nuts, chopped more than whole, nut butter and roasted more than raw, and the measured values run from 18.5 kJ/g for raw almonds to 22.6 kJ/g for pistachios. Whole and raw is the form that gives back the least. That feeds into weight. Adding 42 to 84 g of nuts a day to an energy-restricted diet produced significantly greater weight loss in 4 of 7 trials, and in none of the seven did the nuts make any outcome worse. Notice the amount. That is 1.5 to 3 of the 30 g portions the regulators wrote about, and it went into a diet that was already cutting calories. Nobody tested adding that much on top of what you eat now. Two findings here are quieter than the headlines built on them. When randomised trials in 1690 adults compared sources of polyunsaturated fat, the blood lipid advantage was larger when the fat came from oil than from nuts or from whole diets, at P = 0.03 for LDL cholesterol and P = 0.03 for triglycerides. And the dementia result that gets quoted as a nut result is a diet result. In three cohorts totalling 17,906 adults, the hazard ratio of 0.70 belongs to a plant-based diet index as a whole. Nuts are one of four foods the authors named as the main contributors, next to vegetables, tea or coffee and legumes. What nuts on their own did, nobody measured. Who should be careful, and the first line here is an absence. We hold no card on nut allergy. For a food whose everyday safety question is allergy, that is a real gap, and we would rather name it than let silence read as reassurance. Then the kidneys. A 78-year-old man with existing chronic kidney disease ate 3 to 4 oz of cashews daily for a year, which is 85 to 113 g, three to four ordinary 30 g handfuls every single day, and his serum creatinine went from 2.05 to 6.0 mg/dL, with calcium oxalate deposits on biopsy. That is one man, and a single case cannot tell you how often this happens or at what amount. If you have kidney disease, the size of a daily nut habit is a question for the person treating you. The other safety card we hold is about the industry, not about you. Forty-one studies tested ways to strip aflatoxin out of tree nuts, and cold plasma and UV-C cut it by 70 to 95 percent while damaging the nut, with up to 29 percent of the alpha-tocopherol lost in chemically treated hazelnuts. What those 41 studies measured was the food matrix in a laboratory, almonds, pistachios, Brazil nuts, hazelnuts and walnuts, and how well an industrial method strips a toxin out of it. Nobody in them measured a person, and none of them says how much aflatoxin is in the nuts on your shelf. Now the rest of the holes, and they are large enough that you should know about them before you trust this page. We hold nothing on children, on pregnancy, on older adults or on anyone else for whom the answer might differ. We hold nothing on how much the same nut varies between varieties, growing seasons and storage, so the single number in the table is steadier on the page than it is in a jar. And we hold nothing that turns grams into a portion you would recognise on a plate. Until those are filled, treat this page as the heart and calorie story and take the rest to a clinician.

Sources

1. nuts-r4-04 · systematic review
   The ME of nuts was consistently lower than that predicted by Atwater factors for investigated nut types (almonds, cashews, hazelnuts, pistachios, walnuts, and peanuts).
   Adv Nutr, 2023 · checked 2026-09-23
2. nuts-r4-06 · systematic review
   High-fiber diets and tree nut intake consistently lowered the proportions of DEI and MEI, indicating that dietary composition affects energy absorption efficiency.
   Adv Nutr, 2026 · checked 2026-09-23
3. nuts-r4-05 · systematic review
   ME was influenced by nut type (ME = 22.6 kJ/g for pistachios; ME = 18.5 kJ/g for raw almonds), physical form (flour > chopped > whole nuts), heat processing (butter > roasted > raw) and dose of consumption.
   Adv Nutr, 2023 · checked 2026-09-23
4. nuts-r4-01 · meta-analysis of RCTs
   They also had no effect on systolic (mean difference [MD]: -1.16 mmHg [95% CI, -5.68 to 3.35], I2 = 0%-moderate certainty of evidence) or diastolic (MD: 0.10 mmHg [95% CI, -2.30 to 2.51], I2 = 0%-high certainty of evidence) blood pressure.
   Nutr Rev, 2025 · checked 2026-09-23
5. nuts-r4-02 · meta-analysis of RCTs
   Subgroup analyses showed that the effects of LDL cholesterol (P = 0.03), and TG (P = 0.03) were stronger with oil compared with nut or diet interventions.
   Am J Clin Nutr, 2026 · checked 2026-09-23
6. nuts-r4-03 · meta-analysis of cohorts
   Comparisons between individuals with the highest versus lowest intake of nuts resulted in RRs (95% CIs) for coronary heart disease (CHD), stroke, CVD, CVD mortality and all-cause mortality of 0.81 (0.75-0.87), 0.91 (0.84-0.99), 0.86 (0.82-0.91), 0.74 (0.70-0.78) and 0.77 (0.73-0.81), respectively.
   Arch Cardiovasc Dis, 2026 · checked 2026-09-23
7. nuts-r4-09 · prospective cohorts
   Main contributors to lower risk were higher intake of vegetables, nuts, tea or coffee, and legumes.
   J Prev Alzheimers Dis, 2026 · checked 2026-09-23
8. nuts-r4-10 · scoping review
   While all studies reported that energy restriction resulted in significant weight loss, the addition of nuts to ER diets demonstrated significantly greater weight loss in only approximately half of the included studies (4/7 studies).
   Nutr Res Rev, 2025 · checked 2026-09-23
9. nuts-r4-07 · case report
   On revisiting his history, it was discovered that the patient had been consuming 3-4 oz of cashew nuts daily for the last 1 year.
   Kidney Med, 2026 · checked 2026-09-23
10. nuts-r4-08 · scoping review
    Physical approaches, particularly cold atmospheric plasma and UV-C radiation, achieved robust reduction rates ranging from 70% to 95%; however, these highly energetic treatments often induced lipid oxidation, evidenced by linear increases in malondialdehyde (MDA) levels in matrices such as pistachios.
    Mycotoxin Res, 2026 · checked 2026-09-23
11. nuts-r4-11 · claims register
    POL-HC-6525 Walnuts Walnuts contribute to the improvement of the elasticity of blood vessels Authorised
    EU Register on nutrition and health claims, claim POL-HC-6525, snapshot 2026-09-21 · checked 2026-09-23
12. nuts-r4-12 · claims register
    POL-HC-7666 Nuts - peanuts and tree nuts (almonds, hazelnuts, pecans, pistachios and walnuts), excluding brazil, macadamia and cashew nuts Eating 30g of nuts per day as part of a balanced diet helps maintain heart health. Non-authorised
    EU Register on nutrition and health claims, claim POL-HC-7666, snapshot 2026-09-21 · checked 2026-09-23
13. nuts-r4-13 · qualified health claim
    Supportive but not conclusive research shows that eating 1.5 ounces per day of macadamia nuts, as part of a diet low in saturated fat and cholesterol and not resulting in increased intake of saturated fat or calories may reduce the risk of coronary heart disease.
    FDA, letter of enforcement discretion, macadamia nuts and coronary heart disease · checked 2026-09-23