Almonds

We evaluated 113 human studies on Almonds, including 66 randomised controlled trials and 13 meta-analyses or systematic reviews; the controlled studies include 4,433 participants in total. As of 23 October 2026.

This page as PDF · Summary

1. What Is Almonds? Occurrence, Origin and Forms

Almonds are the seeds of the almond tree (Prunus dulcis), a member of the rose family from the Mediterranean region and Western Asia. They have long been eaten as food, raw, roasted, or processed into flour, paste, oil and almond drink. Studies examine how eating almonds relates to metabolism, body weight and wellbeing.

The almond tree grows in temperate to warm climates, mainly around the Mediterranean, in Western Asia, in California and in Australia. As a food, almonds are available all year round as a snack, a baking ingredient and the base of almond drinks.

After harvest the fruits are dried and the green hull and the hard shell are removed, leaving the seed with its brown skin. Studies test whole almonds (raw or roasted), defatted almond powder, almond flour, almond paste, almond drink, almond oil and almond skin.

Forms and Preparations Studied

• Whole almonds, raw or roasted
• Almond flour and defatted almond powder
• Almond paste (almond butter)
• Almond milk or almond drink
• Almond oil
• Almond skin and almond bran

Not covered on this page:

• Bitter almond (Prunus amara), which contains amygdalin and is not eaten as a food
• Peanut, which is botanically a legume
• Other tree nuts such as walnut, hazelnut, cashew and pistachio
• Almond flavouring (benzaldehyde) as an added aroma compound

2. What Does Almonds Do? Results by Area

How to read the table: counted are human studies whose results concern the area. “High” means at least one meta-analysis and two controlled studies showing improvement, or at least three controlled studies of which three quarters show improvement, with at least 300 participants in total. “Moderate” means at least two controlled studies showing improvement, as a majority, or one meta-analysis alongside controlled studies. “Low” means one controlled study showing improvement. Where an area has no controlled study, the strength of evidence is at most “low” (result “reviews only” or “observational data only”). Everything else is “unclear”. A study counts as showing improvement when at least half of its outcomes in this area improved. Reviews of combination products do not count as meta-analyses; results that only describe side effects or shares of users do not count as an effect. The same study is counted only once.

3. What the Studies Show in Detail

Almonds, Heart and Metabolism (Blood Lipids): Human Studies

In several randomized studies, whole almonds lowered total cholesterol 15161828 and LDL cholesterol 16185183 compared with control diets. A systematic review of almond studies confirms the LDL reduction 2. Triglycerides fell more strongly than under nut-free diets in two studies 2728, but not in others 4183. Blood pressure remained without difference in several randomized studies 20415383; in one study the systolic value fell more strongly 27; liver fat remained unchanged 3783. For LDL particles, the picture was mixed: small particles fell more strongly than under control diets in one study 79, and under statins the pattern shifted toward smaller particles 52. Vascular function improved in one study compared with usual snacks 83, but not in coronary heart disease 53. Several of these studies were funded by the Almond Board of California, an association of the almond industry 162961. Reviews on nuts in general, in which almonds were only a part, also found lower cholesterol levels 1310. Another meta-analysis on almonds, Brazil nuts, and flaxseeds specifically identified lower total and LDL cholesterol for almonds in coronary heart disease 8. What remains open is the effect on clinical endpoints, the influence of the preparation form, and the transferability to people without increased risk.

Further studies in this area: 16 18 20 21 23 27 28 29 30 33 35 36 37 38 39 41 43 49 50 51 52 53 56 57 58 59 61 62 64 65 70 76 78 79 88 89 95 102 104 105 106 108 109 112 113 114 115 116 117

Almonds, Weight, Satiety and Appetite: Human Studies

In adults with overweight, prediabetes or metabolic syndrome, waist circumference decreased more than with control diets in several randomized studies 284376. Body or abdominal fat decreased more in further studies 1851. For body weight and BMI, findings without a difference predominated 2025263679. Also in further randomized studies on body weight and body composition, no advantage was shown 374164. Individual studies reported a higher energy intake 2652 or a higher BMI and waist circumference 54. After an almond snack, adults ate less over the rest of the day than after cookies 31, but hunger and satiety did not differ compared with pastries 66. A meta-analysis on almond-based diet in type 2 diabetes found a lower BMI 12; a meta-analysis on nuts under calorie reduction, mostly almonds, found no advantage for weight and body composition 9. The evidence is overall extensive but mixed; predominantly whole almonds were investigated, raw or roasted, some as a replacement for isocaloric snacks, some as a supplement. Where almonds were exchanged for other foods, part of the effect is attributable to the exchange; the study with cookies as a comparison was funded by the Almond Board of California 25. Weight studies on almond flour, almond butter, almond milk and almond oil are still pending.

Further studies in this area: 22 23 24 26 27 28 31 33 35 36 37 41 43 46 51 52 54 59 64 66 70 72 74 76 78 79 86 87 89 101 104 106 107 109 113 114

Almonds, Inflammation and Oxidative Stress: Human Studies

Two meta-analyses found lower C-reactive protein with almonds 45, and one of them also found lower interleukin-6, but not in obesity or pre-existing conditions 4. Interleukin-6 also fell in randomized trials in obesity and type 2 diabetes 4168. Two meta-analyses also found less malondialdehyde, a marker of oxidative stress 67. Other randomized trials found no differences in inflammatory and oxidative markers 37577680. In one study with almonds alone and together with dark chocolate, oxidative stress did not differ from usual diet 49. The findings come predominantly from small, short studies in overweight or metabolically abnormal adults. The evidence base is therefore extensive but inconsistent, and the transferability to other forms such as almond flour, almond butter or almond milk is open.

Further studies in this area: 37 41 44 46 49 53 56 57 58 59 63 68 70 76 80 84 102 104 105 108 109 110

Almonds, Gut and Microbiome: Human Studies

For the gut and microbiome, the evidence is high, but the results are mixed. The diversity of the gut flora increased in two randomized studies compared with crackers or cookies 4042, but not in others 3234. Butyrate in stool was higher after almonds than with control 32, other short-chain fatty acids remained unchanged 3442. In metabolic syndrome, inflammation markers in stool decreased when they were initially elevated 76. For gastrointestinal complaints and intestinal transit, the studies showed no difference 3234; hard stools were less frequent in one study 34. Mostly whole almonds, raw or roasted, as well as almond flour, almond butter and almond skin were studied. It remains open which preparation and amount most favorably influence the gut flora and whether the observed laboratory changes have clinical significance.

Further studies in this area: 67 69 70 76 84 85 92 93 96 100 101 105 109 113 114

Almonds, Blood Glucose and Insulin: Human Studies

Under calorie reduction, fasting glucose and insulin fell equally with and without almonds 79. For insulin resistance, two randomized trials found an improvement compared with control 1516, two others found no difference 2957; HbA1c remained unchanged in several studies 263637. The post-meal blood sugar rise was smaller in randomized trials after almonds than without almonds or after cookies 2331. A randomized trial with whole almonds before meals in prediabetes improved glucose and insulin values as well as glucose fluctuations compared with control 4348. In contrast, a randomized controlled trial with whole almonds in prediabetes reported worse values for insulin sensitivity, postprandial glucose and fasting insulin 54. A meta-analysis on almond-based diet in type 2 diabetes found an advantage only for HbA1c, not for fasting blood sugar, postprandial blood sugar, GLP-1, HOMA-IR or fasting insulin 12. A meta-analysis on nuts under calorie reduction, mostly almonds, found no difference in blood sugar and insulin 9. The addition of almonds to meals of dried fruit and rice lowered their glycemic index in most cases in a small study 71. The strength of evidence is overall low because the studies are mostly small, short and inconsistent in form, amount and comparison; several comparisons exchanged almonds for other foods, so that part of the effect is attributable to the exchange. For almond flour, almond butter, almond milk and almond oil, no controlled findings are available in this area.

Further studies in this area: 28 29 31 36 37 38 41 43 46 54 57 58 59 71 72 76 79 104 106 109 111 114

Almonds, Cognition and Mood: Human Studies

For cognition and mood, the evidence is low; most controlled studies found no difference. Compared with control diet, executive function and processing speed improved in one study in prediabetes 46. In mild to moderate Alzheimer's dementia, memory, orientation and sleep improved more with ground sweet almonds with a little rock sugar than with the previous treatment alone 47. In healthy older adults, memory improved only in the group with the higher almond amount compared with baseline; a confirmed difference from the snack group was lacking 80. Working memory and mood remained unchanged in older adults with overweight 27, as did anxiety and depression with mild complaints 34. After unusual exertion, mood and fatigue with almonds showed no confirmed difference from cereal bars 45. It remains open whether the findings in prediabetes and Alzheimer's dementia can be confirmed in larger and longer studies.

Further studies in this area: 77 80 113

Almonds, Skin and Skin Care: Human Studies

In healthy women, consumption of whole almonds compared with pretzels increased the minimal erythema dose, that is, the UV amount until skin redness; roughness, melanin index, moisture, and sebum remained unchanged 55. In postmenopausal women, wrinkle severity and width decreased more strongly under almonds than under a calorie-matched snack 60. A larger follow-up study found fewer wrinkles and pigmentation compared with baseline 82; skin moisture and water loss did not differ 6082. In young adults with mild to moderate acne, lesions decreased more strongly under whole almonds than under a calorie-matched cereal-legume snack 75. The evidence is weak: these are few, small randomized controlled studies with a total of 67 participants, mostly against a calorie-matched snack, so part of the effect may be due to the exchange. Mostly whole almonds were studied; for almond flour, almond paste, almond milk, almond oil, and almond bran, no controlled findings are available for the skin.

Almonds: Further Outcomes in Studies

In coronary heart disease, uric acid in the blood fell more with almonds than without almonds 20. After unusual muscle exertion, strength was better maintained with almonds than with control snacks 4563, but not in another study 64; muscle pain subsided in two studies 6364. The studies in this area were mostly small and short; for almond milk, almond skin and almond bran, controlled studies are still pending.

Further studies in this area: 75 81 90 91 94 97 98 99 101 103

4. How Is Almonds Dosed? Preparation, Dose and Duration in Studies

The studies tested predominantly whole almonds, raw, dry roasted or soaked overnight, some with and some without the skin. Individual investigations used almond flour, defatted almond powder, almond butter, almond oil as well as mixtures with other nuts or foods. For almond milk, almond skin and almond bran, only few or no controlled findings are available.

5. Is Almonds Safe? Side Effects in Studies

Gut complaints hardly increased in the studies 3284. Body weight mostly remained unchanged 2536, but in prediabetes BMI and waist circumference increased in one study 54. In a European anaphylaxis registry, almonds were a frequent trigger in adults after hazelnut and walnut 90.

6. What Remains Open: Limits of the Evidence

• Most studies are small and short; clinical endpoints such as heart attack or stroke were not tested.
• The preparation form was rarely compared directly, so it remains unclear whether whole almonds, almond flour, almond butter or almond milk act differently.
• Where almonds were exchanged for other foods, part of the effect is attributable to the exchange.
• Several studies were funded by the almond industry; their findings need independent replication.
• The transferability to people without increased cardiovascular or metabolic risk is open.

7. Bottom Line: What Is Established About Almonds

Almonds are the seeds of the almond tree (Prunus dulcis) from the rose family (Rosaceae) and have long been consumed raw, roasted or as flour, butter, oil and almond milk. Human studies show a mixed picture: several controlled studies with whole almonds found more favorable blood lipid levels, a better waist circumference and a more favorable fat distribution as well as lower inflammation and oxidation markers. Other controlled studies saw no differences in cholesterol, blood pressure, liver fat or body weight, and individual ones even reported less favorable values. For blood sugar, skin and cognition, the evidence is low; randomized studies on almond milk, almond skin and almond bran are still pending. The findings apply to the form and preparation studied in each case; where almonds were exchanged for other foods, part of the effect is attributable to the exchange.