This catalog lists the human trials and pooled analyses in our source registry, with the dose, duration, result and limits of each. It is not a count of every study ever run. A dated search of PubMed, Europe PMC and trial registries finds more small human studies than the registry holds, and the rows marked ‘listed, not graded’ say why we do not count them. Most studies are small and short.
This catalog is the spine of the research desk. Each graded claim elsewhere on the site rests on rows in the table below or on the sources listed on its own page.
The master table
| Trial | Year | n | Dose | Duration | What it found | Limitations |
|---|---|---|---|---|---|---|
| Geng (meta-analysis) | 2022 | 15 RCTs pooled | varies | Not in the abstract | TG, total cholesterol, LDL down; HDL up; subgroup analysis found the reduction only in abnormal lipid metabolism. Null: glucose, BP, BMI | High heterogeneity, the authors' own limitation; lab numbers, not events |
| Guo (meta-analysis) | 2017 | 11 RCTs pooled | varies | Not in the abstract | Same lipid conclusions, separate research group | High heterogeneity for total and LDL cholesterol (I-squared above 75%); lab numbers, not events |
| Moroşan (meta-analysis) | 2026 | 12 RCTs pooled, 901 analyzed | varies: whole berries, purees, powders, extracts and oils | From 4 weeks to 12 weeks | No significant pooled effect on total cholesterol, LDL-C, HDL-C, triglycerides, systolic blood pressure or BMI. Fasting blood glucose and diastolic blood pressure fell significantly, by amounts the abstract calls clinically negligible. Every outcome rated very low certainty (GRADE) | Listed, not graded. We could read only the abstract. Trial sequential analysis did not reach the required information size, and the abstract cites risk of bias, heterogeneity and imprecision |
| Larmo (dry eye) | 2010 | 100 | 2 g/day oil blend, split with meals | 3 months | Smaller tear-osmolarity rise (significant only with covariate adjustment). Worst redness 6% vs 36% (all participants, P=0.04). Worst burning 12% vs 32% (compliers only, P=0.04) | Single trial. Other diary symptoms, TBUT, Schirmer and mOSDI null. Part-funded by Aromtech, the oil's maker, which employed one author during the trial |
| Jarvinen (mechanism) | 2011 | dry-eye cohort | 2 g/day | 3 months | Tear-film fatty acids did NOT differ between groups | Mechanism unanswered |
| Larmo (dry eye, eyelid spray, TOPICAL) | 2019 | 2, 10 and 40 (three parts) | Oil and sodium hyaluronate eyelid spray (topical, not oral) | 1 day, 9 days and 1.5 months, by part | In the 40-person part, the Ocular Surface Disease Index fell more in the eye given the spray than in the untreated eye (P=0.0007) | Listed, not graded. We could read only the abstract. The spray also contains sodium hyaluronate, so it cannot be credited to the oil alone. Same research group as the 2010 trial, so not an independent replication. Funded by Tekes |
| Larmo (vaginal, ORAL) | 2014 | 116 | 3 g/day oral oil | 3 months | About 3x the odds of improved epithelial integrity vs placebo (OR 3.1, 95% CI 1.11-8.95); hormones not measured | Single trial. Compliant-only analysis was non-significant (OR 2.9, 95% CI 0.99-8.35). Lead author at Aromtech Ltd, an oil maker |
| Yu (vaginal, TOPICAL) | 2025 | 368 of 377 | Compound SB vaginal suppository (topical, not oral) | 2 weeks | Both active arms beat no pretreatment on colposcopy readiness (P below 0.001), with no significant difference between suppository and estrogen ointment. No systemic side effects vs estrogen's 5.6% breast discomfort and 4% dizziness or headache | Topical route, different product from Larmo. Compound formula, open-label, no pre-set non-inferiority margin, local vaginal discomfort in both arms, colposcopy-prep setting rather than a dryness-symptom trial |
| De Seta (vaginal gel, COMPOUND product) | 2021 | 60 | Branded gel (Meclon Idra) of sea buckthorn oil, aloe vera, glycyrrhetic acid, hyaluronic acid and glycogen, topical | 14 days, then twice a week to day 90 (12 weeks) | Vaginal Health Index and sexual function scores better than placebo over 12 weeks (Vaginal Health Index significant at day 90), with less vaginal pain and dyspareunia and a lower vaginal pH | Listed, not graded. Five active ingredients, so the result cannot be credited to sea buckthorn alone. We could read only the abstract |
| Yang (skin) | 2009 | 60 (30 oral) | 2 g/day SBA24 oil | 3 months | Hydration +48.6%, elasticity +25.8% on instruments | No placebo arm for oral; brand-linked oil |
| Chan (skin) | 2024 | 40 | 500 mg/day Omegia oil (one softgel) | 12 weeks | Texture -14% vs placebo; hydration +8% vs placebo (significance added by a June 2024 corrigendum); collagen density +10% and pores -15% from baseline only, not vs placebo | Funded by the oil's maker Puredia, with company authors; small n |
| Yang (atopic dermatitis) | 1999 | 49 | 5 g (10 capsules) a day of seed oil, pulp oil or paraffin oil (the placebo) | 4 months | Dermatitis improved significantly in the pulp-oil group (P below 0.01) and the paraffin-oil placebo group (P below 0.001), and not significantly in the seed-oil group (P=0.11). HDL cholesterol rose in the pulp-oil group | Listed, not graded. The abstract reports no comparison between groups, and the placebo group improved too. We could read only the abstract and nobody has repeated it |
| Yang (skin lipids, second report of the 1999 trial) | 2000 | 16 | 5 g a day of seed oil, pulp oil or paraffin oil | 4 months | Skin glycerophospholipid fatty acids stayed largely stable. Seed oil slightly raised docosapentaenoic acid and lowered palmitic acid, without reaching significance | Listed, not graded. A second report from the 1999 trial, counted with it. Fatty-acid measures rather than symptoms. We could read only the abstract |
| Akhtar (sebum, TOPICAL extract cream) | 2010 | 10 healthy volunteers | 1% concentrated fruit extract in a water-in-oil cream, against its base | 8 weeks | Skin sebum secretion lower with the extract cream than with the base (P below 0.05) | Listed, not graded. A fruit extract cream, not the oil, in healthy volunteers. We could read only the abstract and nobody independent has repeated it |
| Khan (skin elasticity, TOPICAL extract emulsion) | 2013 | 11 healthy men | 1% hydro-alcoholic fruit extract in a water-in-oil emulsion, against its base | 7 weeks | Most skin biomechanical parameters on the cutometer were better than with the base (P below 0.05). Extensibility and firmness did not differ | Listed, not graded. A fruit extract emulsion, not the oil. We could read only the abstract and nobody independent has repeated it |
| Khan (skin barrier, TOPICAL extract emulsion) | 2014 | 13 healthy men | 5% fruit extract in an oil-in-water emulsion, against its base | 84 days | Skin hydration (p=0.0003) and transepidermal water loss (p=0.0087) better than with the base | Listed, not graded. A fruit extract emulsion, not the oil, in healthy volunteers. We could read only the abstract and nobody independent has repeated it |
| Khan (melasma, TOPICAL, COMPOUND extract emulsion) | 2013 | 50 patients, two groups of 25 | Emulsions with plant extracts that include sea buckthorn and Cassia fistula, against a placebo emulsion on the other cheek | 12 weeks | Skin melanin, measured with a Mexameter, fell in the patients who used the plant-extract emulsions (p up to 0.05), and the abstract calls the formulations effective on melasma against the placebo | Listed, not graded. Two plant extracts together, and the abstract credits extracts containing catechin without separating the sea buckthorn one, so the result cannot be credited to sea buckthorn alone. We could read only the abstract and nobody independent has repeated it |
| Khan (acne, TOPICAL, COMPOUND extract emulsion) | 2014 | 50 patients, two groups of 25 | Emulsions with 5% plant extracts (sea buckthorn and Cassia fistula), against a placebo emulsion on the other cheek, twice a day | 12 weeks | Skin sebum content fell on the treated side in both extract groups (p up to 0.05), and the abstract calls the formulations effective against the placebo on sebum measurement and on clinical evaluation | Listed, not graded. A single-blind split-face study of two plant extracts together, so the result cannot be credited to sea buckthorn alone. The measure is sebum, not a count of acne lesions. We could read only the abstract and nobody independent has repeated it |
| Kipić (skin hydration, TOPICAL oil emulsion) | 2014 | Not in the abstract | Emulsion with about 40% sea buckthorn oil, topical | Not in the abstract | Skin hydration rose significantly and skin pH did not change significantly in healthy volunteers. A rabbit skin test showed no irritation | Listed, not graded. The abstract gives no number of volunteers and describes no control group. We could read only the abstract and nobody independent has repeated it |
| Boca (psoriasis, TOPICAL extract) | 2019 | 10 | Sea buckthorn extract suspended in corn oil, against pure corn oil on the lesion on the other side of the body, applied twice a day | 8 weeks | Psoriasis Area and Severity Index and Dermatology Life Quality Index scores improved from baseline at 4 and 8 weeks on the treated lesions. The placebo lesions did not change at 4 weeks and were worse by the end | Listed, not graded. Ten patients, single-blind so the observer could be biased, as its authors note, with a short follow-up. An extract in corn oil, not the oil on its own. Nobody independent has repeated it |
| Larmo (infections) | 2008 | about 250 | 28 g/day berry puree | 3 months | No effect on colds; CRP (inflammation marker) decreased | Food dose, healthy adults |
| Morone (constipation) | 2026 | 90 | 500 mg/day fruit extract | 28 days | Spontaneous complete weekly bowel movements 1.5 to 2.6; Bristol stool 1.4 to 3.5; both p below 0.001; no change on placebo. Null: flatulence, SF-12 quality of life | Industry-sponsored (EPO S.r.l.); single 28-day trial; fruit extract, not the oil |
| Mortensen (blood sugar) | 2018 | 18 | Single sucrose meal with sea buckthorn puree | Single meal | Insulin 39.6% lower at 30 min; maximal insulin rise 23.6% lower; glycaemic profile index 44.7% better; postprandial glucose unchanged | Single-blinded; single-meal crossover in overweight men, not a treatment trial; closed access |
| Lehtonen (blood sugar, single meal) | 2010 | 10 healthy men | Dried crushed berries, or extraction residues, in a breakfast with 50 g of glucose | Single meal, on four test days | Whole berries lowered the 30-minute insulin peak by 21.8 mU/l (P=0.039). Oil-free berries also gave a more stable insulin response than the control meal | Listed, not graded. A single-meal crossover in normal-weight men, not a treatment trial. We could read only the abstract and nobody independent has repeated it |
| Ren (blood sugar, impaired glucose regulation) | 2021 | 38 completed | 90 mL a day of sea buckthorn fruit puree, or placebo | 5 weeks per period, 4-week washout, crossover | Fasting plasma glucose fell by a median 0.14 mmol/L on the puree and rose by a median 0.07 on placebo (p=0.045). The 2-hour post-meal glucose area stayed stable | Listed, not graded. Small, from one research group, and a puree rather than an oil. We could read only the abstract and nobody independent has repeated it |
| Larmo (low-dose berries, same cohort as the 2008 infection trial) | 2009 | 229 | 28 g/day berry puree | 3 months | No change in total, HDL or LDL cholesterol or triglycerides; plasma quercetin and isorhamnetin rose; no correlation between flavonol and CRP changes | Second report of the 2008 trial; healthy normolipidemic adults on healthy diets |
| Larmo (metabolomics) | 2013 | 80 | 20 g dried berries, 4 g oil (8 capsules), 14.6 g phenolic extract with maltodextrin, or 100 g frozen bilberries, one diet per period | 33 to 35 days per diet, crossover | Serum metabolite effects differed by baseline metabolic profile | Crossover, overweight women; no placebo diet, bilberries were a fourth diet, not a control; most single-metabolite changes not significant after correction for multiple testing; part-funded by Aromtech, which supplied the oil |
| Lehtonen (berry diets, overweight and obese women) | 2011 | 110 recruited, 80 completed | Four diets in random order: bilberries, sea buckthorn berries, a sea buckthorn phenolic extract or sea buckthorn oil. Amounts are not in the abstract | 33 to 35 days per diet, 30 to 39 days of washout, crossover | Waist circumference fell after the bilberry period (-1.2 cm, P=0.041) and after the sea buckthorn period (-1.1 cm, P=0.008). Vascular cell adhesion molecule fell after bilberries and after the oil, and intercellular adhesion molecule after the extract | Listed, not graded. The same four diets and 80 completers as the 2013 metabolomics row, so it appears to be a second report of one crossover and counts with it. No placebo diet. We could read only the abstract, which calls the effects slightly positive |
| Huang (dose escalation) | 2020 | 13 | Sea buckthorn oils high in cis-palmitoleic acid (380, 760 and 1520 mg a day) or augmented in the trans isomer (120, 240 and 480 mg a day). Grams of oil not given in the abstract | 3 weeks per dose, 4-week washout, crossover | Serum trans-palmitoleic acid rose 26.6% at the top dose of the augmented oil. No significant effects on glucose, insulin, lipids or other clinical measures | Listed, not graded. A dosing study, not powered to detect clinical effects. The abstract reports no carryover or adverse effects. We could read only the abstract and nobody independent has repeated it |
| Gao (liver, cirrhosis) | 2003 | 50 randomized (groups of 30 and 18 analysed in the abstract) | Sea buckthorn extract 15 g three times a day, against vitamin B complex | 6 months | Serum fibrosis markers and total bile acid fell, and liver enzymes normalized sooner. The abstract's comparison with the control group is unclear | Listed, not graded. A brief report, never replicated. People with cirrhosis, vitamin B controls rather than placebo, unequal groups. We could read only the abstract and do not count it in the evidence we grade |
| Singh (nephrotic syndrome) | 2013 | 56 (28 per group) | Hydroalcoholic extract (Badriphal) 350 mg twice a day, added to standard treatment | 3 months | Urine protein fell, and serum albumin and inflammatory markers improved, in the extract group | Listed, not graded. A pilot in an Indian journal, an extract rather than the oil, controls had standard treatment only (no placebo). We could read only the abstract, it was never replicated, and we do not count it in the evidence we grade |
| Vashishtha (hypertension) | 2017 | hypertensive adults | 0.75 ml/day seed oil | 30 days | Reported blood pressure and lipid improvement | RETRACTED 2026. Not usable as evidence |
| Johansson (platelets) | 2000 | 12 | 5 g/day berry oil | 4 weeks | Clear drop in platelet aggregation | 12 healthy men. The highest daily oil dose in the trials we cite, matched by Yang 1999 (5 g a day for 4 months) |
| Eccleston (juice) | 2002 | 20 healthy men | Sea buckthorn juice, or placebo | 8 weeks | No significant change in total cholesterol, LDL-C, platelet aggregation or ICAM-1. HDL-C and triglycerides rose without reaching significance. LDL oxidation susceptibility was moderately lower | Listed, not graded. A juice in healthy men, not an oil. We could read only the abstract and nobody independent has repeated it |
| Suomela (flavonols in porridge) | 2006 | Not in the abstract | Sea buckthorn flavonol aglycones in oatmeal porridge at two doses, with or without a small amount of oil | Not in the abstract | No significant effect on oxidized LDL, C-reactive protein, homocysteine, plasma antioxidant potential or paraoxonase activity. The flavonols were rapidly absorbed | Listed, not graded. Isolated flavonols, not the oil or the berry. We could read only the abstract and nobody independent has repeated it |
| Zhou (hypercholesterolemia, puree) | 2022 | 111 | 90 ml of sea buckthorn puree a day, or placebo | 90 days | No effect on total cholesterol, LDL-C or triglycerides. HDL-C fell over the first 6 weeks and rose over the last 6. Diastolic blood pressure was lower than on placebo with p above 0.05, and hsCRP was moderately lower | Listed, not graded. A null result for lipids, as its authors state. A puree rather than an oil. We could read only the abstract and nobody independent has repeated it |
| Chen (hypercholesterolemia, puree, metabolomics and gut bacteria) | 2022 | 56 | 90 g of sea buckthorn berry puree a day | 90 days, with samples at 0, 45 and 90 days | The plasma metabolic profile changed by day 45 and moved back toward baseline by day 90. Glucose, lactate and creatine were lower at day 90 than at baseline. Butyrate-producing bacteria and some other gut microbes rose and Parasutterella fell | Listed, not graded. One group measured before and after, with no control group, and a puree rather than the oil. The abstract does not say whether these people also took part in the 2022 randomized puree trial above. We could read only the abstract |
| Kopčeková (hypercholesterolemic women, juice) | 2023 | 19 women | 50 mL of 100% sea buckthorn juice a day | 8 weeks | LDL-C, CRP, body weight, body fat and visceral fat fell and HDL-C rose. Triglycerides did not change significantly | Listed, not graded. A single group measured before and after with no control group, so the changes cannot be credited to the juice. We could read only the abstract |
| Yang (juice, fatigue in women) | 2025 | 60 | Sea buckthorn juice or a control juice. The amount is not in the abstract | 8 weeks | Mood scale (POMS2) total mood disturbance was lower and vitality higher with the juice. Ferritin, hemoglobin and erythropoietin were higher than on the control at 4 and 8 weeks, and fatigue after a task load, rated on a visual scale, was lower | Listed, not graded. A juice, not an oil, in premenopausal women with temporary fatigue. We could read only the abstract and nobody independent has repeated it |
| Erglis (coronary syndrome, seed lipid extract) | 2025 | 86 enrolled | Seed lipid extract capsules, 1000 mg twice a day, added to standard therapy | 3 months | Systolic blood pressure fell 2.9 mmHg (p=0.012), LDL-C fell 0.3 mmol/l (p=0.005) and CRP fell 1.0 mg/l (p=0.032) | Listed, not graded. A single-arm pilot with no comparator, in people already taking a statin or ezetimibe. A seed extract, not the berry oil. We could read only the abstract |
| Pop (rheumatoid arthritis) | 2024 | 80 | 1,800 mg a day of sea buckthorn fruit oil added to methotrexate 15 mg a week and diclofenac 50 mg a day, against the two drugs alone | 12 weeks | Disease activity scores (DAS28, SDAI, CDAI) fell more than in the control group (P below 0.001), and more patients had a good or moderate EULAR response | Listed, not graded. The abstract says patients were divided at random 1:1 and names no placebo, so the control group took nothing extra. Small, from one research group, and we could read only the abstract |
| Zugravu (rheumatoid arthritis, probably a second report of the Pop trial) | 2024 | Not in the abstract | 1,800 mg a day of sea buckthorn fruit oil added to methotrexate and diclofenac, against a control group on the two drugs | 12 weeks | Serum malondialdehyde fell more than in the control group. Inflammatory markers did not differ significantly. Pain, physical disability, general condition, disease activity and fatigue were better than in the control group | Listed, not graded. The same dose, drugs and length as the Pop row, so it is probably a second report of one trial and counts with it. The abstract does not say how patients were assigned. We could read only the abstract |
| Abdullahzadeh (burns, TOPICAL) | 2021 | 55 | Sea buckthorn cream against 1% silver sulfadiazine dressings, topical | Until complete healing | Second-degree burns healed in a shorter period with the sea buckthorn cream than with 1% silver sulfadiazine (p below 0.001) | Listed, not graded. Compared with an active dressing, not a placebo. A convenience sample from one hospital in Iran. We could read only the abstract and nobody independent has repeated it |
| Wang (burns, TOPICAL) | 2006 | 151 treated, plus controls | Sea buckthorn oil dressing, changed every other day until healed | Until healed | Less exudate and pain, and faster epithelial growth and wound healing, than vaseline gauze controls (statistically significant) | Listed, not graded. A Chinese-language report with an English abstract, which does not say how patients were allocated or give the number of controls. We could read only the abstract and nobody has repeated it |
| Xin (ear-drum perforation, TOPICAL oil) | 2015 | 199 outpatients: 101 treated, 98 controls | Sea buckthorn oil applied to the perforation once a week, against letting it heal by itself with a weekly check | 2 months of follow-up | Healing at 1 and 2 months was 62.4% and 79.2% with the oil against 29.6% and 57.1% without (P below 0.05). The gain was significant for middle and large perforations and not for small ones, and healing time at 2 months was shorter | Listed, not graded. Randomized in the abstract's words, with no placebo: the control group got no treatment and was only checked. A Chinese-language journal with an English abstract. We could read only the abstract and nobody independent has repeated it |
| Gao (ear-drum perforation, TOPICAL oil patch) | 2017 | 370 patients in a treatment and a control group | A sterile cotton patch with sea buckthorn oil over the perforation, against a sterile cotton patch alone | 2 months | The overall healing rate was significantly higher with the oil patch, and for middle and large perforations the oil raised the healing rate significantly. At 2 months healing generally took less time (P below 0.05) | Listed, not graded. Patients were randomly assigned to the two groups. A dressing for an injury, not a supplement. We could read only the abstract and nobody independent has repeated it |
| Rodhe (hemodialysis) | 2013 | 45 | 4 capsules a day of 500 mg sea buckthorn oil extract (2 g), or placebo | 2 x 8 weeks, 4-week washout, crossover | No significant change in DNA breaks, oxidative DNA lesions, salivary flow or inflammation markers. Plasma phosphate and sodium rose and iron fell | Listed, not graded. A null result in people on hemodialysis at one hospital. We could read only the abstract and nobody independent has repeated it |
| Xiao (children, functional dyspepsia) | 2013 | 120 children, three groups | A Hippophae rhamnoides preparation, domperidone, or both | 8 weeks | Plasma leptin and neuropeptide Y were higher after treatment in the two Hippophae groups than in the domperidone group, and postprandial gastric antrum residue fell | Listed, not graded. Compared with domperidone, not a placebo. The preparation and dose are not described in the abstract. We could read only the abstract and nobody has repeated it |
| Vîrgolici (obese children) | 2013 | 41 obese children, 30 lean controls | 800 mg a day of sea buckthorn pulp oil | 60 days | Total cholesterol, triglycerides, leptin, ceruloplasmin, fasting C peptide, blood pressure and carotid intima-media thickness fell after treatment (all P below 0.05) | Listed, not graded. A before-and-after design with no placebo group. We could read only the abstract and nobody has repeated it |
| Nemes-Nagy (children with type 1 diabetes, COMPOUND product) | 2008 | 30 | Blueberry and sea buckthorn concentrate | 2 months | Erythrocyte superoxide dismutase activity and C peptide were higher and glycated hemoglobin was lower after treatment (all p below 0.05) | Listed, not graded. A two-ingredient product, so the result cannot be credited to sea buckthorn alone. The abstract describes no control group. We could read only the abstract |
| Ramnani (fruit and vegetable shots with inulin, COMPOUND product) | 2010 | 66 healthy volunteers, 22 per group | Two shots, pear-carrot-sea buckthorn or plum-pear-beetroot, each with 5 g a day of Jerusalem artichoke inulin, against a placebo | Not in the abstract | Bifidobacteria were higher on both shots than on placebo (P below 0.0001) and the Lactobacillus and Enterococcus group rose slightly (P=0.042). Other bacterial groups and faecal short-chain fatty acids did not change. Flatulence was slightly higher but mild | Listed, not graded. The active ingredient is the inulin and sea buckthorn is one fruit in one of the two shots, so nothing here can be credited to sea buckthorn. We could read only the abstract |
| Sadhu (polyherbal formulation, Alzheimer-type dementia, COMPOUND product) | 2014 | 109 healthy elderly and 123 patients enrolled, 97 and 104 completed | 500 mg twice a day of a formulation of Bacopa, sea buckthorn (leaves and fruits) and Dioscorea bulbifera extracts. Patients were compared with donepezil 10 mg twice a day, healthy elderly with placebo | 12 months | In the patients, the formulation scored better than donepezil on digit symbol substitution (38.98 vs 35.85, P=0.0001), immediate word recall and attention span, and on the functional activity and depression scales | Listed, not graded. A three-plant formulation, so nothing can be credited to sea buckthorn alone, and the comparator in patients was a drug, not a placebo. We could read only the abstract |
| Yershova (leaf extract, viral infections in children) | 2014 | Not in the abstract | Hyporamin (Erebra), a dry purified extract of sea buckthorn leaves. The dose is not in the abstract | Not in the abstract | The short abstract, in Russian, says the preparation was effective in children with various viral infections | Listed, not graded. A leaf extract, not the berry or the oil. The abstract gives no numbers, no control group and no design, so nothing can be weighed. We could read only the abstract |
The lipid family
Two meta-analyses, Geng 2022 pooling 15 randomized trials and Guo 2017 pooling 11, found that sea buckthorn significantly lowered triglycerides, total cholesterol, and LDL while raising HDL, with the subgroup analysis finding the reduction only in people whose lipids were abnormal to start, though a 2026 meta-analysis and a 111-person trial found no significant effect on lipids. The same Geng analysis found no significant effect on blood glucose, blood pressure, or BMI. Those two positive analyses are the strongest card this plant holds, and the two nulls keep it from being a sure thing. The 2026 meta-analysis pooled 12 randomized trials (901 people analyzed), found no significant pooled effect on total cholesterol, LDL-C, HDL-C or triglycerides and rated every outcome very low certainty, and we could read only its abstract. Zhou 2022 gave 90 ml of sea buckthorn puree or placebo to 111 people with high cholesterol for 90 days and found no change in total cholesterol, LDL-C or triglycerides. The cholesterol page translates the lipid evidence for non-scientists. Matching the doses in this table means counting grams of oil rather than capsules, and our capsules guide scores products on whether that number appears on the label at all.
The dry-eye family
The flagship single trial: 100 adults, 2 grams of blended seed-and-berry oil daily for three months from autumn into winter, randomized, double-blind, placebo-controlled. Tear osmolarity rose less in the sea buckthorn group, and fewer people hit the worst redness score (counting everyone) and the worst burning score (counting compliers only), each significant in one analysis only. The other diary symptoms, tear break-up time, the Schirmer test and the symptom questionnaire showed no significant difference, and we print that. The 2011 follow-up found tear-film fatty acids unchanged, so the mechanism remains open. A second, smaller randomized study from the same group, published in 2019, tested an oil and sodium hyaluronate eyelid spray in three parts of 2, 10 and 40 people and reported a lower Ocular Surface Disease Index score in the sprayed eye of the 40-person part, and we could read only its abstract. It is topical and the spray also contains hyaluronate, so it is a follow-on and not an independent replication. Full dissection on the dry-eye trial page.
The skin family
Two trials, two generations of design. Yang 2009 measured large effects (about 49% hydration, 26% elasticity) with instruments but gave the oral arm no placebo. Chan 2024 measured smaller effects (14% texture and 8% hydration beat placebo, while its 10% collagen density gain did not) against a real placebo, with industry money attached. They triangulate each other, and the skin studies page does the full autopsy. Beside them sit a 1999 trial in atopic dermatitis, where the paraffin placebo group improved as well, and four small studies of topical fruit-extract creams and emulsions in healthy volunteers. We list those and do not grade them.
The women’s health family
Three trials now, and they split by route and by product. Oral: Larmo 2014 randomized 116 postmenopausal women to 3 grams of oil or a placebo daily for three months, 98 finished, and the oil group had about three times the odds of improved vaginal epithelial integrity versus placebo (odds ratio 3.1, 95% CI 1.11 to 8.95), though the trial did not measure how. Topical: a 2025 trial randomized 377 postmenopausal women before a colposcopy to a vaginal sea buckthorn suppository, an estrogen ointment, or no pretreatment, and among the 368 who finished, both treatments made the exam readable and the two of them came out statistically indistinguishable, with none of estrogen’s systemic side effects on the suppository side. That last phrase does a lot of work, so read it carefully. The trial set no non-inferiority margin, so it detected no difference rather than proving the two equal. Keep the routes apart too. The head-to-head with estrogen belongs to a local suppository, a compound formula placed in the vagina, not to the oral capsules Larmo tested. A 2021 randomized trial of 60 postmenopausal women tested a branded vaginal gel that combines sea buckthorn oil with aloe vera, glycyrrhetic acid, hyaluronic acid and glycogen, and its abstract reports better Vaginal Health Index and sexual function scores than placebo over 12 weeks. With five active ingredients, its result cannot be credited to sea buckthorn alone. A registered trial of a sea buckthorn oil cream in 225 women (NCT02937805) completed in 2017, and we found no report of it. The oral and suppository trials are single trials of their kind, and both are honestly framed as support rather than a replacement for anything.
The infection and cardiovascular outliers
Larmo 2008 gave about 250 healthy adults 28 grams of berry puree daily for three months and found no effect on colds, with a mild drop in the inflammation marker CRP. We keep that null in front of every immunity claim on this site because it disciplines all of them.
Vashishtha 2017 reported that hypertensive adults taking 0.75 ml of seed oil daily for 30 days saw blood pressure and lipids improve. That paper is gone. Clinical Nutrition retracted it in 2026, so the one oil trial that reported lower blood pressure no longer counts. Thinner reports exist. A 2025 single-arm pilot of a seed lipid extract reported a 2.9 mmHg fall in systolic pressure, and a 2026 meta-analysis of 12 trials reported a statistically significant but clinically negligible fall in diastolic pressure, which we could read only as an abstract. The pooled Geng 2022 analysis found no significant effect on blood pressure. We kept the retracted row in the table above, marked retracted, because a withdrawal readers can see is more useful than a citation that quietly disappears. Johansson 2000 gave 12 healthy men 5 grams of berry oil daily for four weeks and measured a clear decrease in platelet aggregation. That’s the highest daily oil dose in the trials we cite, matched only by Yang 1999, and the reason our blood-thinner page exists.
The constipation trial
The catalog’s first gut trial arrived in 2026. Morone and colleagues randomized 90 adults with chronic idiopathic constipation to either a 500 mg sea buckthorn fruit extract or a matching placebo, double-blind, for 28 days. Spontaneous complete bowel movements rose from about 1.5 to 2.6 per week on the extract while placebo held flat, and stool consistency normalized on the Bristol scale, both at p below 0.001. Bloating, abdominal pain, and heaviness eased as well. Flatulence did not separate from placebo, and the SF-12 quality-of-life score barely moved, which puts a ceiling on how much a reader should expect a month of this to change their life. The abstract reports no adverse events. The asterisk is the sponsor. The extract came free from EPO S.r.l., the named sponsor, so this is an industry-sponsored study standing alone until an independent group repeats it. We treat it as a real but single, industry-sponsored human result that needs replication, and the constipation page leads with it and its caveats.
The low-dose and crossover studies
Two more reports and one more trial fill out the catalog’s middle. Larmo 2009 is a second report from the 2008 infection trial (229 of its participants had lipid data, 28 grams of berry puree a day for three months), not a separate trial. It found no change in total, HDL or LDL cholesterol or triglycerides, and higher blood levels of the flavonols quercetin and isorhamnetin: nutrition research at food doses, not supplement doses. Larmo 2013 put 80 overweight women through four diets of about a month each, in random order: dried sea buckthorn berries, sea buckthorn oil, a sea buckthorn phenolic extract and frozen bilberries. The diets’ effects on serum metabolites depended on each woman’s baseline metabolic profile, an early hint that who takes the berry shapes what it does. And the Jarvinen 2011 follow-up re-analyzed the dry-eye cohort for mechanism, finding tear-film fatty acids unchanged. None of these stands alone as proof of anything, and each one disciplines the claims the bigger trials make.
The other studies we list and do not grade
Most of the table’s newer rows are small studies we could read only in abstract. We list them so the table matches the registry, and we do not grade on them because nobody independent has repeated them. Two burn studies reported faster healing, a 2021 trial of sea buckthorn cream against 1% silver sulfadiazine in 55 people and a 2006 Chinese report of an oil dressing in 151 burned patients. A 2013 randomized trial in 120 children with functional dyspepsia compared a Hippophae preparation with domperidone, and two uncontrolled studies in children are listed too, one of pulp oil in 41 obese children and one of a blueberry and sea buckthorn concentrate in 30 children with type 1 diabetes. A 2013 crossover in 45 hemodialysis patients found no significant change in oral health or inflammation markers. Huang 2020 is a dose-escalation study of oils chosen for their palmitoleic acid content, not powered for clinical outcomes. Eccleston 2002 gave a juice to 20 healthy men and found no significant change in total cholesterol, LDL-C or platelet aggregation, and Suomela 2006 gave sea buckthorn flavonols in porridge and found no significant effect on oxidized LDL, C-reactive protein or homocysteine. Kopčeková 2023 and Erglis 2025 are uncontrolled studies that reported lower LDL-C.
How the families compare
Read across the families and a grading pattern emerges. Lipids hold the deepest evidence and a split one: two meta-analyses, 15 and 11 pooled trials, point one way, while a 2026 meta-analysis and a 111-person trial found no effect. Dry eyes hold the best single study: 100 people, placebo-controlled, seasonally timed, with a smaller follow-on from the same group. Skin and vaginal health each hold a few positive trials with real caveats. Infections hold a clean null at food doses. And platelets hold a signal, at 5 grams a day, that matters for safety more than benefit. Constipation is the newest and thinnest, a single industry-sponsored extract trial that still needs an independent repeat. A plant whose evidence is this uneven is normal. A website that admits it is rarer.
The trials that do not exist
The fastest way to read this catalog is to notice what is missing from it. No human trial has tested sea buckthorn for cancer, and nothing in the trials we can read touches cognition, sleep, or mood. No placebo-controlled trial has tested it for weight loss, and the pooled BMI null is the closest thing to an answer the science offers, which the weight loss page treats with the seriousness a null deserves. The pooled trials in the 2022 meta-analysis measured blood glucose and found no significant effect. Small studies were designed around blood sugar: a 2021 crossover in people with impaired glucose regulation, where fasting glucose fell by a median 0.14 mmol/L on puree against a rise on placebo (p=0.045), and two single-meal studies. The first, in 18 overweight men, lowered after-meal insulin without moving glucose. The second, in 10 healthy men, found a lower insulin peak with whole berries. None is a treatment trial, and the blood-glucose null says why the blood sugar page stays cautious. Absence of trials is not absence of benefit. It is absence of knowledge, and honest pages price it accordingly.
What would settle the top three questions
Lipids. A trial enrolling thousands of people with abnormal cholesterol, run for years, measuring heart attacks and strokes rather than lab panels. No supplement has ever funded one, so the realistic next step is a 12-month marker trial with follow-up.
Dry eyes. An independent lab, a different climate, about 300 participants across all seasons. One team’s single study is a lead until another team finds the same thing.
Vaginal dryness. For the oral capsules, a straight replication of Larmo 2014 at 3 grams daily, run longer than three months. A 2025 trial already found a topical sea buckthorn suppository no different from estrogen at getting a cervix ready for colposcopy, but that result belongs to a local product, it was not designed to prove equivalence, and it does not carry over to swallowing oil, so the oral question stays open. The vaginal dryness page keeps the two routes separate.
Who paid for all this
The funding map explains the catalog’s shape. Industry connections run through much of it. Yang 2009 used a brand’s oil and ran in a journal outside the PubMed index, Chan 2024 was funded by the brand whose oil it tested, and Morone 2026 tested a fruit extract that its sponsor, EPO S.r.l., supplied free of charge. The 2010 dry-eye trial was funded in part by Aromtech Ltd., the company that made its seed-and-pulp oil, and one author worked for Aromtech during the trial. The 2014 vaginal trial lists its lead author at Aromtech, and the 2013 metabolomics crossover was part-funded by Aromtech, which supplied its capsules and employed its first author. The two lipid meta-analyses carry no such flag in what we can read, and the 2025 suppository trial goes one better by stating that its authors paid the costs themselves and took no outside money. Funding proves nothing in either direction. It tells you where to look for inflation, and the six checks page turns that instinct into a drill.
How to read this catalog
Count the pattern, not the promises. Every positive here is one to three small trials deep. Every null is published next to its positive. When a label or blog quotes any of these trials without its limitation column, you now own the limitation column. The doses behind each row are tabulated on the study dosages page, and how we research documents the grading method itself.
Frequently asked questions
How many human clinical trials of sea buckthorn exist?
The registry lists the human trials and pooled analyses we could read, each with its dose, duration, result and limits. They cover dry eyes, skin, vaginal dryness, blood lipids, infections, platelet function, constipation, burns, atopic dermatitis and a few single-meal studies. Most are small and short, from a single meal to a few months. One of them, Vashishtha 2017 on blood pressure, was retracted in 2026 and no longer counts. A dated search of PubMed, Europe PMC and trial registries finds more small human studies than the registry holds, so this is a floor and not a count of every trial. That is a real but young evidence base, stronger than most supplements and far short of any drug.
What is the strongest sea buckthorn trial evidence?
Blood lipids, where two meta-analyses (15 and 11 randomized trials) found lower triglycerides, total cholesterol and LDL with the reduction only in people with abnormal lipid metabolism, and both report high heterogeneity. A 2026 meta-analysis of 12 randomized trials found no significant pooled effect on any lipid, with very low certainty (we could read only its abstract), and a 2022 placebo-controlled trial in 111 people with high cholesterol found no change in total cholesterol, LDL-C or triglycerides. The strongest single trial is the 100-person dry-eye study, randomized, double-blind, and placebo-controlled.
Why are there so few sea buckthorn trials?
Economics. A drug trial pays for itself through patent exclusivity, and nobody can patent a berry, so no company recovers a nine-figure trial budget from sea buckthorn. What exists comes from small academic groups running modest trials on modest budgets. That is why the catalog is full of small studies instead of a drug-style development program, and why every page on this site grades accordingly.
Illustration.