Herbs & Botanicals · Starflower and GLA seed oil

Borage

Borago officinalis · also called Borage, starflower, bee plant, bee bread, ox's tongue

Borage is a blue-flowered garden herb. Its seeds give a GLA-rich oil sold for arthritis, skin and PMS, while its leaves and flowers go into salads and teas. Trials show some help for rheumatoid arthritis symptoms but no benefit for eczema. The key caution: leaf, flower and tea contain liver-toxic pyrrolizidine alkaloids, and seed oil should only be used if certified free of them.

General information, not medical advice. Last substantive update October 25, 2026 · Not medically reviewed. A named clinician has not signed off on this page. How this page is made

Get your Borage today. Order now from Amazon and have it delivered to your door.Shop Borage on AmazonSupport our free guides: as an Amazon Associate we earn from qualifying purchases.

About Borage

Borage (Borago officinalis) is a garden herb whose seed oil is rich in GLA. Trials of GLA oils show some relief in rheumatoid arthritis, no benefit in eczema, and thin evidence elsewhere. Its leaf and flower contain liver-toxic pyrrolizidine alkaloids, so the safety tab matters more than the benefits.

What it is

Borage is a bristly annual with blue star-shaped flowers, a member of the borage family. Leaves and flowers are eaten fresh or made into tea. The seeds are pressed for an oil sold as borage oil or starflower oil.

Other names include bee plant, bee bread and ox's tongue.

^{6 12}

Leaf and flower versus seed oil

They are not the same product. Leaf and flower contain pyrrolizidine alkaloids (PAs), natural liver toxins. The seed oil is the GLA source, and a hospital reference says it should be used only if certified free of unsaturated PAs.

Evening primrose oil is another GLA oil. Results for it do not automatically apply to borage.

^{19 12 2}

What people use it for

Seed oil is promoted for rheumatoid arthritis, eczema, diabetic nerve pain and PMS. Leaves and flowers are traditionally used as a mild expectorant and in teas.

¹²

What research shows

Rheumatoid arthritis: a Cochrane review pooled seven trials of GLA oils and found less pain and disability, with more side effects that were not statistically different. A 1993 trial of borage seed oil found fewer tender and swollen joints than placebo, but it was small. A longer 2014 trial had no placebo group.

Eczema: borage oil did no better than placebo in a Cochrane review and in a BMJ trial of 140 people.

Diabetic nerve damage: older GLA trials were positive, but their abstracts do not say the GLA came from borage. Asthma: one small trial of an extract improved symptoms but not airway inflammation. PMS: borage was not tested; evening primrose oil failed to beat placebo.

^{2 10 18 1 20 9 13 3}

At a glance

Traditional uses history, not proof

Sweating, cough and phlegm (leaf and flower)
Anti-inflammatory and to promote milk (traditional)
Alternative to evening primrose oil as a GLA source
Salad herb, garnish and tea

Studied for research

Rheumatoid arthritis: some symptom relief from GLA oils (low certainty for borage itself)
Eczema: no benefit over placebo
Diabetic nerve damage: GLA trials, borage not confirmed
Asthma symptoms: one small trial
PMS, menopause, mood: no borage trial

Main compounds

Gamma-linolenic acid (GLA) in seed oil
Pyrrolizidine alkaloids (amabiline, lycopsamine, supinine) in leaf and flower
Nitrate (very high in leaf)
Mucilage and tannins

Common forms

Seed oil capsules
Liquid seed oil
Fresh leaf and flower
Tea or infusion
Liquid extract (one asthma trial)
Skin oil (not covered by our studies)

Research links: Cochrane review: herbal therapy for rheumatoid arthritis (2011) · Cochrane review: evening primrose and borage oil for eczema (2013) · BMJ eczema trial of borage oil (2003) · EFSA opinion on pyrrolizidine alkaloids (2017) · Pyrrolizidine alkaloids in borage leaves (2023) · Memorial Sloan Kettering: Borage · Research Library · Genes & Biomarkers Library

The plant and its history

Accepted scientific name: Borago officinalis L.
Naming authority: L.
Common names: Borage (English); Starflower (English); Bee plant (English); Bee bread (English); Ox's tongue (English)
Botanical family: Boraginaceae (borage family)
Classification: Plantae > Boraginales > Boraginaceae > Borago > Borago officinalis
Native range: Not verified in the sources we read.
Habitat: Not verified in the sources we read; grown in gardens and nurseries.
What the plant looks like: Roots: Not described.; Stems: Not described.; Leaves: Eaten fresh in salads and mixed dishes; non-flowering plants are easily confused with foxglove (reference text).; Flowers: Used as garnish and tea; carry PAs.; Fruit: Not described.; Seeds: Source of GLA-rich oil.
Cultivation: Sold as a garden plant (one foxglove poisoning followed purchase of a mislabelled plant at a nursery).
Harvest and processing: Seed oil is pressed from the seeds; PA-free certification is advised (reference).
Look-alikes and adulterants: Name: Foxglove (Digitalis purpurea); Note: Poisonous; leaves mistaken for borage in poisoning reports. Never eat unidentified leaves.
Sustainability and conservation: Not assessed in the sources read.

Traditional use and history history, not clinical proof

Tradition or era	Documented use	Source
Herbal and traditional medicine (region not stated)	To induce sweating, as an expectorant and anti-inflammatory, to promote lactation, to stimulate adrenal function, and as an alternative source of GLA to evening primrose oil.	s-msk
EU regulatory view	Named as a salad plant in the EU PA public statement; no borage monograph.	s-ema-pa

Traditional use shows how a plant was used. It does not show that a supplement works or is safe today.

What is inside, and how it may work

What the plant contains and how those compounds act in cells and animals. Laboratory and animal mechanisms are not proven effects in people.

Main compounds

Compound	Family	Where it is concentrated	Activity studied	Level of evidence
Gamma-linolenic acid (GLA) PubChem 5280933	omega-6 fatty acid	seed oil Reference says borage oil is about 25% GLA versus about 10% for evening primrose oil (second-hand).	Body can convert it to DGLA, which may reduce leukotrienes and some prostaglandins. ^{2 12 14}	Human data
Amabiline PubChem 442706	1,2-unsaturated pyrrolizidine alkaloid	leaf, flower Not stated in sources read.	Liver-toxic PA named by the hospital reference. ^{12 14}	Expert reference
Lycopsamine and intermedine PubChem 107938	1,2-unsaturated pyrrolizidine alkaloids	leaf, flower Not stated in sources read. Intermedine CID 114843.	Liver-toxic PAs named by the reference; acetylated forms prominent in one leaf analysis. ^{12 19 14}	Expert reference
Supinine PubChem 108053	pyrrolizidine alkaloid	leaf, flower Not stated in sources read.	Listed by the reference among borage PAs. ^{12 14}	Expert reference
Nitrate PubChem 943	inorganic ion	leaf Mean 3968 mg per kg in 15 borage samples.	Can cause methemoglobinemia in infants. ^{11 14}	Human data
Mucilage and tannins	plant polysaccharides and polyphenols	leaf, flower Not measured in sources read.	Mucilage is thought to loosen phlegm; tannins to be mildly astringent (reference). ¹²	Expert reference

Mechanisms, receptors and pathways

Target or pathway	Gene or receptor	What was found	Level
GLA to DGLA to eicosanoids	FADS1/FADS2 (desaturases)	GLA is converted to DGLA, which can block formation of leukotrienes and some prostaglandins. One study found serum GLA and DGLA after borage oil depended on rs174537 genotype (second-hand). ¹²	Mechanism theory
Liver activation of PAs	CYP2A6, CYP3A4, CYP3A5	1,2-unsaturated PAs are converted to reactive pyrroles that bind proteins and injure liver sinusoidal cells. ^{21 24}	Human data
Immune and brain effects	not specified	Cell and animal findings only (second-hand). ¹²	Animal research

Absorption, metabolism and elimination

Absorption: Not studied for borage in the sources read.
Distribution: Not studied.
Metabolism: PAs are activated by liver CYP enzymes (review). GLA is converted to DGLA, varying with genotype (second-hand).
Elimination: Not studied.
Bioavailability: Not established.

^{21 12}

Isolated compounds versus whole extracts

The arthritis and eczema trials used GLA-rich seed oil, not leaf, flower or tea. Leaf and flower carry PAs; seed oil is the part sold as a GLA source and is advised only if PA-free. Results for one part do not apply to another, and results for evening primrose oil should not be assumed for borage.

Nutritional composition

GLA is not an essential nutrient for healthy adults; borage leaf is eaten as a minor salad herb but is not a meaningful nutrient source.

8benefits and uses looked at

24studies and sources checked

Low certaintybest evidence rating found

7lab markers tracked in trials

1 low certainty3 very low certainty3 not enough evidence1 evidence against

What was looked at

The 8 benefits and uses checked for Borage, each with how strong the evidence is. Tap one to open it.

Rheumatoid arthritis (seed oil)Low certainty
Atopic eczema (seed oil, by mouth)Evidence against
Diabetic nerve damage (GLA, source oil unclear)Very low certainty
Asthma and cough (liquid extract)Very low certainty
Premenstrual symptomsNot enough evidence
Menopause symptoms and low moodNot enough evidence
Blood pressure and cholesterol (seed oil, small studies)Not enough evidence
Brain, bone and cancer prevention (lab and animal only)Very low certainty

Lab markers tracked in the studies

Tender and swollen joint countDisease activity score (DAS28)Eczema sign score (SASSAD)Nerve conduction and symptom scorePain score in diabetic neuropathyAsthma control test scoreExhaled nitric oxide and sputum eosinophils

What is Borage good for?

What borage has been studied for and how strong the evidence is. Most uses refer to the seed oil, not the leaf or tea.

Rheumatoid arthritisLow certaintyGLA oils gave some pain and disability relief in a Cochrane review. Borage alone: one small placebo trial and one trial without placebo. Atopic eczemaEvidence againstBorage oil by mouth was no better than placebo in a Cochrane review and a 140-person trial. Diabetic nerve damageVery low certaintyGLA trials were positive but did not name borage as the source and were small or company-funded. Asthma and coughVery low certaintyOne small one-month trial of a borage extract improved symptoms and control scores but not airway inflammation. Premenstrual symptomsNot enough evidenceNo borage trial. Evening primrose oil, a different GLA oil, did not beat placebo. Menopause symptoms and low moodNot enough evidenceA hospital reference says there is no scientific evidence. Blood pressure and cholesterolNot enough evidenceSmall studies are mentioned second-hand; we did not read them. Brain, bone and cancer preventionVery low certaintyLab and animal results only.

Benefits, one by one

Each card is a research question, not a promise. Tap a goal to filter. Open a card to see the studies, doses and limits.

Rheumatoid arthritis (seed oil)

Low certainty

A Cochrane review found moderate evidence that oils rich in GLA, including borage seed oil, give some pain and disability relief in rheumatoid arthritis. The one placebo-controlled borage trial was small and old. A later trial had no placebo group. Borage seed oil is not a replacement for arthritis medicine.

4 studiesTrials used GLA-rich seed oil capsules; no product is recommended her…

Studies, doses and limits

What the trials found

Cochrane (2011) pooled 7 trials of GLA oils (evening primrose, borage, blackcurrant). A 1993 trial of borage seed oil (37 people, 24 weeks) beat cottonseed oil placebo. An 18-month trial (2014) compared borage oil, fish oil and both with no placebo group.

Results

Pooled: less pain and less disability, more side effects that were not statistically different. 1993 trial: tender joints down 36%, swollen joints down 28%, versus no real change on placebo. 2014 trial: all three oil groups improved by a similar amount, so the oil cannot be credited without a placebo.

Effect sizes as reported

Pain mean difference -32.83 on a 100-point scale (95% CI -56.25 to -9.42); adverse events 20% vs 3%, RR 4.24 (0.78 to 22.99).

Why this rating

Cochrane calls it moderate for GLA oils as a group, but the borage-only evidence is one small 1993 placebo trial plus one trial with no placebo; many trials had design flaws. We rate borage itself low.

Form and dose studied

Trials used GLA-rich seed oil capsules; no product is recommended here. Ask a doctor or pharmacist about amounts.

Who was studied

Adults with active rheumatoid arthritis, mostly already on arthritis medicine.

Limits

Pooled three oils; small trials; the 2014 trial had no placebo; authors of the 1993 trial said it should not be viewed as therapy.

Why researchers looked

Borage seed oil is rich in GLA, which may reduce inflammation, and arthritis is the use with the most human trials.

Traditional use

Seed oil promoted as an alternative source of GLA to evening primrose oil (reference text).

Proposed mechanism Pharmacological theory

Borage seed oil is the richest common plant source of gamma-linolenic acid (GLA), an omega-6 fat. The idea is that the body converts GLA to substances that calm inflammation. This is a proposed mechanism, not proof of benefit.

A proposed mechanism does not show that the effect happens in people.

Only oil certified free of unsaturated pyrrolizidine alkaloids is considered acceptable by the reference we read; see the safety tab. Not for pregnancy, breastfeeding or liver disease.

^{2 10 18 12}

Atopic eczema (seed oil, by mouth)

Evidence against

Borage oil taken by mouth did not beat placebo for eczema in Cochrane's review or in a 140-person trial. The evidence is against a benefit.

3 studiesCapsules of borage oil; no benefit shown at the amounts tested.

Studies, doses and limits

What the trials found

Cochrane (2013): 27 trials, 8 of borage oil. BMJ trial (2003): 140 adults and children, 12 weeks.

Results

Neither borage oil nor evening primrose oil improved overall eczema symptoms more than placebo. In the BMJ trial the score fell 3 points on borage oil and 5 on placebo.

Effect sizes as reported

BMJ: difference 1.4 points in favour of placebo (95% CI -2.2 to 5.0). Evening primrose oil pooled difference -2.22 on 0-100 (-10.48 to 6.04).

Why this rating

Two independent sources agree on no benefit; the Cochrane authors said more trials would be hard to justify.

Form and dose studied

Capsules of borage oil; no benefit shown at the amounts tested.

Who was studied

Adults and children with atopic eczema.

Limits

Short trials; the Cochrane review's author had earlier research funded by an evening primrose oil maker (a different author handled that trial). Skin creams were not covered.

Why researchers looked

Borage seed oil's GLA was thought to repair the skin barrier and ease inflammation.

Traditional use

Oil promoted for dermatitis; also used on the skin as a cosmetic oil (not covered by the studies we read).

Proposed mechanism Pharmacological theory

A proposed mechanism does not show that the effect happens in people.

Side effects were mild and mostly stomach-related, similar to placebo. Only oil certified free of unsaturated pyrrolizidine alkaloids is considered acceptable by the reference we read; see the safety tab. Not for pregnancy, breastfeeding or liver disease.

^{1 20 12}

Diabetic nerve damage (GLA, source oil unclear)

Very low certainty

Older trials of GLA found better nerve measures than placebo, and a 2020 trial found GLA about as good as another supplement for pain. Our sources do not say these trials used borage oil, so we cannot say borage helps.

4 studiesGLA capsules; source and amounts are not given in the abstracts.

Studies, doses and limits

What the trials found

Keen 1993 (111 people, 1 year) and Jamal 1990 (22 people, 6 months) both found better nerve results with GLA than placebo. Won 2020 (100 people, 12 weeks) compared GLA with alpha-lipoic acid, with no placebo group.

Results

1993: 13 of 16 nerve measures significantly better than placebo. 1990: symptom scores better than placebo. 2020: pain fell in both groups and GLA was noninferior to alpha-lipoic acid on pain but not on total symptoms.

Effect sizes as reported

Not given as a single number; 2020 pain difference -0.65 (95% CI -1.526 to 0.213).

Why this rating

The abstracts do not name the oil that supplied the GLA, trials are small and old, and the 2020 trial was company-funded and had no placebo arm.

Form and dose studied

GLA capsules; source and amounts are not given in the abstracts.

Who was studied

People with diabetic peripheral neuropathy.

Limits

Source of GLA not stated; no recent placebo-controlled borage trial found; high dropout in the 2020 trial.

Why researchers looked

GLA is needed to build nerve-cell membranes and nerve blood vessels in animal work, so researchers tried it in diabetic neuropathy.

Traditional use

Borage oil is promoted for diabetic nerve pain in reference texts.

Proposed mechanism Pharmacological theory

Proposed: GLA supports nerve cell membranes and blood flow; not shown for borage in the sources read.

A proposed mechanism does not show that the effect happens in people.

Only oil certified free of unsaturated pyrrolizidine alkaloids is considered acceptable by the reference we read; see the safety tab. Not for pregnancy, breastfeeding or liver disease.

^{9 8 23 12}

Asthma and cough (liquid extract)

Very low certainty

One small trial found a borage extract improved asthma symptoms and control scores over one month, but lung tests and airway inflammation did not change. It needs repeating before anyone can rely on it.

2 studiesLiquid extract; no standard product.

Studies, doses and limits

What the trials found

One randomised, double-blind trial of 38 people with moderate asthma, 1 month (Iran).

Results

Asthma control score rose from 10.8 to 15.4; flare-ups fell from 3.6 to 2 a month; spirometry, exhaled nitric oxide and sputum cells did not change significantly.

Effect sizes as reported

ACT +4.6 points; flare-ups 3.6 to 2 a month.

Why this rating

One small, short, single-centre trial; the extract is not described.

Form and dose studied

Liquid extract; no standard product.

Who was studied

Adults 15 to 90 with moderate asthma.

Limits

Small; extract not characterised; objective inflammation measures unchanged.

Why researchers looked

Borage contains mucilage thought to loosen phlegm, and it has been used for coughs.

Traditional use

Traditional expectorant and cough remedy (reference text).

Proposed mechanism Pharmacological theory

Mucilage is thought to coat the airways and loosen phlegm; GLA may lower leukotrienes. Neither was tested in this trial.

A proposed mechanism does not show that the effect happens in people.

Borage leaf and flower carry pyrrolizidine alkaloids; see the safety tab before using any leaf, flower or extract.

^{13 12}

Premenstrual symptoms

Not enough evidence

No trial of borage for PMS was found. Two reviews of herbs for PMS found no benefit for evening primrose oil, a different GLA oil, so we cannot say borage helps.

2 studiesNone tested.

Studies, doses and limits

What the trials found

No borage trial found. Reviews: evening primrose oil no different from placebo.

Results

Reviews of herbs for PMS found no benefit from evening primrose oil; chasteberry and calcium had more support. Borage not tested.

Why this rating

Nothing tested for borage; related oil was negative.

Who was studied

Women with PMS.

Limits

Evening primrose oil is a different plant; borage may differ.

Why researchers looked

GLA oils have long been sold for PMS on the theory that hormone-related prostaglandins are involved.

Traditional use

Promoted for PMS and menopause in consumer texts.

Proposed mechanism Pharmacological theory

Prostaglandin theory only.

A proposed mechanism does not show that the effect happens in people.

Not for pregnancy or breastfeeding.

^{3 22}

Menopause symptoms and low mood

Not enough evidence

A hospital reference says there is no scientific evidence for borage in menopause symptoms or depression.

1 studiesNone tested.

Studies, doses and limits

Why this rating

No trials found; reference says no evidence.

Who was studied

Not studied.

Limits

Nothing to assess.

Why researchers looked

Traditional and online claims.

Traditional use

Promoted for menopause and low mood.

Proposed mechanism Pharmacological theory

No tested mechanism.

A proposed mechanism does not show that the effect happens in people.

Avoid leaf and flower teas for the reasons on the safety tab.

¹²

Blood pressure and cholesterol (seed oil, small studies)

Not enough evidence

The hospital reference mentions small studies in which borage oil (alone or with echium oil) lowered blood pressure or cholesterol. We did not open those studies, so we cannot judge them.

1 studiesSeed oil.

Studies, doses and limits

What the trials found

Summary-only: the reference lists a short study in postmenopausal women with high blood pressure and a borage plus echium oil study in early type 2 diabetes or metabolic syndrome.

Results

Reported lower blood pressure and waist-to-hip ratio, and lower total and LDL cholesterol, in those small studies (second-hand).

Effect sizes as reported

Not available.

Why this rating

Reported second-hand, studies not read.

Form and dose studied

Seed oil.

Who was studied

Postmenopausal women with hypertension; early type 2 diabetes.

Limits

Not independently checked.

Why researchers looked

Omega-6 and other seed oils can change blood fats.

Traditional use

Not a traditional use.

Proposed mechanism Pharmacological theory

Changes in fatty acid levels (theory).

A proposed mechanism does not show that the effect happens in people.

Same safety rules as other borage oil.

¹²

Brain, bone and cancer prevention (lab and animal only)

Very low certainty

Borage extracts and oil have shown effects in cells and animals, for example in Alzheimer's and osteoporosis models. None of this has been tested in people.

1 studiesNone.

Studies, doses and limits

What the trials found

No human study.

Results

Lab and animal results only (second-hand).

Why this rating

Lab and animal only.

Limits

Cell and animal findings often fail in people.

Why researchers looked

Early lab studies seeking new uses for GLA-rich oil and plant extracts.

Proposed mechanism Laboratory research

Reported in animals: less bone inflammation, protection of nerve cells; in cells: anti-cancer effects.

A proposed mechanism does not show that the effect happens in people.

None specific.

¹²

Connections to the rest of health

How Borage connects to the rest of our health knowledge graph. Every link says what kind of connection it is. A safety link (for example to the liver) is a warning, not a benefit. A study that merely mentions a topic never creates a benefit link.

Explore every connection: goals, body systems, organs, symptoms, genes

Kind of connection Connected to Search

A. Wellness goals

Connected to	Kind of connection	Evidence status	Context	Studies
Joint Support		Low certainty	Rheumatoid arthritis (seed oil): A Cochrane review found moderate evidence that oils rich in GLA, including borage seed oil, give some pain and disability relief in rheumatoid arthritis. The one placebo-controlled borage trial was small and old. A later trial had no placebo group. Borage seed oil is not a replacement for arthritis medicine.	^{2 10 18 12}
Healthy Skin		Evidence against	Atopic eczema (seed oil, by mouth): Borage oil taken by mouth did not beat placebo for eczema in Cochrane's review or in a 140-person trial. The evidence is against a benefit.	^{1 20 12}
Blood Sugar Balance		Very low certainty	Diabetic nerve damage (GLA, source oil unclear): Older trials of GLA found better nerve measures than placebo, and a 2020 trial found GLA about as good as another supplement for pain. Our sources do not say these trials used borage oil, so we cannot say borage helps.	^{9 8 23 12}
Breathing Support		Very low certainty	Asthma and cough (liquid extract): One small trial found a borage extract improved asthma symptoms and control scores over one month, but lung tests and airway inflammation did not change. It needs repeating before anyone can rely on it.	^{13 12}
Womens Wellness	Traditional use	Not enough evidence	Premenstrual symptoms: No trial of borage for PMS was found. Two reviews of herbs for PMS found no benefit for evening primrose oil, a different GLA oil, so we cannot say borage helps.	^{3 22}
Menopause Support	Traditional use	Not enough evidence	Menopause symptoms and low mood: A hospital reference says there is no scientific evidence for borage in menopause symptoms or depression.	¹²
Mood Balance	Traditional use	Not enough evidence	Menopause symptoms and low mood: A hospital reference says there is no scientific evidence for borage in menopause symptoms or depression.	¹²
Blood Pressure Support	Preliminary mechanism	Not enough evidence	Blood pressure and cholesterol (seed oil, small studies): The hospital reference mentions small studies in which borage oil (alone or with echium oil) lowered blood pressure or cholesterol. We did not open those studies, so we cannot judge them.	¹²
Cholesterol Balance	Preliminary mechanism	Not enough evidence	Blood pressure and cholesterol (seed oil, small studies): The hospital reference mentions small studies in which borage oil (alone or with echium oil) lowered blood pressure or cholesterol. We did not open those studies, so we cannot judge them.	¹²
Brain Health	Preliminary mechanism	Very low certainty	Brain, bone and cancer prevention (lab and animal only): Borage extracts and oil have shown effects in cells and animals, for example in Alzheimer's and osteoporosis models. None of this has been tested in people.	¹²
Bone Health	Preliminary mechanism	Very low certainty	Brain, bone and cancer prevention (lab and animal only): Borage extracts and oil have shown effects in cells and animals, for example in Alzheimer's and osteoporosis models. None of this has been tested in people.	¹²

B. Body systems

Connected to	Kind of connection	Evidence status	Context	Studies
Bones Joints Muscles	Biological association	Low certainty	Studied for: Rheumatoid arthritis (seed oil) Studied for: Brain, bone and cancer prevention (lab and animal only)	^{2 10 18 12}
Pain Inflammation	Biological association	Low certainty	Studied for: Rheumatoid arthritis (seed oil)	^{2 10 18 12}
Skin Health	Biological association	Evidence against	Studied for: Atopic eczema (seed oil, by mouth)	^{1 20 12}
Nervous System	Biological association	Very low certainty	Studied for: Diabetic nerve damage (GLA, source oil unclear) Studied for: Brain, bone and cancer prevention (lab and animal only)	^{9 8 23 12}
Respiratory Health	Biological association	Very low certainty	Studied for: Asthma and cough (liquid extract)	^{13 12}
Reproductive Health	Biological association	Not enough evidence	Studied for: Premenstrual symptoms Studied for: Menopause symptoms and low mood	^{3 22 12}
Heart Health	Biological association	Not enough evidence	Studied for: Blood pressure and cholesterol (seed oil, small studies)	¹²

I. Biomarkers

Connected to	Kind of connection	Evidence status	Context	Studies
Tender and swollen joint count	Biological association	Low certainty	Measured in trials: lowered. A lab change is not the same as a health benefit.	¹⁰
Disease activity score (DAS28)	Biological association	Low certainty	Measured in trials: lowered. A lab change is not the same as a health benefit.	¹⁸
Eczema sign score (SASSAD)	Biological association	Evidence against	Measured in trials: unchanged. A lab change is not the same as a health benefit.	²⁰
Nerve conduction and symptom score	Biological association	Very low certainty	Measured in trials: improved. A lab change is not the same as a health benefit.	^{9 8}
Pain score in diabetic neuropathy	Biological association	Very low certainty	Measured in trials: lowered. A lab change is not the same as a health benefit.	²³
Asthma control test score	Biological association	Very low certainty	Measured in trials: improved. A lab change is not the same as a health benefit.	¹³
Exhaled nitric oxide and sputum eosinophils	Biological association	Very low certainty	Measured in trials: unchanged. A lab change is not the same as a health benefit.	¹³

J. Genes, receptors and enzymes

Connected to	Kind of connection	Evidence status	Context	Studies
FADS1/FADS2 (rs174537)	Preliminary mechanism	Mechanism theory	May change how borage oil raises GLA and DGLA (second-hand, one study).	¹²
CYP2A6, CYP3A4, CYP3A5	Preliminary mechanism	Human	Activate PAs to toxic forms.	²¹

L. Active compounds

Connected to	Kind of connection	Evidence status	Context	Studies
Pyrrolizidine alkaloids		Moderate certainty	Liver-toxic and carcinogenic in animals; EU limits in foods. Liver-toxic in leaf and flower; seed oil only if certified PA-free. Preparation or compound: leaf, flower; seed oil unless certified free	^{19 21 4}
Nitrate		Low certainty	Linked to infant methemoglobinemia. Preparation or compound: leaf puree
Gamma-linolenic acid (GLA)	Biological association	Human	Body can convert it to DGLA, which may reduce leukotrienes and some prostaglandins. Preparation or compound: seed oil	^{2 12 14}
Amabiline	Biological association	Expert reference	Liver-toxic PA named by the hospital reference. Preparation or compound: leaf, flower	^{12 14}
Lycopsamine and intermedine	Biological association	Expert reference	Liver-toxic PAs named by the reference; acetylated forms prominent in one leaf analysis. Preparation or compound: leaf, flower	^{12 19 14}
Supinine	Biological association	Expert reference	Listed by the reference among borage PAs. Preparation or compound: leaf, flower	^{12 14}
Nitrate	Biological association	Human	Can cause methemoglobinemia in infants. Preparation or compound: leaf	^{11 14}
Mucilage and tannins	Biological association	Expert reference	Mucilage is thought to loosen phlegm; tannins to be mildly astringent (reference). Preparation or compound: leaf, flower	¹²

Each connection is stored with the ingredient, the connected record, the kind of connection, the evidence status, the supporting studies, the preparation or compound involved, and a review date (last reviewed October 25, 2026).

Evidence at a glance: what changed in 3 lab markers

Tap a number to show only those markers. Counts are markers, not people.

Other3

The evidence

The research centre for Borage. Every record here was checked against PubMed or the publisher's page; we never list a study we have not verified.

Evidence by outcome

Rheumatoid arthritis (seed oil) 4 studies · 37 people in trials Low certainty Atopic eczema (seed oil, by mouth) 3 studies · 140 people in trials Evidence against Diabetic nerve damage (GLA, source oil unclear) 4 studies · 233 people in trials Very low certainty Asthma and cough (liquid extract) 2 studies · 38 people in trials Very low certainty Premenstrual symptoms 2 studies Not enough evidence Menopause symptoms and low mood 1 study Not enough evidence Blood pressure and cholesterol (seed oil, small studies) 1 study Not enough evidence Brain, bone and cancer prevention (lab and animal only) 1 study Very low certainty

Bars show how many studies we found for each outcome, not how well it works. People are added up across trials only (reviews are not added, to avoid double counting).

How we rate evidence

High certainty

Several well-run trials or reviews agree; further research is unlikely to change the picture much.

Moderate certainty

Good trials point the same way, but some doubt remains.

Low certainty

Small, short or inconsistent trials. The finding could change with better research.

Very low certainty

Very little or very weak human data. Treat as a possibility only.

Mixed findings

Trials or reviews disagree with each other.

Not enough evidence

Not enough human research to say either way.

Evidence against

Good research found no benefit, or found harm.

We take the reviewers' own risk-of-bias and GRADE statements where they exist, and otherwise judge from study size, design, consistency, directness and whether a product matches what is sold. Many publications do not make evidence high. Method note: We built on our existing summary, read the Memorial Sloan Kettering page, read PubMed abstracts, and checked GBIF and PubChem. EMA, EU regulation and German agency figures are second-hand and flagged. Most studies were read as abstracts only.

Study library: every study we checked 24

Design Research type Plant part Year from Search

Study	Design	People	Product, dose, length	Result
Bamford 2013 Oral evening primrose oil and borage oil for eczema Cochrane Database Syst Rev · PMID 23633319 · DOI	Systematic review	1596 People with atopic eczema (27 trials: 19 evening primrose oil, 8 borage oil)	Oral evening primrose oil and oral borage oil seed oil not given here not given here · vs see results	Neither oil beat placebo for global eczema symptoms. Adverse effects were mild, mostly stomach-related, and the same as with placebo. The authors say these oils are not effective for eczema. A case report warned of possible problems after more than a year of evening primrose oil, and one study found it may increase bleeding on warfarin. Evening primrose oil MD -2.22 (-10.48 to 6.04) on a 0-100 scale; borage not pooled (-10.48 to 6.04) Methods, limits and funding Outcomes: Global eczema symptoms, adverse effects Limitations: Short trials; long-term effects not studied; the bleeding and long-term warnings concern evening primrose oil. Funding and conflicts: The author's earlier research was funded by an evening primrose oil maker (Efamol); a different author handled that trial Read: abstract
Cameron 2011 Herbal therapy for treating rheumatoid arthritis Cochrane Database Syst Rev · PMID 21328257 · DOI	Systematic review	22 People with rheumatoid arthritis (22 herbal trials; 7 on GLA oils)	GLA oils: evening primrose, borage or blackcurrant seed oil (pooled) seed oil not given here not given here · vs see results	Seven studies of GLA oils: lower pain (mean difference -32.83 on a 100-point scale) and less disability (-15.75%). Adverse events were 20% with GLA versus 3% with placebo, not statistically different. Authors: moderate evidence that GLA oils give some relief. Pain mean difference -32.83 (95% CI -56.25 to -9.42); disability -15.75% (-27.06 to -4.44); adverse events RR 4.24 (0.78 to 22.99) (pain -56.25 to -9.42) Methods, limits and funding Outcomes: Pain, disability, adverse events Limitations: Three oils pooled, borage not separated in the abstract; many trials had design flaws. Funding and conflicts: None known Read: abstract
Dante 2011 Herbal treatments for alleviating premenstrual symptoms: a systematic review J Psychosom Obstet Gynaecol · PMID 21171936 · DOI	Systematic review	Women with PMS or PMDD (10 trials)	Herbal remedies including evening primrose oil various not given here not given here · vs see results	Chasteberry was the best studied and beat placebo. Evening primrose oil and St John's wort did not differ from placebo. Borage was not among the remedies named in the abstract. Evening primrose oil: no difference from placebo Methods, limits and funding Outcomes: PMS symptoms Limitations: Evening primrose oil is a different plant; borage was not tested. Funding and conflicts: Not stated in abstract Read: abstract
EFSA Panel on Contaminants in the Food Chain (Knutsen HK et al.) 2017 Risks for human health related to the presence of pyrrolizidine alkaloids in honey, tea, herbal infusions and food supplements EFSA Journal · PMID 32625569 · DOI	Regulatory opinion regulator document	General population	1,2-unsaturated pyrrolizidine alkaloids in foods any not given here not given here · vs see results	Possible concern for human health, especially for frequent and high consumers of tea and herbal infusions. Food supplements based on PA-producing plants could give exposures too close to doses that cause severe short-term toxicity. A reference point of 237 micrograms per kg body weight a day was set for cancer risk. Reference point 237 micrograms/kg body weight/day for carcinogenicity Methods, limits and funding Outcomes: Cancer and acute toxicity risk Limitations: General PA opinion; not about borage specifically. Funding and conflicts: EU agency Read: abstract
European Medicines Agency 2014 Public statement on the use of herbal medicinal products containing toxic, unsaturated pyrrolizidine alkaloids (PAs) European Medicines Agency · Source	Regulatory regulator document	Regulatory statement	Herbal products containing unsaturated PAs any plant part not given here not given here · vs see results	Summary from our earlier notes (document not re-opened): the EU herbal committee says liver toxicity of the toxic PAs is established, children are especially vulnerable, and it names borage only as a salad plant. It is not a borage monograph. not applicable Methods, limits and funding Outcomes: Regulatory position on PAs Limitations: Summary-only here. Not specific to borage seed oil. Funding and conflicts: Regulator Read: summary-only
GBIF 2026 GBIF species match: Borago officinalis (key 2926110) GBIF · Source	Registry or database plant	not applicable	Borage (Borago officinalis) whole plant not given here not given here · vs see results	GBIF API on 2026-10-25: Borago officinalis L. is ACCEPTED (exact match), family Boraginaceae, order Boraginales. not applicable Methods, limits and funding Outcomes: see results Limitations: Taxonomy database record; synonyms not listed in the match call. Funding and conflicts: not applicable Read: full-text
Iraci 2023 Of mixed vegetables and cardiac arrhythmias - Digitalis purpurea confused with Borago officinalis: A case series of accidental digitoxin intoxications J Cardiol Cases · PMID 37521578 · DOI	Case series	2 Two women, aged 35 and 60	Foxglove leaves eaten in place of borage leaf not given here not given here · vs see results	Severe vomiting and nausea about 2 hours after eating a home-made vegetable dish, with ECG changes and high digitoxin levels. One had a heart block and a slow heart rate. Both recovered fully. 2 cases, both recovered Methods, limits and funding Outcomes: Poisoning signs and heart rhythm Limitations: Two cases; identification of the plant was presumed. Funding and conflicts: No competing interests declared Read: abstract
Jamal 1990 The effect of gamma-linolenic acid on human diabetic peripheral neuropathy: a double-blind placebo-controlled trial Diabet Med · PMID 2159860 · DOI	Controlled trial	22 22 people with distal diabetic polyneuropathy	Gamma-linolenic acid vs placebo for 6 months (source oil not stated in abstract) oil (source not stated) not given here not given here · vs see results	Symptom scores and several nerve measures improved significantly more than with placebo. Symptom scores P less than 0.001 vs placebo Methods, limits and funding Outcomes: Symptom score, nerve conduction, temperature thresholds Limitations: Very small, 1990; source of the GLA not stated in the abstract. Funding and conflicts: Not stated in abstract Read: abstract
Keen 1993 Treatment of diabetic neuropathy with gamma-linolenic acid. The gamma-Linolenic Acid Multicenter Trial Group Diabetes Care · PMID 8380765 · DOI	Controlled trial	111 111 people with mild diabetic neuropathy, 7 centres	Gamma-linolenic acid vs placebo for 1 year (source oil not stated in abstract) oil (source not stated) not given here not given here · vs see results	All 16 measures moved in a more favourable direction with GLA than placebo, and 13 differences were statistically significant. Treatment worked better in people with better-controlled diabetes. 13 of 16 parameters significantly better than placebo Methods, limits and funding Outcomes: Nerve conduction and sensation measures Limitations: Source of the GLA not stated in the abstract, so borage is not confirmed. 1993 trial. Funding and conflicts: Not stated in abstract Read: abstract
Leventhal 1993 Treatment of rheumatoid arthritis with gammalinolenic acid Ann Intern Med · PMID 8214997 · DOI	Controlled trial	37 37 adults with rheumatoid arthritis and active joint inflammation	Borage seed oil rich in GLA vs cottonseed oil placebo (24 weeks) seed oil not given here not given here · vs see results	Tender joint count fell 36% and swollen joint count 28% with the borage seed oil; the placebo group showed no significant improvement. No one stopped for side effects. Tender joint score -45%, swollen joint score -41% vs no significant change on placebo Methods, limits and funding Outcomes: Tender and swollen joints, global assessments Limitations: Small, single-centre, 1993, 24 weeks; the authors say it should not be viewed as therapy. Funding and conflicts: Not stated in abstract Read: abstract
Martinez 2013 Methemoglobinemia induced by vegetable intake in infants in northern Spain J Pediatr Gastroenterol Nutr · PMID 23287806 · DOI	Case control	156 78 infants with diet-induced methemoglobinemia and 78 matched controls	Borage and chard purees leaf and stem purees not given here not given here · vs see results	Borage purees were linked to the infant blood disorder (odds ratio 5.2, 95% CI 1.1 to 24.6). Purees stored more than 24 hours and breastfeeding were also linked. Borage had very high nitrate levels (average 3968 mg per kg, 15 samples). OR 5.2 (1.1 to 24.6) for borage; nitrate mean 3968 mg/kg (1.1 to 24.6) Methods, limits and funding Outcomes: Methemoglobinemia, nitrate levels Limitations: Observational; wide confidence interval; 23 years of cases in one area. Funding and conflicts: Not stated in abstract Read: abstract
Memorial Sloan Kettering Cancer Center 2026 Borage: Purported Benefits, Side Effects & More (About Herbs) Memorial Sloan Kettering Cancer Center · Source	Consumer reference plant/human	Patients and clinicians (reference text)	Borage plant and borage seed oil seed oil; leaf and flower not given here not given here · vs see results	Page re-read 2026-10-25. Says borage seed oil is rich in gamma-linolenic acid (GLA); support for rheumatoid arthritis is moderate, two trials do not support eczema use, one small asthma extract trial; oil should be certified free of unsaturated pyrrolizidine alkaloids (0.5 to 1 microgram per gram; German agency advice of no more than 1 microgram a day, quoted second-hand); avoid in pregnancy, breastfeeding and liver disease; one seizure case report, one near-fatal foxglove mix-up, infant methemoglobinemia; cautions with liver-toxic drugs, blood thinners and NSAIDs. Consumer-oncology reference summary; many statements rest on sources we did not open Methods, limits and funding Outcomes: Uses, side effects, warnings, interactions Limitations: A reference page, not a primary study; numbered citations on the page were not opened. Funding and conflicts: None stated; non-commercial hospital reference Read: full-text
Mirsadraee 2016 Effect of Borago Officinalis Extract on Moderate Persistent Asthma: A Phase two Randomized, Double Blind, Placebo-Controlled Clinical Trial Tanaffos · PMID 28210282	Controlled trial	38 38 people with moderate persistent asthma	Borage liquid extract vs matched placebo for 1 month not given here not given here · vs see results	Symptoms and asthma control scores improved significantly (ACT 10.8 to 15.4); flare-ups fell from 3.6 to 2 a month. Lung function, exhaled nitric oxide and sputum cells did not change. Asthma control test +4.6 points; flare-ups per month 3.6 to 2 Methods, limits and funding Outcomes: Asthma control test, symptoms, flare-ups, lung tests, airway inflammation Limitations: Small, one month, one centre; extract not described; clinical scores improved but measured inflammation did not. Funding and conflicts: Not stated in abstract Read: abstract
NIH National Library of Medicine 2026 PubChem name lookups for borage compounds PubChem · Source	Registry or database compounds	not applicable	Gamma-linolenic acid, amabiline, lycopsamine, intermedine, supinine, nitrate seed oil; leaf and flower not given here not given here · vs see results	PubChem API on 2026-10-25 returned CID 5280933 (gamma-linolenic acid), 442706 (amabiline name lookup), 107938 (lycopsamine), 114843 (intermedine), 108053 (supinine), 943 (nitrate). not applicable Methods, limits and funding Outcomes: see results Limitations: Identifier lookup only; says nothing about safety or amounts. Funding and conflicts: not applicable Read: full-text
Negroni 2019 A Case Report of Accidental Intoxication following Ingestion of Foxglove Confused with Borage: High Digoxinemia without Major Complications Case Rep Cardiol · PMID 31871798 · DOI	Case report	1 A 55-year-old woman	Foxglove leaves eaten after being bought as borage leaf not given here not given here · vs see results	Weakness, nausea and vomiting after a savory pie made with five leaves from a plant bought as borage. She had episodes of heart block on monitoring and recovered. 1 case, recovered Methods, limits and funding Outcomes: Poisoning signs and heart rhythm Limitations: Single case report. Funding and conflicts: No conflicts declared Read: abstract
Peloso 2023 Pyrrolizidine Alkaloids in Food on the Italian Market Molecules · PMID 37513219 · DOI	Food survey food samples	602 602 Italian food samples including fresh borage leaves	Foods tested for PAs leaf (fresh borage leaves among the samples) not given here not given here · vs see results	42% of samples contained PAs and 14% exceeded EU limits (Regulation 2023/915, which replaced 2020/2040); dried herbs and teas were most often contaminated. The abstract does not give a separate figure for borage. 42% contaminated, 14% over EU limits across all foods Methods, limits and funding Outcomes: PA contamination vs EU limits Limitations: No borage-only figure in the abstract. Funding and conflicts: No conflict of interest declared Read: abstract
Puri 2007 The safety of evening primrose oil in epilepsy Prostaglandins Leukot Essent Fatty Acids · PMID 17764919 · DOI	Narrative review human/animal	People with epilepsy (re-examination of early reports)	Evening primrose oil (not borage) seed oil not given here not given here · vs see results	The author re-examined two early-1980s reports and judged the link between evening primrose oil and seizures spurious, and suggested removing seizures as a listed side effect. Not applicable Methods, limits and funding Outcomes: Seizure risk Limitations: Review of one oil by one author; not about borage oil; does not settle the borage seizure case report. Funding and conflicts: Not stated in abstract Read: abstract
Reed 2014 Treatment of rheumatoid arthritis with marine and botanical oils: an 18-month, randomized, and double-blind trial Evid Based Complement Alternat Med · PMID 24803948 · DOI	Controlled trial	Adults with rheumatoid arthritis	Borage seed oil alone, fish oil alone, or both (18 months) seed oil not given here not given here · vs see results	No significant differences between the three groups. All three improved at 9 months (DAS28 fall of 1.33 in the borage group) and the gains lasted to 18 months. There was no placebo arm; improvement was similar to matched registry patients. DAS28 change at 9 months: borage -1.33 (-1.83 to -0.84) (-1.83 to -0.84) Methods, limits and funding Outcomes: Disease activity (DAS28, CDAI), RA medicine use Limitations: No placebo group, so improvement cannot be credited to the oil. Funding and conflicts: Not stated in abstract Read: abstract
Sattler 2023 Pyrrolizidine alkaloids in borage (Borago officinalis): Comprehensive profiling and development of a validated LC-MS/MS method for quantification Talanta · PMID 36924638 · DOI	Laboratory analysis plant	Borage leaf material	Borage leaf PA profile leaf not given here not given here · vs see results	Borage leaf contains pyrrolizidine alkaloids and their N-oxides. The EU set maximum levels for PAs in borage (in force since 1 July 2022). In the samples tested, the acetylated forms, which the EU does not regulate, were among the most abundant PAs. Not a trial; no exposure numbers in the abstract Methods, limits and funding Outcomes: PA and PA N-oxide profile Limitations: Leaf material only; says nothing about seed oil or about illness in people. Funding and conflicts: No competing interests declared Read: abstract
Takwale 2003 Efficacy and tolerability of borage oil in adults and children with atopic eczema: randomised, double blind, placebo controlled, parallel group trial BMJ · PMID 14670885 · DOI	Controlled trial	140 140 evaluable adults and children (69 children) with atopic eczema	Borage oil capsules vs placebo, 12 weeks seed oil not given here not given here · vs see results	The score fell from 30 to 27 on borage oil and from 28 to 23 on placebo, a difference of 1.4 points in favour of placebo (P = 0.45). No other measure differed. Treatments were well tolerated. Difference 1.4 points favouring placebo (95% CI -2.2 to 5.0) (-2.2 to 5.0) Methods, limits and funding Outcomes: Eczema sign score (SASSAD) and symptoms Limitations: One hospital; 12 weeks. Funding and conflicts: Not stated in abstract Read: abstract
Teschke 2021 Metabolic Toxification of 1,2-Unsaturated Pyrrolizidine Alkaloids Causes Human Hepatic Sinusoidal Obstruction Syndrome: The Update Int J Mol Sci · PMID 34638760 · DOI	Narrative review human/mechanism	Humans (review)	1,2-unsaturated PAs any not given here not given here · vs see results	PAs with the 1,2 double bond are changed by liver enzymes (CYP2A6, 3A4, 3A5) into reactive forms that injure liver blood-vessel lining cells and can cause hepatic sinusoidal obstruction syndrome: weight gain from fluid, liver failure, transplant or death, even after intake stops. Not applicable Methods, limits and funding Outcomes: Mechanism of liver injury Limitations: Review; not borage-specific; mostly about large accidental intakes. Funding and conflicts: No conflict of interest declared Read: abstract
Whelan 2009 Herbs, vitamins and minerals in the treatment of premenstrual syndrome: a systematic review Can J Clin Pharmacol · PMID 19923637	Systematic review	Women with PMS or PMDD	Herbs, vitamins and minerals including evening primrose oil various not given here not given here · vs see results	Only calcium had good-quality support. No evidence of benefit with evening primrose oil was found. Borage was not named in the abstract. Evening primrose oil: no benefit found Methods, limits and funding Outcomes: PMS symptoms Limitations: Evening primrose oil is a different plant; borage was not tested. Funding and conflicts: Not stated in abstract Read: abstract
Won 2020 gamma-Linolenic Acid versus alpha-Lipoic Acid for Treating Painful Diabetic Neuropathy in Adults: A 12-Week, Double-Placebo, Randomized, Noninferiority Trial Diabetes Metab J · PMID 31701699 · DOI	Controlled trial	100 100 adults with type 2 diabetes and painful neuropathy (73 completed)	Gamma-linolenic acid vs alpha-lipoic acid, 12 weeks (source oil not stated in abstract) oil (source not stated) not given here not given here · vs see results	Both groups improved from baseline. GLA was noninferior to alpha-lipoic acid on pain scores but not on the total symptom score. No serious adverse events. There was no placebo-only group. Pain difference -0.65 (95% CI -1.526 to 0.213) (-1.526 to 0.213) Methods, limits and funding Outcomes: Pain on a visual scale, total symptom score Limitations: Comparison with another supplement, not placebo; 27% dropped out; GLA source not stated. Funding and conflicts: Supported by a grant from Dalin Biotech Korea (a company); funders had no role in design or analysis Read: abstract
Yang 2019 Pyrrolizidine alkaloids-induced hepatic sinusoidal obstruction syndrome: Pathogenesis, clinical manifestations, diagnosis, treatment, and outcomes World J Gastroenterol · PMID 31391770 · DOI	Narrative review	Humans (review)	PA-induced liver injury any not given here not given here · vs see results	PA liver injury typically shows as painful liver enlargement, fluid in the abdomen and jaundice, with abnormal liver tests. Treatments range from fluid control to liver transplant. Not applicable Methods, limits and funding Outcomes: Symptoms and course Limitations: Review; not borage-specific. Funding and conflicts: No conflict of interest declared Read: abstract

What we do not know yet

No recent placebo-controlled trial of borage seed oil alone in arthritis.
Diabetic nerve trials do not name borage as the GLA source.
No borage-specific human liver case data; PA content of finished oils not described in the sources read.
No interaction study with anti-seizure drugs or blood thinners for borage.
No EMA monograph, NCCIH, LiverTox or LactMed entry found; EU and German agency documents not read.

How new research is added

New studies are found by a saved PubMed search and are placed in a review queue. A person reads each one before it changes any evidence rating or safety wording.

Lab markers

What changed in blood or saliva tests during trials. A changed number is not the same as feeling or being healthier, and we give no "normal ranges".

Other 7

▼

Tender and swollen joint count

Lowered in trials

How many joints are sore to press or puffy on examination.

Low certainty

Results and what it means

Why it was measured

The standard bedside measure of active arthritis.

Statistical vs. clinical meaning

Small single trial; supports the oil but is not proof.

Safety implication

None specific.

Study	Result	People	Product
¹⁰	Tender joints down 36% and swollen joints down 28% on borage seed oil; no significant change on placebo.	37 adults with active RA	Borage seed oil vs cottonseed oil, 24 weeks

¹⁰

▼

Disease activity score (DAS28)

Lowered in trials

A score combining joint counts, blood inflammation and the patient's rating; lower is better.

Low certainty

Results and what it means

Why it was measured

Standard way to follow rheumatoid arthritis over time.

Statistical vs. clinical meaning

Improvement cannot be credited to the oil without a placebo arm.

Safety implication

None specific.

Study	Result	People	Product
¹⁸	DAS28 fell 1.33 at 9 months on borage oil alone, with similar falls on fish oil and the combination; no placebo group.	Adults with RA	Borage seed oil, fish oil or both, 18 months

¹⁸

●

Eczema sign score (SASSAD)

Unchanged in trials

A doctor's score of redness, scaling and other eczema signs across six body areas.

Evidence against

Results and what it means

Why it was measured

The main outcome of the BMJ eczema trial.

Statistical vs. clinical meaning

No benefit over placebo.

Safety implication

None specific.

Study	Result	People	Product
²⁰	Score fell 30 to 27 on borage oil and 28 to 23 on placebo; difference 1.4 points favouring placebo (P = 0.45).	140 adults and children with eczema	Borage oil capsules vs placebo, 12 weeks

²⁰

●

Nerve conduction and symptom score

Improved in trials

How fast nerves carry signals, plus a symptom score for numbness and pain.

Very low certainty

Results and what it means

Why it was measured

Used to follow diabetic neuropathy.

Statistical vs. clinical meaning

Borage not confirmed as the GLA source; small old trials.

Safety implication

None specific.

Study	Result	People	Product
⁹	13 of 16 nerve measures significantly better than placebo over 1 year.	111 people with mild diabetic neuropathy	GLA (source not stated)
⁸	Symptom scores and several nerve conduction values better than placebo at 6 months.	22 people with diabetic neuropathy	GLA (source not stated)

^{9 8}

▼

Pain score in diabetic neuropathy

Lowered in trials

Self-rated pain on a 0-10 style line.

Very low certainty

Results and what it means

Why it was measured

Pain is the symptom people notice most.

Statistical vs. clinical meaning

No placebo group, company-funded, 27% dropout.

Safety implication

None specific.

Study	Result	People	Product
²³	Pain fell from baseline in both GLA and alpha-lipoic acid groups; GLA noninferior on pain (difference -0.65, 95% CI -1.526 to 0.213).	100 adults with painful diabetic neuropathy, 73 finished	GLA (source not stated) vs alpha-lipoic acid, 12 weeks

²³

●

Asthma control test score

Improved in trials

A short questionnaire on symptoms and rescue-inhaler use; higher is better controlled.

Very low certainty

Results and what it means

Why it was measured

The main outcome of the borage asthma trial.

Statistical vs. clinical meaning

Small trial; extract unspecified.

Safety implication

See safety tab for leaf and flower alkaloids.

Study	Result	People	Product
¹³	Score rose 10.8 to 15.4 in the borage group after 1 month.	38 adults with moderate asthma	Borage liquid extract vs placebo

¹³

●

Exhaled nitric oxide and sputum eosinophils

Unchanged in trials

Breath and mucus tests that reflect inflammation in the airways.

Very low certainty

Results and what it means

Why it was measured

Shows whether symptoms improved because inflammation fell.

Statistical vs. clinical meaning

Authors say the extract did not suppress asthma inflammation.

Safety implication

None.

Study	Result	People	Product
¹³	No significant change in exhaled nitric oxide or sputum eosinophils and neutrophils.	38 adults with moderate asthma	Borage liquid extract vs placebo

¹³

Safety

What studies and official sources say about side effects, medicines and who should check first. Each point shows what kind of evidence it comes from.

Main points to know 6

Leaf, flower and tea contain liver-toxic pyrrolizidine alkaloids. Borage leaf contains pyrrolizidine alkaloids (PAs) and their N-oxides. The 1,2-unsaturated type is changed by liver enzymes into reactive forms that can block the small liver blood vessels (hepatic sinusoidal obstruction syndrome), with swelling, fluid, liver failure or death. The EU set maximum PA levels for borage as a food from July 2022, and the unregulated acetylated PAs were abundant in one lab analysis. We found no published case of liver injury from borage itself, but the risk rises with length of use and total amount, so regular leaf, flower or tea use is best avoided. Laboratory analysis ^{19 21 24 4 5 12}
Seed oil is different from leaf tea only if it is certified PA-free. The seed oil is the source of GLA. The hospital reference we read says oil may carry small amounts of PAs and should be bought only if certified free of unsaturated PAs (no more than 0.5 to 1 microgram per gram), and quotes a German agency limit of 1 microgram a day. We could not open the German agency or EU documents. If a label has no PA-free statement, we cannot tell what is in it. Consumer reference ¹²
Pregnancy and breastfeeding: avoid. The hospital reference says to avoid borage in pregnancy and breastfeeding (possible birth defects, premature labour, infant blood disorder). The EU committee says children are especially vulnerable to PAs. We did not open the studies behind the birth-defect statement. Consumer reference ^{12 5}
Seizures: one case report, and the evidence on GLA oils conflicts. The hospital reference lists one case of continuous seizure activity in a healthy 41-year-old woman after a week of borage oil. A 2007 review of evening primrose oil (a different GLA oil) judged its link to seizures spurious. We found no trial in people with epilepsy or on anti-seizure medicine, so the question is open. People with epilepsy should ask their prescriber before using any GLA oil. Case report ^{12 17}
Foxglove looks like borage and has caused poisonings. Foxglove leaves were eaten by mistake for borage in two published reports (three women), causing vomiting, abnormal heart rhythm and heart block; all recovered. Never eat wild-picked leaves you cannot identify with certainty. Case series ^{7 15}
Very high nitrates: infants. A Spanish case-control study of 78 infants linked borage purees to methemoglobinemia (blood that carries less oxygen). Borage averaged about 4,000 mg of nitrate per kg. Purees stored over 24 hours and breastfeeding were also associated. Case control ¹¹

Side effects 5

Effect	How often	Severity and recovery	Evidence
Headache, belly pain, nausea, belching, loose stools In the eczema trials, side effects were mild, mostly stomach-related and similar to placebo.	Not established	mild Resolves on stopping (as reported)	Systematic review ^{12 1}
Liver injury (pyrrolizidine alkaloids) Documented for PA-containing plants in general. No borage-specific human case was found. Signs: painful liver, belly swelling, jaundice.	Not established	severe May not reverse; can be fatal	Narrative review ^{21 24 12}
Seizures One reported case after a week of borage oil; cause not established.	Not established	severe Unknown	Case report ¹²
Digitalis poisoning from a plant mix-up Not an effect of borage; caused by foxglove eaten instead.	Not established	severe Recovered in reported cases	Case series ^{7 15}
Infant methemoglobinemia From nitrate in borage purees; stored purees raised the odds.	Not established	severe Treatable in hospital	Case control ¹¹

Liver

Borage leaf and flower contain liver-toxic PAs, and the EU herbal committee says PA liver toxicity is established. Seed oil is said to be safer only when certified PA-free. We did not find a published human liver-injury case for borage or a borage entry in NIH LiverTox, so no case count is claimed. A regulatory risk opinion said food supplements made from PA-producing plants could give exposures too close to doses that cause severe short-term toxicity. ^{21 24 4 5 12}

How often. Unknown for borage; PA liver injury has been reported mainly after large accidental intakes of other plants.

Monitoring. No official routine test advice found. Yellow skin or eyes, dark urine, pale stools, belly swelling or pain should be checked by a doctor promptly.

This is a safety connection. It is not evidence that Borage supports liver health.

Medicines and supplements 5

With	Why it matters	Importance	Evidence	What to do
Liver-toxic medicines (for example anabolic steroids, ketoconazole) prescription	Added liver load from PAs, as reasoned by the hospital reference.	Theoretical; no study.	Consumer reference Very low certainty ¹²	Avoid combining unless your doctor says otherwise.
Blood thinners (warfarin) and antiplatelet drugs prescription	One small study of GLA from evening primrose oil, quoted second-hand, found longer bleeding times in 9 of 12 people. Borage oil has more GLA, so the reference extends the caution. The Cochrane eczema review also notes a study where evening primrose oil may raise bleeding on warfarin.	Possible; by extension, not tested for borage.	Other observational Very low certainty ^{12 1}	Tell your doctor or pharmacist before use; ask about extra blood tests if on warfarin.
Anti-seizure medicines and seizure risk prescription	One seizure case report with borage oil; a 2007 review argues the evening primrose oil link is spurious. No interaction study with anti-seizure drugs found.	Unknown; evidence conflicts and is thin.	Case report Very low certainty ^{12 17}	If you have epilepsy or take anti-seizure medicine, ask your prescriber first.
NSAIDs (aspirin, ibuprofen, COX-2 inhibitors) over-the-counter	The reference says NSAIDs could in theory reduce the effects of borage oil.	Theoretical.	Pharmacological theory Very low certainty ¹²	No action needed beyond telling your pharmacist.
Medicines broken down by liver enzymes CYP3A4 or CYP2A6 prescription	PAs are activated by these enzymes. Whether borage changes drug levels was not tested.	Not tested.	Pharmacological theory Very low certainty ²¹	Mention borage products to your pharmacist.

Who should check first 5

Situation	Advice	Evidence
Pregnancy and breastfeeding	Avoid. Reference advice, with possible birth defects, premature labour and infant blood disorder; PAs are a concern for fetuses and children.	Consumer reference ^{12 5}
Children and infants	Avoid leaf, flower and tea. Do not feed borage puree to infants (nitrates).	Case control ^{11 5}
Liver disease, or taking liver-toxic medicines	Avoid.	Consumer reference ¹²
Epilepsy or seizure history	Ask a prescriber first; evidence is thin and conflicting.	Case report ^{12 17}
Bleeding disorders, warfarin, surgery	Ask a doctor first (bleeding caution by extension from GLA oils).	Consumer reference ^{12 1}

Overdose, product quality and regulators

Summary. The toxic concern is the PAs (amabiline, supinine, lycopsamine, intermedine and their acetyl forms, per the hospital reference), which are liver toxins and, in animals, genotoxic carcinogens. EFSA set a reference point of 237 micrograms per kg body weight a day for cancer risk across PAs.

Animal. EFSA based its cancer reference point on animal data for 1,2-unsaturated PAs (abstract read); no borage-specific animal study was read.

High dose or long term signals. Risk rises with length of exposure and total amount (hospital reference). Long-term GLA oil safety was not studied in the eczema trials.

Overdose response. Liver blood-vessel blockage with fluid retention, liver failure, death in the severe PA cases reported for other plants.

Genotoxicity. 1,2-unsaturated PAs are genotoxic carcinogens in animal studies (EFSA; Italian food survey abstract).

Toxic plant parts. Leaf and flower carry PAs (lab analysis of leaves). Seed oil: PAs possible in small amounts unless certified free (second-hand).

PAs are the main contamination issue; a 2019-2022 Italian survey of 602 foods (including fresh borage leaves) found 42% with PAs and 14% over EU limits, mostly dried herbs and teas. Look-alike foxglove and high nitrates are separate hazards.

PAs in leaf. Borage leaf contains PAs; unregulated acetylated forms were prominent in one analysis. ¹⁹

Food survey. 42% of 602 Italian foods contained PAs and 14% exceeded EU limits; the abstract gives no separate borage figure. ¹⁶

Oil certification. The reference advises only oil certified free of unsaturated PAs. ¹²

Nitrates. Borage averaged about 3,968 mg nitrate per kg in 15 samples. ¹¹

Body	Action	Date
European Food Safety Authority (EFSA) (EU)	2017 opinion: possible concern for frequent tea and herbal infusion drinkers; supplements from PA plants close to severe-toxicity doses; reference point 237 micrograms per kg per day. source	2017

When to get medical help

Yellow skin or eyes, dark urine, pale stools, painful or swollen belly
Vomiting with a slow or irregular heartbeat after eating leaves you cannot identify (possible foxglove)
A seizure
Blue-grey lips or skin, fast breathing or sleepiness in an infant after borage puree
Unusual bleeding or bruising if on blood thinners

Stop the supplement and contact a clinician or emergency service. In the US, Poison Control is 1-800-222-1222.

General information only, not medical advice. Ask a clinician or pharmacist before use, especially with medicines, in pregnancy or with a health condition. See the Natural Remedy Safety guide.

Frequently asked questions

What is borage?

A garden herb with blue star-shaped flowers. Its leaves and flowers are eaten or made into tea, and its seeds are pressed into an oil rich in GLA.

¹²

What is GLA?

Gamma-linolenic acid, an omega-6 fat that the body can turn into substances that may reduce inflammation. Borage seed oil is a rich source.

¹²

Does borage oil help rheumatoid arthritis?

A Cochrane review found moderate evidence that GLA oils give some relief, but for borage itself the evidence is one small 1993 placebo trial and a 2014 trial with no placebo. It does not replace arthritis medicine.

^{2 10 18}

Does borage oil help eczema?

No. Borage oil by mouth did not beat placebo in a Cochrane review or a BMJ trial.

^{1 20}

Does it help diabetic nerve pain?

GLA trials were positive, but their abstracts do not say the GLA came from borage, so we cannot say borage helps.

^{9 8 23}

Is borage tea safe?

Regular use is best avoided. The leaf and flower contain pyrrolizidine alkaloids that can injure the liver, and the EU has set limits for them in borage.

^{19 4 5}

What does PA-free mean?

The oil is certified to contain no more than trace amounts of unsaturated pyrrolizidine alkaloids. The hospital reference we read advises only certified oil. We could not open the original agency limits.

¹²

Can I take it while pregnant or breastfeeding?

No. The reference advises avoiding it, and PAs are a particular concern for unborn babies and children.

^{12 5}

Does borage oil cause seizures or interact with epilepsy medicine?

One seizure case report exists with borage oil. A 2007 review judged the link between evening primrose oil and seizures spurious. No trial in people with epilepsy was found, so ask your prescriber.

^{12 17}

Can borage affect blood thinners?

Possibly. GLA oils may increase bleeding time in a small study of evening primrose oil, and the caution is extended to borage. Tell your pharmacist if you take warfarin.

^{12 1}

Can borage be confused with a poisonous plant?

Yes. Foxglove leaves have been eaten by mistake for borage and caused heart-rhythm problems. Never eat leaves you cannot identify with certainty.

^{7 15}

Is borage puree safe for babies?

Borage is very high in nitrates and was linked to a blood disorder in Spanish infants, especially with purees stored over a day. Avoid it for babies.

¹¹

Explore by health goal

Joint Support Healthy Skin Blood Sugar Balance Breathing Support Womens Wellness Menopause Support Mood Balance Blood Pressure Support Cholesterol Balance Brain Health Bone Health

Editorial transparency

Written by: AlternativeMed editorial team. Research compiled with AI assistance, then checked against PubMed records.
Medical review: None. This page has not been reviewed by a physician or pharmacist.
First published: 2026-10-25
Last substantive update: 2026-10-25
How we research: We built on our existing summary, read the Memorial Sloan Kettering page, read PubMed abstracts, and checked GBIF and PubChem. EMA, EU regulation and German agency figures are second-hand and flagged. Most studies were read as abstracts only.
How we rate evidence: See "How we rate evidence" in the Evidence tab.
Independence: No product is promoted on this page. No affiliate links are shown. If one is added later it will be labelled and kept apart from the evidence.

Sources 24

Each source was opened and checked. Where we read only the abstract, it says so. The small numbers in the text point here.

Bamford JT, Ray S, Musekiwa A, van Gool C, Humphreys R, Ernst E. Oral evening primrose oil and borage oil for eczema. Cochrane Database Syst Rev. 2013. PMID 23633319. DOI 10.1002/14651858.CD004416.pub2. (abstract)
Cameron M, Gagnier JJ, Chrubasik S. Herbal therapy for treating rheumatoid arthritis. Cochrane Database Syst Rev. 2011. PMID 21328257. DOI 10.1002/14651858.CD002948.pub2. (abstract)
Dante G, Facchinetti F. Herbal treatments for alleviating premenstrual symptoms: a systematic review. J Psychosom Obstet Gynaecol. 2011. PMID 21171936. DOI 10.3109/0167482X.2010.538102. (abstract)
EFSA Panel on Contaminants in the Food Chain (Knutsen HK et al.). Risks for human health related to the presence of pyrrolizidine alkaloids in honey, tea, herbal infusions and food supplements. EFSA Journal. 2017. PMID 32625569. DOI 10.2903/j.efsa.2017.4908. (abstract)
European Medicines Agency, HMPC. Public statement on the use of herbal medicinal products containing toxic, unsaturated pyrrolizidine alkaloids (PAs). European Medicines Agency. 2014. (summary-only)
GBIF. GBIF species match: Borago officinalis (key 2926110). GBIF. 2026. (full-text)
Iraci F, Herdeg C, Holzwarth M, Storz MA. Of mixed vegetables and cardiac arrhythmias - Digitalis purpurea confused with Borago officinalis: A case series of accidental digitoxin intoxications. J Cardiol Cases. 2023. PMID 37521578. DOI 10.1016/j.jccase.2023.04.007. (abstract)
Jamal GA, Carmichael H. The effect of gamma-linolenic acid on human diabetic peripheral neuropathy: a double-blind placebo-controlled trial. Diabet Med. 1990. PMID 2159860. DOI 10.1111/j.1464-5491.1990.tb01397.x. (abstract)
Keen H, Payan J, Allawi J, et al.. Treatment of diabetic neuropathy with gamma-linolenic acid. The gamma-Linolenic Acid Multicenter Trial Group. Diabetes Care. 1993. PMID 8380765. DOI 10.2337/diacare.16.1.8. (abstract)
Leventhal LJ, Boyce EG, Zurier RB. Treatment of rheumatoid arthritis with gammalinolenic acid. Ann Intern Med. 1993. PMID 8214997. DOI 10.7326/0003-4819-119-9-199311010-00001. (abstract)
Martinez A, Sanchez-Valverde F, Gil F, et al.. Methemoglobinemia induced by vegetable intake in infants in northern Spain. J Pediatr Gastroenterol Nutr. 2013. PMID 23287806. DOI 10.1097/MPG.0b013e3182849d2b. (abstract)
Memorial Sloan Kettering Cancer Center. Borage: Purported Benefits, Side Effects & More (About Herbs). Memorial Sloan Kettering Cancer Center. 2026. (full-text)
Mirsadraee M, Khashkhashi Moghaddam S, Saeedi P, Ghaffari S. Effect of Borago Officinalis Extract on Moderate Persistent Asthma: A Phase two Randomized, Double Blind, Placebo-Controlled Clinical Trial. Tanaffos. 2016. PMID 28210282. (abstract)
NIH National Library of Medicine. PubChem name lookups for borage compounds. PubChem. 2026. (full-text)
Negroni MS, Marengo A, Caruso D, et al.. A Case Report of Accidental Intoxication following Ingestion of Foxglove Confused with Borage: High Digoxinemia without Major Complications. Case Rep Cardiol. 2019. PMID 31871798. DOI 10.1155/2019/9707428. (abstract)
Peloso M, Minkoumba Sonfack G, Paduano S, De Martino M, De Santis B, Caprai E. Pyrrolizidine Alkaloids in Food on the Italian Market. Molecules. 2023. PMID 37513219. DOI 10.3390/molecules28145346. (abstract)
Puri BK. The safety of evening primrose oil in epilepsy. Prostaglandins Leukot Essent Fatty Acids. 2007. PMID 17764919. DOI 10.1016/j.plefa.2007.07.003. (abstract)
Reed GW, Leung K, Rossetti RG, Vanbuskirk S, Sharp JT, Zurier RB. Treatment of rheumatoid arthritis with marine and botanical oils: an 18-month, randomized, and double-blind trial. Evid Based Complement Alternat Med. 2014. PMID 24803948. DOI 10.1155/2014/857456. (abstract)
Sattler M, Müller V, Bunzel D, Kulling SE, Soukup ST. Pyrrolizidine alkaloids in borage (Borago officinalis): Comprehensive profiling and development of a validated LC-MS/MS method for quantification. Talanta. 2023. PMID 36924638. DOI 10.1016/j.talanta.2023.124425. (abstract)
Takwale A, Tan E, Agarwal S, Barclay G, Ahmed I, Hotchkiss K, Thompson JR, Chapman T, Berth-Jones J. Efficacy and tolerability of borage oil in adults and children with atopic eczema: randomised, double blind, placebo controlled, parallel group trial. BMJ. 2003. PMID 14670885. DOI 10.1136/bmj.327.7428.1385. (abstract)
Teschke R, Vongdala N, Quan NV, Quy TN, Xuan TD. Metabolic Toxification of 1,2-Unsaturated Pyrrolizidine Alkaloids Causes Human Hepatic Sinusoidal Obstruction Syndrome: The Update. Int J Mol Sci. 2021. PMID 34638760. DOI 10.3390/ijms221910419. (abstract)
Whelan AM, Jurgens TM, Naylor H. Herbs, vitamins and minerals in the treatment of premenstrual syndrome: a systematic review. Can J Clin Pharmacol. 2009. PMID 19923637. (abstract)
Won JC, Kwon HS, Moon SS, et al.. gamma-Linolenic Acid versus alpha-Lipoic Acid for Treating Painful Diabetic Neuropathy in Adults: A 12-Week, Double-Placebo, Randomized, Noninferiority Trial. Diabetes Metab J. 2020. PMID 31701699. DOI 10.4093/dmj.2019.0099. (abstract)
Yang XQ, Ye J, Li X, Li Q, Song YH. Pyrrolizidine alkaloids-induced hepatic sinusoidal obstruction syndrome: Pathogenesis, clinical manifestations, diagnosis, treatment, and outcomes. World J Gastroenterol. 2019. PMID 31391770. DOI 10.3748/wjg.v25.i28.3753. (abstract)

Where to buy

Check the label first. Whatever you buy, look for these:

The plant part and the extract type or ratio on the label
A standardized amount of the active compounds, where relevant
A third-party testing mark (USP, NSF or ConsumerLab)
Interactions: ask a pharmacist if you take medicines, are pregnant or have a medical condition

Talk to a pharmacist or doctor before you start any supplement, especially if you take medicines. See the Safety tab for interactions.

Browse Borage on Amazon

Amazon · Several brands to compare

Shop Borage on Amazon

Support our free guides: as an Amazon Associate we earn from qualifying purchases. Full disclosure.

Education, not medical advice. This page is for general information. It does not diagnose or treat any condition and is not a substitute for care from a qualified health professional. Natural does not automatically mean safe. Ask a clinician or pharmacist before changing your care, especially if you take medication, are pregnant or breastfeeding, or have a health condition.