Lavender

Lavandula angustifolia Mill. sits in Lamiaceae, the mint family. NCBI Taxonomy treats it as taxon 39329. IPNI records Philip Miller’s name from the eighth edition of The Gardeners Dictionary (1768) as 449008-1. EMA still lists the older synonym L. officinalis Chaix. It is an evergreen subshrub of dry, sunny Mediterranean ground; NCCIH describes it as native to the region, including France, Spain, Andorra, and Italy. It has narrow grey-green leaves, square young stems, and long bare stalks ending in whorls of small violet two-lipped flowers. The aromatic oil is made in glands on the surface of flowers and leaves. The pharmacopeial drug is the dried flower, which must hold at least 13 ml of essential oil per kilogram, and the medicinal oil is steam-distilled from the flowering tops.

Lavandin, L. × intermedia, is a sterile hybrid of true lavender and spike lavender. It is larger and hardier, can only be propagated from cuttings, and yields far more oil per hectare. One genome study estimated that lavandin accounts for roughly 85% of world lavender-oil production. Its oil is cheaper, more camphoraceous, and has a long history of being sold as true lavender oil. Spike lavender, L. latifolia (the broad-leaved one), is an Iberian species whose oil pairs linalool with large amounts of 1,8-cineole and camphor. L. stoechas, the Spanish, French, or butterfly lavender with showy purple bracts on top of a squat head, is chemically a different animal: its oils are led by fenchone and camphor. EMA’s assessment report warns that L. angustifolia is regularly confused with lavandin and spike lavender, and that flower lots without detailed quality specifications can be mixtures of species.

Common names are a mess. English lavender means L. angustifolia whether or not it grew in England. French lavender can mean L. stoechas, lavandin from Provence, or true lavender from Provence, depending on who is selling; NCCIH even lists French lavender as a common name for L. angustifolia itself. Pharmacopeial lavender oil means L. angustifolia distilled from flowering tops. A cosmetic lavender fragrance can be lavandin or a blend. Spike and stoechas oils are sold for aromatherapy under lavender names too. For anything you rub on skin, put in a bath, or swallow, the Latin binomial and the plant part are the minimum label. Price is a clue: true lavender oil costs more because the plant yields less.

Chemistry is the point. EMA’s assessment report describes true lavender oil as dominated by linalool (roughly 20–50%) and linalyl acetate (25–46%), with smaller amounts of cis-ocimene, terpinen-4-ol, limonene, 1,8-cineole, camphor, lavandulyl acetate, and lavandulol. Linalool is PubChem CID 6549 (C10H18O); linalyl acetate is CID 8294; camphor is CID 2537. The flower also carries rosmarinic and other phenolic acids, the coumarins umbelliferone and herniarin, flavonoids, and up to 13% tannins. Carrasco and colleagues (Planta Medica 2016) measured 37–54% linalool and 21–36% linalyl acetate in Spanish true lavender, against linalool plus 26–32% eucalyptol (1,8-cineole) and 10–18% camphor in spike. Pokajewicz and colleagues (Molecules 2022), growing both species on the same plots, found about 10–13% camphor in lavandin cultivars. The two signature molecules also matter for safety: linalool and linalyl acetate autoxidize in air into skin-sensitizing hydroperoxides. Pure is not the same as fresh.

The name is a small argument. NCCIH repeats the popular derivation from Latin lavare, to wash, and a smell-history reference notes that lavender reportedly scented Roman bathwater. Etymologists are less sure: the Medieval Latin lavendula may come from lividus, bluish, with the washing association attaching later because the plant scented laundry and baths. Either way, the classical record is thinner than gift-shop placards suggest. In the twelfth century Hildegard of Bingen gave lavendula its own chapter in Physica: warm and dry, not worth eating, strong-smelling, deadly to lice if sniffed often, able to clear the eyes and to drive off evil spirits. That is the medieval file in miniature. Lavender was a smell used as medicine, not a dose.

Distillation turned the flower into an oil and then an industry. EMA’s assessment report gathers the early-modern record: distilled lavender water as a sedative nervinum and for toothache and headache, Dodoens in 1608 calling it calming, and the long habit of a few drops of oil on a sugar cube. In England, Ephraim Potter and William Moore began distilling lavender water at Mitcham in Surrey in 1749. Mitcham oil became a premium product until cheaper French oil and suburban housing displaced the fields; the old Potter and Moore distillery passed to W. J. Bush in 1886. In France, Grasse and the lavender growers of Haute-Provence supplied perfumers, British firms among them. The plant’s modern identity as linen, soap, and perfume was built in those stillrooms, not in clinics.

Aromatherapy has a founding story, and lavender is in it. René-Maurice Gattefossé, a Lyon chemist in his family’s perfumery business, was burned in a laboratory explosion in July 1910 and later wrote that a single rinse with lavender essence stopped gas gangrene spreading on his hands. The retellings grew: vats of oil, reflex plunges, scarless miracles. His own account is two sentences. He coined aromathérapie in print in December 1935 and used it as the title of his 1937 book. The regulatory file is quieter. Lavender oil has been marketed as a bath additive since 1976, Germany authorised Silexan capsules in 2009, and EMA’s 2012 monographs settled on traditional use for mild stress and sleep. EMA opened a call for new data for periodic review in 2020. Lavender’s history is a perfume story that acquired a pharmacology chapter late.

EMA did not need a mechanism for a traditional-use monograph and did not claim one. The mechanistic file is mostly Silexan’s. Schuwald, Müller and colleagues (PLoS One 2013) reported that the oil, at nanomolar concentrations they matched to an 80 mg human dose, reduced calcium influx through several voltage-gated calcium channels (N-, P/Q-, and T-type) in synaptosomes, hippocampal neurons, and cell lines. That sounds like pregabalin, and the authors drew the comparison, but Silexan did not bind pregabalin’s α2δ site and its effect was moderate. Several authors were Schwabe employees or consultants. A PET imaging study (Baldinger 2015) summarized by MSKCC reported reduced serotonin-1A receptor binding after Silexan. Linalool is anxiolytic-like in rodents, and MSKCC cites mouse work in which inhaled lavender still worked in animals that could not smell, pointing to serotonergic mechanisms rather than odour alone.

Aromatherapy has a harder mechanism problem. Smell cannot be blinded: participants know when the room smells of lavender, and expectation, attention, massage, and the ritual of a calm room all travel with the scent. MSKCC notes that in massage studies the massage may be doing the work, and a breast-surgery trial it cites found that lavender oil and unscented oil both reduced preoperative anxiety. Endocrine activity is a separate file: Henley 2007 and Ramsey 2019 found weak estrogenic and antiandrogenic activity of lavender oil and some of its components in human cell lines. Whether skin exposure reaches doses that matter in a child is unknown. Oxidation is the third mechanism that matters clinically. Linalool and linalyl acetate turn into sensitizing hydroperoxides on contact with air, so an old bottle is chemically a different product from a fresh one.

Anxiety is where lavender has its best data, and almost all of it is one product. Kasper and colleagues (International Clinical Psychopharmacology 2010) randomized 221 primary-care adults with anxiety disorder not otherwise specified to Silexan 80 mg daily or placebo for ten weeks; Hamilton Anxiety Scale (HAMA) scores fell 16.0 versus 9.5 points, and sleep-quality scores improved. Kasper 2015 (170 outpatients with anxiety-related restlessness) and Kasper 2016 (318 with mixed anxiety and depressive disorder) were also positive with smaller margins, 12.0 versus 9.3 HAMA points in 2015. Kasper’s 2014 International Journal of Neuropsychopharmacology trial in 539 adults with generalized anxiety disorder compared Silexan 160 mg, Silexan 80 mg, paroxetine 20 mg, and placebo for ten weeks: HAMA dropped 14.1, 12.8, 11.3, and 9.5 points. Both Silexan doses beat placebo; paroxetine only trended (p = 0.10) in the main analysis. When the proven drug fails to separate from placebo, “as good as paroxetine” is a weaker sentence than it sounds.

Woelk and Schläfke (Phytomedicine 2010) randomized 77 adults with generalized anxiety disorder to Silexan 80 mg or lorazepam 0.5 mg for six weeks; HAMA fell 11.3 and 11.6 points from a baseline of 25. There was no placebo arm, 0.5 mg is a low starting dose of lorazepam, and Schläfke worked for the sponsor. Pooled analyses agree in direction. Dold and colleagues (European Archives of Psychiatry and Clinical Neuroscience 2023) combined the five placebo-controlled 80 mg trials (1,213 patients) and found Silexan superior on HAMA, with a responder rate ratio of 1.34 and no excess of adverse events or adverse-event withdrawals; the author list includes a Schwabe employee. Donelli’s unfunded 2019 Phytomedicine review put the oral Silexan difference at about 2.9 HAMA points. Yap’s 2019 Scientific Reports network meta-analysis of five studies ranked 160 mg first, yet its own interval for 160 mg versus placebo crossed zero. Modest, consistent, single-sponsor: that is the honest summary. Depression trials, including a sertraline comparison, exist with fewer independent replications.

Aromatherapy is a larger and weaker literature. Donelli 2019 counted 65 randomized trials and judged most to carry a high risk of bias; pooled inhalation studies showed a sizable anxiety reduction (Hedges’ g −0.73), but the designs were too heterogeneous to call that established, and massage studies could not separate lavender from touch. NCCIH says it is unclear whether lavender aromatherapy helps anxiety, stress, depression, pain, or sleep. MSKCC lists small positive sleep studies, such as Lillehei 2015 combining inhaled lavender with sleep hygiene, and notes that in a cancer massage study the massage, not the lavender, seemed responsible. The 2020 Cochrane review of aromatherapy for dementia (Ball and colleagues; 13 trials, 708 participants, lavender in eight) found no convincing evidence of benefit and no evidence of benefit on sleep or mood. Lavender on a pillow is pleasant. As a sleep treatment it is unproven.

EMA’s traditional indications are narrow: tea, tincture, or oral oil for mild symptoms of mental stress and exhaustion and to aid sleep, plus the oil as a bath additive, in adolescents and adults. Everything else is preliminary or marketing. Small trials on depression with lavender tea or tincture, post-herpetic pain, migraine, menstrual cramps, menopausal symptoms, and COVID-19 cough appear in the NCCIH and MSKCC summaries and need replication. Hair-growth, wound-healing, antibacterial, and anticancer claims rest on dishes, animals, blends, or single small studies. Perillyl alcohol is a laboratory and glioma-nasal-spray story, not a reason to drink lavender tea. Lavender oil is not a burn treatment, not an antibiotic, and not a stand-in for an anxiolytic or antidepressant that a clinician has prescribed.

EMA traditional oral oil: lavender oil meeting the European Pharmacopoeia standard, steam-distilled from L. angustifolia flowering tops, 20–80 mg per day by mouth for adolescents, adults, and the elderly. The monograph leaves it to manufacturers to convert that weight into drops for a given dropper, which is why a bottle of loose oil is not a dosing device. As a bath additive: 1–3 g per full bath once daily, at 35–38 °C for 10–20 minutes. Not recommended under twelve. See a doctor or pharmacist if symptoms persist or worsen. Full baths are contraindicated with open wounds, large skin injuries, acute skin disease, high fever, severe infections, severe circulatory disturbances, and cardiac insufficiency.

EMA traditional flower: 1–2 g of dried flower in 150 ml of boiling water as an infusion, three times daily; or a 1:5 tincture in 50–60% ethanol, 2–4 ml three times daily. Same age floor, same pregnancy caution, same warning that it may impair driving. A cup of lavender tea is a mild traditional preparation, not a dose equivalent to an 80 mg oil capsule, and no trial has compared the two. NCCIH considers lavender in normal food amounts likely safe, which is the right frame for culinary lavender in baking or syrups.

Silexan (WS 1265) is a specified essential oil from L. angustifolia flowers in 80 mg soft capsules taken once daily; the trials ran six to ten weeks and also tested 160 mg. In Germany it is sold as Lasea and has been an authorised medicine since 2009 for restlessness due to anxiety in adults over eighteen. A drugstore essential oil is not interchangeable with it: retail oils are not made, tested, or labelled for ingestion, and their linalool, linalyl acetate, and camphor content varies by species and lot. Other lavender capsules are supplements whose oil may or may not resemble the trial material. Read the label for species, oil amount, and manufacturer before assuming a trial applies.

Aromatherapy means diffusers, inhaler sticks, pillow sprays, and massage oils diluted in a carrier. None has a standardized dose. Undiluted oil on skin raises allergy risk, and so does oil that has been open to air: keep bottles tightly closed and treat an old, half-empty bottle as more allergenic than a fresh one. Store every essential oil with a child-resistant closure, away from oral medicines and out of reach. In Australian poison-centre data most essential-oil exposures were accidental, nearly two-thirds involved children under fifteen, therapeutic errors often meant oil mistaken for a liquid medicine, and the authors noted that as little as 5 ml can cause life-threatening toxicity in a child.