Benefits
Lower anxiety ratings in randomized trials
Five randomized, double-blind, placebo-controlled trials of Silexan® 80 mg/day for 10 weeks, pooling 1,213 adults, found larger falls in Hamilton Anxiety Scale scores than placebo, with separation from placebo appearing by about week 2. The size of the advantage is modest: an independent 2019 meta-analysis put it at 2.9 HAMA points (95% CI 0.95 to 4.86), against placebo groups that themselves improved by 8.4 to 9.5 points. A separate 6-week trial in 77 people compared Silexan® with lorazepam 0.5 mg/day, a dose at the bottom of the benzodiazepine anxiolytic range, and both groups improved by about 11 HAMA points; that trial had no placebo group, so it cannot show how much of either result was the drug rather than the placebo response. Everyone in these trials had a diagnosed anxiety condition and a Hamilton score of at least 18. Anxiety disorders are medically managed conditions, and none of this is a reason to start, stop or replace psychiatric treatment.
Sleep improves, but as a knock-on effect of reduced anxiety
In the 221-person trial that made the Pittsburgh Sleep Quality Index a co-primary outcome, Silexan® 80 mg/day lowered PSQI scores by 5.5 points over 10 weeks against 3.8 points on placebo. A mediation analysis of that same dataset, by the trial's own authors, found that 98.4 percent of the sleep benefit was explained by the fall in anxiety, with no measurable direct effect on sleep (p = 0.96). So this is not a sleep aid: it did not sedate anyone, and the people studied had disturbed sleep on top of an anxiety diagnosis rather than primary insomnia. No polysomnography study of oral lavender oil has been published, so nothing can be said about sleep architecture, deep sleep or REM either way.
Depression ratings in trials of anxious and depressed mood
Two randomized, placebo-controlled trials have measured depression ratings directly. In 318 adults with ICD-10 mixed anxiety and depressive disorder, Montgomery and Asberg Depression Rating Scale scores fell 9.2 points on Silexan® 80 mg/day and 6.1 points on placebo over 70 days, with Hamilton anxiety scores falling 10.8 against 8.4. In 498 adults with a mild or moderate episode of major depression, the same 80 mg dose beat placebo by 2.17 MADRS points over 8 weeks, with sertraline 50 mg included as a reference arm and beating placebo by 2.59 points. Both trials were funded and run by the manufacturer and both enrolled people with a diagnosed condition. Depression is medically managed: this is not a treatment for it and not a substitute for talking to a clinician.
Aromatherapy is a different route, and its evidence does not transfer
This page is about what happens when you swallow lavender oil. Inhaled lavender is a separate question with separate data, and the two have never been compared directly: a 2018 systematic review searching PubMed, CINAHL, Cochrane CENTRAL and Embase found no eligible study. The 2019 meta-analysis cited on this page pooled 1,682 people and did find lower anxiety scores with inhaled lavender (Hedges' g 0.73), but rated most of those trials at high risk of bias and found no effect on systolic blood pressure. Whether lavender changes cortisol is unsettled: the studies that measured it are small and their results conflict, and a randomized trial of an 80 mg oral lavender oil capsule found no change in serum cortisol. None of the inhalation evidence tells you anything about an 80 mg capsule.
Mechanism of action
Not a benzodiazepine mechanism
It is worth saying what oral lavender oil does not do. The published laboratory work on Silexan® attributes its calming action to moderate inhibition of voltage dependent calcium channels, a profile its investigators compare with pregabalin, and not to the benzodiazepine binding site of the GABA-A receptor. That fits what has been seen in people: in a manufacturer-run five-way crossover study in 34 evaluable recreational users of CNS depressants, single 80 mg and 640 mg doses were rated the same as placebo on drug liking and produced no sedation, while lorazepam was clearly distinguished from both.
Voltage-gated calcium channel inhibition
In rodent tissue and cell work, lavender oil reduced calcium influx through N-type, P/Q-type and T-type voltage dependent calcium channels at nanomolar concentrations the authors relate to an 80 mg human dose, with the effect in the hippocampus carried mainly by the N-type and P/Q-type channels. This is laboratory and animal evidence for how the oil might work. It has not been shown in the brains of people taking capsules.
A possible serotonin-1A signal, seen on brain imaging
A PET imaging study gave 17 healthy men 160 mg/day of Silexan® or placebo for at least 8 weeks and found reduced serotonin-1A receptor binding potential in the temporal and fusiform gyri, hippocampus, insula and anterior cingulate cortex. The authors describe this as suggesting the serotonin-1A receptor is involved. It is a 17-person imaging study in healthy volunteers, it measured no symptom, and it does not establish partial agonism.
Clinical trials
Randomized, double-blind, multi-centre trial of Silexan® 80 mg/day against lorazepam 0.5 mg/day in 77 adults with generalized anxiety disorder for 6 weeks. There was no placebo arm. Funded by the manufacturer, Dr. Willmar Schwabe GmbH, whose employee is the second author. (Woelk H, Schlafke S. Phytomedicine 2010;17(2):94-9)
77 adults with generalized anxiety disorder. 6 weeks, no placebo group.
Hamilton Anxiety Scale totals fell by 11.3 (SD 6.7) points on Silexan® and 11.6 (SD 6.6) points on lorazepam, from a baseline of 25 points in both groups. Read that carefully: with no placebo group, the trial cannot separate either treatment from the natural course of the condition and the placebo response, which ran to 8.4 to 9.5 HAMA points in the placebo-controlled Silexan® trials. The lorazepam dose of 0.5 mg/day sits at the bottom of the anxiolytic range. The authors reported no sedative effects in this trial, but abuse liability was not measured here; that was studied separately, in healthy volunteers, by the manufacturer. Silexan® is licensed in Germany as the pharmacy medicine Lasea® for restlessness related to anxious mood.
Randomized, double-blind, double-dummy trial in 539 adults meeting DSM criteria for generalized anxiety disorder with a Hamilton Anxiety Scale score of 18 or more. Four arms for 10 weeks: Silexan® 160 mg/day, Silexan® 80 mg/day, paroxetine 20 mg/day as an active reference, and placebo. Two authors are employees of the manufacturer, Dr. Willmar Schwabe GmbH. (Kasper S et al. Int J Neuropsychopharmacol 2014;17(6):859-69)
539 adults with diagnosed generalized anxiety disorder. 10 weeks.
Hamilton Anxiety Scale totals fell by 14.1 points on 160 mg, 12.8 points on 80 mg, 11.3 points on paroxetine and 9.5 points on placebo. Both Silexan® doses beat placebo (p < 0.01). The active reference, paroxetine, did not: it reached only p = 0.10 against placebo in the full analysis set, which is a real caution about how well this trial could tell treatments apart. Response rates were 51.9 percent on 80 mg and 37.8 percent on placebo. Adverse event rates on Silexan® were similar to placebo and lower than on paroxetine.