Benefits
Calmness and state anxiety during demanding lab tasks
In small crossover studies in healthy young adults, single 300 or 600 mg doses of a standardized extract, or 0.3 g of extract in a drink, raised self-rated calmness or lowered state anxiety during stressful computer tasks compared with placebo. Some doses had no effect or the opposite effect, and 600 mg also lowered alertness. The drink study was sponsored by the product makers.
Anxiety and mood scores in adults with long-term health conditions
In placebo-controlled trials, 700 mg a day of extract for 12 weeks in adults with type 2 diabetes and depressive symptoms, and 3 g a day for 8 weeks in adults with stable angina, lowered anxiety scores and mood questionnaire scores more than placebo. A pooled analysis of trials found the same direction, with large differences between studies.
Sleep quality and deep sleep time
In a 30-person crossover trial, 400 mg of a phytosome extract at bedtime for 2 weeks lowered insomnia severity scores and lengthened watch-measured deep sleep versus placebo; the seller's scientific director was an author. 3 g a day for 8 weeks improved sleep scores in adults with angina, but 700 mg a day did not change sleep quality in adults with diabetes.
Memory and attention after single doses
In small crossover studies in healthy young adults, 600 mg of one extract improved attention accuracy and 1,600 mg of dried leaf improved memory scores compared with placebo. Other doses slowed or lowered memory task scores, so effects depended on the dose and the preparation.
Thinking and memory scores in older adults over months
In 42 adults aged 65 to 80 with mild to moderate Alzheimer's disease, 60 drops a day of extract for 4 months improved cognitive scores more than placebo. An extract giving 500 mg rosmarinic acid a day did not change cognitive scores over 24 weeks in 23 people with mild Alzheimer's disease or over 96 weeks in 323 older adults with memory complaints.
Traditional use for occasional bloating and gas
The European Medicines Agency lists lemon balm leaf as a traditional herbal medicine for mild digestive complaints, including bloating and flatulence, based on long-standing use rather than clinical trials. No placebo-controlled trial of lemon balm alone for digestion is cited on this page.
Mechanism of action
GABA transaminase inhibition (lab data)
In a lab assay with rat brain enzyme, a lemon balm extract inhibited GABA transaminase, the enzyme that breaks down the calming messenger GABA. Rosmarinic acid, about 1.5% of the dry leaf, was the main active compound, with ursolic and oleanolic acids also active. This has not been measured in people.
Acetylcholine receptor binding (lab data)
In human brain tissue samples, lemon balm preparations displaced test compounds from nicotinic and muscarinic acetylcholine receptors, but binding strength differed widely between preparations, and the leaf chosen for one human study showed no cholinesterase inhibition.
TSH receptor binding (lab data)
In cell assays, lemon balm extract inhibited the binding of thyroid-stimulating hormone (TSH) to its receptor and lowered TSH-stimulated cAMP production. Whether usual oral doses affect thyroid function in people has not been tested.
Clinical trials
Randomized, placebo-controlled, double-blind, balanced crossover: single doses of 300, 600 and 900 mg of a standardized extract or placebo, 7 days apart, with testing up to 6 hours after dosing. (Kennedy et al. 2002, Pharmacol Biochem Behav)
20 healthy young adults.
600 mg gave a sustained improvement in accuracy of attention, but secondary and working memory factor scores fell at some doses and time points. The lowest dose raised self-rated calmness at the earliest time points, and the 900 mg dose lowered self-rated alertness at all time points.
Randomized, placebo-controlled, double-blind, balanced crossover: single doses of 600, 1,000 and 1,600 mg of encapsulated dried leaf or placebo, 7 days apart, tested at 1, 3 and 6 hours. (Kennedy et al. 2003, Neuropsychopharmacology)
20 healthy young adults.
The 1,600 mg dose improved memory performance and raised self-rated calmness at all time points after dosing. Lower doses slowed timed memory tasks and a rapid visual information processing task, with more slowing at lower doses.
Double-blind, placebo-controlled, randomized, balanced crossover: single 300 mg and 600 mg doses of a standardized extract or placebo, 7 days apart; mood and performance on a 20-minute multitasking stressor 1 hour after dosing. (Kennedy et al. 2004, Psychosom Med)
18 healthy volunteers.
The 600 mg dose lessened the negative mood effects of the stressor, with higher self-rated calmness and lower self-rated alertness than placebo. The 300 mg dose sped up math processing without a loss of accuracy.
Two randomized, double-blind, placebo-controlled crossover studies: 0.3 g or 0.6 g of a standardized extract in a 480 mL iced-tea type drink or a 250 g yogurt, versus placebo, with testing 1 and 3 hours later. Sponsored by the drink maker (Rudolf Wild) and the extract maker (Cognis). (Scholey et al. 2014, Nutrients)
25 healthy young adults (drink study) and 21 healthy young adults (yogurt study).
In the drink study, 0.3 g with fruit sweetener lowered state anxiety and improved working memory at 1 and 3 hours, while 0.6 g with artificial sweetener raised state anxiety at one time point. In the yogurt study, 0.3 g and 0.6 g with fruit sweetener improved immediate word recall at 1 hour only.
Randomized, double-blind, placebo-controlled: 700 mg/day of a hydroalcoholic extract or 700 mg/day of toasted flour for 12 weeks; 60 randomized, 44 completed. (Safari et al. 2023, BMC Complement Med Ther)
Adults aged 20 to 65 with type 2 diabetes and depressive symptoms.
Depression (BDI-II) and anxiety (BAI) scores fell more with lemon balm than with placebo (p < 0.001 and p = 0.04). Sleep quality (PSQI), fasting blood sugar, hs-CRP, body measures and blood pressure did not differ from placebo. No side effects were reported.
Randomized, double-blind, placebo-controlled: 3 g/day of lemon balm capsules or placebo for 8 weeks; outcomes on the DASS-21 scale and the Pittsburgh Sleep Quality Index. (Haybar et al. 2018, Clin Nutr ESPEN)
80 adults with chronic stable angina.
Depression, anxiety, stress and total sleep disturbance scores fell more with lemon balm than with placebo (P < 0.05).
Randomized, double-blind, placebo-controlled crossover: two 200 mg tablets of a phytosome extract or placebo 30 minutes before bed for 2 weeks each, with a 7-day washout; sleep stages from a wrist watch. The scientific director of the company selling the product and two of its consultants were authors. (Di Pierro et al. 2024, Nutrients)
30 adults aged 18 to 65 with tiredness on waking and unrefreshing sleep.
Insomnia Severity Index scores were 2.9 points lower on lemon balm than on placebo (6.8 vs 9.7; p = 0.003). Watch-measured deep sleep rose about 15% and REM sleep fell about 10%. 87% reported better sleep on lemon balm versus 30% on placebo. Anxiety scores did not change.
Randomized, double-blind, placebo-controlled, parallel group trial at three centers in Tehran: a fixed 60 drops/day of extract or placebo for 4 months. (Akhondzadeh et al. 2003, J Neurol Neurosurg Psychiatry)
42 adults aged 65 to 80 with mild to moderate Alzheimer's disease.
Cognitive scores (ADAS-cog and CDR-SB, the main outcomes) improved more than with placebo. Agitation, logged as a side effect rather than tested as an outcome, was more common on placebo (p = 0.03).
Randomized, double-blind, placebo-controlled: extract giving 500 mg rosmarinic acid daily for 24 weeks; safety and tolerability were the primary endpoint. (Noguchi-Shinohara et al. 2020, Sci Rep)
23 people with mild dementia due to Alzheimer's disease.
No serious adverse events and no differences in vital signs or physical and neurologic exams. Cognitive scores did not differ from placebo. A behavior and mood symptom score (NPI-Q) improved by 0.5 points on lemon balm and worsened by 0.7 points on placebo (P = 0.012).
Randomized, placebo-controlled, double-blind: extract giving 500 mg rosmarinic acid daily for 96 weeks, then a 24-week washout; primary endpoint ADAS-cog. (Noguchi-Shinohara et al. 2023, J Alzheimers Dis)
323 older adults with subjective or mild cognitive impairment, without dementia.
Cognitive measures, including the primary ADAS-cog score, did not differ from placebo at 96 weeks. In a subgroup without high blood pressure, the change in Clinical Dementia Rating Sum of Boxes differed between groups (p = 0.036). Hippocampal volume and safety measures did not differ.