Aromatherapy

What it is

Aromatherapy means using concentrated plant oils - commonly lavender, rose, citrus, peppermint or blends - which are inhaled from a diffuser, pad, sachet or pillow, or diluted and rubbed into the skin in massage [1][2]. In patient trials the oils have been used singly or as blends, and lavender has been reviewed on its own [3].

How it might work

Compounds found in many oils, such as linalool and limonene, are proposed to act through the sense of smell and the nervous system [1]. Because a scent is obvious, trials cannot hide whether someone is using an oil, so expectation may explain part of any effect [1].

What the research says

Sleep: the largest recent review pooled 27 randomised trials in 2,072 adults whose insomnia came with another illness. On the 21-point Pittsburgh Sleep Quality Index, where lower is better, inhaled oils improved scores by 2.9 points (24 trials, 1,864 patients) [1]. The benefit appeared on six parts of the questionnaire, but results varied enormously between trials (I2 85% to 97%), funnel plots suggested publication bias and the reviewers called the certainty of evidence average [1]. Directionally consistent results come from other reviews: 34 studies of inhaled oils (average effect size 0.65) [4], 30 studies in adults and older adults (effect size 0.74) [5], 10 studies in older adults (SMD -1.02, reduced to -0.83 after removing one study, with only 2 of 10 trials blinded) [6] and 13 trials in 740 hospitalised cardiac patients (SMD -0.74) [7].

Anxiety and mood: anxiety scores fell in the 27-trial review (9 studies, 802 patients, about 4 points on the anxiety scale) [1], across 32 pooled trials [8], in 20 studies of 1,717 adults awaiting surgery (SMD -0.57) [2] and in pregnant and postpartum women (7 studies, 531 women, SMD -0.40; the benefit was clear before birth but not after) [9]. A review of 32 trials found depressive symptoms improved moderately (SMD -0.56) but found no effect in pregnant or postpartum women [10]. A separate lavender review found lower anxiety (g -0.72) and depression (g -0.43) scores [3].

Dementia: a 2020 Cochrane review of 13 trials (708 residents, all in care facilities) found no convincing benefit for agitation or overall behaviour symptoms, and harms were poorly documented [11]. Newer meta-analyses of 15 trials each found improved agitation (WMD -6.31 on the CMAI scale) and behaviour (WMD -8.07 on the NPI scale) scores after 3 to 4 weeks, with lavender effective in both [12][13].

Pain and other symptoms: a 2024 meta-analysis of 15 cancer trials found no significant effect on pain, although quality-of-life scores improved when aromatherapy was combined with massage [14]. A 2019 review of 17 labour trials found less pain in the transition phase and shorter active labour [15], while an earlier Cochrane review of 2 trials (535 women) found no difference in pain intensity [16]. In cancer patients, a 2026 review of 26 trials (2,912 patients) reported better sleep, fatigue, anxiety and depression scores and fewer vomiting episodes over 24 hours [17], and a 2024 review of 25 trials found nausea and vomiting improved overall - mostly nausea - with peppermint oil the most studied [18].

Specific oils: a 2025 review of 28 trials found Rosa damascena (rose) oil lowered anxiety, stress and sleep scores but had little effect on depression or pulse rate [19]. Used topically rather than inhaled, tea tree oil modestly reduced acne severity in 7 studies (445 people) [20].

Doses used in studies

In the 27-trial review, inhalation lasted a week or less in some studies and more than a month in others; sachets and medicated pillows were common delivery methods [1]. Among individual trials, the review of hospitalised heart patients found larger sleep improvements when more than 3 drops of oil were used and treatment lasted at least 3 days [7], and the surgical-anxiety review found sessions of 20 minutes or less worked at least as well as longer sessions [2].

Safety and interactions

Skin reactions: undiluted or highly concentrated oils are the main trigger of allergic contact dermatitis - about 80 essential oils have caused it [21] - and in a 10-year database of 10,930 dermatitis patients, 8.3% reacted to at least one oil, with massage therapists at higher risk [22]. Swallowing: essential oils are not for eating; a 61-year-old man who drank camphor oil had seizures and was treated in intensive care [23], and concentrated tea tree oil caused serious poisoning signs in 443 dogs and cats [24]. Breathing: volatile oil vapours were linked to breathing symptoms in people with seasonal asthma [25], and menthol-type rubs should not be placed under a small child's nose - a toddler developed severe breathing difficulty after this, and ferret experiments showed the vapour increases airway mucus and slows the cilia that clear it [26]. Dose limits and interactions: the trials reported side effects poorly and did not test interactions with medicines, so no safe maximum can be given - ask a pharmacist or doctor before using oils alongside medicines [11]. Pregnancy and breastfeeding: trials in pregnant and postpartum women found improved anxiety and sleep with moderate-certainty evidence [9]; we found no separate breastfeeding safety data and no dedicated safety testing in pregnancy, so check with a midwife or doctor first.

Caveats and open questions

Every aromatherapy trial is effectively unblinded, so expectation can inflate results [1]. Reviews disagree on dementia [11][12] and on pain [14][15], and a 2025 review of inhaled aromatherapy in cognitive impairment reported improved cognitive-test scores (MMSE +3.89, MoCA +4.11), while the Cochrane review found no significant effect on cognition in the two trials that measured it [27][11]. Most trials are short, so long-term benefits and risks are unknown [1].