《气味科学中的健康密码》课程教学资源（阅读材料）Is lavender an anxiolytic drug? A systematic review of randomised clinical trials

Phytomedicine 19(2012)825-835 Contents lists available at SciVerse ScienceDirect Phytomedicine Phytomedicine ELSEVIER journal homepage:www.elsevier.de/phymed Short communication Is lavender an anxiolytic drug?A systematic review of randomised clinical trials R.Perry*,R.Terry,L.K.Watson,E.Ernst Complementary Medicine,Peninsula Medical School.Veysey Building.Salmon Pool Lane,University of Exeter,Exeter EX2 4SG.United Kingdom ARTICLE INFO ABSTRACT Keywords: Background:Lavender(Lavandula angustifolia)is often recommended for stress/anxiety reliefand believed Lavender to possess anxiolytic effects. Anxiolytic drug Aim:To critically evaluate the efficacy/effectiveness of lavender for the reduction of stress/anxiety. Anxiety Methods:Seven electronic databases were searched to identify all relevant studies.All methods of laven- Stress RCT der administration were included.Data extraction and the assessment of the methodological quality of Systematic review all included trials were conducted by two independent reviewers. Results:Fifteen RCTs met the inclusion criteria.Two trials scored 4 points on the 5-point Jadad scale,the remaining 13 scored two or less.Results from seven trials appeared to favour lavender over controls for at least one relevant outcome. Conclusion:Methodological issues limit the extent to which any conclusions can be drawn regarding the efficacy/effectiveness of lavender.The best evidence suggests that oral lavender supplements may have some therapeutic effects.However,further independent replications are needed before firm conclusions can be drawn. 2012 Elsevier GmbH.All rights reserved. Introduction excessive intake)(Leung and Foster 1996),lavender intake seems to be reasonably safe. Lavender(Lavandula angustifolia)has a long history of medicinal This systematic review is aimed at critically evaluating the data use and is purported to possess anxiolytic,sedative and calming from randomised clinical trials(RCTs)of all types oflavender prepa- properties (Cavanagh and Wilkinson 2002).Most commonly it is rations(oral,olfactory,topical)for the treatment of anxiety. recommended for oral administration.More recently,lavender is also being employed in aromatherapy (Setzer 2009)although spe- Methods cific pharmacological effects of lavender aromatherapy are difficult to distinguish from any innate or learned preferences to this,or any The following electronic databases were searched from their other,odour(Bradley et al.2009). The chemical composition of lavender is complex and several of inception up to December 2010:Medline,EMBASE and PsychInfo (via OVID).AMED and CINAHL(via EBSCO).The Cochrane Library. its constituents (e.g.linalool and linalyl acetate)have been pro- and ISI Web of Knowledge.The search terms used included lavender, posed as being responsible for the perceived anxiolytic effects and anxiety or stress,and derivatives of these terms (see Appendix (Setzer 2009).In animal models,linalool has been found to inhibit GABA(A)binding reception in the CNS inducing a relaxed state A for electronic search strategy).Our own departmental files and the reference lists of all selected articles were searched for further (Brum et al.2001;Hossain et al.2004).Until recently,this activity relevant studies. had not been noted in human studies. Articles were included if they reported an RCT in which human Anxiety is a common disorder (14%of the EU population suf- subjects with or without clinical anxiety were treated with any type fer from one or more anxiety disorders each year (Wittchen et al. of lavender preparation(excluding those containing other ingredi- 2011))but can be severe and debilitating,often requiring med- ents(e.g.Corner et al.1995;Graham et al.2003:Nord and Belew ication.Lavender may provide a gentler treatment option than 2009))and which reported validated measures of anxiety or stress conventional anxiolytic drugs.Apart from rare allergic reactions as an endpoint (e.g.State Trait Anxiety Inventory;STAl,Hamilton (Coulson and Khan 1999)and gastrointestinal complaints (after Anxiety and Depression:HAD),standardised measures of anxiety (e.g.visual analogue scales were acceptable)or physiological mea- sures of stress (e.g.salivary levels of cortisol,heart rate,galvanic skin response).We excluded studies that were not randomised(e.g. Corresponding author.Tel.:+44 0 1392 726044. Willman et al.2009)along with studies comparing two different E-mail address:R.Perry@bath.ac.uk (R.Perry). forms of lavender treatments without a control condition,those 0944-7113/S-see front matter 2012 Elsevier GmbH.All rights reserved. doi:10.1016j-ohymed.2012.02.013

Phytomedicine 19 (2012) 825–835 Contents lists available at SciVerse ScienceDirect Phytomedicine journal h omepage: www.elsevier.de/phymed Short communication Is lavender an anxiolytic drug? A systematic review of randomised clinical trials R. Perry∗, R. Terry, L.K. Watson, E. Ernst Complementary Medicine, Peninsula Medical School, Veysey Building, Salmon Pool Lane, University of Exeter, Exeter EX2 4SG, United Kingdom a r t i c l e i n f o Keywords: Lavender Anxiolytic drug Anxiety Stress RCT Systematic review a b s t r a c t Background: Lavender (Lavandula angustifolia)is oftenrecommended for stress/anxiety relief and believed to possess anxiolytic effects. Aim: To critically evaluate the efficacy/effectiveness of lavender for the reduction of stress/anxiety. Methods: Seven electronic databases were searched to identify all relevant studies. All methods of laven￾der administration were included. Data extraction and the assessment of the methodological quality of all included trials were conducted by two independent reviewers. Results: Fifteen RCTs met the inclusion criteria. Two trials scored 4 points on the 5-point Jadad scale, the remaining 13 scored two or less. Results from seven trials appeared to favour lavender over controls for at least one relevant outcome. Conclusion: Methodological issues limit the extent to which any conclusions can be drawn regarding the efficacy/effectiveness of lavender. The best evidence suggests that oral lavender supplements may have some therapeutic effects. However, further independent replications are needed before firm conclusions can be drawn. © 2012 Elsevier GmbH. All rights reserved. Introduction Lavender (Lavandula angustifolia) has a long history of medicinal use and is purported to possess anxiolytic, sedative and calming properties (Cavanagh and Wilkinson 2002). Most commonly it is recommended for oral administration. More recently, lavender is also being employed in aromatherapy (Setzer 2009) although spe￾cific pharmacological effects of lavender aromatherapy are difficult to distinguish from any innate or learned preferences to this, or any other, odour (Bradley et al. 2009). The chemical composition of lavender is complex and several of its constituents (e.g. linalool and linalyl acetate) have been pro￾posed as being responsible for the perceived anxiolytic effects (Setzer 2009). In animal models, linalool has been found to inhibit GABA(A) binding reception in the CNS inducing a relaxed state (Brum et al. 2001; Hossain et al. 2004). Until recently, this activity had not been noted in human studies. Anxiety is a common disorder (14% of the EU population suf￾fer from one or more anxiety disorders each year (Wittchen et al. 2011)) but can be severe and debilitating, often requiring med￾ication. Lavender may provide a gentler treatment option than conventional anxiolytic drugs. Apart from rare allergic reactions (Coulson and Khan 1999) and gastrointestinal complaints (after ∗ Corresponding author. Tel.: +44 0 1392 726044. E-mail address: R.Perry@bath.ac.uk (R. Perry). excessive intake) (Leung and Foster 1996), lavender intake seems to be reasonably safe. This systematic review is aimed at critically evaluating the data fromrandomised clinicaltrials (RCTs) of alltypes oflavender prepa￾rations (oral, olfactory, topical) for the treatment of anxiety. Methods The following electronic databases were searched from their inception up to December 2010: Medline, EMBASE and PsychInfo (via OVID), AMED and CINAHL (via EBSCO), The Cochrane Library, andISIWeb ofKnowledge. The searchtermsusedincludedlavender, and anxiety or stress, and derivatives of these terms (see Appendix A for electronic search strategy). Our own departmental files and the reference lists of all selected articles were searched for further relevant studies. Articles were included if they reported an RCT in which human subjects with or without clinical anxiety were treated with any type of lavender preparation (excluding those containing other ingredi￾ents (e.g. Corner et al. 1995; Graham et al. 2003; Nord and Belew 2009)) and which reported validated measures of anxiety or stress as an endpoint (e.g. State Trait Anxiety Inventory; STAI, Hamilton Anxiety and Depression; HAD), standardised measures of anxiety (e.g. visual analogue scales were acceptable) or physiological mea￾sures of stress (e.g. salivary levels of cortisol, heart rate, galvanic skin response).We excluded studies that were not randomised (e.g. Willman et al. 2009) along with studies comparing two different forms of lavender treatments without a control condition, those 0944-7113/$ – see front matter © 2012 Elsevier GmbH. All rights reserved. doi:10.1016/j.phymed.2012.02.013

826 R.Perry et al.Phytomedicine 19(2012)825-835 comparing a combined lavender treatment against a no-treatment Muzzarelli et al.2006:Sgoutas-Emch et al.2001;Toda and control(e.g.Hoya et al.2008),uncontrolled trials and those in which Morimoto 2008)investigated the effect of lavender oil inhalation no clinical data were reported (e.g.Buckle 1993;Motomura et al Four trials(Braden et al.2009:Kritsidima et al.2010:Kutlu et al 1999).No language restrictions were imposed.Hard copies of all 2008;Motomura et al.2001)reported a significantly positive effect articles were obtained and read in full by two authors(RP,RT). for at least one anxiety outcome measure.Kritsidima et al.(2010) Data from the articles were independently extracted from all compared lavender oil aromatherapy with no oil (using a candle included studies by two reviewers(RP,RT)according to pre-defined burner)for patients waiting for a dental appointment.They found criteria (Tables 1 and 2).We only reported between-group analy- a significantly greater reduction in State Trait Anxiety Inventory ses from outcomes that conform to our inclusion criteria above (STAl)compared to the control group(p<0.001)but no significant To assess methodological quality,the Jadad scalel was used(Jadad between-groups difference in the Modified Dental Anxiety Scale et al.1996).To supplement the Jadad score,using Verhagen et al. (MDAS).Braden(Braden et al.2009)found a lavender aromather- (1998)and Boutron et al.(2005)as a guide,additional information apy group (lavender hybrid)had significantly less self-reported pertaining to risk of bias was extracted (Table 2).Discrepancies anxiety than either control group (p<0.01)for pre-operative anxi- were resolved through discussion between the authors.A meta- etv analysis was considered but deemed to be not appropriate because Both Motomura et al.(2001)Kutlu et al.(2008)induced stress- of the clinical and statistical heterogeneity of the primary studies. ful situations in healthy volunteers.Lavender odour was released in the experimental conditions but not the control conditions Results Motomura et al.(1999)found that anxiety levels were associated with reduced mental stress in the lavender condition although no The literature search identified 440 potentially relevant titles significant differences between conditions were found for physio- and abstracts.Fifteen RCTs involving 1565 participants met the logical measures.Kutlu et al.(2008)used an exam to induce anxiety inclusion criteria(Fig.1).Where possible,between-group analy- and found the lavender group had significantly lower anxiety scores ses of the main anxiety outcomes are presented in Table 1.The on the STAl than the control group (p<0.001)after the 60-min included studies were published between 1995 and 2010.origi- exam.Unfortunately as baseline anxiety scores would be likely to nated from six countries and were all written in English.Sample be very high prior to an exam,baseline anxiety scores were not sizes ranged from 16 to 340.The majority of trials used Lavandula taken,so it is impossible to establish change over time.Both trials angustifolia unless otherwise stated. suffer major methodological issues (Table 2) Eight trials(Bradley et al.2009:Howard and Hughes 2008;Kutlu Four inhalation trials (Howard and Hughes 2008;Muzzarelli et al.2008;Morris 2002:Motomura et al.2001:Sgoutas-Emch et al.2006;Sgoutas-Emch et al.2001:Toda and Morimoto 2008) et al.2001:Toda and Morimoto 2008:Xu et al.2008)used healthy did not demonstrate an effect.Muzzarelli et al.(2006)used patients volunteers with assumed normal levels of anxiety in which anx- waiting for an gastrointestinal endoscopy.Reductions in within- iety was induced for the purpose of the study.One trial (Soden group anxiety levels were reported in both the lavender(Provencal et al.2004)assessed the efficacy of lavender in cancer patients lavender)and grapeseed inhalation groups,however,no between- with high levels of anxiety and depression.Three studies (Braden group analyses were performed which limits its conclusiveness et al.2009;Kritsidima et al.2010;Muzzarelli et al.2006)looked at Longer inhalation time(more than 5 min)was suggested for future pre-procedural anxiety (e.g.,dental appointment,gastrointestinal trials. endoscopic procedure,pre-operation).One trial(Dunn et al.1995) The remaining trials(Howard and Hughes 2008;Sgoutas-Emch involved patients in an Intensive Care Unit (ICU)and two studies et al.2001:Toda and Morimoto 2008)of lavender inhalation focussed upon the value of lavender for individuals with gener- assessed efficacy in healthy volunteers.Stress was induced in a alised or sub-syndromal anxiety disorder(according to DSMIV or variety of ways e.g.arithmetic tasks(Sgoutas-Emch et al.2001; ICD-10)(Kasper et al.2010:Woelk and Schlafke 2010). Toda and Morimoto 2008)and an arousal task(Howard and Hughes The methodological quality of the included trials was variable 2008).In Toda's(Toda and Morimoto 2008)study it was found that but generally poor(Table 2):Jadad scores ranged from 0 to 4.with during the experimental period,neither group showed any signifi- just two trials(Kasper et al.2010;Woelk and Schlafke 2010)achiev- cant variation in levels of salivary cortisol (stress hormone).In the ing a score of 4 points.The majority of trials scored less than 2 lavender group.levels of chromogranin A(CgA is a novel stress Different methods of lavender administration were tested.Eight marker found in saliva (Nakane et al.1998,2002))that had been trials (Braden et al.2009:Howard and Hughes 2008:Kritsidima elevated at the end of the arithmetic task were statistically lower et al.2010:Kutlu et al.2008:Motomura et al.2001:Muzzarelli 10 min later whereas they were still elevated in the control group. et al.2006;Sgoutas-Emch et al.2001:Toda and Morimoto 2008) However,there was no significant difference between the groups at assessed the efficacy of lavender aromatherapy.Two trials(Dunn 5 and 10min implying that lavender may only help stress levels to et al.1995;Soden et al.2004)employed aromatherapy massage, drop quicker.There were no significant between group differences one used an oil-dripping technique(Xu et al.2008),one involved in subjective ratings of stress. bathing in lavender oil(Morris 2002)and three used oral capsules Sgoutas-Emch et al.(2001)compared four groups:two received (Bradley et al.2009:Kasper et al.2010:Woelk and Schlafke 2010). lavender aromatherapy and two groups did not receive aromather- The results of the 15 trials will be described in more detail in the apy.One group from each condition knew about their group following section,according to the method of administration. allocation,whilst the two other groups did not.The aim was to see if knowledge of treatment impacted on results.In fact,no sig- Inhalation of lavender nificant differences between the four groups were found on any measure.Interestingly,Group 3 had significantly higher levels of Eight trials (Braden et al.2009:Howard and Hughes 2008; anxiety prior to task than other three groups yet this group's anxiety Kritsidima et al.2010:Kutlu et al.2008:Motomura et al.2001: levels went down following the arousal task.In general,the authors felt a more stress-provoking task might be required in future trials. Howard and Hughes (2008)employed an experimental stress 1 Jadad score(1 for randomisation.1 for sequence generation and allocation con- task to compare lavender with tea tree oil odour(against a no odour cealment,1 for stating it is double blind and 1 for description of blinding and condition)in an attempt to compensate for previous experiments appropriateness and 1 for withdrawal stating number and reasons per group). that do not test lavender against another strong odour.They also

826 R. Perry et al. / Phytomedicine 19 (2012) 825–835 comparing a combined lavender treatment against a no-treatment control(e.g.Hoya et al. 2008),uncontrolledtrials andthose in which no clinical data were reported (e.g. Buckle 1993; Motomura et al. 1999). No language restrictions were imposed. Hard copies of all articles were obtained and read in full by two authors (RP, RT). Data from the articles were independently extracted from all included studies by two reviewers (RP, RT) according to pre-defined criteria (Tables 1 and 2). We only reported between-group analy￾ses from outcomes that conform to our inclusion criteria above. To assess methodological quality, the Jadad scale1 was used (Jadad et al. 1996). To supplement the Jadad score, using Verhagen et al. (1998) and Boutron et al. (2005) as a guide, additional information pertaining to risk of bias was extracted (Table 2). Discrepancies were resolved through discussion between the authors. A meta￾analysis was considered but deemed to be not appropriate because of the clinical and statistical heterogeneity of the primary studies. Results The literature search identified 440 potentially relevant titles and abstracts. Fifteen RCTs involving 1565 participants met the inclusion criteria (Fig. 1). Where possible, between-group analy￾ses of the main anxiety outcomes are presented in Table 1. The included studies were published between 1995 and 2010, origi￾nated from six countries and were all written in English. Sample sizes ranged from 16 to 340. The majority of trials used Lavandula angustifolia unless otherwise stated. Eighttrials (Bradley et al. 2009; Howard and Hughes 2008; Kutlu et al. 2008; Morris 2002; Motomura et al. 2001; Sgoutas-Emch et al. 2001; Toda and Morimoto 2008; Xu et al. 2008) used healthy volunteers with assumed normal levels of anxiety in which anx￾iety was induced for the purpose of the study. One trial (Soden et al. 2004) assessed the efficacy of lavender in cancer patients with high levels of anxiety and depression. Three studies (Braden et al. 2009; Kritsidima et al. 2010; Muzzarelli et al. 2006) looked at pre-procedural anxiety (e.g., dental appointment, gastrointestinal endoscopic procedure, pre-operation). One trial (Dunn et al. 1995) involved patients in an Intensive Care Unit (ICU) and two studies focussed upon the value of lavender for individuals with gener￾alised or sub-syndromal anxiety disorder (according to DSMIV or ICD-10) (Kasper et al. 2010; Woelk and Schlafke 2010). The methodological quality of the included trials was variable but generally poor (Table 2); Jadad scores ranged from 0 to 4, with justtwo trials (Kasper et al. 2010;Woelk and Schlafke 2010) achiev￾ing a score of 4 points. The majority of trials scored less than 2. Different methods of lavender administration were tested. Eight trials (Braden et al. 2009; Howard and Hughes 2008; Kritsidima et al. 2010; Kutlu et al. 2008; Motomura et al. 2001; Muzzarelli et al. 2006; Sgoutas-Emch et al. 2001; Toda and Morimoto 2008) assessed the efficacy of lavender aromatherapy. Two trials (Dunn et al. 1995; Soden et al. 2004) employed aromatherapy massage, one used an oil-dripping technique (Xu et al. 2008), one involved bathing in lavender oil (Morris 2002) and three used oral capsules (Bradley et al. 2009; Kasper et al. 2010; Woelk and Schlafke 2010). The results of the 15 trials will be described in more detail in the following section, according to the method of administration. Inhalation of lavender Eight trials (Braden et al. 2009; Howard and Hughes 2008; Kritsidima et al. 2010; Kutlu et al. 2008; Motomura et al. 2001; 1 Jadad score (1 for randomisation, 1 for sequence generation and allocation con￾cealment, 1 for stating it is double blind and 1 for description of blinding and appropriateness and 1 for withdrawal stating number and reasons per group). Muzzarelli et al. 2006; Sgoutas-Emch et al. 2001; Toda and Morimoto 2008) investigated the effect of lavender oil inhalation. Four trials (Braden et al. 2009; Kritsidima et al. 2010; Kutlu et al. 2008; Motomura et al. 2001) reported a significantly positive effect for at least one anxiety outcome measure. Kritsidima et al. (2010) compared lavender oil aromatherapy with no oil (using a candle burner) for patients waiting for a dental appointment. They found a significantly greater reduction in State Trait Anxiety Inventory (STAI) compared to the control group (p < 0.001) but no significant between-groups difference in the Modified Dental Anxiety Scale (MDAS). Braden (Braden et al. 2009) found a lavender aromather￾apy group (lavender hybrid) had significantly less self-reported anxiety than either control group (p < 0.01) for pre-operative anxi￾ety. Both Motomura et al. (2001) Kutlu et al. (2008) induced stress￾ful situations in healthy volunteers. Lavender odour was released in the experimental conditions but not the control conditions. Motomura et al. (1999) found that anxiety levels were associated with reduced mental stress in the lavender condition although no significant differences between conditions were found for physio￾logical measures. Kutlu et al.(2008) used an exam to induce anxiety and found the lavender grouphad significantly lower anxiety scores on the STAI than the control group (p < 0.001) after the 60-min exam. Unfortunately as baseline anxiety scores would be likely to be very high prior to an exam, baseline anxiety scores were not taken, so it is impossible to establish change over time. Both trials suffer major methodological issues (Table 2). Four inhalation trials (Howard and Hughes 2008; Muzzarelli et al. 2006; Sgoutas-Emch et al. 2001; Toda and Morimoto 2008) did not demonstrate an effect. Muzzarelli et al.(2006) used patients waiting for an gastrointestinal endoscopy. Reductions in within￾group anxiety levels were reported in both the lavender (Provencal lavender) and grapeseed inhalation groups, however, no between￾group analyses were performed which limits its conclusiveness. Longer inhalation time (more than 5 min) was suggested for future trials. The remaining trials (Howard and Hughes 2008; Sgoutas-Emch et al. 2001; Toda and Morimoto 2008) of lavender inhalation assessed efficacy in healthy volunteers. Stress was induced in a variety of ways e.g. arithmetic tasks (Sgoutas-Emch et al. 2001; Toda and Morimoto 2008) and an arousaltask (Howard and Hughes 2008). In Toda’s (Toda and Morimoto 2008) study it was found that during the experimental period, neither group showed any signifi- cant variation in levels of salivary cortisol (stress hormone). In the lavender group, levels of chromogranin A (CgA is a novel stress marker found in saliva (Nakane et al. 1998, 2002)) that had been elevated at the end of the arithmetic task were statistically lower 10 min later whereas they were still elevated in the control group. However,there was no significant difference between the groups at 5 and 10 min implying that lavender may only help stress levels to drop quicker. There were no significant between group differences in subjective ratings of stress. Sgoutas-Emch et al. (2001) compared four groups; two received lavender aromatherapy and two groups did not receive aromather￾apy. One group from each condition knew about their group allocation, whilst the two other groups did not. The aim was to see if knowledge of treatment impacted on results. In fact, no sig￾nificant differences between the four groups were found on any measure. Interestingly, Group 3 had significantly higher levels of anxietyprior to task thanother three groups yetthis group’s anxiety levels went down following the arousaltask. In general,the authors felt a more stress-provoking task might be required in future trials. Howard and Hughes (2008) employed an experimental stress task to compare lavender with tea tree oil odour (against a no odour condition) in an attempt to compensate for previous experiments that do not test lavender against another strong odour. They also

Table 1 RCTs of lavender. First author(year) Study design Condition(sample Experimental Control treatment Main anxiety Main results Author's conclusion Adverse events 【ref]country size randomised: treatment outcome (between-group analysed) measurements analysis) Kritsidima(2010) RCT,single-blind,2 Patients waiting for Odour of 5 drops of No odour in (1)State Trait (1)Experimental Although anxiety about No adverse events (Kritsidima, parallel groups. dental lavender oil in 10 waiting room using Anxiety Indicator group showed less future dental visits occurred in either Newton,and placebo-control appointment cc water using a water in a candle (STAl-6 item) STAI(p<0.001) seems to be unaffected, group Asimakopoulou. (340:340) candle burner. burner lavender scent reduces 2010)UK placed in waiting state anxiety in dental room (2)Modified Dental (2)No sig. Datlents Anxiety Scale difference in MDAS (MDAS) Muzzarelli(2006) RCT Single-blind 2 Patients Inhalation of 3 Inhalation of 3 STAI(state only) No between-group Although not Not reported groups (within undergoing a drops of 10% drops of 100% differences statistically be Force.Sebold groups) colonoscopy or lavender oil+90% grapeseed oil in a reported effective,patients 2006)U5A esophagogastroduo- grapeseed oil)in a 100 ml cup for reported the lavender denoscopy 100ml cup for 5min scent to be pleasant (118:118) 5min Braden(2009) RCT,single-blind,3 Pre-operative Group 1 Standard Group 2 Standard Pre-operative Experimental Lavedin was associated Not reported 8 (Braden, groups anxiety(150:150) care+lavender care without anxiety (100 mm group had sig.less with significantly Reichow,Halm aromatherapy (1 aromatherapy VAS) anxiety (p<0.01) lower anxiety on 2009)U5A inhalation of Group 3 Standard than the control operating room undiluted lavandin care+jojoba oil groups transfer,suggesting it then taped to gown aromatherapy is a simple,low-risk and topical (inhalation and and cost effective application-one topical application intervention drop on pulse as in Group 1) point) Motomura(2001) RCT Healthy volunteers Group 1 Inhalation Group 2 stressful (1)Cox McKay's Sig.difference in Lavender odourants Not reported (Motomura, (4242) of 3 ml lavender oil situation Stress/arousal stress/arousal were associated with Sakurai,and via diffuser for (soundproof room) adjective checklist scores between reduced mental stress Yotsuva 2001) 20 min during a for 20 min 1902012) (2)Blood pressure stress and Japan stressful situation Group 3 (3)Heart rate stress/lavender (soundproof room) non-stressful condition and situation between stress only and non-stress condition(p<0.01) No sig.differences between conditions for physiological measures Toda(2008)(Toda RCT Healthy volunteers Inhalation of Control (no odour) (1)Cortisol levels (1)No Lavender aroma has a Not reported and Morimoto (30:30) lavender for in saliva between-groups stress relief effect 2008)Jap3n 10min.Lavender (2)CgA levels in differences oil infiltrated onto saliva reported filter paper held (3)10-point VAS (2)In the lavender 10cm from nose group CgA was significantly lower 10min later but not in control gp. 3】No between-group differences reported 等

R. Perry et al. / Phytomedicine 19 (2012) 825–835 827 Table 1 RCTs of lavender. First author (year) [ref] country Study design Condition (sample size randomised: analysed) Experimental treatment Control treatment Main anxiety outcome measurements Main results (between-group analysis) Author’s conclusion Adverse events Kritsidima (2010) (Kritsidima, Newton, and Asimakopoulou, 2010) UK RCT, single-blind, 2 parallel groups, placebo-control Patients waiting for dental appointment (340:340) Odour of 5 drops of lavender oil in 10 cc water using a candle burner, placed in waiting room No odour in waiting room using water in a candle burner (1) State Trait Anxiety Indicator (STAI-6 item) (1) Experimental group showed less STAI (p < 0.001) Although anxiety about future dental visits seems to be unaffected, lavender scent reduces state anxiety in dental patients No adverse events occurred in either group (2) Modified Dental Anxiety Scale (MDAS) (2) No sig. difference in MDAS Muzzarelli (2006) (Muzzarelli, Force, Sebold 2006) USA RCT Single-blind 2 groups (within groups) Patients undergoing a colonoscopy or esophagogastroduo￾denoscopy (118:118) Inhalation of 3 drops of 10% lavender oil + 90% grapeseed oil) in a 100 ml cup for 5 min Inhalation of 3 drops of 100% grapeseed oil in a 100 ml cup for 5 min STAI (state only) No between-group differences reported Although not statistically be effective, patients reported the lavender scent to be pleasant Not reported Braden (2009) (Braden, Reichow, Halm 2009) USA RCT, single-blind, 3 groups Pre-operative anxiety (150:150a) Group 1 Standard care + lavender aromatherapy (1 inhalation of undiluted lavandin then taped to gown and topical application – one drop on pulse point) Group 2 Standard care without aromatherapy Group 3 Standard care + jojoba oil aromatherapy (inhalation and topical application – as in Group 1) Pre-operative anxiety (100 mm VAS) Experimental group had sig. less anxiety (p < 0.01) than the control groups Lavedin was associated with significantly lower anxiety on operating room transfer, suggesting it is a simple, low-risk and cost effective intervention Not reported Motomura (2001) (Motomura, Sakurai, and Yotsuya 2001) Japan RCT Healthy volunteers (42:42a) Group 1 Inhalation of 3 ml lavender oil via diffuser for 20 min during a stressful situation (soundproof room) Group 2 stressful situation (soundproof room) for 20 min Group 3 non-stressful situation (1) Cox & McKay’s Stress/arousal adjective checklist (2) Blood pressure (3) Heart rate Sig. difference in stress/arousal scores between stress and stress/lavender condition and between stress only and non-stress condition (p < 0.01) No sig. differences between conditions for physiological measures Lavender odourants were associated with reduced mental stress Not reported Toda (2008) (Toda and Morimoto 2008) Japan RCT Healthy volunteers (30:30a) Inhalation of lavender for 10 min. Lavender oil infiltrated onto filter paper held 10 cm from nose Control (no odour) (1) Cortisol levels in saliva (2) CgA levels in saliva (3) 10-point VAS (1) No between-groups differences reported (2) In the lavender group CgA was significantly lower 10 min later but not in control gp. (3) No between-group differences reported Lavender aroma has a stress relief effect Not reported

Table 1 (Continued) First author(year) Study design Condition(sample Experimental Control treatment Main anxiety Main results Author's conclusion Adverse events ref]country size randomised: treatment outcome (between-group analysed) measurements analysis) Kutu(2008) RCT Healthy volunteers Inhalation of Control (no odour) STAI(20-item) Lavender group Lavender aroma "Some"suffered (Kutlu,Yilmaz, (95:95) incenses of during 60 min showed sig.lower inhalation decreases discomfort due to and Cecen 2008) lavender(10 per examination STA1(p<0.001) exam anxiety odour and were Turkey classroom)during than control group excluded 60min (not accounting for examination baseline） Sgoutas-Emch RCT 4 groups Healthy volunteers Group 1 Inhalation Group 2 No (1)STAI(6-item) No sig.differences Although Not reported (2001) (80:62) of lavender via lavender (but told (2)Autonomic between groups on aromatherapy by itself (Sgoutas-Emch. humidifier(with there was) Perception any measure may not be the most Fox,Preston, prior knowledge) Group 4 No Questionnaire effective treatment in Brooks,and Group 3 Inhalation lavender (and not (APQ) reducing acute stress. Serber 2001)USA of lavender(no told anything) (3)Perceived Stress ones knowledge of the prior knowledge) scae5】 procedure may relate (4)Maths Anxiety to how one responds Scale (MAS) (5)Blood pressure (BP) Perry (6)Heart rate(HR) Howard(2008) RCT DB Healthy volunteers Group 1 Two Group 2 Two (1)STAl (20-item) No sig. Previous associations Not reported (Howard and (96:92) inhalations of 3 inhalations of 3 (2)Galvanic skin between-group of lavender aroma with Hughes 2008) drops lavender oil drops tea tree oil response(GSR) difference for assisted relaxation may Ireland in a screw top jar atter an arousal aroma on ether have been influenced after an arousal task.The jar measure by expectancy biases. task.The jar remained open and relevant remained open. Group 3 no odour expectancies are easily following an manipulable arousal task Dunn(1995) RCT,open,3 groups (1)Aromatherapy Massage may offer a 2012) Patients admitted Group 1 Group 2 Massage (1)Anxiety Not due to the (Dunn,Sleep,and to Intensive Care aromatherapy with carrier oil fo behaviour scores was associated useful therapy for therapy Collett 1995)UK Unit(122:66 massage with 30 min up to 3 (4-point scale)at with sig.g乎eater nurses to consider received all three lavender oil diluted sessions the end of an reductions of when planning the reatment sessions) to 1%for 30 min up Group 3 Bed rest aromatherapy perceived anxiety psychological are of to3 sessions session at time 1(Chi2, patients in ICU.The (2)Blood pressure P-0.05) addition of lavender (BP) (2)No significant essential oil appears to (3)Heart rate (HR) ditterence between enbance the eftects of (4)Number of groups on any massage. breaths per minute physiological measure Soden(2004) RCT DB Patients with Aromatherapy Massage with inert Anxiety and No sig. The addition of Not reported (Soden,Vincent, cancer(42:36) massage with carrier oil(sweet depression (HAD) between-groups lavender oil did not Craske,Lucas, lavender oil mixed almond)30 min 1x difference appear to increase the and Ashley 2004) with carrier oil week for 4 weeks beneficial effects of UK (sweet almond)to massage in patients a dilution of 1 with advanced cancer 30 min 1x week for however,patients with 4 weeks high levels of psychological distress responded best to these therapies

828 R. Perry et al. / Phytomedicine 19 (2012) 825–835 Table 1 (Continued) First author (year) [ref] country Study design Condition (sample size randomised: analysed) Experimental treatment Control treatment Main anxiety outcome measurements Main results (between-group analysis) Author’s conclusion Adverse events Kutlu (2008) (Kutlu, Yilmaz, and C¸ ec¸ en 2008) Turkey RCT Healthy volunteers (95:95) Inhalation of incenses of lavender (10 per classroom) during 60 min examination Control (no odour) during 60 min examination STAI (20-item) Lavender group showed sig. lower STAI (p < 0.001) than control group (not accounting for baseline) Lavender aroma inhalation decreases exam anxiety “Some” suffered discomfort due to odour and were excluded Sgoutas-Emch (2001) (Sgoutas-Emch, Fox, Preston, Brooks, and Serber 2001) USA RCT 4 groups Healthy volunteers (80:62) Group 1 Inhalation of lavender via humidifier (with prior knowledge) Group 3 Inhalation of lavender (no prior knowledge) Group 2 No lavender (but told there was) Group 4 No lavender (and not told anything) (1) STAI (6-item) (2) Autonomic Perception Questionnaire (APQ) (3) Perceived Stress Scale (PSS) (4) Maths Anxiety Scale (MAS) (5) Blood pressure (BP) (6) Heart rate (HR) No sig. differences between groups on any measure Although aromatherapy by itself may not be the most effective treatment in reducing acute stress, ones knowledge of the procedure may relate to how one responds Not reported Howard (2008) (Howard and Hughes 2008) Ireland RCT DB Healthy volunteers (96:92) Group 1 Two inhalations of 3 drops lavender oil in a screw top jar after an arousal task. The jar remained open. Group 2 Two inhalations of 3 drops tea tree oil after an arousal task. The jar remained open Group 3 no odour following an arousal task (1) STAI (20-item) (2) Galvanic skin response (GSR) No sig. between-group difference for aroma on either measure Previous associations of lavender aroma with assisted relaxation may have been influenced by expectancy biases, and relevant expectancies are easily manipulable Not reported Dunn (1995) (Dunn, Sleep, and Collett 1995) UK RCT, open, 3 groups Patients admitted to Intensive Care Unit (122:66 received all three treatment sessions) Group 1 aromatherapy massage with lavender oil diluted to 1%, for 30 min up to 3 sessions Group 2 Massage with carrier oil for 30 min up to 3 sessions Group 3 Bed rest (1) Anxiety behaviour scores (4-point scale) at the end of an aromatherapy session (2) Blood pressure (BP) (3) Heart rate (HR) (4) Number of breaths per minute (1) Aromatherapy was associated with sig. greater reductions of perceived anxiety at time 1 (Chi2, p = 0.05) (2) No significant difference between groups on any physiological measure Massage may offer a useful therapy for nurses to consider when planning the psychological are of patients in ICU. The addition of lavender essential oil appears to enhance the effects of massage. Not due to the therapy Soden (2004) (Soden, Vincent, Craske, Lucas, and Ashley 2004) UK RCT DB Patients with cancer (42:36) Aromatherapy massage with lavender oil mixed with carrier oil (sweet almond) to a dilution of 1%, 30 min 1× week for 4 weeks Massage with inert carrier oil (sweet almond) 30 min 1× week for 4 weeks Anxiety and depression (HAD) No sig. between-groups difference The addition of lavender oil did not appear to increase the beneficial effects of massage in patients with advanced cancer however, patients with high levels of psychological distress responded best to these therapies Not reported

Woelk(2010) RCT,DB.active Generalised 80 mg/day of an oral Lorazepam (1)Hamilton (1)HAMA Silexan is as effective 20 pps suffered (Woelk and control, anxiety disorder Lavender extract 0.5 mg/day)for 6 Anxiety Scale decreased by4 and well-tolerated as from 26 cases of Schlafke 2010) double-dummy (77:69) (Silexan)for 6 weeks weeks (HAMA) (lavender)versus lorazepam in adults AEs in the Silexan Germany (2)Penn State 46%(Lorazepam) with GAD group (mainly Worry Inventory (2)and (3)PSWS gastrointestinal (PSWS) and SAS decreased disorders)18 pps (3)Zung to a similar extent suffering from 19 Self-Rating Anxiety in both groups AEs in the Scale(SAS) lorezepam group (mainly fatigue/sedative effects) Kasper (2010) RCT DB pc Subsyndromal' 80 mg/day of an oral 1x day placebo Hamilton Anxiety HAMA decreased Lavender oil is both 30 pps(Silexan (Kasper.Gastpar. anxiety disorder lavender oil extract (0.08 mg of Scale(HAMA) by 59.3%with efficacious and safe for group)reported 55 Iler,Volz,ller. (221:212) (Silexan)for 10 weeks lavender so lavender oil versus the relief of AEs and 35pps and Dienel 2010) identical for smell) 35.4%with placebo (unspecified)anxiety (placebo group) Germany for 10 weeks (pc0.01) disorder reported 68 AEs. Four serious but not attributable 8 events occurred and all AEs that Perry resulted in withdrawal were not related to treatment Bradley(2009) RCT DB Healthy volunteers (1)Gelatine capsules Placebo gelatine (1)STAI-Y state Trend for200μl Lavender has anxiolytic Not reported (Bradley.Brown, (97:96) with sunflower oil and capsule with just anxiety scale lavender had more effects under Chu,and Lea lavender (100 ul)1x sunflower oil 1x (3)Heart rate(HR) of an eftect in low conditions of low 元ed 2009)UK (2)Gelatine capsules (4)Galvanic skin but not under high anxiety,but may not with sunflower oil and response(GSR) anxiety on both extend to conditions of avender200】】x physiological and high anxiety self-report measures (p<0.05) 1825 Xu(2008)(XL RCT single-blind 3 Healthy volunteers Ayurvedic oil-dripping (1)Ayurvedic (1)STAI at end of (1)No sig. The complicated Not reported Uebaba.Ogawa, groups(within (16:16) (sesame oil containing oil-dripping with procedure difference between pharmaco-physio- Tatsuse,Wang. groups) 0.3%lavender oil)1 sesame carrier oil (2)Heart rate(HR) lavender and psychologic action of Hisajima. session onlv 2]No ou-dnpping carner oil-dripping ayruvedic oil treatment Venkatraman but significant may provide a useful 2008)Japan difference between model for future both oil-dripping pharmaco-physio- conditions and the psychotherapy. control in STAI (2)HR-no significant differences found between groups Morris (2002) RCT single blind Healthy volunteers 10 min bath/day using 10 min bath/day UWIST Mood No between-group The results are Not reported (Morris 2002)UK (40:40) 3 pipettes of lavender using only Adjective Checklist differences encouraging and oil (20%)+grapeseed grapeseed oil reported on tense further work is needed oil (80%)added to bath (100元)for14days arousal scale of the water for 14 days UWIST DB:double blind:RCT:randomised controlled trial:pp:participants:pc:placebo-controlled:sig.:significant:AFs:adverse events. Assumed there were no drop-outs due to short trial duration (not directly reported) 臣

R. Perry et al. / Phytomedicine 19 (2012) 825–835 829 Woelk (2010) (Woelk and Schlafke 2010) Germany RCT, DB, active control, double-dummy Generalised anxiety disorder (77:69) 80 mg/day of an oral Lavender extract (Silexan) for 6 weeks Lorazepam 0.5 mg/day) for 6 weeks (1) Hamilton Anxiety Scale (HAMA) (2) Penn State Worry Inventory (PSWS) (3) Zung Self-Rating Anxiety Scale (SAS) (1) HAMA decreased by 45% (lavender) versus 46% (Lorazepam) (2) and (3) PSWS and SAS decreased to a similar extent in both groups Silexan is as effective and well-tolerated as lorazepam in adults with GAD 20 pps suffered from 26 cases of AEs in the Silexan group (mainly gastrointestinal disorders), 18 pps suffering from 19 AEs in the lorezepam group (mainly fatigue/sedative effects) Kasper (2010) (Kasper, Gastpar, ller, Volz, ller, and Dienel 2010) Germany RCT DB pc ‘Subsyndromal’ anxiety disorder (221:212) 80 mg/day of an oral lavender oil extract (Silexan) for 10 weeks 1× day placebo (0.08 mg of lavender so identical for smell) for 10 weeks Hamilton Anxiety Scale (HAMA) HAMA decreased by 59.3% with lavender oil versus 35.4% with placebo (p < 0.01) Lavender oil is both efficacious and safe for the relief of (unspecified) anxiety disorder 30 pps (Silexan group) reported 55 AEs and 35pps (placebo group) reported 68 AEs. Four serious but not attributable events occurred and all AEs that resulted in withdrawal were not related to treatment Bradley (2009) (Bradley, Brown, Chu, and Lea 2009) UK RCT DB Healthy volunteers (97:96) (1) Gelatine capsules with sunflower oil and lavender (100 l) 1× (2) Gelatine capsules with sunflower oil and Lavender (200 l) 1× Placebo gelatine capsule with just sunflower oil 1× (1) STAI-Y state anxiety scale (3) Heart rate (HR) (4) Galvanic skin response (GSR) Trend for 200 l lavender had more of an effect in low but not under high anxiety on both physiological and self-report measures (p < 0.05) Lavender has anxiolytic effects under conditions of low anxiety, but may not extend to conditions of high anxiety Not reported Xu (2008) (Xu, Uebaba, Ogawa, Tatsuse, Wang, Hisajima, Venkatraman 2008) Japan RCT single-blind 3 groups (within groups) Healthy volunteers (16:16a) Ayurvedic oil-dripping (sesame oil containing 0.3% lavender oil) 1 session only (1) Ayurvedic oil-dripping with sesame carrier oil (2) No oil-dripping (1) STAI at end of procedure (2) Heart rate (HR) (1) No sig. difference between lavender and carrier oil-dripping but significant difference between both oil-dripping conditions and the control in STAI (2) HR – no significant differences found between groups The complicated pharmaco-physio￾psychologic action of ayruvedic oil treatment may provide a useful model for future pharmaco-physio￾psychotherapy. Not reported Morris (2002) (Morris 2002) UK RCT single blind Healthy volunteers (40:40) 10 min bath/day using 3 pipettes of lavender oil (20%) + grapeseed oil (80%) added to bath water for 14 days 10 min bath/day using only grapeseed oil (100%) for 14 days UWIST Mood Adjective Checklist No between-group differences reported on tense arousal scale of the UWIST The results are encouraging and further work is needed Not reported DB: double blind; RCT: randomised controlled trial; pp: participants; pc: placebo-controlled; sig.: significant; AEs: adverse events. a Assumed there were no drop-outs due to short trial duration (not directly reported).

Table 2 Methodological quality of trials. First author(date) Was the Was the Was the Were Who was Was the Was the Was the In addition, Was Was an a Was the ladad score Where treatment randomisa- treatment groups blinded? trial method of number ot were analysis priori therapeutic (Max. relevant. allocation tion allocation similar at described blinding with- reasons conducted power time score-5) how many described procedure concealed? baseline on as double- described drawals/ given for on the caloulation equivalent items in as random- described prognostic blind? and appro- dropouts in each group? Intent-to- described? between Section 4 of ized? and was it ndicators？ priate? each group treat groups? the herbal- appropri- men- group? specific ate? tioned? CONSORT statement were described fully and partly (F/P)? Kritsidima(2010) Yes Not NR Yes Patient No(SB) Yes No drop NA Yes Yes Yes Perry described outs Muzzarelli(2006) Yes Yes NR Yes Patient No(SB) NA No drop NA Yes* No Yes outs Braden(2009) Yes Yes R Nof Patient No(SB) NA NRd NR Yes Motomura(2001) Yes Not Unclear No blinding No NA NR NR NR No S 1 described Toda(2008) Yes Not NR Unclear Unclear No NA NRd Not by NRb Yes described (HV) group Kutu(2008) Yes No NR Unclear No blinding No NA No Not by Unclear No Yes 0 (HV) group Sgoutas-Emch(2001)Yes Not NR No! Participants No NA Yes by Not by No Yes 192012) blind in 2 乎oup group conditions 825 Howard(2008 Yes No NR Unclear Participants:Yes Unclear Yes by Not by No Yes 2 835 (HV) outcome group group assessors Dunn(1995) Yes Not Yes Yes Outcome No NA Not by Not by NR Ye Yes 1 described assessork 乎up group Soden (2004) Yes Not Yes Noi Participants:Yes No Not by Not b NR Yes 1 described outcome group group assessors Woelk (2010) Yes Yes NR Unclear Patient, Yes Yes! Yes by Yes by Yes Yes 5/1 outcome group group assessor unclear Kasper(2010) Yes Yes Outcome Yes Yes (IP) Yes by Yes by Yes 21 assessor, group group patient Bradley(2009) Yes No NR Yes(HV) Unclear Yes Unclear NR No NR9 Yes 63

830 R. Perry et al. / Phytomedicine 19 (2012) 825–835 Table 2 Methodological quality of trials. First author (date) Was the treatment allocation described as random￾ized? Was the randomisa￾tion procedure described and was it appropri￾ate? Was the treatment allocation concealed? Were groups similar at baseline on prognostic indicators? Who was blinded? Was the trial described as double￾blind? Was the method of blinding described and appro￾priate? Was the number of with￾drawals/ dropouts in each group men￾tioned? In addition, were reasons given for each group? Was analysis conducted on the intent-to￾treat group? Was an a priori power calculation described? Was the therapeutic time equivalent between groups? Jadad score (Max. score = 5) Where relevant, how many items in Section 4 of the herbal￾specific CONSORT statement were described fully and partly (F/P)? Kritsidima (2010) Yes Not described NR Yes Patient No (SB) Yes No drop outs NA Yesa Yes Yes 1 Muzzarelli (2006) Yes Yes NR Yes Patient No (SB) NA No drop outs NA Yesa No Yes 1 Braden (2009) Yes Yes NR Noc Patient No (SB) NA NRd NR NRb Yes Yes 1 Motomura (2001) Yes Not described NR Unclear No blinding No NA NRd NR NRb No Noe 1 Toda (2008) Yes Not described NR Unclear (HV) Unclear No NA NRd Not by group NRb No Yes 1 Kutlu (2008) Yes No NR Unclear (HV) No blinding No NA No Not by group Unclear No Yes 0 Sgoutas-Emch (2001) Yes Not described NR Nof Participants blind in 2 conditions No NA Yes by group Not by group No No Yes 1 Howard (2008 Yes No NR Unclear (HV) Participants; outcome assessors Yes Unclear Yes by group Not by group NR No Yes 2 Dunn (1995) Yes Not described Yes Yes Outcome assessorg No NA Not by group Not by group NR Yesh Yesi 1 Soden (2004) Yes Not described Yes Noj Participants; outcome assessors Yes No Not by group Not by group NR Yesk Yes 1 Woelk (2010) Yes Yes NR Unclear Patient, outcome assessor unclear Yes Yesl Yes by group Yes by group Yes No Yes 4 5/1 Kasper (2010) Yes Yes Unclearm Yesn Outcome assessor, patient Yes Yes (IP) Yes by group Yes by group Noo Yesp Yes 4 2/1 Bradley (2009) Yes No NR Yes (HV) Unclear Yes Unclear NR No NRq No Yes 1 6/3

R.Perry et al Phytomedicine 19 (2012)825-835 831 divided each condition into three different cognitive primes (assist relaxation,inhibit relaxation or no information)to test whether it was expectancy of treatment that brought about a reduction in stress levels.The lavender condition did not have a significant effect on relaxation but the type of prime they received did have a near- significant impact(inhibiting prime).The authors concluded that it was the priming condition,rather than the lavender,which exerted the observed effects. Aromatherapy massage Two trials assessed the effectiveness of lavender using aro matherapy massage.Dunn et al.(1995)found a significant effect on observed or self-reported anxiety levels of lavender aromatherapy (lavender vera)massage compared to undisturbed bed rest(p=0.05) in patients in an Intensive Care Unit.This effect was only found after the first massage session.However,it is unsurprising to find an effect of something over nothing.More importantly,no signif- icant difference was found between the lavender massage group 鱼 and the carrier oil massage group.Attempts to blind the outcome assessors were made by using different nurses to conduct pre-post assessments.Methodological issues limit the conclusiveness of the findings.The authors recognised that sense of smell may be dis- torted in ICUs due to face-masks and nasal tubes,which would impact on the results. Soden et al.(2004)compared anxiety scores between massage groups with and without lavender in cancer patients.No signifi- 星 cant differences were found within or between groups following massage,despite the non-lavender group having higher anxiety at baseline (HAD score (>19)(p=0.03)).Both the non-lavender mas- sage group and control group were underpowered at baseline. Oral lavender Three trials(Bradley et al.2009;Kasper et al.2010;Woelk and Schlafke 2010)assessed oral lavender.In a high quality inves- C85)0N tigation,Woelk and Schlafke (2010)assessed the efficacy of an oral lavender preparation(Silexan)compared to lorazepam,a stan- dardised anti-anxiety drug for general anxiety disorder and found no significant difference in anxiety levels as measured by the Hamilton Anxiety Rating Scale(HAMA).Between-group Cls did not demonstrate a significant difference,but within groups difference indicates that both treatments reduced anxiety levels,implying that lavender was as effective an anxiolytic as lorazepam.A one- week washout period prior to study commencement was used to prevent the effects of other drugs confounding the results.Analy sis was conducted on both the intention-to-treat and per-protocol sample.Any adverse events(AE)were reported;none were serious but patients suffered marginally more AEs in the lavender group (mainly gastrointestinal problems)than the lorazepam group (mainly fatigue).Although this trial would have benefited from a third arm placebo group,it was well conducted and well reported and scored 4 on the Jadad scale. In a placebo-controlled,double-blind trial,which scored 4 on the Jadad scale,Kasper et al.(2010)tested the efficacy of oral lavender oil capsules(Silexan)over 10 weeks for'subsyndromal'anxiety dis- order (according to DMS-IV or ICD-10).Anxiety levels decreased significantly more in the intervention group (p<0.01)compared to the control according to the HAMA.Although authors reported conducting intention-to-treat analysis,data from only 212 of the 221 randomised patients were analysed (LOCF),which resulted in (8002c)nx a slightly underpowered study (N=110 needed in each condition) after five participants were excluded following the initial recruit- ment phase.Both these trials missed out on the final Jadad point for not clearly reporting whether treatment allocation was concealed

R. Perry et al. / Phytomedicine 19 (2012) 825–835 831 Xu (2008) Yes No NR Yesr Participant blind No (SB) NA NRd NRb No No Yes 0 Morris (2002) Yes Yes NR Unclears Participant blind No (SB) NA No drop outs No Yesa No Yes 2 SB – single blind, IP- indistinguishable placebo, NR – not reported, OT – open trial, NA – not applicable; HV – Healthy volunteers. a Due to no drop-outs. b Assumed there were no drop outs due to short duration of intervention thus ITT conducted. c Significantly smaller number in control group had history of anxiety/panic disorder. d Dropouts were not directly reported. e Therapeutic time was equivalent between the two stressful conditions but not the non-stressful control. f Group 3 had sig. higher levels of anxiety at baseline. g Attempts were made to blind outcome assessor. h Fell underpowered after baseline. i Therapeutic time equivalent between massage groups only. j Sig. more cases of anxiety/depression (HAD score > 19) in massage group than the aromatherapy group. k Underpowered. l Unclear how outcome assessors were blinded. m Pharmacy-controlled. n CGI1 scores were higher in placebo group but did not reach significance. o Reported as ITT but wasn’t in its strictest sense (221 randomised but 112 analysed). p Fell below power after 5 patients excluded. q 96 in methods and 97 in table. r Same participants (cross-over trial). s Baseline measures reported but no statistical analysis provided. divided each condition into three different cognitive primes (assist relaxation, inhibit relaxation or no information) to test whether it was expectancy of treatment that brought about a reduction in stress levels. The lavender condition did not have a significant effect on relaxation but the type of prime they received did have a near￾significant impact (inhibiting prime). The authors concluded that it was the priming condition, rather than the lavender, which exerted the observed effects. Aromatherapy massage Two trials assessed the effectiveness of lavender using aro￾matherapy massage. Dunn et al. (1995) found a significant effect on observed or self-reported anxiety levels of lavender aromatherapy (lavender vera)massage comparedtoundisturbedbedrest(p = 0.05) in patients in an Intensive Care Unit. This effect was only found after the first massage session. However, it is unsurprising to find an effect of something over nothing. More importantly, no signif￾icant difference was found between the lavender massage group and the carrier oil massage group. Attempts to blind the outcome assessors were made by using different nurses to conduct pre-post assessments. Methodological issues limit the conclusiveness of the findings. The authors recognised that sense of smell may be dis￾torted in ICUs due to face-masks and nasal tubes, which would impact on the results. Soden et al. (2004) compared anxiety scores between massage groups with and without lavender in cancer patients. No signifi- cant differences were found within or between groups following massage, despite the non-lavender group having higher anxiety at baseline (HAD score (>19) (p = 0.03)). Both the non-lavender mas￾sage group and control group were underpowered at baseline. Oral lavender Three trials (Bradley et al. 2009; Kasper et al. 2010; Woelk and Schlafke 2010) assessed oral lavender. In a high quality inves￾tigation, Woelk and Schlafke (2010) assessed the efficacy of an oral lavender preparation (Silexan) compared to lorazepam, a stan￾dardised anti-anxiety drug for general anxiety disorder and found no significant difference in anxiety levels as measured by the Hamilton Anxiety Rating Scale (HAMA). Between-group CIs did not demonstrate a significant difference, but within groups difference indicates that both treatments reduced anxiety levels, implying that lavender was as effective an anxiolytic as lorazepam. A one￾week washout period prior to study commencement was used to prevent the effects of other drugs confounding the results. Analy￾sis was conducted on both the intention-to-treat and per-protocol sample. Any adverse events (AE) were reported; none were serious, but patients suffered marginally more AEs in the lavender group (mainly gastrointestinal problems) than the lorazepam group (mainly fatigue). Although this trial would have benefited from a third arm placebo group, it was well conducted and well reported and scored 4 on the Jadad scale. Inaplacebo-controlled,double-blindtrial, whichscored4 onthe Jadad scale, Kasper et al. (2010) tested the efficacy of oral lavender oil capsules (Silexan) over 10 weeks for ‘subsyndromal’ anxiety dis￾order (according to DMS-IV or ICD-10). Anxiety levels decreased significantly more in the intervention group (p < 0.01) compared to the control according to the HAMA. Although authors reported conducting intention-to-treat analysis, data from only 212 of the 221 randomised patients were analysed (LOCF), which resulted in a slightly underpowered study (N = 110 needed in each condition) after five participants were excluded following the initial recruit￾ment phase. Both these trials missed out on the final Jadad pointfor not clearly reporting whether treatment allocation was concealed

832 R.Perry et al.Phytomedicine 19(2012)825-835 Total number retrieved from electronic search: 440 Duplicates removed: Excluded (Animal study): 192 56 Abstracts studied for inclusion: 192 Excluded:Not Anxiety or Stress: 31 Excluded:Not Lavender 21 Excluded:Not RCT: 120 Excluded:Not solely Lavender:5 22 事 Relevant articles to include: 15 Fig.1.Flow chart of the study selection process. Bradley et al.(2009)also tested oral lavender in a double blind any condition.This single-blind,within-groups design only used 16 RCT by inducing anxiety in healthy participants using extracts from participants.A lack of power calculation also calls into question the two anxiety-provoking films.They tested different doses of laven- robustness of the findings. der (200 ul and 100 ul)versus placebo and found a significantly greater decrease in state anxiety after the neutral film clip after Bathing in lavender both doses of lavender,but not following the anxiety-inducing film clip which implies that lavender may have an effect on general anx- Morris(2002)found no significant between-group difference in iety states but not when the levels of anxiety get too high.Similar anxiety scores (using the UWIST tense score)in this double-blind effects were found for the physiological measures (HR and GSR) RCT using healthy volunteers randomised to a lavender oil bath or Methodological reporting was often unclear (Table 2)and report- non-scented bath;although bathing helped reduce arousal scores ing of numbers was confused(different numbers reported in tables in both groups. than in text). Discussion Dripping oil Our review included 15 RCTs of lavender for anxiety or stress. Xu et al.(2008)tested Shirodhara (a form of Ayurveda medicine The majority of trials used either the STAl or HAD as the main that involves pouring oil over the forehead or the 'third eye'(Anon self-reported outcome measure,whilst blood pressure,heart rate 2011))on healthy volunteers and found no significant difference and galvanic skin response were reported as physiological outcome in STAI between oil dripping conditions but a significant differ- measures.Seven trials(Bradenet al.2009;Bradleyetal.2009;Dunn ence between both oil dripping conditions and the control (supine et al.1995:Kasper et al.2010;Kritsidima et al.2010:Kutlu et al. position only).STAI was only measured at the end of the proce- 2008:Motomura et al.2001)showed results appearing to favour dure.No significant differences were found in heart rate between the lavender intervention for at least one relevant outcome and one

832 R. Perry et al. / Phytomedicine 19 (2012) 825–835 Total number retrieved from electronic search: 440 Duplicates removed: 192 Abstracts studied for inclusion: 192 Relevant articles to include: 15 Excluded (Animal study): 56 Excluded: Not Anxiety or Stress: 31 Excluded: Not RCT: 120 Excluded: Not Lavender: 21 Excluded: Not solely Lavender: 5 22 Fig. 1. Flow chart of the study selection process. Bradley et al. (2009) also tested oral lavender in a double blind RCT by inducing anxiety in healthy participants using extracts from two anxiety-provoking films. They tested different doses of laven￾der (200 l and 100 l) versus placebo and found a significantly greater decrease in state anxiety after the neutral film clip after both doses of lavender, but not following the anxiety-inducing film clip which implies thatlavender may have an effect on general anx￾iety states but not when the levels of anxiety get too high. Similar effects were found for the physiological measures (HR and GSR). Methodological reporting was often unclear (Table 2) and report￾ing of numbers was confused (different numbers reported in tables than in text). Dripping oil Xu et al. (2008) tested Shirodhara (a form of Ayurveda medicine that involves pouring oil over the forehead or the ‘third eye’ (Anon 2011)) on healthy volunteers and found no significant difference in STAI between oil dripping conditions but a significant differ￾ence between both oil dripping conditions and the control (supine position only). STAI was only measured at the end of the proce￾dure. No significant differences were found in heart rate between any condition. This single-blind, within-groups design only used 16 participants. A lack of power calculation also calls into question the robustness of the findings. Bathing in lavender Morris (2002) found no significant between-group difference in anxiety scores (using the UWIST tense score) in this double-blind RCT using healthy volunteers randomised to a lavender oil bath or non-scented bath; although bathing helped reduce arousal scores in both groups. Discussion Our review included 15 RCTs of lavender for anxiety or stress. The majority of trials used either the STAI or HAD as the main self-reported outcome measure, whilst blood pressure, heart rate and galvanic skin response were reported as physiological outcome measures. Seven trials (Braden et al. 2009; Bradley et al. 2009; Dunn et al. 1995; Kasper et al. 2010; Kritsidima et al. 2010; Kutlu et al. 2008; Motomura et al. 2001) showed results appearing to favour the lavender intervention for atleast one relevant outcome and one

R.Perry et al Phytomedicine 19 (2012)825-835 833 comparative trial by Woelk and Schlafke(2010)found thatlavender Furthermore,the majority of the included trials did not pro- was as effective as the anxiolytic drug lorazepam.Of the remaining vide reasons by group for dropouts.This is a frequent phenomenon studies,the findings of one RCT(Muzzarelli et al.2006)could not be in research of this type and there is a common misconception evaluated because it did not provide between-group comparisons that "natural"means safe (Ernst 2007a).Researchers investigating of change in outcome from baseline.The remaining six trials did not botanical products should comply with the CONSORT guidelines demonstrate significant differences between groups.However,all for the reporting of herbal products (Gagnier et al.2006).None results should be considered in the context of their methodological of the included trials of oral lavender (Bradley et al.2009;Kasper limitations. et al.2010:Woelk and Schlafke 2010)provided information that The results suggest limited specific effects of lavender inhalation met more than 9 of the 15 CONSORT statement criteria regarding and massage on anxiety measures.Although reductions in anxiety the extraction and preparation of herbs.Future trials should report measures were observed,methodological issues limit the extent to such information to allow adequate between-trial comparison of which firm conclusions can be drawn.On the other hand,promising the results to be made. evidence regarding ingested lavender preparations was found. The reporting of many of the trials was generally poor;all were Safety described as randomised but only five gave details on randomisa- tion process(Braden et al.2009;Kasper et al.2010;Morris 2002; The current data suggests that lavender is relatively safe, Muzzarelli et al.2006:Woelk and Schlafke 2010)and two on treat- although the gastrointestinal effects of oral lavender being taken in ment allocation concealment(Dunn et al.1995;Soden et al.2004) the long term haven't been assessed.Of a total of 15 randomised tri- Five trials(Braden et al.2009:Dunn et al.1995:Kasper et al.2010: als only 5 trials(Dunn et al.1995:Kasper et al.2010:Kritsidima etal. Kritsidima et al.2010:Soden et al.2004)included power calcu- 2010:Kutlu et al.2008;Woelk and Schlafke 2010)reported infor- lations,although one was underpowered at baseline (Soden et al mation on adverse events.One(Dunn et al.1995)had AEs that were 2004)and two became marginally underpowered following par- not related to the therapy.The remaining ten did not mention AEs at ticipant drop out(Dunn et al.1995;Kasper et al.2010).Most trials all.It is good practice for all trials to report on adverse events even if reported that participants were similar on prognostic indicators, unrelated to the therapy.In these RCTs,most of the AEs mentioned although this was assumed rather than tested in the majority of tri- were minor ailments.Kutlu et al.(2008)had "some"participants als that used healthy participants.Three trials(Braden et al.2009; reporting discomfort from the odour thus were excluded.In the oral Sgoutas-Emch et al.2001:Soden et al.2004)reported differences in lavender trials,Kasper et al.(2010)reported slightly more adverse anxiety levels at baseline.In addition,the trials differed greatly in events from the placebo group than the lavender group;the most terms of the conditions treated,and independent replications are frequently reported AEs were related to infections and infestations. missing for most. followed by gastrointestinal disorders and nervous system disor- Five trials were described as double-blind (Bradley et al.2009: ders.Woelk and Schlafke (2010)reported slightly more adverse Howard and Hughes 2008;Kasper et al.2010;Soden et al.2004; events in the lavender group than the lorazepam group but again Woelk and Schlafke 2010).although in two (Bradley et al.2009: none were described as serious.Since lavender oil showed no seda- Woelk and Schlafke 2010),blinding details were not clearly pro tive effects and has no potential for drug abuse,lavender appears to vided.Five trials(Braden et al.2009;Kritsidima et al.2010;Morris be an effective and well-tolerated alternative to benzodiazepines 2002:Muzzarelli et al.2006:Xu et al.2008)were described as for amelioration of generalised anxiety.Lavender should,however, single-blind with the patient/participant being blinded.As laven- be used with caution particularly as lavender oil can also be a pow der is a particularly distinctive and recognisable odour,blinding is erful allergen and,as a precaution,ingestion should be avoided unrealistic in most of the trials (i.e.trials comparing lavender aro- during pregnancy due to emmenagogue effects (Ernst 2002)and matherapy versus no odour(Bradley et al.2009;Soden et al.2004). when breastfeeding. or alternative odour(Howard and Hughes 2008)).Blinding was pos- Given funding for CAM research is difficult to obtain and our sible in trials that used oral lavender capsules(Kasper et al.2010; review did not identify convincing evidence for lavender aro- Woelk and Schlafke 2010).where smell would not be detected. matherapy,massage or oil-dripping.additional research should Specific odours in general may provide the calmative effects via focus on replication of oral lavender.Future trials of oral laven- associative memory evocation rather than any direct pharmaco- der would benefit from adopting a good trial design and stringent logical action.This possibility needs to be addressed in future trials reporting to enable replication:this would include adopting a perhaps by comparing lavender with an alternative odours in addi- randomised design with allocation concealment,triple blind (if tion to 'no odour'conditions,as considered in Howard and Hughes possible)and indistinguishable placebos.Intention-to-treat anal- (2008).Also,a single-blind design would mean the outcome asses- yses and a priori power calculations should be conducted and sor was aware of the condition and given that they reported on reported.All withdrawals,dropouts and adverse events should be treatment effectiveness in one trial(Dunn et al.1995).blinding of fully reported giving the number and reason by group.It is essen- outcome assessor is essential to reduce the effect of demand char- tial to get good tolerability and safety data for all modes of lavender acteristics or the Hawthorne effect.Failure to blind may therefore application,thus longer-term follow ups would be required for oral have increased the risk of bias. lavender before it is recommended as an alternative treatment for Withdrawals/dropouts were generally under-reported.In par- anxiety. ticular,the trials that induced an anxiety state were particularly poor at reporting this information and in many cases it was Limitations assumed there were no dropouts due to the short duration of the tri- als,although this was not explicitly stated.Three trials(Kritsidima As described above,the methodological issues observed in the et al.2010:Morris 2002:Muzzarelli et aL 2006)used intention-to- included studies are notable and limit the extent to which the anxi- treat analyses (ITT),two (Kritsidima et al.2010:Muzzarelli et al. olytic properties of lavender can be evaluated.The methodological 2006)were the result of not having any dropouts.A further four issues of the lavender inhalation trials are most difficult to manage; trials(Braden et al.2009;Morris 2002:Motomura et al.2001:Toda trying to ensure that consistent levels of lavender odour is released and Morimoto 2008)also appeared to use ITT analyses due to no between groups,or when attempting successful replication of such drop-outs but again this was only assumed due to the short dura- trials,can be particularly hard to control.In addition,although the tion of the trial. search strategy was thorough,some clinical trials may not have

R. Perry et al. / Phytomedicine 19 (2012) 825–835 833 comparative trial byWoelk and Schlafke (2010)found thatlavender was as effective as the anxiolytic drug lorazepam. Of the remaining studies,the findings of one RCT (Muzzarelli et al. 2006) could not be evaluated because it did not provide between-group comparisons of change in outcome from baseline. The remaining six trials did not demonstrate significant differences between groups. However, all results should be considered in the context of their methodological limitations. The results suggestlimited specific effects oflavender inhalation and massage on anxiety measures. Although reductions in anxiety measures were observed, methodological issues limit the extent to which firm conclusions can be drawn. On the other hand, promising evidence regarding ingested lavender preparations was found. The reporting of many of the trials was generally poor; all were described as randomised but only five gave details on randomisa￾tion process (Braden et al. 2009; Kasper et al. 2010; Morris 2002; Muzzarelli et al. 2006; Woelk and Schlafke 2010) and two on treat￾ment allocation concealment (Dunn et al. 1995; Soden et al. 2004). Five trials (Braden et al. 2009; Dunn et al. 1995; Kasper et al. 2010; Kritsidima et al. 2010; Soden et al. 2004) included power calcu￾lations, although one was underpowered at baseline (Soden et al. 2004) and two became marginally underpowered following par￾ticipant drop out (Dunn et al. 1995; Kasper et al. 2010). Most trials reported that participants were similar on prognostic indicators, although this was assumed rather than tested in the majority of tri￾als that used healthy participants. Three trials (Braden et al. 2009; Sgoutas-Emch et al. 2001; Soden et al. 2004) reported differences in anxiety levels at baseline. In addition, the trials differed greatly in terms of the conditions treated, and independent replications are missing for most. Five trials were described as double-blind (Bradley et al. 2009; Howard and Hughes 2008; Kasper et al. 2010; Soden et al. 2004; Woelk and Schlafke 2010), although in two (Bradley et al. 2009; Woelk and Schlafke 2010), blinding details were not clearly pro￾vided. Five trials (Braden et al. 2009; Kritsidima et al. 2010; Morris 2002; Muzzarelli et al. 2006; Xu et al. 2008) were described as single-blind with the patient/participant being blinded. As laven￾der is a particularly distinctive and recognisable odour, blinding is unrealistic in most of the trials (i.e. trials comparing lavender aro￾matherapy versus no odour (Bradley et al. 2009; Soden et al. 2004), or alternative odour (Howard and Hughes 2008)). Blinding was pos￾sible in trials that used oral lavender capsules (Kasper et al. 2010; Woelk and Schlafke 2010), where smell would not be detected. Specific odours in general may provide the calmative effects via associative memory evocation rather than any direct pharmaco￾logical action. This possibility needs to be addressed in future trials, perhaps by comparing lavender with an alternative odours in addi￾tion to ‘no odour’ conditions, as considered in Howard and Hughes (2008). Also, a single-blind design would mean the outcome asses￾sor was aware of the condition and given that they reported on treatment effectiveness in one trial (Dunn et al. 1995), blinding of outcome assessor is essential to reduce the effect of demand char￾acteristics or the Hawthorne effect. Failure to blind may therefore have increased the risk of bias. Withdrawals/dropouts were generally under-reported. In par￾ticular, the trials that induced an anxiety state were particularly poor at reporting this information and in many cases it was assumedthere werenodropoutsdue to the shortdurationofthe tri￾als, although this was not explicitly stated. Three trials (Kritsidima et al. 2010; Morris 2002; Muzzarelli et al. 2006) used intention-to￾treat analyses (ITT), two (Kritsidima et al. 2010; Muzzarelli et al. 2006) were the result of not having any dropouts. A further four trials (Braden et al. 2009; Morris 2002; Motomura et al. 2001; Toda and Morimoto 2008) also appeared to use ITT analyses due to no drop-outs but again this was only assumed due to the short dura￾tion of the trial. Furthermore, the majority of the included trials did not pro￾vide reasons by group for dropouts. This is a frequent phenomenon in research of this type and there is a common misconception that “natural” means safe (Ernst 2007a). Researchers investigating botanical products should comply with the CONSORT guidelines for the reporting of herbal products (Gagnier et al. 2006). None of the included trials of oral lavender (Bradley et al. 2009; Kasper et al. 2010; Woelk and Schlafke 2010) provided information that met more than 9 of the 15 CONSORT statement criteria regarding the extraction and preparation of herbs. Future trials should report such information to allow adequate between-trial comparison of the results to be made. Safety The current data suggests that lavender is relatively safe, although the gastrointestinal effects of oral lavender being taken in the long term haven’t been assessed. Of a total of 15 randomised tri￾als only 5 trials (Dunnet al. 1995;Kasper et al. 2010;Kritsidima et al. 2010; Kutlu et al. 2008; Woelk and Schlafke 2010) reported infor￾mation on adverse events. One (Dunn et al. 1995) had AEs that were not related to the therapy. The remaining ten did not mentionAEs at all. Itis good practice for alltrials to report on adverse events even if unrelated to the therapy. In these RCTs, most of the AEs mentioned were minor ailments. Kutlu et al. (2008) had “some” participants reporting discomfortfrom the odour thus were excluded. In the oral lavender trials, Kasper et al. (2010) reported slightly more adverse events from the placebo group than the lavender group; the most frequently reported AEs were related to infections and infestations, followed by gastrointestinal disorders and nervous system disor￾ders. Woelk and Schlafke (2010) reported slightly more adverse events in the lavender group than the lorazepam group but again none were described as serious. Since lavender oil showed no seda￾tive effects and has no potential for drug abuse, lavender appears to be an effective and well-tolerated alternative to benzodiazepines for amelioration of generalised anxiety. Lavender should, however, be used with caution particularly as lavender oil can also be a pow￾erful allergen and, as a precaution, ingestion should be avoided during pregnancy due to emmenagogue effects (Ernst 2002) and when breastfeeding. Given funding for CAM research is difficult to obtain and our review did not identify convincing evidence for lavender aro￾matherapy, massage or oil-dripping, additional research should focus on replication of oral lavender. Future trials of oral laven￾der would benefit from adopting a good trial design and stringent reporting to enable replication: this would include adopting a randomised design with allocation concealment, triple blind (if possible) and indistinguishable placebos. Intention-to-treat anal￾yses and a priori power calculations should be conducted and reported. All withdrawals, dropouts and adverse events should be fully reported giving the number and reason by group. It is essen￾tialto get good tolerability and safety data for all modes of lavender application, thus longer-term follow ups would be required for oral lavender before it is recommended as an alternative treatment for anxiety. Limitations As described above, the methodological issues observed in the included studies are notable and limit the extent to which the anxi￾olytic properties of lavender can be evaluated. The methodological issues ofthe lavender inhalation trials are most difficultto manage; trying to ensure that consistent levels of lavender odour is released between groups, or when attempting successful replication of such trials, can be particularly hard to control. In addition, although the search strategy was thorough, some clinical trials may not have

834 R.Perry et al.Phytomedicine 19(2012)825-835 been identified.However,our systematic and detailed search strat- Appendix A(Continued egy should have assisted in identifying all trials and in reducing bias. Publication bias is a problem in all medical research (Easterbrook Searches et al.1991)and is particularly problematic in alternative medicine 40 lofinda.ti,ab. (Ernst 2007b:Ernst and Pittler 1997).Collectively these limitations 41 ostoghodous.ti,ab. render our systematic review less than conclusive as one might postokhodous.ti.ab have hoped. Silexan.ti,ab. 44 spigandos.ti.ab Conclusions 45 exp Lavandula/ 46 1 or 2 or3 or 4or 5 or 6or 7or8 or9 or 10or 11 or 12 or 13 or 14or 15 or 16 or 17 or 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26 or 27 Only few RCTs of lavender,for anxiety/stress are available using or 28 or 29 or 30 or 31 or 32 or 33 or 34 or 35 or 36 or 37 or 38 or 39 diverse administration methods (ie.oral,olfactory and as a mas- or 40 or 41 or 42 or 43 or 44 or 45 sage oil).The evidence for oral lavender is promising but,until Adaptation reaction.ti,ab. Adjustment disorders.ti,ab. independent replications emerge with long-term follow-up data 48 Adiustment reaction tLab. remains inconclusive.The use of more widely used forms of laven- 50 Agoraphobis.ti.ab. der administration(aromatherapy,inhalation,massage etc.)is not 51 Angst.ti,ab. currently supported by good evidence of efficacy.Future trials, 52 Anguish.ti.ab. 53 Antianxiety.ti.ab well-reported and adopting stringent methodology (e.g,adhering 54 anti-anxierv.tab to PRISMA guidelines).in combination with in vitro pharmaco- 55 Anxiets.ti.ab. logical research,would help to elucidate the therapeutic value of 56 Anxiolytics.ti,ab lavender as an anxiolytic. 57 Anxouss tiab. 58 Combat Disorders.ti,ab 59 Distress.tab. Conflict of interest 60 NeurosS.ti,ab. Neurotic.ti,ab. None 62 Obsessive-Compulsive Disorders.ti,ab. 63 Panic attack$.ti,ab. 64 Panic Disorders.ti.ab. Funding 65 PhobiaS.ti.ab. 66 Phobic DisorderS.ti.ab. None. 67 Stress.ti,ab. Worry.ti.ab. 69 exp adjustment disorders/or exp anxiety disorders/or exp neurotic Appendix A.Medline-lavender for stress disorders/ 70 exp Anxiety/ Searches 71 exp Stress.Psychological/ 72 exp Anti-Anxiety Agents/ Al birri.ti,ab. 73 47 or 48 or 49 or 50 or 51 or 52 or 53 or 54 or 55 or 56 or 57 or 58 or 59 2 alhucema tiab or 60 or 61 or 62 or 63 or 64 or 65 or 66 or 67 or 68 or 69 or 70 or 71 3 arva neh.ti,ab. 0r72 4 espic.ti,ab. 74 46 and 73 5 fi rigla.ti,ab 6 firigla.tiab. 7 frigous.ti.ab hzama.ti.ab. khazama.ti,ab References 10 khirii.ti,ab. Anon,2011.Shirodhara Massage.http://www.mindbodysoul.tv/health/shirodhara- Kkhouzamaa tlab khouzami.ti.ab. massage.html (accessed 25.05.11) Boutron,I..Moher,D.,Tugwell,P..Giraudeau,B..Poiraudeau,S.,Nizard,R..Ravaud,P. 13 khuzama tiab 14 2005.A checklist to evaluate a report of a nonpharmacological trial (CLEAR NPT) Kleiner Speik.ti,ab. was developed using consensus.Journal ofClinical Epidemiology 58,1233-1240 L$x intermedia.ti,ab Braden,R.Reichow,S.,Halm,MA.,2009.The use of the essential oil lavandin 6 L$angustifoliaS.ti,ab. to reduce preoperative anxiety in surgical patients.Journal of Perianesthesia 17 L$dentata.ti.ab Nursing24(6).348-355. 18 Is latifolia tiab Bradley.B.F..Brown,S.L.Chu,S..Lea.R.W..2009.Effects of orally administered Ls multinda ti ab lavender essential oil on responses to anxiety-provoking film clips.Human Psy- 20 L$officinalis.ti,ab. chopharmacology 24(4).319-330. L$spica Loisel.ti,ab. Brum.LF,Elisabetsky.E.Souza,D..2001.Effects of linalool on((3)H)MK801 and 22 L$stoechas.ti.ab. f(3)H,muscimol binding in mouse cortical membranes.Phytotherapy Research L$vera DC.ti.ab. 54D-4D5 24 L$vulgaris LamS.ti.ab. Buckle.J..1993.Aromatherapy:does it matter which lavender essential oil is used Nursing Times 89.32-35. 25 Lasea.ti.ab. 2 Cavanagh,H.M.,Wilkinson.J.M..2002.Biological activities of lavender essential oil. Lavanda tiab. Phytotherapy Research 16(4).301-308. 27 lavande.ti,ab. Corner.J,Cawl ey.N..Hildebrand,S.,1995.An evaluation of the use of massage 8 Lavandin.ti,ab and essential oils in the wellbeing of cancer patients.International Journal of 29 lavandino tiab Palliative Nursing 1.67-73. 0 lavando.ti.ab Coulson,LH..Khan,A.S.1999.Facial pillo titis due to lavender oil allergy. 31 Lavandula.ti,ab. Consult Dermatitis 41,111. 32 Lavandulae.ti,ab. Dunn,C.Sleep.J.Collett.D..1995.Sensing an improvement a an experimental study 3 lavanulae tLab. to evaluate the use of aromatherapy,massage and periods of rest in an intensive 34 lavenda.ti,ab. care unit lournal of Advanced Nursing 21.34-40. 5 Lavendelti.ab Easterbrook,P.I.,Berlin,J.A.,Gopalan.R.Matthews,D.R,1991.Publication bias in 6 Lavender.ti,ab. clinical research.Lancet 337,867-872. 37 Lavenol.ti,ab. Ernst.E.2002.Herbal medicinal products during pregnancy:are they safe?BJOG: 38 lawanda.ti.ab. An International Journal of Obstetric and Gynaecology 109.227-235. 39 Lofi nda.ti,ab. Ernst,E..2007.'First,do no harm'with complementary and alternative medicine Trends in Pharmacological Sciences 28(2).48-50

834 R. Perry et al. / Phytomedicine 19 (2012) 825–835 been identified. However, our systematic and detailed search strat￾egy shouldhave assistedinidentifying alltrials andinreducing bias. Publication bias is a problem in all medical research (Easterbrook et al. 1991) and is particularly problematic in alternative medicine (Ernst 2007b; Ernst and Pittler 1997). Collectively these limitations render our systematic review less than conclusive as one might have hoped. Conclusions Only few RCTs of lavender, for anxiety/stress are available using diverse administration methods (i.e. oral, olfactory and as a mas￾sage oil). The evidence for oral lavender is promising but, until independent replications emerge with long-term follow-up data, remains inconclusive. The use of more widely used forms of laven￾der administration (aromatherapy, inhalation, massage etc.) is not currently supported by good evidence of efficacy. Future trials, well-reported and adopting stringent methodology (e.g., adhering to PRISMA guidelines), in combination with in vitro pharmaco￾logical research, would help to elucidate the therapeutic value of lavender as an anxiolytic. Conflict of interest None. Funding None. Appendix A. Medline – lavender for stress Searches 1 Al birri.ti,ab. 2 alhucema.ti,ab. 3 arva neh.ti,ab. 4 espic.ti,ab. 5 fi rigla.ti,ab. 6 firigla.ti,ab. 7 frigous.ti,ab. 8 hzama.ti,ab. 9 khazama.ti,ab. 10 khirii.ti,ab. 11 khouzamaa.ti,ab. 12 khouzami.ti,ab. 13 khuzama.ti,ab. 14 Kleiner Speik.ti,ab. 15 L$ x intermedia.ti,ab. 16 L$ angustifolia$.ti,ab. 17 L$ dentata.ti,ab. 18 L$ latifolia.ti,ab. 19 L$ multifida.ti,ab. 20 L$ officinalis.ti,ab. 21 L$ spica Loisel.ti,ab. 22 L$ stoechas.ti,ab. 23 L$ vera DC.ti,ab. 24 L$ vulgaris Lam$.ti,ab. 25 Lasea.ti,ab. 26 Lavanda.ti,ab. 27 lavande.ti,ab. 28 Lavandin.ti,ab. 29 lavandino.ti,ab. 30 lavando.ti,ab. 31 Lavandula.ti,ab. 32 Lavandulae.ti,ab. 33 lavanulae.ti,ab. 34 lavenda.ti,ab. 35 Lavendel.ti,ab. 36 Lavender.ti,ab. 37 Lavenol.ti,ab. 38 lawanda.ti,ab. 39 Lofi nda.ti,ab. Appendix A (Continued ) Searches 40 lofinda.ti,ab. 41 ostoghodous.ti,ab. 42 postokhodous.ti,ab. 43 Silexan.ti,ab. 44 spigandos.ti,ab. 45 exp Lavandula/ 46 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26 or 27 or 28 or 29 or 30 or 31 or 32 or 33 or 34 or 35 or 36 or 37 or 38 or 39 or 40 or 41 or 42 or 43 or 44 or 45 47 Adaptation reaction.ti,ab. 48 Adjustment disorder$.ti,ab. 49 Adjustment reaction.ti,ab. 50 Agoraphobi$.ti,ab. 51 Angst.ti,ab. 52 Anguish.ti,ab. 53 Antianxiety.ti,ab. 54 anti-anxiety.ti,ab. 55 Anxiet$.ti,ab. 56 Anxiolytic$.ti,ab. 57 Anxious$.ti,ab. 58 Combat Disorder$.ti,ab. 59 Distress.ti,ab. 60 Neuros$.ti,ab. 61 Neurotic.ti,ab. 62 Obsessive-Compulsive Disorder$.ti,ab. 63 Panic attack$.ti,ab. 64 Panic Disorder$.ti,ab. 65 Phobia$.ti,ab. 66 Phobic Disorder$.ti,ab. 67 Stress.ti,ab. 68 Worry.ti,ab. 69 exp adjustment disorders/or exp anxiety disorders/or exp neurotic disorders/ 70 exp Anxiety/ 71 exp Stress, Psychological/ 72 exp Anti-Anxiety Agents/ 73 47 or 48 or 49 or 50 or 51 or 52 or 53 or 54 or 55 or 56 or 57 or 58 or 59 or 60 or 61 or 62 or 63 or 64 or 65 or 66 or 67 or 68 or 69 or 70 or 71 or 72 74 46 and 73 References Anon, 2011. Shirodhara Massage. http://www.mindbodysoul.tv/health/shirodhara￾massage.html (accessed 25.05.11). Boutron, I., Moher, D., Tugwell, P., Giraudeau, B., Poiraudeau, S., Nizard, R., Ravaud, P., 2005. A checklist to evaluate a report of a nonpharmacological trial (CLEAR NPT) wasdevelopedusing consensus.Journal ofClinical Epidemiology58,1233–1240. Braden, R., Reichow, S., Halm, M.A., 2009. The use of the essential oil lavandin to reduce preoperative anxiety in surgical patients. Journal of Perianesthesia Nursing 24 (6), 348–355. Bradley, B.F., Brown, S.L., Chu, S., Lea, R.W., 2009. Effects of orally administered lavender essential oil on responses to anxiety-provoking film clips. Human Psy￾chopharmacology 24 (4), 319–330. Brum, L.F., Elisabetsky, E., Souza, D., 2001. Effects of linalool on ((3)H) MK801 and {(3)H} muscimol binding in mouse cortical membranes. Phytotherapy Research 15, 422–425. Buckle, J., 1993. Aromatherapy: does it matter which lavender essential oil is used. Nursing Times 89, 32–35. Cavanagh, H.M., Wilkinson, J.M., 2002. Biological activities of lavender essential oil. Phytotherapy Research 16 (4), 301–308. Corner, J., Cawley, N., Hildebrand, S., 1995. An evaluation of the use of massage and essential oils in the wellbeing of cancer patients. International Journal of Palliative Nursing 1, 67–73. Coulson, I.H., Khan, A.S., 1999. Facial pillow dermatitis due to lavender oil allergy. Consult Dermatitis 41, 111. Dunn, C., Sleep, J., Collett, D., 1995. Sensing an improvement an experimental study to evaluate the use of aromatherapy, massage and periods of rest in an intensive care unit. Journal of Advanced Nursing 21, 34–40. Easterbrook, P.J., Berlin, J.A., Gopalan, R., Matthews, D.R., 1991. Publication bias in clinical research. Lancet 337, 867–872. Ernst, E., 2002. Herbal medicinal products during pregnancy: are they safe? BJOG: An International Journal of Obstetric and Gynaecology 109, 227–235. Ernst, E., 2007. ‘First, do no harm’ with complementary and alternative medicine. Trends in Pharmacological Sciences 28 (2), 48–50