The above review of literature and plays suggests common and consistent uses of lavender, certainly in England and the United States, and the primary psychological effect seems to be that of calming. This section examines the scientific research to see if there is a basis to this effect, that is, why lavender and its components might be calming, how it affects the brain and psychological states, and discusses issues related to the odour perception and, finally, examines whether these effects can be seen in practice. One important proviso is that lavender oils can vary in composition, and this may affect the results obtained in studies, yet many of the following studies do not specify the composition of the lavender oil, so that results may not be comparable (or repeatable).
Doing research on the psychological effects of odour on people is difficult and this is one reason why research on the psychological effects of odour has been limited. Many mechanisms may be involved in producing psychological effects. Once an effect is found, it is very difficult to know why it is occurring, for example, an odour may exert its effects through a mixture of pharmacological and various psychological mechanisms, for example, hedonic, learning and social mechanisms; all possibly operating at the same time and maybe even in different directions.
More specifically, psychological responses to odour may involve an interaction of an individual's experience, the particular situation, and the individual's current state of mind and this combination makes both interpretation of results and prediction difficult. A change in one of these may change the psychological response to an odour, even though the odour remains the same. These issues will be discussed in more detail later, although research does show that there may be a primary pharmacological mechanism involved in the calming effect of lavender.
Unlike psychological mechanisms, the pharmacological mechanism does not involve any perception of lavender odour, and the psychological effects are due to the odour compounds entering the body and acting directly on the brain, that is, via the blood stream through absorb-tion through the lungs or olfactory mucosa. These effects depend on the specific molecular properties of lavender, and are therefore likely to give similar effects across all people.
The main terpenoid components of lavender oil, for example, linalool and linanyl acetate, are lipophilic and similar terpenes are found to suppress cell action potentials, possibly through direct steric interaction with cell membranes, thereby increasing their volumes and thus slowing the inflow of calcium ions and reducing sodium ion permeability (Teuscher et al, 1990). Such general suppression of cell electrical activity suggests the possibility of a light sedative or anaesthetic effect.
Elisabetsky et al. (1995) noted that plant extracts with a high linalool content are used as a home-made anticonvulsant remedy in Brazil. More specifically, they found that linalool competitively inhibited glutamate binding in the cells in the cortex of rats and suggested that this may cause the sedative effect. Linalool also inhibited glutamate-related epilepsy, and gave protection against pentylenetrazol (PTZ), quinolinic acid and transcorneal electroshock-induced convulsions (Elisabetsky et al., 1999).
However, the reduction in the number of convulsions caused by PTZ induced kindling, was not linked to changes in the glutamate binding levels that accompany kindling, so linalool may act by a variety of mechanisms to prevent convulsions other than glutamate binding inhibition. For example, Yamada et al. (1994) found that inhaled lavender oil reduced pentetrazol, nicotine and electroshock induced convulsions in mice, and suggested that these effects may be due to an augmentation of GABA-ergic action. Evidence for this mechanism comes from Aoshima and Hamamoto's (1999) investigation of the potentiation of GABAa receptors expressed in Xenopus oocytes by lavender oil components. These increased response in the presence of GABAA at low concentrations, possibly through binding to the potentiation-site in GABAA receptors, thus increasing the affinity of GABAA to the receptors.
This effect resembles the potentiation of GABAA receptors by benzodiazepine, barbiturate, steroids and anesthetics, which is associated with their anxiolytic, anticonvulsant and sedative effects.
These reductions in cell activity due to lavender oil and linalool have also been observed in non-CNS cells and tissues, reflecting a general spasmolytic or 'calming' effect. Mazzanti et al. (1998) found that linalool inhibited acetyl choline and barium chloride induced contractions of isolated guinea-pig ileum, and similarly, Lis-Balchin and Hart (1999) showed that lavender oil and linalool reduced the twitch response to nerve stimulation in isolated rat diaphragm phrenic nerve preparations, and found that the mechanism was post synaptic, not atropine-like and likely to be mediated through cAMP and not through cGMP.
Reflecting this pharmacological rather than psychological action, Buchbauer et al. (1993) found that a 78 per cent decrease in the motility of mice resulted from inhalation of ambient lavender oil for one hour, with similar effects for the main components, linalool (73 per cent) and linalyl acetate (63 per cent); these three being the most effective sedatives of forty-two oils and their components tested. When mice were first agitated by caffeine injection, inhaled lavender oil reduced motility by 92 per cent. They also note that the ability to sedate related to increasing lipophilic properties. This is consistent with Teuscher et al.'s (1990) suggestion that the effect is due to steric interaction with cell membranes. Inhalation of lavender oil led to a serum level comparable to an intravenous injection - this may be brought about by rapid absorption through both the nasal and lung mucosa (Buchbauer et al, 1991). As might be expected, given the highly lipophilic nature of the brain, the components were transported across the blood brain barrier, being found in the cortex of test animals (Buchbauer et al. 1993a).
Paralleling the quotes in literature and plays, these findings suggest that lavender odour may act as a light intranasally-introduced sedative. Speculatively, such effects may also account for the use of lavender to scent sheets, that is, to ensure a good night's sleep as well as any moth repellent properties.
More significantly, these studies also identify how the effect may be caused and this knowledge should now allow its optimisation, for example, though breeding of high linalool lavenders and perhaps through chemical modification. Assuming that the active components in lavender can enter the brain, the next question is whether lavender actually affects brain activity, and this is addressed in the next section.
There have been several studies of the brain's responses to odour. These have used the changing electrical activity picked up by scalp electrodes in response to lavender odours as a measure of brain activity (EEG). EEGs have the advantage of being objective rather than subjective measures, although interpretation of the results may be complex (reflecting the trade-off of reliability and validity between the two types of measure).
The CNV measure (contingent negative variation) has been used to investigate the effect of lavender. The CNV is an upward shift in EEG waves which occurs when people are expecting something to happen, for example, between a signalling tone followed by a light, which the subject then switches off. Since the CNV is found to be reduced by sedatives and increased by stimulants, Torii et al. (1988) investigated the effect of lavender on CNV in trained Japanese perfumers. They found that it reduced CNV, but unlike sedatives, had no effect on reaction time or heart rate, possibly due to the low amounts absorbed.
These results were replicated by Lorig and Roberts (1990) in untrained American subjects. They comment that this suggests that CNV is robust across cultures. However, they also suggest that the effects may also involve cognitive mediation or mood induction due to previous experience with the odour as well as due to the direct physiological or pharmacological effects on the brain. This was because they found that the CNV also changed depending on what subjects' thought the odour was, even though the odour presented was different. They also used concentrations of lavender that the subjects found 'arousing and distracting' and suggest that this may also have led to the lowered CNV.
Changes in different frequency bands of the EEG to lavender and other odours have also been measured. Diego et al. (1998) examined how lavender oil affected the spectra of the EEG compared to rosemary oil. A group breathing lavender showed increased frontal alpha- and beta-2 waves, suggesting an increase in drowsiness, and this group also rated themselves as more relaxed. The study lacked a no-odour control group so perhaps taking part on the study alone without odour might have given similar results. Diego et al. (1998) did not report changes in theta waves. In contrast, Klemm et al. (1992) reported that lavender and other odours were associated with a general increase in theta waves, which is associated with increased drowsiness. However, the typical alpha wave increase during drowsiness was not observed, making these results hard to interpret.
A main finding by Klemm et al. (1992), which also contrasts with Diego et al.'s (1998) results, was that EEG responses to the odours varied greatly between subjects. They suggest that the difference observed in EEG responses may be due to subjects visualising the object or situations that the odours evoked, since these are likely to be different across subjects.
This evocation of specific objects or situations relates to the suggestion that all olfactory percepts are, in fact, associates (Booth, 1979). Certainly, odour words are virtually always names of the typical source object, thus, odours may be nothing more than carriers of the meaning of their associates, by evoking recognition of the source and by evoking any affect that may be associated with it (Kirk-Smith and Booth, 1987). As EEG measures neocortical activity, it can be considered to reflect people's affect and thoughts, and these will be influenced by individual associations with each odour. Klemm et al. (1992) comment that unless such individual thoughts and affect are accounted for, it may not be appropriate to look for consistent EEG responses across subjects.
Overall, these studies suggest that brain responses to lavender are in the direction of increased calmness rather than excitement. However, to investigate the pharmacological effect alone, rather than to have a confounding with perceptually-related effects, would require levels of lavender below awareness to be given for long enough for sufficient amounts to enter the brain.
Factors such as subject expectation, the experimental situation and its interpretation and personal experience of lavender may all influence objective EEG responses to lavender. In other words, EEG responses to lavender also depend on psychological factors unrelated to the chemical characteristics of lavender. These are especially important in understanding the psychological effects of lavender in laboratory conditions, and these are the subjects of the next section.
As a stimulus, the fragrance of lavender might be expected to have some effect on behaviour, cognition, affect, memory and mood. Several studies have investigated how lavender affects such psychological factors.
Ludvigson and Rottman (1989) compared the effects of lavender and cloves compared to a no-odour control on various tasks; cognitive - using arithmetical tests, vocabulary using synonyms, and analogies to test reasoning skills; memory - where twenty-four words were presented, half of the subjects were asked if each word was pleasant or not, the other half were asked if the word contained repeated letters, then all were asked to recall the words after 5 min; affect -whether they liked doing the tasks and would return to do similar studies and mood. Their study was somewhat complex with two sessions, where the three groups received any of the three other conditions in a second session a week later.
They found that cognitive functioning, particularly for arithmetic, was reduced by lavender in the first session, but not the second. There was no effect on mood or memory in either session. People with lavender in the first session found the session more enjoyable than the other groups; however, people who had no odour in the first session and lavender in the second session said they would not return. Ludvigson and Rottman (1989) comment that the intensity of the lavender odour was quite strong, so this may have caused a distraction in the first session and the decreased performance, and been more of an annoyance to those in the second session rather than first, where it was novel. The total time taken to do the various tasks was at least 25 min, so it may be that initial perceptual effects were mixed with later pharmacological effects as the lavender entered the brain.
In complete contrast, Diego et al. (1998) found that subjects performed arithmetical tests faster and more accurately following exposure to lavender odour for 3 min than rosemary odour. Since there were no control groups and no assessments of the perceived strengths and pleasantness of either odour were given it is hard to interpret these results.
Knasko (1992) studied the influence of odour pleasantness by comparing the effect of ambient lavender, lemon and dimethyl sulphide on mood and a standard creativity test in two 1 h sessions 1 week apart, counterbalanced as to control for one experimental condition. She found that lavender had no effects on either mood or creativity, though DMS did tend to lower mood ratings. She comments that the lavender concentration was moderate compared to Ludvigson and Rottman (1989) where a reduced performance was observed.
Degel and Koster (1999) compared the effects of lavender and jasmine on creativity and on arithmetical and concentration tasks over 45 min, but used imperceptible or unnoticed levels of odour, thus removing distracting and expectational cues. They found that lavender reduced the number of errors made in the arithmental and concentration tasks compared to the control and jasmine groups, that is, in this study the improvement is likely to be an effect of the lavender odour rather than due to some general expectation effect. They suggest that doing these tasks is stressful and that the sedative effect of lavender may have improved performance by reducing arousal; also that the effect would therefore not be shown if the subjects had not been stressed.
In interpreting these results, several methododological points need to be borne in mind. In many studies of the effects of odour on human response, a main variable that frequently arises is whether the fragrance is perceived as pleasant or unpleasant, which also tends to correlate with the perceived strength of the odour (i.e. too much of a good thing is not liked).
The importance of the hedonic aspect is not surprising. Pleasantness or unpleasantness is related to the value of an odour in evoking recognition of the source, and approach or avoidance of the source. As such, hedonic responses play a fundamental part in an organism's survival behaviour. As well as whether the odour is pleasant or not, performance also may be affected depending on how the odour is attended to, particularly if the odour is perceived as an annoying distraction or as incongruent to the task.
In general, the issue behind many studies on the psychological effects of odour is how or whether a response or performance is changed when people are in an unpleasant or pleasant condition; and this will depend on the particular task, the particular conditions and people's perceptions of them. Since all these parametres can be changed (and inevitably differ between studies), this means that psychological responses to odours are malleable, and highly dependent on the particular situation.
For example, in several of the studies mentioned above lavender was used in concentrations high enough to be considered distracting or unpleasant by subjects, that is, at levels which subjects would not choose to use themselves. As reflected in the quotes from literature and plays, lavender, in self-administered amounts, is seen as a particularly pleasant odour and worthy of carrying around on handkerchiefs.
The level used also relates to arousal. Any sensory stimulus will cause some arousal by definition, and whether this arousal improves or decreases performance on a task will depend on the level of the stimulus, for example, too much may decrease performance, and too little will have little effect, and the pre-existing level of arousal, for example, if someone is underaroused, the stimulation may improve performance, but if they are already optimally aroused, the same stimulation may reduce it.
Given these factors, it is quite possible that lower and pleasanter levels of lavender might have given very different results. Olfaction is also different from the other senses in that the sense adapts, that is, after a while a smell is not perceived. This suggests that its perceptual effect may only last until it has adapted, and that intermittent exposure or varying concentrations might be needed to maintain an effect.
The impact of factors independent of the odour itself on the response to the odour can be both dramatic and counter-intuitive. For example, Rotton et al. (1978) hypothesised that the presence of an unpleasant odour (butyric acid) would create a negative mood, which would, in turn, cause a reduction in people's rated attractiveness of others. This follows the well-researched result that putting people in a bad mood makes them rate others worse. Contrary to their hypothesis, subjects in the presence of the bad odour actually rated (fictitious) others as more attractive compared to subjects who rated without the odour. However, it was discovered that the subjects had thought that the rated person was also suffering the bad odour, that is, was next door doing the same study. This perception may have increased positive feelings towards a 'fellow sufferer'. They repeated the study with the changed instructions that the imaginary rater was not present, and found that subjects now rated them lower. The conclusion of this study is that the psychological response caused by an odour was completely reversed simply because of a different belief about where a person was.
In other studies, the different responses are due to the different tasks imposed. Baron (1993) found that when women job applicants wore fragrance for a mock job interview, male interviewers liked them less and rated them lower in intelligence and friendliness than female applicants not wearing perfume. Baron suggests that the men may have felt that the perfume distracted them from objective assessment. In contrast, Kirk-Smith and Booth (1990) found that in the presence of perfume, men rated photographed women as sexier and softer compared to a no-perfume situation. In this experiment, there was no pressure on subjects to perform a task, and the low levels of odour were unlikely to be distracting.
Knasko et al. (1990) also examined how peoples' expectations about odours influences their psychological responses in a test in which three groups were told that either an unspecified neutral, pleasant or unpleasant odour was present. In fact, no odours were present at all. Even so, people in the 'pleasant odour' condition reported the least ill-health symptoms and those in the 'unpleasant odour' condition reported the most symptoms. Those in the 'pleasant odour' condition also reported a better mood, presumably because people have an implicit expectation that pleasant odours improve mood. In this case, merely thinking that a pleasant odour was present had the same effect as if the pleasant odour had been present (Knasko, 1992).
Individual experience can also affect response. Lavender was the favourite fragrance of an occupational therapist and an acquaintance of the author. She had begun a routine job that she found stressful, so she decided to make her life more pleasant by introducing the smell of lavender into her workplace. She later changed jobs, but now found that every time she smelt lavender it reminded her of the stressful job, even though she still found the smell pleasant. In this case, the hedonics are positive, but the affective associations are now negative.
The conclusion that can be drawn from examples such as these is that it is very difficult to make simple generalisations concerning the effects of any fragrance, such as lavender, on psychological responses which are based on the immediate perceptual effects, rather than the longer term pharmacological effects. A pharmacological effect is more likely to affect people similarly, but the additional psychological mechanisms will create complex effects at the individual level.
As for the EEG responses, psychological responses to a particular odour may depend not only on its pleasantness and congruity with a situation, but also on the demands of the situation and the perceptions, state of mind and the experience of the person. As stated above, it is quite possible that simply changing any one of these many variables will change the psychological response to the odour.
These considerations suggest that the time course of effects needs to be examined in lavender since initial perceptual effects may be replaced or accompanied by later pharmacological effects. The hedonic aspects also need to be carefully accounted for since variations in concentration may account for many of the effects observed. Furthermore, in real-life circumstances, through repeated use by someone the characteristic smell of lavender at moderate concentrations may 'signal' the later sedative effects, and thus the smell might acquire, through classical conditioning, a calming effect by itself, for example, Schiffman and Siebert (1991) claim that an apricot fragrance paired with a relaxed state after progressive relaxation later 'triggers' the relaxed state by itself.
In passing, it may be noted that such dependence of odour response on so many other variables almost eliminates the possibility that traditional 'sterilised' types of laboratory-based experimentation will uncover the responses that are normal in the 'real-life' situation. As soon as you impose rigid control by the method of removing the usual contextual stimuli, you lose the determinants of the response. This leads to a consideration of the effects of lavender in applied situations, which, incidentally, correspond more to the real-life usage referred to in the literature and plays.
Applications of the psychological effects of lavender
The sedative effects of lavender identified above suggest that it might offer a less invasive, and more economic, means of improving sleep and reducing stress and anxiety than current drugs, and without their side-effects, for example, benzodiazepines may cause residual daytime sedation, performance decrements and anterograde amnesia (Vogel, 1992). Accordingly, clinical studies have been carried out to examine this possibility.
Based on Buchbauer et al.'s (1993) results on the sedative effects of lavender on mice, Bradshaw et al. (1998) found that lavender straw present during transportation halved the incidence of travel sickness symptoms from the 30 per cent measured in pigs bedded with wheat straw, although stress levels, as measured by salivary cortisol, were not changed. They suggest that this could either be due to the sedative effect of lavender or to the distracting effect of the novel odour.
Moving to humans, Hardy et al. (1995) conducted a trial on four insomniac psychogeriatric patients, under long-term drug treatment (i.e. with Temazepam, Promazine or Hemineverine). After 2 weeks monitoring of sleep, patients were taken off their drugs for 2 weeks. For a final
2 weeks an unobtrusive level of lavender oil was introduced using an odour diffuser for three one to one and half hour periods each night. As expected, patients slept less when the drugs were taken away. However, after introduction of the ambient lavender oil, sleep returned to the same level as under medication. This appeared to be a pure pharmacological effect; placebo and expectancy effects can be excluded as these patients had Alzheimer's disease. Patients were also reported to be less restless during sleep and to have better waking moods than under medication. Additionally, the lavender oil masked the unpleasant odours typically present in psychogeriatric wards.
In another clinical trial, Brooker et al. (1997) conducted a single-case experimental design across another four patients with senile dementia and behaviour problems. Each patient had between 8—12 half-hour trials over a 3-month period with four conditions randomised across time (ambient lavender, lavender oil and massage, massage, and no treatment). Behaviour was then recorded every minute for the hour after treatment. Although the clinical impressions were that all patients appeared to benefit from treatments, in this study significant improvements in behaviour were only observed for both lavender aroma and massage in one patient. Like Klemm et al. (1992), they comment that because of differences between individuals, the detailed study of individual responses to such interventions may be more suitable than group designs.
In contrast to these two small studies, Dunn et al. (1995) randomly allocated 122 patients to a massage, an aromatherapy massage with 1 per cent lavender oil or a rest period group, each receiving 1—3 treatment sessions of 15—30 min over a 5-day period. Physiological (BP, HR and resp.rate), psychological (mood and anxiety) and behavioural (motor and facial) measures were taken before and after each session. In the first session the lavender and massage group reported less anxiety than the control group, however, no other differences were observed on the other measures. They suggest that perhaps a higher concentration of lavender might have shown more effects.
Finally, Romine et al. (1999) used ambient lavender to investigate recovery after brisk walking for 2 min, with BP, mean arterial pressure, pulse pressure and HR taken pre- and postexercise and then after a 10 min rest in a dimly lit room with ten subjects given no odour and ten subjects given lavender which was put in the air by an electrical pot-pourri device. Although all measures were all lowered in the lavender condition, no significant differences were found. They suggest that larger samples should be used to give the statistical power to clarify these findings. Dale and Cornwell (1994) compared the effects of lavender oil, synthetic lavender oil and distilled water in reducing perineal discomfort after childbirth when used in bath water over ten days. No significant differences between groups were found.
These studies give mostly negative results regarding the effects of lavender in real-life application. In addition to the confounding issues discussed in the laboratory studies, these trials illustrate the additional problems of conducting such trials, for example, standard measures may be insensitive or inappropriate since people with different conditions may react differently, also it may be difficult to control for expectational effects since therapists have a vested interest in showing that their interventions work. Perhaps these account for the lack of consistent effects in contrast to both the experimental evidence for sedative properties and the folklore for them given in the literature and plays.
Was this article helpful?