Out of the woods: psychophysiological investigations on wood odors to estimate their suitability as ambient scents

Wood has been an important raw material for centuries, and humans encounter it in the form of furniture and building materials or in products of daily use. The odors emanating from such wood-based materials can add another facet to their functionality by influencing human well-being and mood. Moreover, wood odors can be used in odor-enhanced environments to foster relaxation and stress recovery. The current study was conducted to assess the subjective evaluation of, and physiological responses to, wood odors and individual wood odor constituents. The subjects smelled the odors of pine needles oil, cedar wood oil, α-pinene, and thymoquinone and evaluated the odor percept and their subjective feelings. Heart rate, finger temperature, and skin conductance were monitored throughout the experiment. The results showed that wood odors were distinct regarding their pleasantness, familiarity, and intensity. All odors enhanced the subjective feelings of happiness and content. Though no odor-specific physiological effects became evident, certain physiological measures correlated with self-reports of happy and relaxed feelings and the intensity ratings of the odors. The present study provides a basis for further studies on the use of wood odors as ambient scents and helps understanding the influence of natural odors on humans.


Introduction
Humans are continuously surrounded by stimuli triggering their senses of hearing, vision, touch, taste, and smell. Among these, humans tend to underestimate their sense of smell despite its many functions. Odor-active compounds are involved in cellular communication, carry information about our conspecifics, and are important quality parameters of food and consumer products (Doty 1986;Solov'yov et al. 2012;Stockhorst and Pietrowsky 2004). Furthermore, odors can influence human physiology, behavior, mood, and emotions (Alaoui-Ismaïli et al. 1997a;Bensafi et al. 2004;Herz 2009;Lehrner et al. 2000). Recent studies showed that being in nature has positive physiological and psychological effects leading to stress reduction, enhanced well-being, and a strengthened immune system (Hansen et al. 2017;Kardan et al. 2015;Wyles et al. 2019). Odors experienced in natural environments are thought to contribute to these effects.
The influence of wood essential oils and wooden environment on humans has been investigated in several studies. For instance, Matsubara et al. (2011) studied the impact of the essential oil of Siberian fir (Abies sibirica (Ledeb.)) on human physiology during and after a visual task. They showed that the arousal level in the recovery period was reduced by the smell of the essential oil, as indicated by electroencephalography and electrocardiography. In another study, the subjective feelings of well-being were enhanced in a room with walls made of Japanese cedar compared with a control room. The participants felt calmer and more comfortable in the wooden room and rated the smell as being significantly more pleasant (Matsubara and Kawai 2014). Comparing the feelings of comfort between a forest and a city environment, the positive impact of wood on human psychophysiology has also been reported by Park et al. (2007). The cerebral activity and salivary cortisol levels indicated that participants who were sent temporarily to the forest area were more relaxed than those who were sent to the city area. These effects may be induced by odorants and volatiles stemming from woods and plants in the forest, which generally have a pleasant odor but are also known to have an antioxidant activity and to increase cell viabilities (Ka et al. 2005).
From the above-cited studies, it appears that wood odors exert their positive effects on humans irrespective of their origin and application, i.e., irrespective of whether one perceives the odor from natural environments like forests, from essential oils, or from wooden products. The relaxing effects and the generally positive perception of wood odors open up many possibilities to purposefully use them, for instance to promote human well-being and recovery from stressful situations. Previous investigations on the influence of nature-related stimuli on human restoration, subjective well-being, and health mainly focused on visual and auditory stimuli (Richardson et al. 2017). Recently, however, it has been shown that adding an olfactory stimulus increases the recovery potential of a natural environment (Sona et al. 2019), underlining the importance of natural odors in our surroundings. The benefits of wood odors could be used in different environmental settings, for instance at home to restore from work faster, in public transportation to increase the comfort, or in shops to enhance consumers' comfort and relaxation during shopping.
Taken together, natural and wood odors exert positive effects on human psychology and physiology. They may enhance comfort in public places or work environments. However, there is a multitude of wood odors, and to date, it is unclear which wood odors should be used for such applications. In the current study, the smell percept, the subjective feelings, and physiological responses elicited by different wood odors were assessed to achieve first insights into wood odor perception. Essential oils from cedar and pine wood as well as two pure substances present in cedar and pine wood were used in this study. Cedar and pine wood are employed for a range of products like furniture or pencils, and the two pure substances, thymoquinone with its pencil-like smell and the resin-like smelling α-pinene, have a major impact on the overall odor character of cedar and pine wood, respectively (Schreiner et al. 2017(Schreiner et al. , 2018. Lavender oil was used as an odorous control stimulus due to its well-known pleasant and relaxing odor (Alaoui-Ismaïli et al. 1997b;Millot and Brand 2001). The insights gained by these investigations were expected to provide hints toward the selection of wood odors for future experiments regarding their performance as ambient scents. It was hypothesized that wood odors are pleasant and lead to enhanced positive feelings. Furthermore, the pleasantness perception was expected to be reflected in the physiological responses of the participants (Bensafi et al. 2002;Delplanque et al. 2009;Matsubara et al. 2011;Matsubara and Kawai 2014;Park et al. 2007).

Participants
All participants were informed about the research aims and methods and gave written informed consent prior to their entry in the study. Twenty healthy women (mean age = 29 ± 9 years) participated in the experiment. Normal olfactory function was ascertained prior to the experiment by using the MONEX-40 identification test (Freiherr et al. 2012). Exclusion criteria were smoking, pregnancy, age under 18 or above 40 years, and allergies, because odor perception can change in all these cases (Katotomichelakis et al. 2007;Larsson et al. 2000;Ochsenbein-Kölble et al. 2007). All participants were right-handers, which was tested with the Edinburgh Laterality Inventory (Oldfield 1971). Moreover, the subjects were asked to complete the Montreal Cognitive Assessment (MoCA) test. All participants passed the test meaning that they had a score of 26 or higher (average ± SD: 28.7 ± 1.6) (Nasreddine et al. 2005).

Odor stimuli and odor presentation
The following odor stimuli were used: cedar wood oil (Takasago, Zülpich, Germany), pine needles oil (Dragonspice Naturwaren, Reutlingen, Germany), α-pinene ((1S,5S)-2,6,6-trimethylbicyclo[3.1.1]hept-2-ene), thymoquinone (2-isopropyl-5-methyl-p-benzoquinone) (both Sigma-Aldrich, Steinheim, Germany), and lavender oil (Symrise, Holzminden, Germany). They were diluted in propylene glycol (1,2-propanediol, Sigma-Aldrich, Steinheim, Germany) resulting in 10% solutions of pine needle oil and cedar wood oil, a 5% solution of lavender oil and 0.5% solutions of α-pinene and thymoquinone. Of these solutions, 10 mL was filled in the glass reservoirs of the olfactometer. As odorless control stimulus, 10 mL of water was used. The solution of thymoquinone was absorbed on fiber material and placed in a plastic jar, since the odor quality (pencil-like) was not emitted in the solution. A possible reason could be its poor solubility in polar solutions (Odeh et al. 2012). All odors were tested by selected trained panelists prior to the experiment regarding odor intensity and odor quality and were evaluated as sufficiently intense.
Odors were presented with a computer-controlled air-dilution olfactometer (16-40 S/N B05, Whiff LLC, USA) for 2 s within a continuous airflow of 2.0 L per minute (Lundström et al. 2010). A respiration-triggered olfactory stimulation (RET-ROS) introduced by Hoffmann-Hensel and Freiherr (2016) was used to synchronize olfactory stimulus onset with inspiration onset. The inter-stimulus interval was 20 s. Experiment presentation software E-Prime 2.0 (Psychology Software Tools Inc, (Schneider et al. 2002)) was connected to the software LabChart (see below), allowing the transmission of the odor onset signal into LabChart.

Physiological measurements
Physiological measures included skin conductance (SC), heart rate (HR), skin temperature (ST), and respiration and were recorded in conjunction with the PowerLab 16/30 system and LabChart 8.0 (ADInstruments Ltd., Oxford, UK). The sampling rate for all physiological measurements was 1 kHz.
Skin conductance was measured exosomatically with alternating current (cf. Boucsein et al. 2012) and was recorded in µS using GSR Amp FE116, GSR Finger Electrodes (MLT116F, ADInstruments Ltd., Oxford, UK). Electrodes were positioned on the second phalanx volar sites of the index and middle finger of the nondominant hand.
Heart rate was measured using a piezo-electric pulse transducer (TN1012/ST, ADInstruments Ltd., Oxford, UK), which was positioned on the thumb of the nondominant hand.
Measures of skin temperature were taken with a skin temperature probe (MLT422/A, MLT409/A, ADInstruments Ltd., Oxford, UK) connected to a thermistor pod (ML309, ADInstruments Ltd., Oxford, UK). The sensor was placed on the distal phalanx of the ring finger of the non-dominant hand.
Respiration parameters were recorded with the PowerLab Spirometer Pod (Pow-erLab Spirometer Pod ML311, ADInstruments Ltd., Oxford, UK), which was connected to the olfactometer using a 2-m Teflon tubing and applied to the subjects by a 2.1-m oxygen nasal cannula (medisize Deutschland GmbH, Neunkirchen-Seelscheid, Cat. No. 15000 4025MT, phthalate-free). With this system, the respirationbased positive and negative pressure in the connected tubing was measured and recorded as pressure differences in Volt (V).

Subjective evaluations
The participants were asked to rate their subjective state of alert, active, relaxed, calm, happy, and content feeling as well as the familiarity of each odor on a visual analogue scale (0 = not at all to 10 = extremely). The questions were asked as follows: "The following questions refer to your state of mind. Active? Alert? Calm? Content? Happy? Relaxed?" and "How familiar is the smell to you?".

Procedure
Upon arrival, subjects were comfortably seated in an artificially lighted test room. The room was ventilated prior to the experiment and the temperature was controlled by air-conditioning (22-25 °C). Auditory cues were excluded by soundproof headphones with noise-canceling function. First, subjects were asked to complete the questionnaire about their handedness and the MoCA test. Then, the recording system and the olfactometer were connected to the subject, and the odors were presented in three blocks. Each block consisted of all six odors, resulting in three presentations of each odor. Odor presentation order was randomized within each block for each subject. Directly after each odor presentation of 2 s, the subjects rated the pleasantness and intensity of the odor on a 10-point visual analogue scale (for pleasantness: 0 = very unpleasant to 10 = very pleasant; for intensity: 0 = no perception to 10 = strong perception). In a second part of the experiment, the subjects smelled each odor in randomized order one more time and then rated their subjective feeling (according to the procedure given in "Subjective evaluations" section).
For an overview of the experimental procedure, see Fig. 1.

Data analysis
Skin conductance (SC) was analyzed by continuous decomposition analysis (CDA) according to Benedek and Kaernbach (2010) using Ledalab (V.3.4.9, http:// www. ledal ab. de/ softw are. htm). Thereby, the variable ISCR (integrated skin conductance response: area of the phasic driver) was analyzed in the response window of 8 s after stimulus presentation. The minimum amplitude value was set to .02 µS. Heart rate (HR) was analyzed in beats per min (bpm) generating difference scores by subtracting the mean of the 1 s prestimulus value from the mean of the 8 s after stimulus presentation (adapted from Bensafi et al. 2002;Jennings et al. 1981).
Skin temperature (ST) in °C was analyzed by subtracting the 1 s prestimulus mean from the mean of the 8 s after odor presentation.
The data were not normally distributed for which reason nonparametric tests were used. All statistical analyses (Shapiro-Wilk test, Friedman test, Wilcoxon test, and Spearman correlation test) were calculated using IBM SPSS Statistics 25. The level of significance was p < .05 for each of the statistical analyses. Spearman correlations were calculated between the physiological measures and the subjective evaluations as well as for the physiological measures and the intensity and pleasantness ratings, respectively. The correlations were calculated for the mean of all odors as well as for each odor individually.
Wilcoxon tests for familiarity ratings revealed that pine needle oil was rated as more familiar than α-pinene (p = .015) and the control (p = .005). Lavender oil was rated as significantly more familiar than the control (p = .005; Fig. 3). The ratings show that the odors lavender oil, pine needle oil, cedar wood oil, and α-pinene were recognized as familiar by the panelists (median ratings: 8, 7, 6.5, and 6, respectively), whereas thymoquinone and the control were rated as neutral with median ratings of 5.
Pleasantness ratings showed that cedar wood oil (median: 3) was perceived as significantly less pleasant than the other odors (Wilcoxon; p < .05 in each case). Lavender (median: 6) was perceived as most pleasant and thereby significantly more 1 3 Wood Science and Technology (2020) 54:1385-1400 Fig. 2 Median ratings of perceived intensity (0 = no perception to 10 = strong perception). Significant differences (p < .05, Wilcoxon test) are indicated by different letters Fig. 3 Median ratings of familiarity (0 = very unfamiliar to 10 = very familiar). Significant differences (p < .05, Wilcoxon test) are indicated by different letters 1 3 pleasant than thymoquinone (p = .010) and α-pinene (p = .035). The control (median: 5) was also significantly more pleasant than thymoquinone (p = .013). The median pleasantness rating of pine needle oil was 4.5. The results are displayed in Fig. 4.

Subjective evaluations of activation and mood
The median subjective ratings of activation (items: feeling alert, active, calm, and relaxed) and mood (items: feeling happy and content) are shown in Table 1. The Friedman tests revealed no significant effects regarding the alert, active, calm, and relaxed feeling. In contrast, a trend could be seen for the item "content" (χ 2 (5) = 9.82, p < .10), and the Friedman test failed to be significant for the item "happy" (χ 2 (5) = 9.05, p = .11). With median ratings of 8 and 7, respectively, α-pinene and lavender evoked a higher subjective content feeling than the control (p < .05; median rating: 5.5). Moreover, the odors of α-pinene and lavender evoked a higher subjective happy feeling than cedar (median ratings: 7, 7, and 6.5, respectively; p < .05, Wilcoxon).

Correlations between subjective evaluations and physiological measures
Spearman correlations over all stimuli revealed a significantly positive correlation between HR change and the ratings of feeling happy (r s = .258, p = .006) and the ratings of feeling relaxed (r s = .200, p = .033). A negative correlation was significant between HR change and the intensity ratings (r s = − .186, p = .048). Significant Spearman correlations between subjective evaluations and physiological measures on an individual stimulus level are listed in Table 2. No other correlations were significant.

Wood odors: their perceived pleasantness and impact on subjective ratings of activation and mood
It is generally believed that natural and wood odors are perceived as pleasant, and this notion has been supported by previous reports, for example, for lavender odor, cedar wood odor (Dravnieks et al. 1984), and pine odor (Herz and Cupchik 1992). The hypothesis that wood odors are perceived as pleasant odors was, however, not confirmed in the present study. Whereas the odor of lavender was rated as being pleasant, the wood odors pine, thymoquinone, and α-pinene were rated as neutral, and cedar odor was rated as unpleasant. From this, it is deduced that the pleasantness of wood odors depends on their individual odor character, which differs among wood species. In previous studies, it was shown that the odor character of cedar wood is mainly pencil-like and earthy, while pine wood was described as more resin-and sawdust-like (Schreiner et al. 2018(Schreiner et al. , 2020. Such qualitative differences in wood odors depend on the species and may lead to a different pleasantness percept. Indeed, Degel and Köster (1998) also reported rather low pleasantness ratings for cedar wood odor and for an ambient woody odor (ratings of 31 and 38 on a scale ranging from 0 to 100), which supports the current findings of wood odors being of diverse pleasantness. The odors did not induce significantly different subjective activation levels (feeling alert, active, calm, and relaxed). In contrast to that, the subjective ratings of feeling happy and content were influenced by the olfactory events. The participants felt more content after presentation of α-pinene and lavender than after presentation of no odor (the control condition) and happier compared to the cedar odor. While only being significant for those comparisons, on a descriptive base all wood odors resulted in a higher median rating of subjective content and happy feeling than the odorless control, with the exception of pine, which was rated equally to the control in terms of happy feeling. Overall, it appears that wood odors induce subjective happy and content feelings, which is in line with previous work (Song et al. 2019). Among the here used wood odors, α-pinene made the participants most happy and content on a subjective level. α-Pinene was, besides thymoquinone, one of the odorants that were administered as single components in the present study. It is a major wood volatile, especially occurring in Pinus species like Scots pine (Granström 2005). This odorant alone appears sufficient to generate increased subjective mood. Similar results were obtained for jasmine tea odor and (R)-(−)-linalool. In a study on the sedative effects of jasmine tea odor, (R)-(−)-linalool, the main odor component in jasmine tea, elicited the same effects on the heart rate and mood states of participants as the jasmine tea odor itself (Kuroda et al. 2005).
In sum, the present results highlight that different wood odors, in this case the odors of cedar wood oil, pine needles oil, thymoquinone, and α-pinene, evoke different pleasantness perceptions. This is important to consider when selecting wood odors for application to odor-enhanced environments, as careful selection of wood odors appears necessary to ensure achieving the desired outcomes. The results also point toward enhanced happy and content feelings after experiencing natural and wood odors as short-time sensory events. Thereby, α-pinene as an individual odoractive compound from wood odors appears as potent as, or even more potent than, complex natural wood odors, at least pine and cedar wood odors.

Physiological responses to wood odors
Earlier studies showed that subjective evaluations of olfactory events are linked to autonomic nervous system responses. For instance, pleasantness ratings were negatively correlated to heart rate changes in a study by Bensafi and colleagues. The more pleasant the odor, the less pronounced was the heart rate increase (Bensafi et al. 2002). Moreover, Delplanque et al. (2009) reported that unpleasant odors elicited greater skin conductance responses than pleasant odors. Hedblom et al. (2019) also found a significant relationship between the perceived pleasantness of an odor and the skin conductance level in a multisensory experiment of stress reduction in different environments (urban, park, and forest area). In the present study, there was no significant influence of perceived odor pleasantness on the physiological parameters. A possible reason is that the pleasantness ratings of the presented odors were in a rather narrow range (from 3 to 6), and that the odors were closely related to each other, all being odors from nature. A broader pleasantness range or different odor categories might be necessary for physiological measures to serve as indicators of subjective odor pleasantness perception in an event-related design (Pichon et al. 2015).
Spearman correlations revealed no significant relation between the evaluations of content feeling and the physiological parameters. In contrast to that, the subjective happiness correlated with the measured change in heart rate. The happier the subjects were, the more the heart rate increased after perceiving the odor. This stands in line with numerous previous studies showing that heart rate increases with happiness (reviewed by Kreibig 2010). Considering the intensity ratings, a significantly negative correlation with changes in heart rate emerged from the present data. Different odor intensities might enhance stimulus relevance, or entail a different sniffing behavior. Only a few authors have considered the impact of odor intensity on physiological responses to olfactory events, with conflicting outcomes (Bensafi et al. 2002;Glass et al. 2014).
Correlations between subjective ratings and physiological measures were also calculated for each stimulus individually as these relationships may depend on the respective odor character. The results did indeed differ depending on the odor stimulus. The happier the subject felt after smelling lavender, or the more pleasant the cedar and α-pinene odor was rated, the lower was the difference score of the finger temperature. This is in contrast to what is expected from the literature, namely an increase in finger temperature with positive valence (de Wijk et al. 2012). Heart rate changes were positively correlated with ratings of relaxed, happy, and content feelings after smelling the odor of pine needle oil. These results are only partly in line with previous research (Bensafi et al. 2002;Hoffmann-Hensel and Freiherr 2016). Yet, evidence in this field of research is often not consistent across studies, potentially due to the use of different methods of odor application and calculation of dependent variables.
In sum, subjective evaluations of happy and relaxed feelings, and intensity ratings, were reflected in cardiovascular measures. Moreover, the results point toward odor-specific correlations between physiological parameters and self-reports. Future studies should investigate in more detail to which extent these relationships depend on the odor stimulus, and determine the underlying mechanisms. Overall, the stimuli were not, or only marginally differentiated by means of physiological measures. Therefore, the physiological responses to the olfactory events did hardly add any valuable information of wood odor perception in the present study. Future studies should explore alternative methods to gain insights into the impact of wood odors on human physiology in event-related designs or in long-term exposure.

Implications for the use of wood odors as ambient scents
Various studies showed that wood odors exhibit calming effects. On the one hand, smelling the odors of pine tree essential oils or being in a wooden room resulted in faster restoration from mental fatigue (Matsubara et al. 2011;Matsubara and Kawai 2014). On the other hand, environments connected to wood, such as forests, urban nature or green neighborhoods, support physiological relaxation, reduce stress, and improve health perception (Hunter 2019;Kardan et al. 2015;Park et al. 2007). In the present study, the wood odors did not influence the subjective evaluation of activation (feeling active, alert, calm, and relaxed). However, content and happy feelings were significantly enhanced by the odors of lavender and α-pinene. Among the investigated wood odors, α-pinene seems to be eligible for further testing as ambient scent as it affected subjective mood the most. The different wood odors were also rated differently in terms of pleasantness. Pine needle odor was perceived as the most pleasant among the wood odors, whereas cedar wood odor was perceived as the least pleasant. Accordingly, pine needle odor and α-pinene might be promising odor stimuli to be used in future studies aiming at evaluating the effects of long-time exposure to wood odors, as it would be the case in odor-enhanced environments.
It is still open to question, however, whether the experimental procedure applied here can be used to estimate psychophysiological effects of wood odors in environmental exposure situations, for example, for room scents. In the above-cited studies, the effects of wood odor and wood environment were tested using exposure periods of several days to weeks whereas the present study focused on the immediate effects of the wood odors when presented as olfactory events. Nevertheless, subjective evaluation of perceptual dimensions of wood odors is an important step before testing odors in more complex study designs. Whereas the current study focused on olfactory stimuli as sole stimuli and not in combination with congruent auditory or visual stimuli, future studies should also explore how different wood odors are perceived when presented along with auditory, visual, or tactile stimuli. Furthermore, stress might be induced in participants to specifically address the potential of the here investigated wood odors to enhance recovery from stressful situations. A recent study on stress reduction by urban green spaces, for instance, used a multisensory setup with combined visual, auditory, and olfactory stimuli. After evoking a stress response in the participants, the skin conductance level showed significant effects for the stress recovery in forests in comparison to urban environments, as the skin conductance levels were lowered faster in forests (Hedblom et al. 2019). It would be interesting to evaluate whether these effects are modulated by the kind of wood odor used. The impact of wood odors might additionally be increased when presenting a natural setup including, for example, visible trees and plants to enhance the immersion into the simulated environment (Sona et al. 2019).
In sum, the present study provided insights into the perception of wood odors and is suggested to be used as an upstream step toward finding the most beneficial odors for more elaborated studies targeting the impact of wood odors on human behavior in everyday situations. For instance, specific wood odors and odorants could be used as room scents to influence human behavior by enhancing mood and restoration in work or shopping environments. Thereby, the present study showed that not only complex wood odors might be selected, but that single odorants could be used as room scents. Furthermore, it suggests α-pinene as most promising to enhance mood. These implications provide important knowledge for future studies investigating consumer behavior. Future studies will demonstrate whether responses to olfactory events can predict effectiveness of odors in more ecologically valid contexts.

Conclusion
The aim of this research was to characterize wood odors on a perceptual and psychophysiological level. Based on subjective odor evaluation, it is concluded that pleasantness ratings of wood odors depend on the specific character of the individual wood odor. Moreover, self-reports of happy and content feelings raised upon smelling α-pinene and lavender compared to the control condition and cedar odor. Regarding the physiological reactions of the panelists, this present study showed that intensity ratings of the odors and subjective happiness ratings correlated with difference scores of heart rate. To determine the effectiveness of wood odors as product scents or ambient odors, their impact on consumer and human behavior in general should be explored by further research. From the present study, using an event-related design, α-pinene seems to be a key substance with regard to wood odor and is suggested to be selected as an olfactory stimulus in future studies.