Forest bathing diminishes anxiety in undergraduate students: a pilot study in the Valdivian temperate rainforest

ABSTRACT Urban life has generally resulted in lifestyles characterised by high stress and anxiety, all which has increased the pervasiveness of mental disorders. University students, a group highly vulnerable to these problems, are likely to develop mental pathologies associated with anxiety and substance abuse. Forest bathing is one nature-oriented intervention aimed at improving physical as well as psychological well-being by reducing anxious symptomatology and stress. Here we evaluated the healing impact of forest bathing on the stress and anxiety of undergraduate students at one local university. We also included others key aspects of psychological mental health that have received scarce attention in forest bathing research like mindfulness, wellbeing and self-regulated learning. Twenty-one students belonging to the Universidad Austral de Chile were randomly assigned to the city or forest groups and took part in an intervention consisting in a low-impact 30 minute walk and a 15-minute contemplation afterwards. Five instruments to evaluate overall mental health (and blood pressure) were applied before and after the intervention. Our results showed that a rather short session of forest bathing is enough to reduce anxiety in undergraduate university students. The other measures of psychological or physiological health showed no statistical differences before and after the intervention (between city and forest) or time effects (within the forest or city treatment). A potential reason for this might be related to the short duration of the intervention itself. Our results are encouraging regarding mental health therapy, as this type of intervention could easily be implemented on campus.


Introduction
Urban life has generally resulted in lifestyles characterised by high stress and anxiety, with scarce moments of peace and feelings of well-being in daily life (Sundquist et al. 2004).University students are a group vulnerable to develop mental pathologies associated with anxiety and substance abuse (Eisenberg et al. 2007).Considering this background, several authors have sought to conduct nature-based preventive interventions in order to reduce the negative effects of heavy academic workloads while also strengthening the development of positive emotions (Regehr et al. 2013;Pascoe et al. 2020).
Forest bathing (FB) is one nature-oriented intervention aimed at inducing states of relaxation to improve physical and mental health as well as overall well-being (Miyazaki 2018;Wen et al. 2019).In university students, research has shown that FB induces a reduction in stress and negative feelings and emotions (Rajoo et al. 2019) together with a feeling of vitality that benefits the parasympathetic system and physiological and mental well-being (Bielinis et al. 2019;Lyu et al. 2019).Although most studies have examined people from East Asian countries, similar positive results have been observed in Europe and the Middle East (Vujcic and Tomicevic-Dubljevic 2018;Olafsdottir et al. 2020).The physiological effects of FB have been more variable, depending on particularities of the methodological setup, but with positive effects in terms of reduced blood pressure and heart rate, mostly observed when contemplative practices are incorporated (Lyu et al. 2019;Rajoo et al. 2019).Unfortunately, most research on FB has been conducted in Asia and Europe, with nearly no studies in South America and therefore, the generality of these findings remains yet to be stablished.
South America is a subcontinent characterised by astonishing numbers of biodiversity, high endemicity and natural ecosystems that host multiple biodiversity hotspots (Marchese 2015).Among those is the Chilean winter rainfall Valdivian forest biodiversity hotspot, one of the 35 key terrestrial areas in the globe for biodiversity conservation (Myers et al. 2000).This hotspot has a large latitudinal range (ca.3,000 km), encompassing different ecosystems, vegetational formations and climates (Myers 2011).Prior to the COVID-19 pandemic, Chile showed rather poor mental health indicators (Basso-Musso et al. 2012;González-Suitt et al. 2019).,Although these indicators worsened during the pandemic as a consequence of lockdowns, quarantine, and social distancing (Langer et al. 2022).Therefore, under the umbrella of a pilot study, we evaluated the healing impact of a brief FB session (less than one hour) on the stress and state anxiety of undergraduate students at one local university.In addition, we included others key aspects of psychological mental health that have received scarce attention in FB research like mindfulness, wellbeing and self-regulated learning.Specifically, we would like to test if a brief FB session, which promotes attentional awareness, has a positive effect on mindfulness skills without any specific mindfulness practice or training (e.g.breathing meditation).Moreover, although FB has demonstrated positive effects on aspects of well-being such as positive emotions (e.g.Bielinis et al. 2018), it has not been evaluated whether they also have the same effect in areas of wellbeing related to personal development.Furthermore, we included self-regulated learning as a relevant outcome of academic performance given its positive association with wellbeing (e.g.Tavakolizadeh et al. 2012;Davis and Hadwin 2021).Therefore, it is worth exploring the effect of FB on both variables which has not yet been assessed in the frame of FB.Finally, considering that the effect on blood pressure in university students presented mixed results (Rajoo et al. 2019;Song et al. 2019) we decided to measure it.

Methods
A pre-post pilot study with control group design was used to evaluate the psychological and physiological effects of FB compared with an urban group in the city of Valdivia, Chile.The main scope of conducting a pilot study was to collect preliminary data which could show the relevance of the topic for stakeholders and gives insights on novel research questions (Van Teijlingen and Hundley 2002).

Participants and procedure
The study was conducted on a Saturday morning during the second semester of 2019 in the city of Valdivia.Specifically, this place is a 10-minute walk from the "Isla Teja" campus, encompassing more than 60 hectares mostly composed of temperate Valdivian forest and areas next to two navigable rivers (i.e. the rivers Cruces and Cau-Cau).r'sforests, and was authorized by the Ethics Committee of the Valdivian Health Service.Twenty-one students (mean age: 21.4 ± 2.3 years, range: 18-25) belonging to the Universidad Austral de Chile (UACh) were randomly assigned to the city (N = 11; 5 females and 6 females) or forest (N = 10; 5 females and 5 males) groups.
Each participant signed an informed consent before taking part in the intervention, which consisted in a low-impact 30-minute walk in either the city centre or a nearby forest along pre-established routes (see Figures S1 and figure S2 in Appendix for the routes) and a 15-minute contemplation afterwards.The FB session was conducted in the Arboretum of the UACh, which is a green space aimed for education and recreation of the general population (http:// arboretumuach.cl/).Specifically, this place is a 10-minute walk from the "Isla Teja" campus, encompassing more than 60 hectares mostly composed of temperate Valdivian forest and areas next to two navigable rivers (i.e. the rivers Cruces and Cau-Cau).Participants were asked to sleep for eight hours the night before and to eat a balanced breakfast as well as to turn off their mobile phones throughout the intervention.Throughout the 30-minute walk, two instructors per group sought to keep the participants' attention focused on the here and now, directing their senses towards their actions and the environment.The activities were conducted in silence.At the end of the route, participants were given yoga mats to sit down or lie down comfortably and a 15minute guided contemplation of a panoramic view of a river or a typical city street corner was conducted.Psychological tests (and blood pressure) were applied before and after the intervention.

Instruments
Five instruments were applied to evaluate overall mental health.An adapted and locally validated (Véliz Burgos 2012) version of the Ryff (1989) was used to assess psychological wellbeing.The scale comprises 29 statements that evaluate self-acceptance, positive relations with others, autonomy, environmental mastery, purpose in life, and personal growth.Respondents rated statements on a Likert scale ranging from 1 ("Strongly disagree") to 5 ("Strongly agree").Mindfulness was evaluated using the Spanish version (Soler et al. 2012) of the Mindful Attention Awareness Scale (MAAS) (Brown and Ryan 2003), which comprises 15 statements related to daily life experiences rated on a Likert scale ranging from 1 ("Almost always") to 6 ("Almost never").Anxiety was evaluated with the Chilean version (Burgos-Fonseca and Gutiérrez-Sepúlveda 2013) of the State-Trait Anxiety Inventory (STAI) (Speilberger et al. 1970), which evaluates state anxiety (i.e. a transitory emotional state consisting of feelings of apprehension, nervousness, and physiological sequelae such as an increased heart rate or respiration) and anxiety as a predisposed behavior (i.e.trait).In this study, only the state anxiety scale was applied, which is composed of 20 statements rated on a Likert scale ranging from 1 ("Not at all") to 5 ("Very much so").Selfregulated learning was evaluated using the Spanish version (Bruna et al. 2017) of the Inventory of Learning Self-Regulation Processes (IPAA), composed of 12 items grouped into three dimensions ─forethought, performance, and self-reflection─ rated on a Likert ranging from 1 ("Never") to 5 ("Always") (Zimmerman 2002).Finally, Perceived stress (PSS-14) was evaluated using the locally validated (Carvajal et al. 2017) scale developed by Cohen et al. (1983).It comprises 14 questions ─rated on a Likert scale ranging from 1 ("Never") to 5 ("Very often")─ that evaluate the degree to which a variety of daily experiences can be regarded as stressful.
Systolic and diastolic blood pressure were taken using an automatic monitor (Élite Plus OMRON HEM-7320) whose smart armband reduces the risk of incorrect measurements due to misplacement (Mitienda.cl 2019).

Statistical analyses
We used a linear mixed modeling approach employing maximum likelihood to evaluate the effects of the intervention treatment (i.e.City versus Forest) and time (Pre-intervention versus Post-intervention) in the different psychological and physiological variables measured, taking into account that we had repeated measurements for each individual.For all analyses, the intervention treatment and time were set as fixed factors and the ID of each individual was included as a random factor.We employed model selection using an information-theoretical approach (Anderson and Burnham 2002) to evaluate the impact of (i) the null model, (ii) intervention treatment, (iii) time, (iv) intervention treatment plus time, and (v) intervention treatment in interaction with time, in all the psychological and physiological variables measured.Likert scores for each statement for each mental health variable were averaged for each individual before analyses, which resulted in normally distributed variables.Systolic and diastolic blood pressures were also averaged before analyses.All analyses were performed in R version 3.6.3(2021) employing packages lme4 (Bates et al. 2015) and AICcmodavg (Mazerolle et al. 2020).

Results
Irrespectively of the variable considered, there was not a single best model (i.e.Akaike weight of 0.95 or more) for any of them (Appendix Table S1).Therefore, we estimated four model-averaged effect sizes for each group: Pre and Post-intervention treatment effects (i.e.city versus forest) and time effects within the forest treatment and time effects within the city treatment (Appendix Table S2).Of the five psychological variables studied, only STAI and IPAA showed effect sizes different from 0 (Appendix Table S2).All six components of RYFF had no significant effect sizes (results not shown).Regarding STAI, there were no differences in treatment effects before the intervention, but the mean STAI was reduced in the forest group after the intervention.Mean STAI did not show any time effects in the city group, but it was reduced in the forest group after the intervention (Figure 1, Appendix Table S2).There was also an effect of treatment on IPAA, with the city group having lower values than the forest group both before and after the intervention (Figure 1, Appendix Table S2), which indicates that the differences were already present before the onset of the experiments.The remaining psychological variables (RYFF, MAAS, and PSS) showed very modest effect sizes and were in no case statistically different from 0 (Figure 1, Appendix Table S2).Mean arterial blood pressure also exhibited very small effect sizes, with no apparent impact of the intervention treatment or time effects (Figure 1, Appendix Table S2).

Discussion
Our results showed that a rather short session (i.e. 30 minute walk plus 15 minutes of contemplative activity) of FB is enough to reduce state anxiety in undergraduate university students, which is in agreement with previous research in other countries (Park et al. 2014;Song et al. 2019).This is an encouraging result regarding mental health therapy, as this type of intervention would not be unfeasible for a student with regular intensive academic duties.
The other measures of psychological or physiological (MAP) health showed no statistical differences before and after the intervention (between city and forest) or time effects (within the forest or city treatment).A potential reason for this might be related to the short duration of the intervention itself, which may be insufficient to generate changes in specific variables such as mindfulness, wellbeing, self-regulate learning and stress perception.Despite the fact that FB promotes attentional awareness through keeping the participants' attention focused on the here and now, directing their senses towards their actions and the environment (for this reason we used the Mindful Attention Awareness Scale, MAAS), our results suggest that for an improvement in mindfulness skills specific mindful practices or training (e.g.breathing meditation or body scan) should be incorporated during the FB (e.g.Muro et al. 2021).Moreover, our results suggest that changes in eudaimonic-based components of well-being such us personal growth may need longer FB interventions where a process of self-knowledge could be guided and maintained.Thus, current research shows that brief FB activities improved aspects of the wellbeing related to the decrease of negative emotions and the increase of sense of vitality (e.g.Bielinis et al. 2018).
Furthermore, Olafsdottir et al. (2020) suggest that longer doses or a combination of multiple events must be used to effectively observe changes in psychological traits or attain long-term changes in states.In any case, a 30-minute walk in nature with a short contemplation afterwards might be enough to reduce anxiety levels and even the cortisol levels associated with stressful situations (Tyrväinen et al. 2014).Regarding blood pressure (MAP), as was reported by Song et al. (2019), we found no major effects between groups as a consequence of the intervention.However, it should be noted that any effects of the FB intervention on MAP may have been affected by the physical exertion of the walks themselves.One way to disentangle these effects (i.e. the walk from the intervention), is to carry out a factorial design that includes walk (yes/no) completely independent from the intervention.
The present pilot study is not free from limitations.The sample size is rather small, and therefore the generalisation of these results should be made with caution.Furthermore, we did not include a more accurate evaluation of environmental conditions during the study (e.g.humidity, wind speed) which could impact the personal experience of FB.A future RCT protocol may consider additional physiological variables ─such as cortisol─, which would have made it possible to compare with the self-report measures as a more integral assessment.Additionally, we were not aware of the academic situation of the participants at the moment of the study, which could have helped us to understand if they were under a particular stressful period (e.g.exams) which would increase the amount of anxiety among them (Olafsdottir et al. 2020).This may explain between-group differences observed at IPAA on pre-test, which did not allow us to start from equivalent scores.Therefore, it could give information about differences in self-learning process as we found in this study previous to the FB session which leads to nonconclusive results.Additionally, we did not include a participant's profile of medical conditions which could impact mental health.For instance, whether or not smokers were included and whether or not participants have preexisting medical conditions which may have an impact on the blood pressure measure.All these aspects should be considered in future research.Specifically, the effect of FB on academic performance and its relationship with the wellbeing of students requires further special attention for mental health promotion.
In summary, the observed state anxiety reduction after a brief FB intervention, which does not require any special training to be implemented, can be a significant argument to encourage students to become involved in these activities.In addition, the very short time required to carry it out, means that this type of intervention can be completed even between classes.Furthermore, the forest place is almost on campus (10-minute walk), and thus, it was explicitly selected considering how feasible it would be for students to visit it autonomously while also gauging how realistic it would be for them to use it on a regular basis.Actually, a recent study found that both a guided or self-guided forest healing program produced psychological benefits using campus forests (Shin et al. 2023).We expect that the preliminary data presented in this study boost FB research in South America, a region where the relatively easy access to forests and nature emerges as a major opportunity to study the effects of FB on the health and well-being of university students.