Forest Bathing Is Better than Walking in Urban Park: Comparison of Cardiac and Vascular Function between Urban and Forest Parks

Forest bathing is beneficial for human health. To investigate whether walking in forest or urban parks affects cardiovascular functions (CVFs), the present study was conducted in five forest trails in the Xitou Experimental Forest and in five urban parks in Taipei city. We recruited 25 adult volunteers for an observational pilot study in forest parks (n = 14) and urban parks (n = 11). CVFs were assessed by measuring the arterial pressure waveform using an oscillometric blood pressure (BP) device. The baseline and paired differences of systolic BP (SBP), central end SBP, heart rate, left ventricle (LV) dP/dt max and cardiac output in participants were lower before and after walking in a forest park than those in an urban park. In addition, the systemic vascular compliance and brachial artery compliance of those who walked in a forest park were significantly higher compared with those in an urban park. Linear mixed models demonstrated lower levels of SBP by 5.22 mmHg, heart rate by 2.46 beats/min, and cardiac output by 0.52 L/min, and LV dP/dt max by 146.91 mmHg/s among those who walked in forest compared to those in an urban park after controlling covariates. This study provides evidence of the potential beneficial effects of walking exercise in forest parks on CVFs.


Introduction
Cardiovascular diseases (CVD) are the leading causes of morbidity and mortality worldwide [1]. The increasing rate of CVD has become a major focus of epidemiological studies, public health policies, and prevention studies. Researchers have reported that multiple factors increase the risk of developing CVD, such as hypertension, dietary habits, smoking habits, genetics, diabetes, environmental pollution, and psychosocial factors [2][3][4]. In addition to major CVD risk factors, the development of CVD is strongly connected to lifestyle factors such as psychosocial stress, unhealthy dietary habits, use of tobacco, and physical inactivity. More than three-fourths of all CVD mortality may be preventable with adequate changes in lifestyle [5]. Lifestyle changes, such as exercise habits, are one

Design and Participants
The present study was designed with the aim of determining the health effects of walking exercise on CVFs in a forest environment compared with those in an urban park environment. We recruited 25 adult volunteers for an observational pilot study of forest park and urban park. Participants with a mean age of 54.21 years (n = 14) in forest park and 42.36 years (n = 11) in urban park were recruited for this study. All participants provided informed consent before undergoing cardiovascular health examinations. The health examinations of participants were conducted from 21 to 26 June 2016, in National Taiwan University (NTU) Xitou Experimental Forest, Nantou County, Taiwan, and urban park examinations were conducted from 22 to 23 August 2016, in Taipei City, Taiwan.
Fourteen participants joined a 6-day/5-night forest walking exercise program in Xitou forest from 21 to 26 June 2016. During the forest walking, all participants had to maintain dietary control and walking exercise. All participants had to undergo CVFs examinations in the morning at a wood house before the forest walking. The participants walked for 1.5 h in a Phyllostachys edulis forest field as a forest walking exercise. Then, all participants came back to start at the wood house and undergo CVFs examinations again after taking a rest for 20 min. The full walking distance was approximately 2.4 km. All participants walked on five different forest trails (Japanese cedar forest, Taiwan red cedar forest, greensward, Taiwania cryptomerioides forest, and Phyllostachys edulis forest) for 1.5 h in the morning in NTU Xitou Experimental Forest. On the sixth day, after the health examinations were completed, all participants finished forest walking exercise and returned to Taipei City.
The health examinations of participants in urban parks in Taipei City were conducted from 22 to 23 August 2016. Of the 11 participants who were enrolled for the urban park examination, four also participated in the forest walking exercise program in Xitou forest. During the urban park walking exercise, all participants had to maintain dietary control and walking exercise. On the first day, all participants underwent CVFs examinations in the morning at the 228 Peace Park before walking in urban park. The participants walked for 1.5 h in Memorial Park as a walking exercise. Then, all participants finished urban park walking and underwent CVFs examinations again after a 20-min break. All participants went to the Chiang Kai Shek Memorial Hall Park and underwent CVFs examinations before and after walking exercise in urban park. In the afternoon, all participants went to the Pao An Temple Park and underwent CVFs examinations before and after walking exercise in urban park. On the second day, all participants had to go to the Sun Yat Sen Memorial Hall Park in the morning and the Long Shan Temple Park in the afternoon for urban park examinations.
This study was approved by the 37th meeting (30 January 2013) of the research ethics committee of the NTU Hospital, Taipei City, Taiwan. All participants provided written informed consent before undergoing a series of detailed examinations and questionnaires.

Site Descriptions of the Forest and Urban Park Environments
The Xitou experimental forest site of the NTU covers 2349 ha, mainly populated with natural hardwood forest and some plantations containing predominantly conifers. The forest ranges in elevation from 500 to 2025 m. The climate is subtropical, with an average annual rainfall of 2590 mm between 1941 and 2010. The average annual air temperature and relative humidity from 2011 to 2015 were 17 • C and 88%, respectively, according to the Xitou monitoring station of the Taiwan Central Weather Bureau (TCWB). The five sites of forest exercise training were Phyllostachys edulis forest, Japanese cedar forest, Taiwan red cedar, greensward, and a Taiwania cryptomerioides forest located approximately 1150-1300 m within the Xitou forest (Figures 1a and 2).

Phyllostachys edulis
forest Japanese cedar forest Taiwan red cedar forest Greensward Taiwania cryptomerioides forest  Location of study plot in forest trails and urban parks. The five forest trails in NTU Xitou Experimental Forest, Nantou County, Taiwan (a) and urban parks marked (b) in Arabic numerals were (1) Phyllostachys edulis forest, (2) Japanese cedar forest, (3) Taiwan red cedar forest, (4) greensward, (5) Taiwania cryptomerioides forest; the five urban parks in Taipei City, Taiwan (6)   Taipei City is the largest metropolitan area in the country. The urbanized area covers about 271.8 km 2 with estimated population of 2,597,000. Taipei City has a monsooninfluenced, humid, subtropical climate in which summer is hot, humid, and accompanied by occasional heavy rainstorms and typhoons. According to the TCWB, the annual average air temperature and relative humidity from 1991 to 2020 was 23.5 • C and 74.8%, respectively.

Exposure Assessments
The instruments used for environmental monitoring were real-time recording on the five forest trails and in the five urban parks. The concentration and size distribution, as well as real-time mass concentration of particulate matter (PM; PM 10 , PM 2.5 , PM 1 ), were monitored using a DustTrak aerosol monitor (model 8533; TSI Inc., Shoreview, MN, USA). The total volatile organic compounds (TVOC) concentration was measured with a ppbRAE3000 photoionization detector (RAE Systems Inc., San Jose, CA, USA). The concentrations of CO and CO 2 , temperature, and relative humidity were monitored using an IAQ monitor (model 2211; Kanomax, Andover, NJ, USA).

Cardiovascular Function Assessments
The internet was used for data transmission from the data collection site to a central analysis center, and BP was determined by pressure waveform changes according to Bernoulli flow effects. Vascular compliance and peripheral resistance of the brachial artery were derived by incorporating the arterial pressure signals from a standard cuff sphygmomanometer using a physical model. This method was previously validated against invasive [28] and non-invasive measurements [29]. Brachial artery compliance (BAC) was calculated at mean arterial pressure by theoretical design [28]. Brachial artery distensibility and resistance were derived from the modified formula of BAC, which was associated with cardiovascular risk factors in healthy young adults in the Bogalusa Heart Study [30,31]. This method has also been used to derive other hemodynamic parameters, including cardiac hemodynamics, such as the maximum rate of left ventricular pressure rise (LV dP/dt max), stroke volume (SV), and cardiac output [32] were also proved its application in our recent study [33].

Statistical Analysis
Means ± standard deviation (SD) of continuous variables between groups were compared using the Student's two-tailed t-test or the Mann-Whitney U-test if not in normal distribution. The analysis of variance was used to analyze the differences among group means. Environmental factors, including PM, TVOC, CO, CO 2 , temperature, and humidity in five forest parks and urban parks were compared. The goal of the analysis was to test for the mean differences and to quantify them. We used the paired t-test to evaluate each individual's mean difference of cardiac and vascular hemodynamics measured before and after walking exercise in five forest parks and five urban parks. Regarding CVFs changes after walking in different environments, the differences between in the forest and urban environments were tested by using two sample t-test. All statistical analyses were performed with SPSS statistical software (version 19; IBM SPSS Statistics, 2010, Chicago, IL, USA).
Linear mixed-effects models were used to estimate the changes on CVFs after walking (forest bathing) in the Xitou Experimental Forest compared to those changes of cardiovascular functions after walking in urban parks. The individual characteristics was set as random effects, and estimated by restricted maximum likelihood, and fixed effects were model-based after controlling the categories of age, gender, BMI, and hypertension status.

Results
General characteristics of study participants in forest and urban parks are summarized in Table 1. The mean age of participants in forest and urban parks were 54.2 and 42.36 years, respectively. The enrollees in forest and urban parks were mostly males-71.43% and 72.73%, respectively. The prevalence of hypertension (140 mmHg in systolic BP or 90 mmHg in diastolic BP) in forest and urban parks was 28.57% and 36.36%, respectively. Diabetes mellitus in adults accounted for 28.57% of the participants, and 42.86% were diagnosed with hyperlipidemia in forest parks. Overall, the demographic characteristics of study participants in forest and urban parks were not significantly different.

Environmental Characteristics
The comparison of ambient air quality monitoring between forest and urban parks are presented in Table 2. Levels of PM 1 , PM 2.5 , PM 10 , total PM, TVOC, CO, and temperature were significantly lower in the forest environment than in urban parks. There were significant differences in air pollution between forest and urban parks. However, the relative humidity in forest parks was significantly higher than in urban parks.
The environmental monitoring results in five different forest trails are presented in Table 3. Based on analysis of variance, there were significant differences among the five different forest trails according to results of the F-test of analysis of variance. The PM 1 , PM 2.5 , PM 10 , and total PM levels, TVOC, CO 2 , temperature, and relative humidity were significantly different among the five forest trails. However, there were no significant differences in CO level among the five different forest trails.
Ambient air quality monitoring data among the five urban parks are presented in Table 4. There were significant differences among the five urban parks. The PM 1 , PM 2.5 , PM 10 , and total PM levels were significantly lower in Long Shan Temple Park. However, 228 Peace Park had significantly higher levels than other parks. There was significantly better air quality in TVOC, CO, and CO 2 , lower temperature, and higher relative humidity in the Xitou forest environment compared to those of urban parks.  Table 3. Ambient air quality monitoring among five different forest trails.

Cardiovascular Function
Brachial systolic BP (SBP), central end-cSBP (cSBP), and pulse pressure (PP) among forest participants were significantly lower after walking exercise in forest parks (Table 5). Cardiac function including heart rate (HR), LV dP/dt max, LV contractility, cardiac output, cardiac index (CI), SV, and SV index among forest participants were significantly lower after walking exercise in forest parks. Systemic vascular compliance (SVC) among forest participants was significantly higher in forest parks. Brachial artery resistance (BAR) among forest participants was significantly lower after walking exercise in forest parks. However, among urban participants, there were no significant differences after walking exercise in urban parks.
Comparison of CVFs changes after walking exercise in forest and urban parks showed in Table 6. The effects after walking exercise among forest participants in forest parks are as follows: Decreased 3.78, and 4.62 mmHg in the SBP and cSBP, respectively; decreased 1.36 and 5.29 mmHg in the mean artery pressure and PP, respectively; and decreased 1.3 beats/min, 97.18 mmHg/s, and 0.25 L/min in the HR, LV dP/dt max, and cardiac output, respectively. There were statistically significant differences between the crude effects in forest and urban parks.
In order to estimate the health effects on CVFs after walking exercise, linear mixedeffects models were used to adjust random effects of individual characteristics and controlling factors of age, gender, BMI, and hypertension status. The estimated effects of forest park vs. urban parks using linear mixed models were shown in Table 7. Brachial SBP and cSBP of participants in forest parks were significantly lower than that in urban parks. The same we found that the heart rate, LV dP/dt max, LV contractility, cardiac output, CI, SV, and SV index in forest parks were significantly lower than that in urban parks. Linear mixed models showed lower levels of SBP by 5.22 mmHg, heart rate by 2.46 beats/min, cardiac output by 0.52 L/min, and LV dP/dt max by 146.91 (mmHg/s), respectively, while those walking in the forest compared to those in urban parks after controlling covariates.

Discussion
This study is the first in travel medicine to show the potentially beneficial health effects, in terms of CVFs, after walking exercise of participants in forest parks compared to those measured in urban parks. This study provided a novel finding of concurrent CVFs evaluation and real-time environmental monitoring, which showed that improved air quality is accompanied by a better cardiovascular response to walking exercise in forest parks than in urban parks. Our previous study showed that those living in a forest environment may demonstrate improved cardiovascular health, such as significantly lower levels of total cholesterol, low-density lipoprotein cholesterol, fasting glucose, and carotid intima-media thickness than those living in an urban environment [34].
Studies regarding certain activities in a forest environment, such as breathing, relaxation, meditation, and walking exercise have shown beneficial health effects [35]. In the present study, we found that walking exercise has more significant effects on CVFs with exposure to forest parks. SBP and cSBP of participants in forest parks were significantly lower than those in urban parks. It is well known that BP is influenced by multiple factors, relative contributions are probably quite different from person to person. For example, lack of sleep at night, anxiety, having noise within the measured environment, not taking rest before the measurement, and lifestyle factors (e.g., diet, coffee, smoking, alcohol, and exercise) affect BP components. Our findings are consistent with previous studies that exposure to a forest environment by walking or even just staying in the forest environment can decrease BP, cortisol, HR, and sympathetic nerve activity [35,36].
Many studies have confirmed that participants with the widest PP have the greatest risk of mortality associated with CVD [37]. PP has also been reported to be a significant indicator of myocardial infarction [38]. Furthermore, PP is a strong indicator of cardiovascular risk even among normotensive persons [39]. Our study shows that participants' PP and HR before, after, and paired difference after walking exercise in forest parks were lower than those in urban parks. This study also indicates that exposure to forest environments might decrease PP and HR levels in humans. On the basis of these studies, walking exercise in forest parks, as compared with that in urban parks, can be considered more effective in providing relaxation. Cardiac function, including dP/dt max, LV contractility, cardiac output, and SV, decreased significantly in participants in forest parks compared with those in urban parks. This implies that in humans, cardiac burden is reduced after walking exercise in a forest park. Many studies have shown that contact with a forest environment might make people calmer and more relaxed than in an urban area [40]. Exposure to a forest environment is now receiving increased attention for its effects of reducing stress and providing a feeling of relaxation. Intra-individual paired difference comparisons of SVC, BAC, and BAR before and after walking exercise of participants in forest parks had statistically significant differences from the corresponding differences in urban parks.
Assessments of the impact on cardiac function before and after exercise training are very important for hypertensive patients or general people to improve health. Many studies have reported that exercise training can provoke important changes in the CVFs in humans and animals with hypertension [41][42][43][44]. Study reported that low-intensity exercise training, which is performed on a motor treadmill for 60 min, 5 times per week for 18 weeks, at 55% maximum oxygen uptake (VO 2max ), can significantly decrease BP, HR, and cardiac output in hypertensive rats than in a sedentary condition or a high-intensity exercise training (88% VO 2max ). The decrease in cardiac sympathetic tone in low-intensity exercise trained in hypertensive rats may cause hemodynamic changes (decreased BP, HR, and cardiac output), consequently attenuate the cardiovascular load in hypertension [42]. Study reported that low-intensity exercise training in older hypertensive patients, who were required walking at home for 60 min 3 times per week for 9 months at 50% VO 2max , revealed a significant reduction in arterial BP, HR, and cardiac output after training [43]. In the present study, walking approximately 2.4 km for 1.5 h in forest trails has more significant reduction on cardiac and vascular workload in a forest environment compared to that in urban parks. Particularly, HR, LV dP/dt max, LV contractility, cardiac output, and SV before and after exercise of participants in forest parks were significantly lower than those of participants in urban parks.
The participants recruited for the observational pilot study all lived in Taipei City and walk regularly in urban parks, with an average walking time of 1.5 h/day. However, the participants in the urban park walking group did not demonstrate improved CVFs. CVFs of participants who walk regularly in urban parks may be affected by traffic air pollution [45,46], sports type and exercise intensity [47], higher ambient temperature in summer [48], and urban heat island effect [49]. The noise levels in the two different environments are different. The city noises (e.g., road traffic noise and engineering construction) exposures engender physiological reactions of stress [50]. Based on the above discussion, we considered inappropriate urban weather conditions, lack of park's greenness or inadequate tree canopy coverage in urban area might attenuate the beneficial healthy effects of urban parks on CVFs. It is well known that increasing tree canopy coverage or park's green area can reduce ambient air pollution and temperature in urban area. Thus, improvements in urban green spaces and canopy coverage are a very important issue. Urban forest or afforestation in the city should be encouraged to increase the urban greenness.
There are some possible explanations in our findings to understand cardiovascular phenomenon in forest parks. First, HR, LV contractility, cardiac output, and SV in forest parks decreased after walking exercise, indicating that physiological relaxation on reduction of cardiovascular activity in human body [51][52][53]. Second, exposure to forest environments might enhance the immune response of NK cells (CD3 − /CD56 + ) and activating NK cells (CD3 − /CD56 + /CD69 + ) in humans. The percentage of NK cells was significantly higher in the forest group (19.5 ± 9.1%) than in the urban group (16.4 ± 8.4%). After the fiveday in Xitou forest bathing, forest environments and plant-released BVOCs inhaled by walkers, the percentage of activating NK cells of walkers increased significantly from 0.83 ± 0.39% to 1.72 ± 0.1% in our previous studies [15]. Third, our recent studies showed that long-term traffic-related air pollution, concentrations of particulate matters and NO 2 in Taipei City were closely and independently associated with inflammation, and thrombotic biomarkers [54] Thus, better air quality in forest environments may improve cardiovascular health [34]. Forth, better green landscapes in forest environment may benefit CVFs [55][56][57].
This study has several strengths. The strength of this study demonstrates the consistent health effects of walking exercise in different forest trails on BP components, cardiac function and hemodynamic parameters simultaneously compared with those in urban parks. This study corroborated that the findings for before and after walking exercise, and the paired difference between these cardiovascular hemodynamics revealed the significantly better health effects in forest parks. The unique findings of this study also can be supported by concurrent measurements of environmental parameters, including lower air pollution parameters and higher levels beneficial phytoncides.
However, our study had some limitations. First, our sample size was relatively small, which may cause true effects on the CVFs changes to be undetected. However, the CVFs measurements of the participants were taken in both hands and in five forest trails or five urban parks. With the increased number of measurements, we had a higher chance to detect differences in CVFs changes between walking in forest park areas and urban park areas using linear mixed-effects models. Second, the age, gender, and diseases of our participants in forest and urban parks not evenly distributed. However, Table 1 showed that these characteristics of the two groups were not significantly heterogenous. We also fitted the observed CVFs changes after walking exercises with linear mixed-effects models to adjust potential confounding factors, especially age, gender, BMI and hypertension. Third, there was no control group walking through the urban area to investigate if the parks or forests improve the health condition in comparison to urban area. Fourth, the air temperature and relative humidity in Xitou Experimental Forest and the urban parks were statistically significantly different (10 • C and 30%, respectively). The different weather conditions, including air temperature, relative humidity, atmospheric pressure and wind speed, that occurred during the research in urban and forest parks have not been considered.
Fifth, although our results indicate positive health effects of walking exercise in forest parks compared with urban parks, we could not infer that the health effects are attributable to BVOCs, because BVOCs from tree leaves in forest environments were not investigated in this study. We believe that levels of BVOCs with healthy potential inhalable during forest bathing are higher in forests than urban parks. The health benefits of BVOCs found in forest environments and the mechanisms mediating the effects of phytoncides on the cardiovascular system are very important in public health and environmental sciences.
Perspectives of this study also indicated that further studies designed to demonstrate the potential role of phytoncides on CVFs are warranted.

Conclusions
This study indicates the potential health effects on CVFs of participants who walked in a forest environment. Our results support that a better urban design with increased urban green area and canopy coverage, as well as afforestation in urban cities, may improve the health of citizens. The government should actively develop parks and recreational areas through urban greening plans. Conducting a large-scale cohort study with more participants and investigating other mechanisms including human bioclimatology, and sympathetic/parasympathetic functions, are warranted in future. Funding: This study was financially supported by grant 105-C02 of the Experimental Forest, National Taiwan University. This study also supported by grant MG-287 from the National Taiwan University Hospital.
Institutional Review Board Statement: This research has been approved by Research Ethics Committee of the National Taiwan University Hospital (201212173RIND). The study participants provided their written informed consent before received a series of detailed examinations and questionnaires.
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.