
2023 Impact Factor
Mild cognitive impairment (MCI) can be defined as a transitional state between healthy aging and dementia [1] and is characterized by a decline in cognitive function and memory beyond that associated with typical aging, while activities in daily lives are unaffected [2]. MCI has a prevalence of between 3% and 19% in the general elderly population, and 20−50% of patients usually develop dementia within 3 years [3]. Since early intervention at the prodromal stage of the disease represents the best treatment option for dementia [4], timely treatment of MCI is essential. Nevertheless, only few medical treatments are known to effectively retard disease progress, highlighting the need for a multidisciplinary approach for treating MCI.
As an alternative therapy that has recently been gaining interest, forest therapy has been proposed as a potential treatment for palliating MCI symptoms. Forest therapy is defined as “immune-strengthening and health-promoting activities utilizing various elements of the forest, such as fragrance and scenic view” [5] and is known to have the potential to improve physical and mental wellbeing [6-9]. Notably, several recent studies have suggested the therapeutic benefits of forest therapy for elderly individuals with MCI or other types of cognitive decline, by demonstrating enhanced neural and parasympathetic nervous activity [10,11], decreased depression levels [11], and improved mini-mental state examination (MMSE) scores [12] in patients who underwent forest therapy sessions. However, the restorative potential of forest therapy upon MCI requires further investigation because only little is known about the underlying therapeutic mechanism. Particularly, because forest therapy sessions are generally composed of various types of activities and elements (i.e., physical activities and olfactory/visual stimulations) [5], it is required to clarify the therapeutic effects of each element of forest therapy on MCI patients.
In this context, phytoncides, which represent the olfactory-related therapeutic component of the forest environment, are considered one of the vital factors that induce the therapeutic benefits of forest therapy. Phytoncides are volatile organic substances derived from trees and plants which are emitted to prevent damage caused by bacteria, fungi, or insects [13]. The benefits of phytoncides have been reported with regard to mental wellbeing, such as emotion regulation, attention, and cognitive/executive functions, when applied to normal populations in an aromatherapeutic manner [14-16]. Such evidence suggests that phytoncides also hold the therapeutic potential towards MCI, because one of the critical symptoms of MCI is cognitive dysfunction related to major disturbances in the prefrontal cortex (PFC) [17,18]. Bae et al. [19] presented supportive evidence for this hypothesis by showing that the cognitive functions of MCI-induced (Aβ-injected) rats that inhaled phytoncides were protected. This finding, along with results that highlighted the cognitive elevation effect of phytoncides in a normal population [14-16], cautiously suggests that phytoncides have therapeutic benefits for human patients with MCI.
Here, based on previous findings regarding the effects of phytoncides, we conducted a double-blind, randomized controlled pilot study to investigate their therapeutic benefits in patients with MCI. To our knowledge, there has been no reported study to verify the cognitive elevation effect of phytoncides in patients with MCI. Thus, in the present pilot study, we aimed to examine the cognitive and neurophysiological changes after phytoncide inhalation in patients with MCI. To accomplish this goal, we imposed Stroop tasks on patients with MCI before and after 30-minute phytoncide inhalation, comparing their changes in task performance and hemodynamic response of task-related prefrontal regions. We used functional near-infrared spectroscopy (fNIRS), a brain imaging tool that measures brain activation via relative changes in oxyhemoglobin (HbO2), in terms of evaluating changes in hemodynamic response measured while patients with MCI performed the Stroop task. According to brain imaging studies demonstrating that the dorsolateral and ventrolateral parts of the PFC are specifically sensitive to Stroop tasks depending on the task type, treatment intervention, clinical or cognitive conditions, and age [20-25], the present study focused on fNIRS-measured neurophysiological changes in the dorsolateral PFC (DLPFC) and ventrolateral PFC (VLPFC).
All experiments were conducted in accordance with relevant guidelines and regulations. The research was carried out within the psychiatric facilities of Chuncheon Sacred Heart Hospital, a teaching hospital affiliated with Hallym University, College of Medicine, Republic of Korea. The study is registered in the Clinical Trials Registry of Korea (registration: KCT0007317; date: 19/05/2022). This study was approved by the Institutional Review Board of Chuncheon Sacred Heart Hospital, Republic of Korea (Approval No. 2021-06-011). All the participants provided written informed consent.
Individuals aged over 60 years, who had received clinical diagnoses of MCI, were recruited through promotional efforts at local hospitals with psychiatric units and community mental health centers.
Inclusion criteria for the study encompassed the following: (a) meeting the diagnostic criteria for mild neurocognitive disorder as per the Diagnostic and Statistical Manual of Mental Disorders, 5th edition (DSM-5) [26]; (b) a confirmed diagnosis of MCI in line with the criteria established by Petersen [27], which includes memory complaints, the ability to perform daily activities normally, general cognitive function within the expected range, memory decline beyond age-related expectations, and the absence of dementia; and (c) scoring between 0.5 and 2 on the Clinical Dementia Rating Scale [28] or between 2 and 5 the Global Deterioration Scale [29].
Exclusion criteria were as follows: (a) present use of cognitive enhancers (donepezil, rivastigmine, galantamine, and memantine); (b) history of neurological conditions including head trauma, epilepsy, or Parkinson’s disease; (c) prior psychiatric conditions align with specific DSM-5 criteria, including bipolar disorder, substance abuse, ongoing alcohol dependence, and personality disorders; (d) history of endocrine disease, asthmatic condition, or clinically severe diseases that affect brain functions, such as cardiovascular function; (e) any defects in olfaction or vision (e.g., color-blindness); (f) Korean-illiterate; and (g) participation in another clinical study within a month. The use of psychiatric and general medications, except cognitive enhancers, was allowed during study participation.
Licensed clinical psychologists and board-certified clinical psychiatrists performed comprehensive, in-person diagnostic assessments. These evaluations involved conducting face-to-face interviews with patients and a thorough review of the medical history, presenting symptoms, and psychosocial functioning, incorporating all accessible sources of information. Of the 30 participants screened, 21 right-handed participants (10 males and 11 females) aged between 63 and 89 years met the initial criteria.
The phytoncide essence oil used in this study included distilled oil extracted from cypress leaves using steam distillation. The oil passed the safety standard conformity test and was stored at a constant temperature of 23°C before it was applied to the masks. Gas chromatography-mass spectrometry was used to analyze the chemical composition of the phytoncide oil. The results of this analysis are presented in Table 1.
The Stroop task is used to assess cognitive performance and neurophysiological changes in patients with MCI [30]. The Stroop task is designed to estimate cognitive functions, mainly including the executive functions such as interference control, inhibition, and response flexibility by presenting conflict-inducing stimuli (e.g., the word “blue” written in red ink) to the participant [31]. As the quality of the task primarily focuses on assessing executive functions, PFC regions represent the underlying significant neural correlates of Stroop tasks [20-22].
In the present experiment, the participants performed a one-block incongruent Stroop task consisting of 100 Korean words of colors printed in unmatched colors (e.g., the word “blue” printed in red ink), before and after the inhalation stage. Participants were instructed to orally verify the ink color of as many words as possible in 60 seconds, and their performance was evaluated as a score based on the number of correct answers given in the limited time period.
The 21 participants were randomly stratified into two groups: the “saline (no odor) group” (n = 10) as the control group and the “phytoncide group” (n = 11) as the phytoncide inhalation group. Random assignment was performed by a research assistant who was not involved in this study. Regardless of group, all participants were informed that they were subjected to inhalation of an aromatic oil, which is known to be helpful in restoring cognitive function in MCI. When a participant entered the inhalation stage, the research assistant responsible for the group assignment delivered either a phytoncide or a saline-applied mask depending on the group condition. Fragrance was applied by dropping 1 ml of undiluted phytoncide essence oil for the phytoncide group or 1 ml of saline solution for the saline group evenly on both edges of a dental mask. The mask covered both the mouth and nose of the participants, and the participants were instructed to sit comfortably without talking or moving while breathing spontaneously with their eyes fully closed for 30 minutes. The experimental room used in inhalation stage was maintained at a constant temperature of 23°C with a relative humidity of 50%.
Before and after the inhalation stage, the participants performed a one-block incongruent Stroop task consisting of 100 Korean words of colors printed in unmatched colors. While performing the task, the hemodynamic response of their bilateral DLPFC and VLPFC was measured using an fNIRS device placed on the forehead.
The hemodynamic responses of the bilateral DLPFC and VLFPC measured during the Stroop tasks were recorded using a high-density NIRS device (NIRSIT; OBELAB Inc.). The device consists of 24 source laser diodes and 32 photodetectors, enabling hemodynamic response analysis of 48 channels distributed in the PFC regions (Fig. 1). In this study, since our region of interest (ROI) mainly lies in the bilateral DLPFC and VLPFC, channels only included as DLPFC (right: 1, 2, 3, 5, 6, 11, 17, 18; left: 19, 20, 33, 34, 35, 38, 39, 43) and VLPFC (right: 4, 9, 10; left: 40, 44, 45) regions according to the Broadman Area were analyzed (Fig. 1B).
To minimize artifacts due to environmental noise-related light and physiological noise, the detected light signals from the channels were filtered using a 0.005−0.1 Hz band-pass filter. The rejection threshold for the signal-to-noise ratio was 30 dB, and the relative hemodynamic responses of unrejected channels were calculated using the modified Beer-Lambert law [32]. Since the measurement for each Stroop task block was conducted for 60 seconds, the calculated HbO2 values for each channel were block-averaged individually before grand-averaging for each group. The HbO2 level in a certain channel or region represents the neural activation of a specific PFC region during the task. The representative means and standard deviations of HbO2 levels from channels within each of the four ROI regions (right DLPFC: 1, 2, 3, 5, 6, 11, 17, 18; left DLPFC: 19, 20, 33, 34, 35, 38, 39, 43; right VLPFC: 4, 9, 10; left VLPFC: 40, 44, 45) were calculated.
Owing to limitations of the sample size and data distribution, nonparametric statistical analysis was performed on the data. For confirming the demographic and clinical homogeneity of two groups, χ2 and Mann–Whitney U tests were used. To evaluate the Stroop task performance and hemodynamic changes in the PFC regions, generalized estimating equations (GEE) were applied with regard to multiple comparisons of Stroop task scores and HbO2 values for each of the 4 ROI regions. GEE analyses were conducted by setting “inhalation (pre-inhalation; post-inhalation)” as the within-group factor and “group (saline; phytoncide)” as the between-group factor, showing the main effects or interactions predicted by “inhalation” and “group.” Furthermore, Wilcoxon signed-rank tests were applied for the within-group comparison of the task scores and HbO2 values of each group to clarify the difference between before (pre-) and after (post-) inhalation. All statistical analyses were performed using the IBM SPSS Statistics version 23 (IBM Co.). Values of p < 0.05 were considered significant.
The demographic and clinical information of the participants is presented in Table 2. Statistical analysis showed that there was no significant difference in demographics and clinical conditions between the saline and phytoncide groups.
GEE analysis showed that there was a significant main effect of “inhalation” (Wald χ2 = 9.649, p = 0.002) on the Stroop task performance. Although not statistically significant, an interactional trend was also found between “group” and “inhalation” (Wald χ2 = 2.272, p = 0.132). Table 3 and Figure 2 present the results of within-group comparisons of Stroop task performance of each group before (pre-) and after (post-) the inhalation stage. While no significant difference was found in the task performance of the saline group, there was an increase in the number of correct responses by 6.64 words (p = 0.009) after inhalation in the phytoncide group. The results indicated that the Stroop task performance of the phytoncide group improved after the 30-minute inhalation of phytoncide.
With regard to the task-related HbO2 levels in the four ROI regions (right/left DLPFC; right/left VLPFC), the hemodynamic response during the Stroop task is shown in Figure 2. GEE multiple comparison results showed a significant main effect of “inhalation” (Wald χ2 = 4.713, p = 0.030) and interaction of “group × inhalation” (Wald χ2 = 5.373, p = 0.020) in left VLPFC, suggesting that the change in HbO2 level in left VLPFC might have depended on group and inhalation (Table 4). No other significant effects were observed in the other ROI regions.
Table 5 presents additional results of within-group comparisons in each group, showing a reduction in HbO2 levels in the left VLPFC of the phytoncide group after inhalation (p = 0.042). The post-inhalation HbO2 level in the left VLPFC of the phytoncide group decreased by 0.023 μM compared to the pre-inhalation level (Figs. 3, 4). There were no other significant pre- vs. post-inhalation differences shown in other regions or in the saline group.
In the present pilot study, we aimed to determine the effect of inhaled phytoncides on the cognitive function of patients with MCI. By comparing Stroop task performance and task-related hemodynamic changes before and after 30-minute inhalation of saline (no-odor) or phytoncide, we found improved task performance and attenuated hemodynamic responses of the left VLPFC in the phytoncide group.
Regarding the behavioral results, patients with MCI who inhaled phytoncide showed enhanced Stroop task performance after inhalation by answering 6.64 more correct answers compared to those answered pre-inhalation, whereas no significant difference was found between the task scores in patients who inhaled saline (no-odor). By imposing a cognitive load via conflict-inducing (incongruent) stimuli, the Stroop task allows the evaluation of executive functions such as interference control, inhibition, and response flexibility [33]. With regard to MCI, owing to the deficiency in interference control and inhibition [34,35], patients with MCI exhibit poor performance in the Stroop task compared to that of the normal population [36]. Since our data showed an increase in Stroop task performance in the phytoncide group, it could be cautiously inferred that the acute improvement of cognitive function in patients with MCI could have been facilitated by phytoncide inhalation.
Such inference may also be validated by our obser-vations of fNIRS-measured hemodynamic changes in the PFC, exhibiting a notable reduction of HbO2 level in the left VLPFC area of phytoncide group. According to the Compensation-Related Utilization of Neural Circuits Hypothesis (CRUNCH) [37], populations with poor cognitive functions are likely to recruit more cerebral resources to compensate for their cognitive deficits [37-39]. Several fMRI studies have reported abnormal activation patterns of PFC regions during Stroop tasks in elderly populations or patients with MCI, implying that increased disruptions in cognitive functions are associated with the increased occurrence of PFC overactivation during the Stroop task [20,21]. Moreover, another brain imaging study using fNIRS found that older adults are more likely to show higher activation in the DLPFC and VLPFC during the Stroop task than younger adults, implying a compensatory hyperactivation to compensate for their cognitive decline [22]. In this context, our findings of task-related hypoactivation in the left VLPFC can be presumed to represent a temporal alleviation of cognitive dysfunction in patients with MCI induced by phytoncide inhalation. Therefore, while patients with MCI recruited additional cerebral resources to compensate for their cognitive deficits before inhaling phytoncide, the need for compensation diminished after phytoncide inhalation owing to the restoration of cognitive function, exhibiting subsided neural activation in the left VLPFC. Taken together with performance improvement in the Stroop task, a possible explanation for our neurophysiological findings is that phytoncide inhalation has beneficial effects on cognitive function in patients with MCI.
Concerning the trait of the Stroop task and related brain regions, the hypoactivation of the left VLPFC observed here may refer to mitigation of conflict adjustment or working memory dysfunction in patients with MCI. VLPFC has been considered to play a key role in mediating conflict-driven adjustments in cognitive control [40]. In the Stroop task, in particular, the VLPFC interacts with the DLPFC to select representations within the working memory on which task performance should be based [41]. In accordance with brain imaging studies showing compensatory hyperactivation of the PFC in the Stroop task performed by cognitively defective populations [20-22], a plausible inference could be made that increased VLPFC activation in the Stroop task of patients with MCI stems from their cognitive deficits in conflict adjustment and working memory. Owing to such deficits, patients with MCI need to exert more cognitive effort and recruit additional cerebral resources to their task-related prefrontal regions while performing the Stroop task [21]. In this regard, the decline in left VLPFC HbO2 levels shown in the present study may presumably represent a temporal alleviation of dysfunction in cognitive conflict adjustment or working memory of patients with MCI mediated by phytoncide inhalation.
Overall, the findings of the present study appear to support the cognitive elevation effect of phytoncide inhalation in patients with MCI. However, further investigation is required, because the underlying mechanism through which phytoncide enables the cognitive enhancement in MCI remains unexplored. One plausible explanation is that the underlying process is the neuromodulatory effect of phytoncides on acetylcholinesterase (AChE) activation. Bae et al. [19] showed the neuromodulatory effect of phytoncide by observing the reduction in AChE activity and protection of cognitive functions in Aβ-injected rats that inhaled phytoncide. From this perspective, the improvement in cognitive task performance of patients with MCI shown herein may have been mediated by the AChE-modulating effect of the inhaled phytoncide. Another potential cause of the phytoncide-derived cognitive enhancement is the mood-regulatory effect. Studies reviewing the effects of forest-derived volatile organic compounds have reported a positive mood-regulatory effect of phytoncides in various populations [14,42]. Given that a positive mood is likely to improve performance in many cognitive tasks [43-45], the improved Stroop task performance of patients with MCI in the present study might have been a consequence of the positive regulation of mood induced by phytoncide inhalation. In summary, we propose that further investigations regarding the effects of phytoncides on cognitive functions should consider the AChE-modulatory or mood-regulatory effect.
To the best of our knowledge, the present pilot study is the first to investigate the effect of inhaled phytoncides on the Stroop task performance of patients with MCI under experimental conditions. A beneficial effect of phyton-cide on cognitive function in MCI was shown based on improved Stroop task performance after phytoncide inhalation, and this improvement was supported by the decreased hemodynamic response observed in the left VLPFC of phytoncide-exposed individuals. Our findings broaden the therapeutic potential of phytoncide and forest therapies, suggesting an alternative clinical application for treating patients with MCI.
The present pilot study, however, had a few limitations. First, the inhalation protocol might not have been sufficient to verify the therapeutic effects of the phytoncide. A previous study conducted by Bae et al. [19] placed rats into a vaporization system cage distributing phytoncide vapor for 2 consecutive hours for 30 days, whereas the participants in the present study were only exposed to phytoncide in a single session of 30 minutes. Hence, a long-term multi-session protocol should be used to clarify the outcomes of this study. Future research could explore varying durations and frequencies of exposure, similar to the study by Bae et al. [19], which exposed rats to phytoncide vapor for two hours daily over 30 days. Such approaches would aid in more accurately assessing the potential therapeutic effects of phytoncides. Secondly, the results of the current study do not definitively exclude the potential influence of olfactory activation on the observed outcomes. The control group in our investigation was not subjected to any aromatic stimuli, having been fitted with masks treated with saline solution only. This absence of olfactory stimulation in the control group might have contributed to the observed differences between the groups. It is therefore imperative for future research to incorporate additional odoriferous substances in the control condition. Such an approach would facilitate a more precise elucidation of the unique effects of phytoncides in contrast to other aromatic compounds. Finally, owing to the small sample size and nonparametric statistical analysis, the generalizability of the findings may be limited. Future studies should overcome these limitations by employing larger sample sizes and parametric statistical methods, thereby enhancing the reliability of the results.
This research was funded by the ‘R&D Program for Forest Science Technology (No. 2021402B10-2123-0101),’ provided by Korea Forest Service (Korea Forestry Promotion Institute), the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Government of Korea (No. 2021R1I1A3058026), the National Research Foundation of Korea (NRF) grant funded by the Ministry of Science, ICT, and Future Planning (MSIP), Government of Korea (No. 2017R1A2B4008920), and Hallym University Research Fund.
The authors take thankful pleasure in acknowledging the unsparing assistance of all participants.
No potential conflict of interest relevant to this article was reported.
Conceptualization: Do Hoon Kim, Seungchan Park. Data acquisition: Hansol Kim, Jiheon Kim, Seungchan Park. Formal analysis: Seungchan Park, Jiheon Kim. Supervision: Do Hoon Kim. Writing – original draft: Seungchan Park. Writing – review & editing: Jiheon Kim, Hansol Kim.
![]() |
![]() |