Relaxing Effect Induced by Forest Sound in Patients with Gambling Disorder

Round 1

Reviewer 1 Report

The paper presents a small but interesting study on the effect induced by “forest sound” in people affected by gambling disorder (GD). The study aims to give a contribution to the wide research field related to the healing effects for humans deriving from their contact with nature, in this case related particularly with forest bathing.

The study is well structured and the methods and tools used are adequate. Anyway, the link between the study results and the benefits of forest in GD therapy is very weak. As highlighted by the authors, the study has several limitations, but the main one is the extreme simplification of the model.

First, the forest environment is replaced by the sound of a “murmuring creek in the forest”. The sound is only one of the several constituents of the landscape, and the creek (not always present in forests) is only one of the possible sounds of the forest. It is very hard to consider the sound of a “murmuring creek in the forest” as a substitute of the forest environment.

Second, the study found an improvement in depression-related symptoms using “forest sounds” and, because the literature highlighted that GD is related with depression, the authors concluded that the “forest sounds” can be effectively used as a therapeutic tool against GD. This sort of “Aristotelian syllogism” is not demonstrated and the “cause-effect” relationships seem simplistic.

Third. If we assume to use “forest sounds” as a therapy for patients with GD (as suggested by authors, lines 206-209), what about the several related variables? How many minutes, how many times a day? For how many weeks? The same sound or different sounds? Could there be a paradoxical reaction? Etc….

So, although the study is rigorous, the authors should generally limit the goals and mitigate the scope of results, in order to make clear to the readers that the study is a very preliminary one that can be only considered as a very initial step in the field.

Specific comments

Lines 55-56. Authors should mitigate the scope of their statement.

Lines 86-87. Authors should clearly declare why they decided to use only the sounds of forest and why they used “murmuring creek” sound.

Lines 99-104. Authors should better describe the procedure adopted for the experiment.

Author Response

Dear Referee 1:

Thank you for your insightful review. We have revised the manuscript based on your comments and have summarized the revisions and corrections as follows.

We deeply appreciate your thorough review and look forward to receiving your response regarding our revised report.

---------------------------------------------------------------------------------------------------

Referee 1’s Comments and Suggestions for Authors

1) The paper presents a small but interesting study on the effect induced by “forest sound” in people affected by gambling disorder (GD).

→ We greatly appreciate your comment about the interest in our study.

2) The study aims to give a contribution to the wide research field related to the healing effects for humans deriving from their contact with nature, in this case related particularly with forest bathing.

The study is well structured and the methods and tools used are adequate. Anyway, the link between the study results and the benefits of forest in GD therapy is very weak. As highlighted by the authors, the study has several limitations, but the main one is the extreme simplification of the model.

First, the forest environment is replaced by the sound of a “murmuring creek in the forest”. The sound is only one of the several constituents of the landscape, and the creek (not always present in forests) is only one of the possible sounds of the forest. It is very hard to consider the sound of a “murmuring creek in the forest” as a substitute of the forest environment.

→Your comment makes an excellent point. The "murmuring creek in the forest" is one of several components of the forest environment. To avoid any misunderstanding on this point, we added the follow sentence:

====== (page 6, line 198-202)

The rich and dynamic forest environment comprises a near-infinite number of sensory and auditory components. To simplify this diverse environment for targeted assessment within this study design, the number of forest bathing inputs was reduced to 1. We therefore focused on the "murmuring creek" sound as one of the environmental elements of the forest, and we observed the responses by gamblers to this specific forest sound alone.

======

3) Second, the study found an improvement in depression-related symptoms using “forest sounds” and, because the literature highlighted that GD is related with depression, the authors concluded that the “forest sounds” can be effectively used as a therapeutic tool against GD. This sort of “Aristotelian syllogism” is not demonstrated and the “cause-effect” relationships seem simplistic.

→ Thank you for your valuable insight on this point. In response, we emphasize that individuals with GD were able to relax by listening to the naturally derived stimuli of the "murmuring creek." The purpose of the study is not to simplify the causal relationship. To avoid any misunderstanding on these points, we added the following text:

====== (page 7, line 229-230)

We consider it a valuable finding that study participants with GD, as well as the healthy

subjects, showed a relaxation response to the sounds of the forest elements.

======

We also revised the following text:

====== (page 7, line 257-259)

Original: It is noteworthy that patients with GD can obtain a relaxing effect via forest

related stimuli, suggesting that this effect can be used as a method to suppress the

progression of symptoms.

Revision → It is noteworthy that individuals with GD can experience a relaxation effect through exposure to forest-related stimuli, which suggests that this result may be effective against the progression of symptoms.

======

4) Third. If we assume to use “forest sounds” as a therapy for patients with GD (as suggested by authors, lines 206-209), what about the several related variables? How many minutes, how many times a day? For how many weeks? The same sound or different sounds? Could there be a paradoxical reaction? Etc….

So, although the study is rigorous, the authors should generally limit the goals and mitigate the scope of results, in order to make clear to the readers that the study is a very preliminary one that can be only considered as a very initial step in the field.

→We appreciate your recognition of the rigor and questioning of the limitations. This study confirms that individuals with GD can relax with exposure to forest-related stimuli. Findings were that even 1 minute of sound stimulation alone produced relaxation changes, so the examination of effective degree, frequency, and type of stimulation is indeed a topic for future research.

To date, studies on biologic responses to natural stimuli have not yet revealed which stimulus is the most significant. Within this complex field, research has not yet yielded sufficient findings to fully understand the relationship between specific stimuli and biologic responses. Therefore, we acknowledge that this study is a very preliminary experiment, while also understanding that conclusions on this relationship have not yet been reported in this field. Taking these points into account, we have revised the following text.

====== (page 7, line 246-253)

Original: The final limitation was shortness in terms of stimulus; the results were compared before and after only 1 min of forest and city auditory stimulation. Future studies examining multiple days of sound stimulation and the duration of responses to natural stimuli are needed.

Revision → The final limitation of this study was short exposure time to the stimulus. The degree of relaxation was compared before exposure and after only 1 minute of exposure to forest and to city auditory stimulation; therefore, these results serve as only preliminary findings for exposure time. The relationship between specific stimuli and biologic responses remains complex to study and a difficult topic to fully understand, and more extensive investigation is required to examine the effective degree of and time for exposure to stimuli, the frequency of exposure to stimuli, and the type of stimulation.

======

5) Specific comments

Lines 55-56. Authors should mitigate the scope of their statement.

→Thank you for the comment. This part is based on the published literature, but we revised the sentence according your advice.

====== (page 2, line 55-56)

Original: Therefore, improving depressive symptomatology and preventing depression are important strategies for alleviating GD symptoms.

Revision →Therefore, both improving depressive symptomatology and preventing depression may serve as part of the key strategies for alleviating GD symptoms.

======

6) Lines 86-87. Authors should clearly declare why they decided to use only the sounds of forest and why they used “murmuring creek” sound.

→Thank you for your query. To design our study, we reviewed research that serves as a model case (Jo et al., Int J Environ Res Public Health 2019), which documented that a relaxation response was produced in healthy individuals by exposure to the stimulation of a “murmuring creek” sound. Because the forest is so complex with manifold components, assessment of exposure effects quickly complicates the study. Therefore, we focused on one sound only—the "murmuring creek" sound as an environmental element of the forest.

We added the follow sentence.

====== (page 6, line 198-202)

The rich and dynamic forest environment comprises a near-infinite number of sensory and auditory components. To simplify this diverse environment for targeted assessment within this study design, the number of forest bathing inputs was reduced to 1. We therefore focused on the "murmuring creek" sound as one of the environmental elements of the forest, and we observed the responses by gamblers to this specific forest sound alone.

======

7) Lines 99-104. Authors should better describe the procedure adopted for the experiment.

→Thank you for this important comment on process. We added the following sentence according to this advice.

====== (page 3, line 95-101)

Study participants were provided with a description of the purpose and outline of the study while in a waiting room. The study was able to be performed for 2 patients simultaneously because the auditory stimulus was provided while the participants were in a seated position with closed eyes. After exposure, they entered the experimental room, where sensors for the physiologic measurements were fitted while they received a description of the measurement procedure. The present experiment was conducted after the participants first practiced listening to the sound of ocean waves, which served as a dummy auditory stimulus.

======

We also revised the following text:

====== (page 3, line 104)

We performed a within-participants experiment.

======

Reviewer 2 Report

This study investigated effects of forest sound on relaxing outcomes for patients with gambling disorder (GD). As the author stated, GD is becoming a serious problem in the word wide, and thus, this topic is interesting.

However, there are several serious issues not to overcome.

1.What is the rationale to use the forest sounds? While the reviewer acknowledge that this research group has been investigating mainly forest therapy etc., but it should be considered whether forest therapy can be applied to all diseases or not. What is the evidence for benefits of forest sound use to GD. Should be expanded the rationale of this study based on previous literatures. Similarly, what is the hypothesis?

2.To the best of the reviewer’s knowledge, there are very serious problems to conduct this study. First, sample size is quite low as expected. Can the author calculate post-poc power analysis to determine enough sample size? In general, physiological and psychological responses in any patients would be varied compared with healthy individuals.

3.The authors appear not to understand NIRS principles and usage. Why did not the author measure baseline values. Did the author calculate NIRS changes just after start of hearing sound without steady state baseline? The NIRS device which was used in this study can evaluate relative changes from baseline. The authors conducted the study randomly, and hence, the endpoint of the first trial (either control or forest) would be the starting point of the second trial? If so, there is a serious problem, and make it difficult to interrupt the results. The authors seemed to ignore the effects of prior stimulus (again, at the end of control or forest). Aslo, MBL principles of this NIRS are strongly affected by skin blood flow. The authors should add this limitation. HbO2 is just only oxygenated hemoglobin levels.,which should be affected by cutaneous circulation, completely NOT an indirect index of cerebral blood flow. The reviewers evalauted just oxygenation levels at forehead, not in the brain.

4.The timeline is very short. As mentioned above, the reviewer would doubt NIRS data without baseline, or no consideration of just prior effect of stimulus.

5.In general, indirect sympathetic and para-sympathetic nerve activity using HRV metrics would require at least 5 min (see the rev. Eut J Heart J in 1996). 1 min data acquisition is very short and incorrect to evaluate SNA.

6.Did the authors perform the study for two patients, simultaneously (From the Fig 1A)? If so, this may produce additional effect by each other.

Author Response

Dear Referee 2:

Thank you for your insightful review. We have revised the manuscript based on your comments and have summarized the revisions and corrections as follows. We deeply appreciate your thorough review and look forward to receiving your response regarding our revised report.

---------------------------------------------------------------------------------------------------

Referee 2’s Comments and Suggestions for Authors

1) What is the rationale to use the forest sounds? While the reviewer acknowledge that this research group has been investigating mainly forest therapy etc., but it should be considered whether forest therapy can be applied to all diseases or not. What is the evidence for benefits of forest sound use to GD. Should be expanded the rationale of this study based on previous literatures. Similarly, what is the hypothesis?

→ Thank you for your inquiry on the rationale. To design our study, we reviewed research that serves as a model case (Jo et al., Int J Environ Res Public Health 2019), which documented that a relaxation response was produced in healthy individuals by exposure to the stimulation of a “murmuring creek” sound. Because the forest is so complex with manifold components, assessment of exposure effects quickly complicates the study. Therefore, we focused on one sound only—the "murmuring creek" sound as an environmental element of the forest.

Furthermore, we emphasize that individuals with GD were able to relax by listening to the naturally derived stimuli of the "murmuring creek." To data, no studies have provided data on the effects of naturally derived stimuli on GD; therefore, as a first step in this research, we conducted this study with a small group of participants and submitted our findings as a Brief Report.

 We added the following text:

====== (page 6, line 198-202)

The rich and dynamic forest environment comprises a near-infinite number of sensory and auditory components. To simplify this diverse environment for targeted assessment within this study design, the number of forest bathing inputs was reduced to 1. We therefore focused on the "murmuring creek" sound as one of the environmental elements of the forest, and we observed the responses by gamblers to this specific forest sound alone.

======

We also added the following text:

====== (page 7, line 229-230)

We consider it a valuable finding that study participants with GD, as well as the healthy subjects, showed a relaxation response to the sounds of the forest elements.

 ======

2) To the best of the reviewer’s knowledge, there are very serious problems to conduct this study. First, sample size is quite low as expected. Can the author calculate post-poc power analysis to determine enough sample size? In general, physiological and psychological responses in any patients would be varied compared with healthy individuals.

→Your comment is right on target. As a matter of fact, it was difficult to recruit GD patients who were willing to participate in this experiment. For this reason, we performed the study on a small number of participants as preliminary research and presented our results as a Brief Report. For future studies, we are planning to increase the number of participants with gambling addiction for a much larger sample size.

We have revised the sentence as follows:

====== (page 7, line 239-241)

Original: Furthermore, we could only recruit 12 participants; thus robust statistical evidence could not be provided.

Revision → Furthermore, it was difficult to recruit individuals with GD who were willing to participate in the experiment, and ultimately only 12 participants were able to be included; therefore, robust statistical evidence could not be provided.

======

3) The authors appear not to understand NIRS principles and usage. Why did not the author measure baseline values. Did the author calculate NIRS changes just after start of hearing sound without steady state baseline? The NIRS device which was used in this study can evaluate relative changes from baseline.

→Thank you for suggesting that we amplify this point. Because absolute values cannot be determined with this specific measurement system, the baseline was set to 0 before each measurement and then the relative change from the baseline was evaluated.

We have added the sentence as follows:

====== (page 3, line 122-123)

The baseline was set to 0 before each measurement and then the relative change from the baseline was evaluated.

======

The authors conducted the study randomly, and hence, the endpoint of the first trial (either control or forest) would be the starting point of the second trial? If so, there is a serious problem, and make it difficult to interrupt the results. The authors seemed to ignore the effects of prior stimulus (again, at the end of control or forest).

→Thank you for the important comment. After the first auditory stimulus, the participants spent 2 minutes undergoing subjective assessment and then had 1 minute to rest. Thus, until the second auditory stimulus was performed, there was approximately a 5-minute interval. The heart rate at "1 minute of rest" before auditory stimulation was 73.4 beats per minute during the first stimulation, and 72.9 beats per minute at the second stimulation. Because no significant difference was observed in heart rate (p = 0.315), we determined that there was no residual effect of the first stimulus by taking these intervals.

Aslo, MBL principles of this NIRS are strongly affected by skin blood flow. The authors should add this limitation. HbO2 is just only oxygenated hemoglobin levels.,which should be affected by cutaneous circulation, completely NOT an indirect index of cerebral blood flow. The reviewers evalauted just oxygenation levels at forehead, not in the brain.

→Thank you for this suggestion. In general, oxygen–hemoglobin (oxy–Hb) is known to be strongly affected by changes in skin blood flow resulting from exercise and other factors (Miyazawa et al. Eur J Appl Physiol 2013 ). However, the auditory stimulation was performed in a seated position without exercise in this experiment.

In addition, the phenomenon is observed that both oxy–Hb and deoxyhemoglobin (deoxy–Hb) concentrations change in to the same direction when the skin blood flow changes. However, this study we confirmed the results that the increase in deoxy–Hb concentrations in context of the the decrease in oxy–Hb suggest that the skin blood flow caused by movement or exercise is unlikely to have been affected in this case.

With regard to the effect of skin blood flow, a study by Takahashi et al. (NeuroImage 2011) used a verbal fluency task (VFT) to evaluate prefrontal cortex activities. They observed, "most studies using near-infrared spectroscopy (MIRS) have reported major activation in the front pole, but those using PET or fMRI have not," and they investigated the relationship between NIRS and skin blood flow during VFT. The results suggested that that a major part of the task-related changes in the oxy–Hb concentration in the forehead results from task-related changes in the skin blood flow.

In contrast, the study by Sato et al. (NeuroImage, 2013) assessed participants who performed the working memory task. The changes on near-infrared spectroscopy (NIRS), on functional magnetic resonance imaging (fMRI), and in skin blood flow were measured simultaneously, and yielded the following results:

1) A significant correlation was observed between "temporal changes in the prefrontal NIRS-Hb signals in the activation area" and "blood-oxygen-level-dependent ( BOLD ) signals in the gray matter," confirming the validity of the NIRS signal for this assessment.

2) Although a significant correlation was also observed between oxy–Hb concentration and skin blood flow, they demonstrated that the relation between "NIRS-Hb signals and BOLD signals" is stronger than that of "NIRS-Hb signals and skin blood flow." The conclusion was that NIRS can be used to measure hemodynamic signals originating from prefrontal cortex activation.

In context of these study findings and conclusions, we consider the following points to be valid for our study:

1) Although skin blood flow changes occur with "exercise," in this study it can be ruled out that skin blood flow affected the results of NIRS signals by taking into account the changes in the deoxy–Hb concentrations.

2) The validity of the NIRS signal was confirmed by the fMRI results, according to the results of simultaneous NIRS, fMRI, and skin blood flow measurement experiments by Sato et al. (NeuroImage, 2013).

3) However, it is true that the changes in skin blood flow affect the NIRS signal; therefore, for future studies, we believe that we should measure NIRS signals and skin blood flow simultaneously.

In this regard, we added the following text:

====== (page 6, line 206-209)

It is reported that NIRS-determined cerebral oxy–Hb may be affected by changes in forehead skin blood flow during dynamic exercise [27]. However, in this study, the auditory stimulation was provided in a seated position without exercise or physical movement.

======

We also added the following text:

====== (page 7, line 242-244)

The influence of skin blood flow on NIRS measurements remains an ongoing debate [27,35,36]; therefore, for future studies, we believe that we should measure NIRS signals and skin blood flow simultaneously to show the correct evaluation of prefrontal activities.

======

References:

27. Miyazawa, T.; Horiuchi, M.; Komine, H.; Sugawara, J.; Fadel, PJ.; Ogoh, S. Skin blood flow influences cerebral oxygenation measured by near-infrared spectroscopy during dynamic exercise. Eur. J. Appl. Physiol. 2013, 113(11), 2841–8.
28. Takahashi, T.; Takikawa, Y.; Kawagoe, R.; Shibuya, S.; Iwano, T.; Kitazawa, S. Influence of skin blood flow on near-infrared spectroscopy signals measured on the forehead during a verbal fluency task. Neuroimage. 2011, 57, 991–1002.
29. Sato, H.; Yahata, N.; Funane, T.; Takizawa, R.; Katura, T.; Atsumori, H.; Nishimura, Y.; Kinoshita, A.; Kiguchi, M.; Koizumi, H.; Fukuda, M.; Kasai, K. A NIRS-fMRI investigation of prefrontal cortex activity during a working memory task. Neuroimage. 2013, 83, 158–173.

4) The timeline is very short. As mentioned above, the reviewer would doubt NIRS data without baseline, or no consideration of just prior effect of stimulus.

→ We appreciate this concern. In response, as detailed previously, absolute values cannot be determined with this measurement system, so the baseline was set to 0 before each measurement, and then the relative change from the baseline was evaluated. Furthermore, we believe that the effects of the order are offset by the counterbalance with respect to the stimulus order.

5) In general, indirect sympathetic and para-sympathetic nerve activity using HRV metrics would require at least 5 min (see the rev. Eut J Heart J in 1996). 1 min data acquisition is very short and incorrect to evaluate SNA.

→We appreciate for your important comment. As indicated in the reference you provided, the 5-minute measurement time in heart rate variability (HRV) is necessary to measure very low frequency (VLF; 0.0033–0.04 Hz). However, in our study, we calculated the low frequency (LF; 0.04–0.15 Hz) versus the VLF. In addition, several methods for frequency analysis have been reported. However, in this study, we used the maximum entropy method (MEM), which is effective for calculating spectra from short data lengths.

This study was designed to investigate acute physiologic responses to sound stimulation, so we considered that exposure to 5 minutes of recording is too long for this purpose. A measuring duration of 60 seconds is a result of compromise.

We added the following text:

====== (page 4, line 131-132)

In this study, HRV recording was analyzed every 60 seconds to investigate acute physiologic responses to sound stimulation.

======

6) Did the authors perform the study for two patients, simultaneously (From the Fig 1A)? If so, this may produce additional effect by each other.

→We agree with your point. In fact, it would be more appropriate to conduct this experiment for each participant alone. However, in this case, we performed the study for 2 patients simultaneously, taking into account the fact that the auditory stimulus was administered while the participant’s eyes were closed.

We added the following text.

====== (page 3, line 96-98)

The study was able to be performed for 2 patients simultaneously because the auditory stimulus was provided while the participants were in a seated position with closed eyes.

======

Reviewer 3 Report

This paper examines a small group of people with gambling disorder and assesses their response to relaxing sounds. In particular they want to relieve the symptoms of depression.

This pilot study has points of interest, but lacks a control group. This limitation must be discussed, in addition to those that the authors have listed.

The sample is too modest to draw definitive conclusions.

Author Response

Dear Referee 3:

Thank you for your insightful review. We have revised the manuscript based on your comments and have summarized the revisions and corrections as follows.

We deeply appreciate your thorough review and look forward to receiving your response regarding our revised report.

---------------------------------------------------------------------------------------------------

Referee 3’s Comments and Suggestions for Authors

1) This paper examines a small group of people with gambling disorder and assesses their response to relaxing sounds. In particular they want to relieve the symptoms of depression.

→Your comment is exactly right for the intent of our study.

2) This pilot study has points of interest, but lacks a control group. This limitation must be discussed, in addition to those that the authors have listed.

→Thank you for your comment.

This study is a "within-participant experiment," and the control auditory stimulation is "city sounds."

We added the sentence as follows.

====== (page 3, line 104)

We performed a within-participant experiment.

======

3) The sample is too modest to draw definitive conclusions.

→Your comment is correct regarding the scope of the study. As a matter of fact, it was difficult to recruit GD patients who were willing to participate in this experiment. For this reason, we performed the study on a small number of participants as preliminary research and presented our results as a Brief Report. For future studies, we are planning to increase the number of participants with gambling addiction for a much larger sample size.

We have revised the sentence as follows:

====== (page 7, line 239-241)

Original: Furthermore, we could only recruit 12 participants; thus robust statistical evidence could not be provided.

Revision → Furthermore, it was difficult to recruit individuals with GD who were willing to participate in the experiment, and ultimately only 12 participants were able to be included; therefore, robust statistical evidence could not be provided.

======

Round 2

Reviewer 2 Report

The authors does not understnad my comments.

Baseline measurment of NIRS is needed. In other words, the autrhos should hav set as o at the first second of baseline (e.g., 5 min). That is, the authos should show 6 min 'if you took 5 min between experimens'.

Because, there are possible changes during 5min period of baseline; but it is impossible to show.

One min stimuls period is not enough to obtain HRV.

The authors does not understand NIRS principles, especially, this device. It is a serious problem to interrupt the results.

Collectively, the reviewer cannot understand and would doubt if NIRS and HRV data are reliable.