Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.
Advertisement
Scientific Reports volume 15, Article number: 8493 (2025)
Metrics details
Gastrointestinal (GI) symptoms are common and can affect children’s social lives. This study investigated the effects of exposure to a rich natural environment on GI symptoms, salivary cortisol levels, salivary amylase levels, and the gut microbiota in young children. Children aged 5–6 years from four kindergartens in Japan were randomly assigned to two groups: a nature childcare group and a regular childcare group. The children were exposed to their respective conditions once weekly for one month. Before and after the intervention, GI symptoms were detected using the Children’s Somatization Inventory to calculate a ‘GI score’ and categorize participants into GI and control groups (primary outcome measure). Fecal examinations were performed for gut microbiota using 16 S-rRNA analysis, salivary cortisol and amylase levels were quantified, and the Child Behavior Checklist was administered. The two groups had similar GI symptoms, salivary cortisol and amylase levels, and behavioral characteristics. Following the intervention, significant differences in the GI score, abdominal pain, constipation, Shannon index value, and salivary cortisol and amylase levels (p < 0.05) were observed between the two childcare groups. Spending free and abundant time in nature during early childhood could help maintain digestive system homeostasis, increase gut microbiota diversity, and reduce cortisol levels.
Gastrointestinal symptoms are common and can influence the social life of young children1. Irritable bowel syndrome (IBS) can cause recurrent abdominal pain, and the symptoms associated with IBS can lead to early social maladjustment, which can manifest in several ways, including refusal to go to school. Gastrointestinal symptoms affect up to 40% of the general population worldwide2. Thus, gastrointestinal symptoms are prevalent, and focusing on these symptoms is essential. Although precise data on the prevalence of gastrointestinal symptoms in young children are unavailable, a previous study from our laboratory reported a 58% prevalence of gastrointestinal symptoms in seven-year-old children3. Gastrointestinal symptoms thus appear to be at least as common in children as in adults. These symptoms are associated with stress responses, cortisol, amylase, and gut bacteria4.
Regarding stress responses, animal models and clinical populations provide evidence that early life stressors can lead to an aberrant response of the hypothalamic-pituitary-adrenocortical (HPA) axis and the arousal/sympathetic nervous system and increase the incidence of stress-related disorders after puberty5,6. Salivary amylase and cortisol are two widely used proxy measures of stress7,8, and increased baseline salivary cortisol levels have been observed in patients with IBS relative to controls7. A dysregulated innate immune system has been suggested to result from childhood trauma and lead to an increased vulnerability to psychopathology and somatic disorders across the lifespan9. A body of evidence also suggests that sleep disturbances associated with childhood adversity have a causal role in elevating an individual’s vulnerability to stress in adulthood10. Responsive and predictable caregiver behaviors can improve child outcomes11. If early environmental stressors lead to stress vulnerabilities later in life, the question that naturally follows is whether enrichment of a child’s psychosocial environment can reduce the risk of such stress vulnerabilities in adulthood.
The gut microbiome, the community of microbes and associated genes residing in the gastrointestinal tract, has been postulated to contribute to neurobiological and behavioral development in children12. Evidence primarily from animal studies supports communication between the microbiome and the central nervous system, with effects on social, explorative, and affective behavior. This is mediated through several mechanisms, including the coordination between the neuroendocrine and immune systems, stimulation of the vagus nerve, and metabolism of neurotransmitters13. A human intervention trial manipulated urban environmental biodiversity in children to evaluate its effects on the commensal microbiome and immunoregulation; the study found that biodiversity intervention enhanced immunoregulatory pathways, manifested in increased plasma TGF-β1 levels, proportion of regulatory T cells, and plasma IL-10:IL-17 A ratio14. An environment set up to allow children to freely interact with nature comfortably could positively impact the influence of the microbiome on childhood outcomes. However, medical data substantiating this hypothesis are scarce13,14.
Given the above, reducing gastrointestinal symptoms is vital for children’s physical and mental health. However, no studies have linked these factors in young children. We hypothesized that childcare provided in a biodiverse natural environment could improve the health of children in early childhood with potentially long-lasting effects. This study aimed to investigate the effects of various levels of exposure to a rich natural environment (once weekly for one month) on gastrointestinal symptoms, salivary cortisol, salivary amylase, and gut microbiota. We verified the hypothesis that exposure to a rich natural environment reduces gastrointestinal symptoms in young children. Furthermore, we tested a secondary hypothesis that exposure to a rich natural environment reduces salivary cortisol and salivary amylase levels and a third hypothesis that this exposure increases the diversity of the gut microbiota.
This randomized trial was conducted in four kindergartens in Tohoku/Hokkaido, Japan. This study was approved by the Ethical Review Board of Tohoku University School of Medicine (approval number 2022-1-489) and registered with the University Hospital Medical Information Network (study number UMIN000047462, date of first trial registration: 11/04/2022). All experiments were performed in accordance with the Ethics Committee Tohoku University Graduate School of Medicine guidelines and regulations. Informed consent was obtained from the legal guardians of all study participants for inclusion in the study as well as for publication of their information/images in an online open access publication.
The inclusion criterion for this study was 5–6-year-old healthy children attending kindergarten. The exclusion criteria were a history of antibiotic treatment within the previous month, history of internal medicine or neurological diseases, and history of mental illness. After obtaining written informed consent from the children’s parents and the participating children, the children were randomly assigned to one of two groups: a nature childcare group or a regular childcare group. Figure 1 shows a flowchart of the study. Of the 157 initially enrolled children, 27 did not provide consent; this was due to the children and their families either being too busy to take part or changing their mind regarding study participation. The remaining 130 children were assessed for eligibility and randomized. At the time the intervention trial was conducted, all children had been in childcare for between 29 and 30 months. All children received childcare from Monday to Friday for six hours per day, which was standardized because they all attended public kindergartens.
Study flow chart.
Cluster randomization was performed on a class-by-class basis within the kindergarten, and the children were with their class teachers and classmates to avoid unnecessary increases in stress. Randomization was performed by investigators not directly participating in the intervention trial using the RAND function in Microsoft Excel (Office 365, version 21.0; Microsoft Corporation, Redmond, WA, USA); those with values ≥ 0.5 were assigned to the intervention group, and the other classes inevitably became controls. These calculations were performed under strict control using a personal computer that others could not operate.
This study employed waitlist control; in the regular childcare group, the participants were initially (for the first month) only exposed to the kindergarten environment, while in the nature childcare group, an additional period was allotted for a nature walk (Figs. 2 and 3). Children spent 4 h (between 09:00 and 13:00) in a natural environment. The natural environments provided as part of this study were grassland and forest landscapes located within a 5–10-minute walk from the children’s kindergarten so as not to disrupt the children’s kindergarten routine. Table 1 shows the timetable of activities that the children were exposed to in the regular and nature childcare groups. Further information regarding the intervention is provided in the supplementary material. The children were exposed to the respective conditions once a week for a month (a total of four times; Fig. 4). After one month, the children in the regular childcare group had the same experience as those in the nature childcare group. This was to keep the children’s experiences fair, as requested by their teachers.
Description of the study groups and the interventions.
Pictures illustrating the children in the nature intervention group (upper) and the regular childcare group (lower).
Study protocol.
The children engaged in activities along with trekking. These activities were similar to those provided during the daily childcare program. This was to prevent the activity from increasing curiosity or stress. Children from the four kindergartens had an equal opportunity to engage in all activities. The activities comprised the following:
Crafts: Children made crafts using fallen leaves and branches.
Science: Children used microscopes to examine objects they found in nature.
The children read picture books and participated in a treasure hunt in nature.
Painting: The children drew freely and worked on prints.
Before the intervention, gastrointestinal symptoms were detected using the Children’s Somatization Inventory (CSI)15. The version of this questionnaire used in this study comprises 35 items that evaluate symptoms among children in the preceding two weeks; the 35-item version was utilized as it has been used in previous research in the field3. The following gastrointestinal symptoms of the questionnaire were grouped into a ‘GI score’: nausea and abdominal discomfort, abdominal pain, nausea after eating, bloating, constipation, diarrhea, and vomiting. If the GI score was ≥ 1, participants were categorized into the ‘GI group,’ and if the GI score was 0, they were classified as ‘Controls.’ The CSI has good reliability16. The CSI was administered and scored by the children’s parents, who were not formally notified of their child’s allocated group. Child behavior was analyzed using the Child Behavior Checklist for ages 4–1817. Fecal examination was performed for gut microbiota using 16 S-rRNA analysis, with the V1-V2 regions as targets (further information about the methodological details is provided below and in the supplementary material). Fecal analyses were performed by a professional vendor, resulting in a set of QIIME 2 bacterial flora analyses. Gut microbiota alpha diversity was analyzed using the Shannon index, calculated using QIIME 2 version 2020.8 (https://qiime2.org/)18. The Shannon index measures species diversity and evaluates the number of species and their relative abundance. Beta-unweighted and -weighted UniFrac analyses were also conducted using QIIME 2 to determine diversity metrics. The statistical tests performed are described in the ‘Statistical analysis’ section below. The salivary cortisol and salivary amylase levels were also measured using the methodology reported in previous studies19,20. Salivary samples were collected at 9 a.m. and 11 a.m. in both study groups before meals.
Data collected on the following factors after the intervention were the primary and secondary endpoints:
Primary endpoint: CSI GI scores.
Secondary endpoint: gut microbiota (fecal examination).
Secondary endpoint: salivary cortisol and amylase levels.
Secondary endpoint: Child Behavior Checklist (CBCL 4–18 for PorT).
Fecal samples were collected using Mykinso fecal collection kits containing a guanidine thiocyanate solution (Cykinso, Inc., Tokyo, Japan) and stored at 4℃ for one week. DNA was extracted from the fecal samples using an automated DNA extraction machine (GENE PREP STAR PI-1200 A, Kurabo Industries Ltd., Osaka, Japan) according to the manufacturer’s protocol. The V1–V2 region of the 16 S rRNA gene was amplified using forward (16S_27Fmod: TCG TCG GCA GCG TCA GAT GTG TAT AAG AGA CAG AGR GTT TGA TYM TGG CTC AG) and reverse (16S_338R: GTC TCG TGG GCT CGG AGA TGT GTA TAA GAG ACA GTG CTG CCT CCC GTA GGA GT) primers using the KAPA HiFi HotStart ReadyMix (Roche). To sequence the 16 S amplicons on the Illumina MiSeq platform, we attached the dual-index adapters using the Nextera XT Index kit. Each library was diluted to 5 ng/µL, and equal volumes of the libraries were mixed to a concentration of 4 nM. The DNA concentration of the mixed libraries was quantified through qPCR using the KAPA SYBR FAST qPCR Master Mix (KK4601, KAPA Biosystems, Wilmington, MA, USA) with primers 1 (AAT GAT ACG GCG ACC ACC) and 2 (CAA GCA GAA GAC GGC ATA CGA). The library preparations were conducted according to the Illumina 16 S library preparation protocol (Illumina, San Diego, CA, USA). Libraries were sequenced using the MiSeq Reagent Kit v2 (500 cycles) and 250 bp paired-ends.
The paired-end reads of the partial 16 S rRNA gene sequences were analyzed using QIIME 2 (version 2020.8)5. The steps for data processing and assignment based on the QIIME 2 pipeline were as follows: (1) DADA2 for joining paired-end reads, filtering, and denoising; and (2) assigning taxonomic information to each amplicon sequence variant using naive Bayes classifier in QIIME 2 classifier with the 16 S gene of V1-V2 region data of SILVA (version 138)6 to determine the identity and composition of the bacterial genera.
The sample size was determined based on previous studies in the field and calculated using Power Analysis in SPSS version 29 (IBM Corp., Armonk, NY, USA) with the following parameters: significance level (α) = 0.05, power (β) = 0.8, and effect size (d) = 0.5 (moderate correlation). Sample sizes of at least 64 participants and 34 pairs were required based on this calculation for the two-group comparison and pre- and post-intervention comparison t-tests, respectively. The statistical analyses were performed using the same software. For normally distributed items, the t-test was used to compare the mean values, whereas the Mann–Whitney U test was used for non-normally distributed data. Chi-squared testing was performed to cross-tabulate sex, childcare, and maternal educational background. Alpha coefficients were calculated to assess the internal consistency of all 30 questionnaire items of the CSI and for the 7 GI score items alone. The Child Behavior Checklist scores were used to generate T-scores for participants for analysis. All participants were analyzed according to their originally assigned groups, and there was no crossover. Further analyses were focused on participants with confirmed gastrointestinal symptoms. Significant differences were verified with two-way ANOVA and post-hoc testing using the Bonferroni test. Mixed effects models were also created to facilitate differentiation between- and within-individual effects. P values < 0.05 were considered statistically significant.
For the microbiological statistical analyses, non-rarefied data were used, and the data were converted to relative abundance. Statistical analyses were performed using QIIME2 version 2020.8, with the package QIIME2 feature-table.relative-frequency used for the relative abundance conversion. For the beta diversity analyses, differences in Bray–Curtis distances between groups were assessed using a non-parametric permutation-based multivariate analysis of variation (PERMANOVA) test with 999 permutations21 using the q2-diversity plugin in QIIME2, and principal coordinate analysis plots were generated. DADA2 was used for quality filtering, noise removal, and chimera removal. No samples had insufficient reads (number of 16 S rRNAs), and the final sample size was the total sample size of all subjects.
Recruitment commenced on July 5, 2022, and finished on September 28, 2022. The follow-up period ended on November 30, 2022. Of the 130 participants randomized in this study, 64 were allocated to the nature childcare group (31 males and 33 females, mean age 5.59 ± 0.50 years) and 66 to the regular childcare group (32 males and 34 females, mean age 5.47 ± 0.50 years; Table 2). The groups were well matched at baseline, with no difference between the childcare groups in terms of sex and maternal educational background. Some randomized children did not provide completed questionnaires (parental refusal to submit the questionnaire), stool samples (constipation or child refusal), or saliva samples (absence during the saliva collection period; Fig. 1). Alpha coefficients for the total CSI items and just the 7 GI score items alone were both 0.776, confirming internal consistency. Before the intervention, the GI score derived from the CSI was non-significantly higher in the nature childcare group (2.48 ± 3.44) than in the regular childcare group (1.43 ± 2.32; Mann–Whitney U test result = 2131.0, Z = 1.104, effect size = 0.10, p = 0.269; Fig. 5). However, following the intervention, gastrointestinal symptoms in the nature childcare group were significantly lower than those in the regular childcare group (Mann–Whitney U test result = 1201.5, Z = − 4.136, effect size = 0.37, p < 0.001), particularly abdominal pain (Mann–Whitney U test result = 1565.5, Z = − 2.720, effect size = 0.24, p = 0.007) and constipation (Mann–Whitney U test result = 1692.0, Z = 0.146, effect size = 0.13, p < 0.001). Most symptoms in the nature childcare group decreased after the intervention, whereas no large change or clear pattern was observed in the symptoms among the regular childcare group (Supplementary Table S1). Supplementary Fig. S1 shows the changes in pre- and post-intervention scores at the individual participant level.
Comparison of GI scores before and after the intervention in the two study groups.
Table 3 presents the Shannon index data. No significant difference was observed in the overall before-and-after comparison or for each of the two groups. Furthermore, there was no difference in the Shannon index value between the nature and regular childcare groups before the intervention; however, the Shannon index value was higher in the nature childcare group than in the regular childcare group following the intervention (t-test value = 2.552, p = 0.013, effect size = 0.55, 95% confidence interval = − 0.347, − 0.043; Fig. 6). The beta-unweighted and -weighted UniFrac analyses (Supplementary Fig. S2), as well as the PERMANOVA testing (beta-unweighted group: pseudo-F = 0.983, p = 0.514; beta-weighted group: pseudo-F = 0.864, p = 0.688; Supplementary Fig. S3), confirmed no significant differences between the nature and regular childcare groups.
Comparison of Shannon index scores before and after the intervention in the two study groups.
The participants’ salivary cortisol and amylase levels are presented in Table 4; Fig. 7. Cortisol and amylase levels before the intervention were not significantly different between the two groups. However, after the intervention, salivary cortisol and amylase levels were significantly lower in the nature childcare group than in the regular childcare group.
Levels of salivary cortisol (top) and amylase (bottom) by childcare group before and after the intervention.
To facilitate the differentiation of between- and within-individual effects in our longitudinal study, we performed mixed model analyses incorporating child-level variability and kindergarten as random effects, and childcare group (nature versus regular), time, and a childcare group × time interaction term as fixed effects. This analysis was performed to evaluate the effects of these variables on GI scores, salivary cortisol, salivary amylase, and Shannon index values. Pairwise comparisons were conducted based on data obtained before and after the intervention (including the 9 a.m. saliva sample as the ‘before’ data and the 11 a.m. saliva sample as the ‘after’ data). These analyses, the results of which are shown in Supplementary Tables S2–S5, confirmed the lack of an effect of kindergarten on GI scores, salivary cortisol, salivary amylase, and Shannon index values. However, effects were seen for time (F = 13.326, df = 122, p < 0.001) and a childcare group × time interaction (F = 12.890, df = 122, p < 0.001) on GI scores of the childcare group (F = 4.860, df = 122, p = 0.029). Salivary cortisol values were affected by a childcare group × time interaction (F = 20.744, df = 125, p < 0.001). Salivary amylase values were affected by time (F = 8.248, df = 125, p = 0.005) and a childcare group × time interaction (F = 13.991, df = 125, p < 0.001). There were no significant fixed effects on the Shannon index. To ensure that other potential confounders did not influence the results, the mixed effects models were repeated with the addition of four covariates: maternal age, maternal educational background, and baseline time spent outdoors (the latter as two separate covariates: from waking time to 12 p.m., and between 12 p.m. and 6 p.m.). The results of this analysis identified the same significant variables as those identified in the models without covariates, with no new additional statistically significant results identified (data not shown).
Supplementary Table S6 presents the results of the Child Behavior Checklist. There were no significant differences in any element of the Child Behavior Checklist between the nature and regular childcare groups before and after the intervention. However, when comparing the results before and after the intervention within each childcare group (intragroup comparisons), significantly lower scores for the total score, introversion scale, extroversion scale, social withdrawal, and social issues were observed in the nature and regular childcare groups (Supplementary Table S6). A significantly lower attention issue score was noted following the intervention in the nature childcare group but not in the regular childcare group, whereas significantly lower misbehavior and aggressive behavior scores were noted in the regular childcare group but not in the nature childcare group. Supplementary Fig. S4 shows the changes in pre- and post-intervention Child Behavior Checklist scores at the individual participant level.
Mixed effects models were also developed for the Child Behavior Checklist scores, incorporating child-level variability and kindergarten as random effects, and childcare group (nature versus regular), time, and a childcare group × time interaction term as fixed effects. The results are summarized in Supplementary Tables S7–S17 and confirmed the lack of an effect of kindergarten on any of the measures of the Child Behavior Checklist. However, time showed an effect on Total Problems (F = 35.069, df = 80.427, p < 0.001), Internalizing Problems (F = 19.416, df = 112.059, p < 0.001), Externalizing Problems (F = 25.013, df = 109.740, p < 0.001), Depressed (F = 15.574, df = 122.000, p < 0.001), Anxious/Depressed (F = 7.419, df = 115.910, p = 0.007), Social Problems (F = 15.331, df = 118.927, p < 0.001), Attention Problems (F = 6.803, df = 116.050, p = 0.010), Rule-breaking Behavior (F = 9.214, df = 118.838, p = 0.003), and Aggressive Behavior (F = 11.751, df = 111.505, p = 0.001). There was also an effect of childcare group × time interaction on Somatic Complaints (F = 5.653, df = 120.653, p = 0.019).
The current study aimed to investigate the effect of exposure to a rich natural environment on gastrointestinal symptoms, salivary cortisol, salivary amylase, and gut microbiota in young children, hypothesizing that exposure to a nature childcare intervention would reduce gastrointestinal symptoms, salivary cortisol, and salivary amylase, as well as increase the diversity of gut microbiota relative to those associated with exposure to regular childcare. We confirmed that most of these hypotheses were true by studying a group of approximately five-year-old children. However, an increase in gut microbiota diversity was only observed using a measure of alpha diversity (Shannon index) in the nature group following the intervention, and there were no significant changes in beta diversity. Another interesting finding was that no significant difference was observed when data from the entire cohort (including children with and without gastrointestinal symptoms) and data from only those with gastrointestinal symptoms were compared. This study’s findings are noteworthy and emphasize the need to optimize children’s long-term outcomes and potential. Given the improvements in gastroenterological symptoms and the biochemical findings, these data provide promising evidence of an approach to improve childhood psychosocial outcomes.
Children in the nature childcare group had fewer gastrointestinal symptoms than those in the regular childcare group (the primary endpoint). The proportion of children with gastrointestinal symptoms in our study was 55.56% in the nature group and 58.33% in the regular group, confirming the high prevalence of gastrointestinal symptoms in children, consistent with the 58% prevalence rate reported in a previous study of seven-year-old children3. Disorders of gut-brain interaction (DGBI, formerly known as functional gastrointestinal disorder) in children and adolescents profoundly impact children’s health22. Gastrointestinal symptoms (abdominal pain, constipation, diarrhea, or psychosocial distress) in a child reflect which components of the brain-gut axis are affected and to what extent. Some children with IBS have rectal hyperalgesia23. Visceral hypersensitivity may be related to psychological distress (anxiety, depression, impulsiveness, anger)24, with elevated salivary cortisol or amylase levels. Increased mucosal proinflammatory cytokines have been demonstrated in this context and may be induced due to acute infectious gastroenteritis (postinfectious IBS)22. Alterations in the gut microbiome have been demonstrated in pediatric patients with IBS25,26. Although this study was not limited to children with DGBI, the nature of the childcare intervention may have had a positive effect on gut health in children, although the beta diversity indices in the current study did not provide objective evidence of a change in microbiota driving the observed changes. Nevertheless, the other findings of this study warrant further investigation into the factors driving the change in GI scores observed.
This study also provides evidence indicating a reduction in the levels of the stress hormone cortisol and amylase associated with care taking place in nature. Patients with IBS are characterized by increased levels of adrenocorticotropic hormone and cortisol secretion following the administration of corticotrophin-releasing hormone27. The baseline salivary cortisol level is increased in patients with IBS compared to that in controls, and the cortisol awakening response is blunted7. Changing gut microbiota in germ-free mice greatly influences adrenocorticotropic hormone and cortisol secretion to reduce stress28. Therefore, although relaxed mood and altered gut microbiota in the natural environment could be hypothesized to have accounted for the reduced salivary cortisol in the nature childcare group, the lack of strong evidence of altered gut microbiota in the current study challenges this. Changes in salivary amylase are also a marker of stress response8. Concordant changes in cortisol and amylase levels support the reliability of the stress reduction identified in the nature childcare group in the present study.
This study is unique but builds on previous studies that indicated a link between the gut microbiome and child development12,13,14. A previous interventional study on urban environmental biodiversity in Finnish children focused on the microbiome and immunoregulatory pathways14. The intervention diversified both the environmental and skin gammaproteobacterial communities, which, in turn, were associated with increases in plasma TGF-β1 levels and the proportion of regulatory T cells. They also found that children with naturally diverse environments had a more diverse gut microbiota14. However, the study was not a randomized controlled trial like the current study. Furthermore, the current study expanded the parameters to incorporate gastrointestinal symptoms, gut microbiota, and biomarkers of stress response (salivary cortisol and amylase).
The composition of the human gut microbiota is linked to health and disease29. The diversity of gut microbiota is decreased in pathological conditions, especially in IBS30. Therefore, our findings of increased diversity in gut microbiota in the nature childcare group after the intervention and the decreased GI scores are meaningful. A previous study demonstrated that nature childcare centers showed more diversity in skin Proteobacteria alpha-diversity than standard childcare centers14. The presence of fecal Faecalibacterium prausnitzii was negatively associated with proinflammatory cytokine IL-17 A levels in the nature-oriented group only. It was suggested that the environmental microbiota may have adhered to the clothes, hair, face, lips, skin, hands, and fingers of the children in the nature care group. They can be ingested in the gastrointestinal tract via the oral route, resulting in an increased diversity of the gut microbiota.
Scores on many components of the Child Behavior Checklist, including the total score, decreased after the intervention in the nature and regular childcare groups, with no significant difference between the two groups. However, the results reflect the characteristics of the two groups. In normal childcare, scrambles and fights over playground equipment often occur, and the teacher usually corrects the child’s aggression or misbehavior in such instances. One month of these instructions and corrections may have influenced the Child Behavior Checklist scores. However, children in the nature childcare group spent more time in nature than in a regular nursery. Nature is a vast environment; therefore, fights for playground equipment are unlikely. Children require attention in a natural environment because they can encounter small animals and flowers and need to be aware of dangerous terrains, insects, and flowers. These differences in the experiences of the children in the nature and regular childcare groups may explain the study’s results.
This study had several limitations. First, although parents were not formally notified about which group their child was allocated to, they may have discovered this through discussions with their children. Thus, the outcome assessment of the CSI scoring may not have been truly blinded. However, this limitation would not have influenced the analysis of gut microbiota and salivary data, as this was not conducted by the parents. A second limitation was the inability to control the home environment, which is unavoidable in intervention trials such as the current study. Third, the current study did not provide long-term follow-up data to confirm the trends in GI symptoms, gut microbiota, and salivary amylase and cortisol levels. Fourth, data on laxative use were not collected as it is uncommon to use these medications in very young children such as those in the current study. Fifth, we did not collect data from participants on delivery method relating to their birth, the use of other medications, or pet ownership; this should be addressed in future studies. Finally, the difference in the time of saliva sample collection before (9 a.m.) and after (11 a.m.) the intervention could have influenced the results of the study; however, this was unavoidable due to the nature of the kindergarten curriculum. The strengths of this study include its randomized trial design and the collection of data from a range of outcome measures. In conclusion, our study supports the idea that spending time freely and abundantly in nature during early childhood could help maintain homeostasis of the digestive system and reduce the levels of the stress hormone cortisol. This may reduce the incidence of IBS. Further studies with larger populations and longer follow-up periods are required to confirm these findings and to explore the potential mechanisms driving these changes, particularly given the beta diversity results observed in this study.
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.
Ansems, S. M. et al. Qualitative study evaluating the expectations and experiences of Dutch parents of children with chronic gastrointestinal symptoms visiting their general practitioner. BMJ Open. 13, e069429 (2023).
Article PubMed PubMed Central Google Scholar
Sperber, A. D. et al. Worldwide prevalence and burden of functional gastrointestinal disorders, results of Rome Foundation Global Study. Gastroenterology 160, 99–114e3 (2021).
Article PubMed MATH Google Scholar
Seino, S. et al. Enhanced auditory brainstem response and parental bonding style in children with gastrointestinal symptoms. PLoS One. 7, e32913 (2012).
Article ADS PubMed PubMed Central MATH Google Scholar
Cherpak, C. E. Mindful eating: A review of how the stress-digestion-mindfulness triad may modulate and improve gastrointestinal and digestive function. Integr. Med. (Encinitas). 18, 48–53 (2019).
PubMed Google Scholar
Pervanidou, P. & Chrousos, G. P. Early-life stress: From neuroendocrine mechanisms to stress-related disorders. Horm. Res. Paediatr. 89, 372–379 (2018).
Article PubMed Google Scholar
Berhe, O., Gerhardt, S. & Schmahl, C. Clinical outcomes of severe forms of early social stress. Curr. Top. Behav. Neurosci. 54, 417–438 (2022).
Article PubMed MATH Google Scholar
Groeger, D. et al. Interactions between symptoms and psychological status in irritable bowel syndrome: An exploratory study of the impact of a probiotic combination. Neurogastroenterol Motil. 35, e14477 (2023).
Article PubMed MATH Google Scholar
Strahler, J., Mueller, A., Rosenloecher, F., Kirschbaum, C. & Rohleder, N. Salivary alpha-amylase stress reactivity across different age groups. Psychophysiology 47, 587–595 (2010).
Article PubMed Google Scholar
de Koning, R. M., Kuzminskaite, E., Vinkers, C. H., Giltay, E. J. & Penninx, B. W. J. H. Childhood trauma and LPS-stimulated inflammation in adulthood: Results from the Netherlands study of depression and anxiety. Brain Behav. Immun. 106, 21–29 (2022).
Article PubMed Google Scholar
Simon, L. & Admon, R. From childhood adversity to latent stress vulnerability in adulthood: The mediating roles of sleep disturbances and HPA axis dysfunction. Neuropsychopharmacology 48, 1425–1435 (2023).
Article PubMed PubMed Central MATH Google Scholar
Shonkoff, J. P. Rethinking the definition of evidence-based interventions to promote early childhood development. Pediatrics 140, e20173136 (2017).
Article PubMed Google Scholar
Flannery, J. E. et al. Gut feelings begin in childhood: The gut metagenome correlates with early environment, caregiving, and behavior. mBio 11, e02780–e02719 (2020).
Article PubMed PubMed Central MATH Google Scholar
Borre, Y. E. et al. Microbiota and neurodevelopmental windows: Implications for brain disorders. Trends Mol. Med. 20, 509–518 (2014).
Article PubMed MATH Google Scholar
Roslund, M. I. et al. Biodiversity intervention enhances immune regulation and health-associated commensal microbiota among daycare children. Sci. Adv. 6, eaba2578 (2020).
Article ADS PubMed PubMed Central Google Scholar
Walker, L. S., Garber, J. & Greene, J. W. Somatization symptoms in pediatric abdominal pain patients: Relation to chronicity of abdominal pain and parent somatization. J. Abnorm. Child. Psychol. 19, 379–394 (1991).
Article PubMed Google Scholar
Walker, L. S., Beck, J. E., Garber, J. & Lambert, W. Children’s somatization inventory: Psychometric properties of the revised form (CSI-24). J. Pediatr. Psychol. 34, 430–440 (2009).
Article PubMed Google Scholar
Achenbach, T. M. (ed) Manual for the Child Behavior Checklist/4–18 and 1991 Profile (Department of Psychiatry, University of Vermont, 1991).
Bolyen, E. et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat. Biotechnol. 37, 852–857 (2019).
Article PubMed PubMed Central MATH Google Scholar
Eid, B. Alterations in cortisol profiles among mothers of children with ASD related to poor child sleep quality. Healthc. (Basel). 10, 666 (2022).
MATH Google Scholar
Morgans, R., Orme, P., Bezuglov, E., Di Michele, R. & Moreira, A. The immunological and hormonal responses to competitive match-play in elite soccer players. Int. J. Environ. Res. Public. Health. 19, 11784 (2022).
Article PubMed PubMed Central Google Scholar
Anderson, M. J. A new method for non-parametric multivariate analysis of variance. Austral Ecol. 26, 32–46 (2001).
MATH Google Scholar
Hyams, J. S. et al. Childhood functional gastrointestinal disorders: Child/adolescent. Gastroenterology 150, 1456–1468 (2016).
Article MATH Google Scholar
Lorenzo, D. C. et al. Visceral hyperalgesia in children with functional abdominal pain. J. Pediatr. 139, 838–843 (2001).
Article PubMed MATH Google Scholar
Iovino, P. et al. Irritable bowel syndrome in childhood: Visceral hypersensitivity and psychosocial aspects. Neurogastroenterol. Motil. 21, 940–e74 (2009).
Article MathSciNet PubMed MATH Google Scholar
Saulnier, D. M. et al. Gastrointestinal microbiome signatures of pediatric patients with irritable bowel syndrome. Gastroenterology 141, 1782–1791 (2011).
Article PubMed MATH Google Scholar
Rajilic-Stojanovic, M. et al. Global and deep molecular analysis of microbiota signatures in fecal samples from patients with irritable bowel syndrome. Gastroenterology 141, 1792–1801 (2011).
Article PubMed MATH Google Scholar
Fukudo, S., Nomura, T. & Hongo, M. Impact of corticotropin-releasing hormone on gastrointestinal motility and adrenocorticotropic hormone in normal controls and patients with irritable bowel syndrome. Gut 42, 845–849 (1998).
Article PubMed PubMed Central Google Scholar
Sudo, N. et al. Postnatal microbial colonization programs the hypothalamic-pituitary-adrenal system for stress response in mice. J. Physiol. 558, 263–275 (2004).
Article PubMed PubMed Central MATH Google Scholar
Almeida, A. et al. A new genomic blueprint of the human gut microbiota. Nature 568, 499–504 (2019).
Article ADS PubMed PubMed Central MATH Google Scholar
Tanaka, Y. et al. Omics profiles of fecal and oral microbiota change in irritable bowel syndrome patients with diarrhea and symptom exacerbation. J. Gastroenterol. 57, 748–760 (2022).
Article PubMed PubMed Central MATH Google Scholar
Download references
This study was funded by a grant from the Japan Society for the Promotion of Science (JSPS; Grant No. 21K02331). The funding agency of the study had no role in the study design, data collection, data analysis, data interpretation, or writing of the manuscript. We want to acknowledge Cykinso, Inc., for conducting the intestinal microbiota analyses reported in this manuscript. The research described in this study was presented at the 64th Annual General Meeting of the Japanese Society of Psychosomatic Medicine and Academic Lectures in Japan. Editorial support in the form of medical writing, assembling tables, creating high-resolution images based on authors’ detailed directions, collating author comments, copyediting, fact-checking, and referencing was provided by Editage and Cactus Communications.
Department of Behavioral Medicine, Tohoku University Graduate School of Medicine, 2-1 Seiryo-Machi, Aoba-Ku, Sendai, 980-8575, Miyagi, Japan
Chikako Shibata, Tomohiko Muratsubaki, Emiko Aizawa, Satoshi Watanabe, Motoyori Kanazawa & Shin Fukudo
Department of Exercise Education for Children, Faculty of Sports Science, Sendai University, Sendai, Japan
Chikako Shibata
Department of Early Childhood Education, Koriyama Women’s University Junior College, Koriyama, Japan
Suguru Shibata
Department of Health and Nutrition, Faculty of Human Sciences, Sendai Shirayuri Women’s College, Sendai, Japan
Emiko Aizawa
Research Center for Accelerator and Radioisotope Science, Tohoku University, Sendai, Japan
Shin Fukudo
Department of Psychosomatic Medicine, Japanese Red Cross Ishinomaki Hospital, Ishinomaki, Japan
Shin Fukudo
Department of Psychosomatic Medicine, Tohoku University Hospital, Sendai, Japan
Tomohiko Muratsubaki, Motoyori Kanazawa & Shin Fukudo
You can also search for this author in PubMed Google Scholar
You can also search for this author in PubMed Google Scholar
You can also search for this author in PubMed Google Scholar
You can also search for this author in PubMed Google Scholar
You can also search for this author in PubMed Google Scholar
You can also search for this author in PubMed Google Scholar
You can also search for this author in PubMed Google Scholar
C.S.: conceptualization, methodology, validation, investigation, writing – original draft; T.M.: formal analysis; S.S.: investigation; E.A.: formal analysis; S.W.: formal analysis; M.K.: validation; S.F.: supervision. All authors had full access to all data in the study and had the final responsibility for the decision to submit the manuscript for publication. C.S. and S.S. accessed and verified the data reported in this manuscript.
Correspondence to Chikako Shibata.
The authors declare no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Below is the link to the electronic supplementary material.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Reprints and permissions
Shibata, C., Muratsubaki, T., Shibata, S. et al. A randomized controlled trial of environmental richness on gastrointestinal symptoms, salivary cortisol, and gut microbiota in early childhood. Sci Rep 15, 8493 (2025). https://doi.org/10.1038/s41598-025-86618-3
Download citation
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-025-86618-3
Anyone you share the following link with will be able to read this content:
Sorry, a shareable link is not currently available for this article.
Provided by the Springer Nature SharedIt content-sharing initiative
Advertisement
© 2025 Springer Nature Limited
Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.