Jukuri, open repository of the Natural Resources Institute Finland (Luke) All material supplied via Jukuri is protected by copyright and other intellectual property rights. Duplication or sale, in electronic or print form, of any part of the repository collections is prohibited. Making electronic or print copies of the material is permitted only for your own personal use or for educational purposes. For other purposes, this article may be used in accordance with the publisher’s terms. There may be differences between this version and the publisher’s version. You are advised to cite the publisher’s version. This is an electronic reprint of the original article. This reprint may differ from the original in pagination and typographic detail. Author(s): Mika Saarenpää, Marja I. Roslund, Noora Nurminen, Riikka Puhakka, Laura Kummola, Olli H. Laitinen, Heikki Hyöty, Aki Sinkkonen, Title: Urban indoor gardening enhances immune regulation and diversifies skin microbiota — A placebo-controlled double-blinded intervention study Year: 2024 Version: Published version Copyright: The Authors 2024 Rights: CC BY-NC-ND 4.0 Rights url: http://creativecommons.org/licenses/by-nc-nd/4.0/ Please cite the original version: Mika Saarenpää, Marja I. Roslund, Noora Nurminen, Riikka Puhakka, Laura Kummola, Olli H. Laitinen, Heikki Hyöty, Aki Sinkkonen, Urban indoor gardening enhances immune regulation and diversifies skin microbiota — A placebo-controlled double-blinded intervention study, Environment International, Volume 187, 2024, 108705, ISSN 0160-4120, https://doi.org/10.1016/j.envint.2024.108705. Environment International 187 (2024) 108705 Available online 26 April 2024 0160-4120/© 2024 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Full length article Urban indoor gardening enhances immune regulation and diversifies skin microbiota — A placebo-controlled double-blinded intervention study Mika Saarenpaa a,b, Marja I. Roslund b, Noora Nurminen c, Riikka Puhakka a, Laura Kummola c, Olli H. Laitinen c, Heikki Hyoty c, Aki Sinkkonen b,* a Ecosystems and Environment Research Programme, Faculty of Biological and Environmental Sciences, University of Helsinki, Niemenkatu 73, 15140 Lahti, Finland b Natural Resources Institute Finland, Turku and Helsinki, Finland c Faculty of Medicine and Health Technology, Tampere University, Arvo Ylpon Katu 34, 33520 Tampere, Finland A R T I C L E I N F O Handling Editor: Marti Nadal Keywords: Biodiversity hypothesis Hygiene hypothesis Skin microbiota Urban gardening Cytokine Immune-mediated disease A B S T R A C T According to the hygiene and biodiversity hypotheses, frequent exposure to environmental microbiota, especially through soil contact, diversifies commensal microbiota, enhances immune modulation, and ultimately lowers the risk of immune-mediated diseases. Here we test the underlying assumption of the hygiene and biodiversity hypotheses by instructing volunteers to grow edible plants indoors during the winter season when natural exposure to environmental microbiota is low. The one-month randomized, placebo-controlled double-blind trial consisted of two treatments: participants received either microbially diverse growing medium or visually similar but microbially poor growing medium. Skin microbiota and a panel of seven immune markers were analyzed in the beginning of the trial and after one month. The diversity of five bacterial phyla (Bacteroidetes, Plancto- mycetes, Proteobacteria, Cyanobacteria, and Verrucomicrobia) and one class (Bacteroidia) increased on the skin of participants in the intervention group while no changes were observed in the placebo group. The number of nodes and edges in the co-occurrence networks of the skin bacteria increased on average three times more in the intervention group than in the placebo group. The plasma levels of the immunomodulatory cytokine interleukin 10 (IL-10) increased in the intervention group when compared with the placebo group. A similar trend was observed in the interleukin 17A (IL-17A) levels and in the IL-10:IL-17A ratios. Participants in both groups re- ported high satisfaction and adherence to the trial. The current study provides evidence in support of the core assumption of the hygiene and biodiversity hypotheses of immune-mediated diseases. Indoor urban gardening offers a meaningful and convenient approach for increasing year-round exposure to environmental microbiota, paving the way for other prophylactic practices that might help prevent immune-mediated diseases. 1. Introduction Autoimmune and immune-mediated diseases (e.g., asthma and al- lergies) are becoming more and more prevalent in developed countries (Lerner et al., 2016; To et al., 2012). Exposure to diverse environmental microbiota, especially through soil, is important for the natural devel- opment and functioning of the immune system and, ultimately, for human health (Ege et al., 2012; Fyhrquist et al., 2014; Noverr & Huff- nagle, 2005; Ottman et al., 2019; Rook, 2009; Stein et al., 2016; Stiemsma et al., 2015; Valkonen et al., 2015; von Hertzen & Haahtela, 2006). According to the hygiene hypothesis and its variant, the biodiversity hypothesis, this is linked to biodiversity loss and reduced microbial exposure (Aerts et al., 2018; Civitello et al., 2015; Haahtela, 2019; Haahtela et al., 2021; Hanski et al., 2012; Lehtimaki et al., 2017; Rohr et al., 2020; Rook, 2009; Ruokolainen et al., 2017; von Hertzen et al., 2011; von Hertzen & Haahtela, 2006). Beneficial microbial exposure has decreased among urban dwellers (Gupta et al., 2020; Parajuli et al., 2018; Shan et al., 2020), and people living in urban areas have distinct microbiota compositions compared to their rural coun- terparts (Hanski et al., 2012; Lehtimaki et al., 2017, 2018; Ruokolainen et al., 2015, 2020). Compelling evidence suggests that home surroundings play a key * Corresponding author. E-mail addresses: mika.saarenpaa@helsinki.fi (M. Saarenpaa), marja.roslund@luke.fi (M.I. Roslund), noora.nurminen@tuni.fi (N. Nurminen), riikka.puhakka@ helsinki.fi (R. Puhakka), laura.kummola@tuni.fi (L. Kummola), olli.laitinen@tuni.fi (O.H. Laitinen), heikki.hyoty@tuni.fi (H. Hyoty), aki.sinkkonen@luke.fi (A. Sinkkonen). Contents lists available at ScienceDirect Environment International journal homepage: www.elsevier.com/locate/envint https://doi.org/10.1016/j.envint.2024.108705 Received 23 January 2024; Received in revised form 26 March 2024; Accepted 25 April 2024 Environment International 187 (2024) 108705 2 role in determining microbiota-related health outcomes. Forest and agricultural land cover during the first year of life is inversely associated with the risk of atopic sensitization and type 1 diabetes (Hanski et al., 2012; Nurminen et al., 2021; Ruokolainen et al., 2015). Biodiverse and green living areas correlate with respiratory health (Donovan et al., 2018; Liddicoat et al., 2018). Diverse yard vegetation and outdoor nature-related activities have been linked to health-related changes in the gut microbiota composition (Parajuli et al., 2020; Sobko et al., 2020). Living in rural or farm-like conditions is directly associated with diverse commensal microbiota and a well-functioning immune system (Ege et al., 2012; Kirjavainen et al., 2019; Stein et al., 2016; von Mutius & Radon, 2008). Rural second homes and active contacts with urban green space have been linked to differences in commensal microbial communities (Saarenpaa et al., 2021; Selway et al., 2020), and the number of visits to urban green space is associated with reduced need for asthma and psychotropic medication (Turunen et al., 2023). In line with comparative studies on disease incidence, biodiversity intervention tri- als (i.e., introduction of microbiologically diverse vegetation and soil to everyday living environment) have been shown to enrich commensal microbiota and enhance immune modulation among urban citizens, at least temporarily, but as notified by several authors, many of the studies have been of pilot character (Hui et al., 2019a; Roslund et al., 2020, 2021, 2022, 2023; Tischer et al., 2022). Even though built urban space is known to reduce transfer of health- associated microbiota indoors (Hui et al., 2019b; Parajuli et al., 2018; Zhao et al., 2024), only a handful of studies have concentrated on ways to modify indoor exposure to rich environmental microbiota in densely built areas. A single study suggested that indoor green walls enrich skin microbiota and may enhance immunomodulation (Soininen et al., 2022), and a few intervention trials indicated that short-term contact with microbiologically rich soil products supports diverse skin micro- biota (Gronroos et al., 2019; Nurminen et al., 2018; see also Shaffer & Lozupone, 2018). Gardening has successfully been used as a form of therapy for centuries, and recently it has been scientifically proven to provide substantial health benefits ranging from alleviated depression and stress to elevated cognitive functions (Soga et al., 2016). While some studies have looked into whether urban gardening affects the human microbiota (Gascon et al., 2020; Mhuireach et al., 2023), and a single study has even recorded plasma IL-6 levels among breast cancer survi- vors to understand the effect of home gardening on inflammation (Bail et al., 2018), no study has, to our knowledge, investigated the impact of indoor gardening on commensal microbiota and immunomodulation simultaneously in the context of the hygiene and biodiversity hypothe- ses of immune-mediated diseases (see Rook & Lowry, 2022). Here we present the results of a test of the core idea on which the hygiene and biodiversity hypotheses were built on, i.e., whether expo- sure to soil-based microbiota causes changes in commensal microbiota and enhances immune system modulation. For this, we provided vol- unteers with growing medium and crop species for indoor gardening. To understand the role of soil microbiota exposure in the shifts of immune marker levels in plasma, we built a randomized double-blind trial. Half of the volunteers received microbially rich growing medium and the other half visually similar but microbially poor peat-based growing medium. As far as we know, this is the first randomized double-blind clinical trial examining the effects of indoor gardening on both the microbiota and immune system in parallel. The detailed hypotheses were that i) urban indoor gardening with microbially rich growing medium (intervention group) causes changes in the skin microbial communities while gardening with peat-based growing medium (pla- cebo group) causes no or minuscule changes; ii) the co-occurrence net- works of the skin bacteria are more complex in the intervention than in the placebo group after but not before the intervention; and that iii) immune modulation is enhanced in the intervention but not in the placebo group. 2. Materials and methods 2.1. Participants Participants of this trial were recruited from the cities of Lahti (population > 120 000) and Hyvinkaa (population > 45 000) in Finland. The trial followed the recommendations of the Finnish Advisory Board on Research Integrity and was approved by the ethics committee of the Pirkanmaa Hospital District (approval code: R19077, date: 19.8.2019). Written informed consent in accordance with the Declaration of Helsinki was signed by all participants. Exclusion criteria included age under 18 years, severe doctor-diagnosed immune deficiency (e.g., HIV, antibody deficiency), immunosuppressive medication, immune-mediated disease (e.g., colitis ulcerosa, Crohn’s disease, rheumatoid arthritis), leuko- penia, tetanus antibody deficiency, diabetes, mental or memory disor- der, cancer diagnosis, rash or eczema, daily smoking, owning indoor pets, and living on an operating farm. Participants were placed either in an intervention or placebo group through a stratified randomization (variables: age, reported gender). All participants were asked to fill out questionnaires assessing their living conditions and lifestyle both during their everyday life and the trial (e. g., medication and food supplements, diet and changes in it, outdoor activities and hobbies, contact with soil). Adherence to the trial, nega- tive and positive experiences related to the trial, and hand-washing and hand-sanitizing habits were also recorded. All participants using anti- biotics during or six months before the trial or anti-inflammatory drugs during the trial were excluded from the analyses. Samples from thirteen placebo and fifteen intervention group participants were used in the analyses (Table 1). Table 1 Demographic characteristics of the study participants. Placebo group Intervention group No of participants 13 15 Gender Female 10 11 Male 3 4 Age <35 3 4 35–60 6 6 >60 4 5 Average 47 46 Range 29–72 29–70 Type of residence Apartment 5 6 Rowhouse 4 3 Detached house 4 6 Gardening during growing season Outdoor vegetable garden 5 3 Outdoor flower bed 2 2 Indoor vegetables 2 1 Daily 1 0 Weekly 5 6 Monthly 1 0 Rarely 2 6 Never 4 3 Outdoor recreation Walking Daily 8 5 Weekly 3 6 Monthly 0 4 Rarely 2 0 Never 0 0 Cycling Daily 3 4 Weekly 3 3 Monthly 2 5 Rarely 5 1 Never 0 2 Other (e.g., berry picking, swimming, hiking, fishing) Daily 5 3 Weekly 3 6 Monthly 3 4 Rarely 2 2 Never 0 0 Diet Omnivorous 13 15 Vegetarian/vegan 0 0 Changes during trial No: all No: all M. Saarenpaa et al. Environment International 187 (2024) 108705 3 2.2. Planters and crop species Participants were provided with a plastic planter (size 70x14x18 cm), spray bottle, desk lamp with a bulb (20 W 3000 K 2000 lm), seeds and plants, and growing medium (Fig. 1). The placebo group used commercial horticultural peat (Luonnonturve by Kekkila, Finland) mixed with heat-expanded lightweight clay pebbles (Leca by Saint- Gobain, Finland) and inorganic fertilizer sticks (Substral by Transmeri, Finland). Horticultural peat was chosen because of its wide use and low bacterial diversity (1.3  108 16S rRNA sequences per gram soil; Roslund et al., 2022; Fig. 2). The intervention group used the same horticultural peat and clay pebble blend mixed with a microbiologically rich compost-based mixture (3.5  109 16S rRNA sequences per gram mixture; Hui et al., 2019a) developed and used in our previous studies (Hui et al., 2019a; Nurminen et al., 2018). This mixture contains sieved composted materials such as tree bark and mulch, dung, deciduous leaf litter, peat, agricultural sludge, and Sphagnum moss. Both groups were provided with the same seven crop species: lettuce (Lactuca sativa), white mustard (Sinapis alba), radish (Raphanus rapha- nistrum subsp. sativus), garlic (Allium sativum), ginger (Zingiber offici- nale), pea (Pisum sativum), and fava bean (Vicia faba). This set of species was chosen as some of them can be harvested shortly after sowing as microgreens (mustard) and sprouts (pea, fava bean), and some later as leaves (lettuce), taproots (radish), bulbs (garlic), and rhizomes (ginger). Both groups were given the same instructions. First, a layer of clay pebbles was placed at the bottom of the planter to improve drainage. The planter was then filled either with the peat-based or compost-based growing medium. The seeds, rhizomes, and bulbs were planted in different sections of the planter according to the instructions. These first steps were completed barehanded. The planter was then placed close to a window and watered, and the lamp was set up nearby. As the trial was conducted during the winter season, in February and March, grow lights were needed. Daily tasks included both soil and dietary exposure: Monitoring the moisture level by inserting a bare finger deep into the growing medium, watering and misting with the spray bottle, harvest- ing, and resowing. 2.3. Skin and growing medium samples Skin swab samples for bacterial analyses were collected at the beginning of the trial (0 mo) and at one month (1 mo) by a trained nurse. The back of the hand (2x2 cm) was swabbed with ten horizontal and ten vertical strokes with a cotton swab wetted with saline buffer (0.1 % Tween 20 in 0.15 M NaCl). The cotton tips were placed in separate sterile plastic tubes and stored at 80 C until further processing. Growing medium samples were collected into separate zip lock bags with sterilized plastic spoons by a university researcher before the start of the trial, and they were stored at 80 C until further processing. Bacterial DNA was extracted from the skin samples with the Fast DNA Spin Kit for Soil (MP Biomedicals, Santa Ana, CA, USA) and from the growing medium samples with the PowerSoil DNA Isolation Kit (Qiagen, Hilden, Germany) as per the manufacturers’ standard pro- tocols. DNA concentrations were quantified using the Quant-iT Pico- Green dsDNA Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA). The V4 region of the 16S rRNA gene was amplified with the 505F and 806R primers. Sterile water was used as a negative control during the DNA extraction and no-template control was used during the PCR. Cupriavidus necator JMP134 (DSM4058) was used as a positive control. All samples were sequenced at the Institute for Molecular Medicine Finland (FIMM, Helsinki, Finland) using the Illumina MiSeq platform (2x300 bp, V3 reagent kit). 2.4. Blood samples Blood samples were used to analyze the levels of anti-inflammatory cytokines interleukin 10 (IL-10) and transforming growth factor beta (TGF-β1), which have been connected to a lower risk of several immune- mediated diseases (Burmeister & Marriott, 2018; Li et al., 2006; Opal & DePalo, 2000; Prud’homme and Piccirillo, 2000); proinflammatory cy- tokines interleukin 17A (IL-17A), interleukin 1 beta (IL-1β), and tumor necrosis factor (TNF-α), which have been associated with the risk of immune-mediated diseases (Dinarello, 2000; Honkanen et al., 2010; Kuwabara et al., 2017); and multifunctional cytokines interleukin 6 (IL- 6) and interleukin 21 (IL-21), which have both anti-inflammatory and proinflammatory properties (Mehta et al., 2004; Rose-John, 2012). The blood samples were collected at the beginning of the trial (0 mo) and at one month (1 mo) by a trained nurse. Venous blood was drawn into two Vacutainer CPT Mononuclear Cell Preparation tubes containing sodium citrate (BD Biosciences, NJ, USA). The tubes were centrifuged as per the manufacturer’s instructions to separate the plasma. The plasma samples were stored at 80 C until further processing. IL-17A, IL-10, IL- 1β, IL-6, IL-21, and TNF-α concentrations were measured from the plasma samples using the Milliplex MAP High Sensitivity T Cell Panel kit (Merck KGaA, Darmstadt, Germany) with the Bio-Plex 200 system (Bio- Rad Laboratories, Hercules, CA, USA) and Bio-Plex Manager software (version 4.1, Bio-Rad Laboratories, Hercules, CA, USA). TGF-β1 Fig. 1. Gardening equipment provided for the participants consisted of a plastic planter, lamp, bulb, spray bottle, crop species, and growing medium. M. Saarenpaa et al. Environment International 187 (2024) 108705 4 concentrations were measured using ELISA (BioVendor, Czech Republic). 2.5. Bioinformatics Paired-end sequence data (.fastq) from the skin and growing medium (i.e. soil) samples were processed using mothur (version 1.39.5, htt ps://www.mothur.org, accessed on 27 July 2022; Schloss et al., 2009) following the protocols by Schloss & Westcott (2011) and Kozich et al. (2013) as described in our earlier studies (Nurminen et al., 2018; Par- ajuli et al., 2020). General quality of the sequencing data was good; per sequence quality scores for all samples were (Q score) > 30. The overall number of reads was 4 639 172 in skin samples and 3 449 602 in soil samples. There were 56 575  25 692 and 49 280  8 770 (mean  SD) reads per skin and soil sample, respectively. The sequences were aligned using the mothur version of SILVA bacterial reference (version 132; Pruesse et al., 2007). Less abundant (10 sequences across all experi- mental units) operational taxonomic units (OTUs) were removed to avoid PCR or sequencing artifacts. Contaminant OTUs were removed as described by Roslund et al. (2021). Both the skin and soil samples were subsampled to the lowest sequence counts (5 907 and 33 805, respec- tively; 1000 iterations). Good’s coverage index (rarefaction curve analysis) was 0.91  0.06 for soil and for skin 0.96  0.06 (mean  SD). 2.6. Statistics All statistical analyses and data visualizations were performed using the R statistical software environment (version 4.1.2, R Foundation, Vienna, Austria; R Core Team, 2020). Following packages were used: phyloseq (version 1.38.0; McMurdie & Holmes, 2013) for community composition analyses and diversity calculations, vegan (version 2.5–7; Oksanen et al., 2019) for diversity calculations, ggplot2 (version 3.3.5; Wickham, 2016) for visualizations, MKinfer (version 0.6; Kohl, 2020) for permutation tests, lme4 (version 1.1–29; Bates et al., 2015) and rsq (version 2.5; Zhang, 2018) for linear mixed-effects models (LMM), and cooccur (version 1.3; Griffith et al., 2016) and visNetwork (version 2.1.0; Thieurmel, 2021) for co-occurrence network analyses. Statistical analyses of the skin and growing medium bacterial com- munities were conducted at different taxonomic levels (i.e., OTU, genus, family, order, class, phylum). At the OTU level, analyses were conducted within the most abundant phyla and classes (relative abundance > 0.025 % across all samples). The data were transformed to proportions by dividing the reads for each operational taxonomic unit (OTU) in a sample by the total number of reads in that sample (McKnight et al., 2019). The cytokine concentration data were log10 transformed to normalize the distributions. In the cytokine analyses, changes in the log10 cytokine concentrations were used. These were calculated by extracting the starting value (0 mo) from the one-month value (1 mo). Bacterial Shannon and Simpson diversity indices, observed richness, and relative abundances were compared between the intervention and placebo group as well as between the time points using the Student’s t- test or Mann–Whitney U test. The Student’s t-test was used when the data were normally distributed based on the Shapiro–Wilk test, and the Mann–Whitney U test was used when the data were not normally distributed. Paired tests were used when different time points were compared within the groups. Changes in the cytokine levels were compared between the intervention and placebo group using a permu- tation t-test. LMMs were used to study the relationship between the bacterial and cytokine variables. Participants were used as a random factor in all models. Bacterial co-occurrence network analyses were conducted at different taxonomic levels (OTU, genus, family, order, class, phylum) using binary presence-absence data, and the number of nodes and edges was recorded. p-values were corrected using the Ben- jamini–Hochberg method (Benjamini & Hochberg, 1995). 3. Results 3.1. Anti-inflammatory cytokine IL-10 increased in the intervention group Compared to the placebo group, log10 IL-10 levels increased in the intervention group during the trial (Fig. 3a, Table 2). A similar differ- ence was observed in log10 IL-17A levels and log10 IL-10:IL-17A ratios (Fig. 3b–c, Table 2). No statistically significant differences were observed in log10 TNF-α, IL-21, IL-6, IL-1β or TGF-β1 levels (Fig. 3d–h, Table 2). After the Benjamini–Hochberg p-value correction, only dif- ferences in the log10 IL-10 levels remained significant (Fig. 3a, Table 2). 3.2. Skin bacterial diversity increased in the intervention group The Shannon diversity of the phyla Proteobacteria, Bacteroidetes, Fig. 2. Horticultural peat used by the placebo group had lower OTU richness (a) and Shannon diversity (b) than the compost-based mixture used by the inter- vention group. M. Saarenpaa et al. Environment International 187 (2024) 108705 5 Fig. 3. Log10 IL-10 (a) and IL-17A (b) concentrations and log10 IL-10:IL-17A ratio (c) increased in the intervention group during the trial. No differences between the groups were observed in log10 TNF-α (d), IL-21 (e), IL-6 (f), IL-1β (g) or TGF-β1 (h) concentrations. Boxplots show medians (thicker line), upper and lower hinges (box), values 1.5 times the interquartile range (whiskers), and values outside hinges (data points, outliers). M. Saarenpaa et al. Environment International 187 (2024) 108705 6 Planctomycetes, Cyanobacteria, and Verrucomicrobia and of the class Bacteroidia increased (q < 0.10) in the intervention group during the trial (Fig. 4, Table 3). No differences in the Shannon diversities between the time points were observed in the placebo group (Fig. 4, Table 3). There were no differences (q > 0.10) at order, family or genus levels, neither between time points nor between treatments. More than 60 % of the skin bacterial community belonged to the 15 most abundant genera, while uncultured genera covered less than one tenth of the skin bacterial community (Fig. 5). The most abundant genera were always Staphylo- coccus, Streptococcus, Ralstonia, and Corynebacterium (Fig. 5). 3.3. Nodes and edges increased in the co-occurrence networks The topological parameters of the skin bacterial co-occurrence net- works were compared between the intervention and placebo group (Table 4). The complexity of the networks grew in both groups as the number of nodes (individual taxa or OTUs) and edges (correlations) increased at most of the taxonomic levels. Although these increases were multiple times higher in the intervention group at most levels, they were most pronounced at the OTU and genus levels. At the genus level, the number of nodes and edges in the intervention group increased by 146 and 6236, respectively, and in the placebo group by 50 and 677, respectively (Table 4, Fig. 6). The proportion of positive correlations increased in both groups at most of the taxonomic levels. After the trial, the skin bacterial co-occurrence networks in the intervention group had more nodes and edges than the networks in the placebo group. 3.4. Adherence to the trial Adherence to the trial was 100 % in both groups, and most partici- pants cared for their crops on a daily basis. A few participants reported taking short breaks during the weekends (Table A.1), but these breaks did not take place right before sampling. Participants in both groups reported minimal soil contact unrelated to the trial (Table A.1), for example through their outdoor gardens or outdoor recreation, most probably due to the winter season. 4. Discussion The results of this biodiversity intervention trial support the core assumption on which the hygiene and biodiversity hypotheses were built on. During the trial, skin bacterial diversity became higher and co- occurrence networks more complex in the intervention group than in the placebo group, and these differences were reflected in changes in the immune response. Since the diversity of five bacterial phyla and one bacterial class increased in the intervention but not in the placebo group during the exposure, the results support the first hypothesis that handling microbially rich growing medium enriches the skin micro- biota. The second microbially oriented hypothesis was supported as the number of co-occurrence nodes and edges were higher in the interven- tion group than in the placebo group after the trial. As between- treatment differences in changes in the immunomodulatory cytokine levels were evident, the trial is also in accordance with the hypothesis that immune modulation was enhanced in the intervention but not in the placebo group. The current study indicates that daily contact with microbially rich soil—instead of any soil—is vital for immune modulation. However, while the plasma levels of the anti-inflammatory IL-10 increased in the intervention group, the underlying trigger is unclear. Although shifts in IL-10 levels have previously been associated with Gammaproteobacteria on skin (Fyhrquist et al., 2014; Hanski et al., 2012; Roslund et al., 2020, 2022), the finding was not repeated in the current study (LMM: Gam- maproteobacteria ~ IL-10, p > 0.05, data not shown). This may be related to the mode of exposure: while in previous studies participants have been exposed to the environmental microbiota mainly externally, in the current study the participants were exposed also via gastrointes- tinal tract while consuming the crops. This might cause shifts in the gut microbiota, and the potential gut microbiota differences might affect the immune response. Indeed, edible plants contain microbiota that are hypothetically linked to human immune response (Wicaksono et al., Fig. 3. (continued). Table 2 Differences in log10 cytokine concentrations between the groups were tested with a permutation t-test (9999 permutations). p-values, q-values, and signifi- cance after the correction are given for all cytokines and for IL-10:IL-17A ratio. Cytokine p-value q-value Significant with FDR 0.1 IL-10 0.0046 0.0368 Yes IL-17A 0.0312 0.1248 No IL-10:IL-17A 0.0479 0.1277 No TNF-α 0.1141 0.2118 No IL-21 0.1324 0.2118 No IL-6 0.2709 0.3612 No IL-1β 0.3668 0.4085 No TGF-β1 0.4085 0.4085 No M. Saarenpaa et al. Environment International 187 (2024) 108705 7 Fig. 4. Shannon diversity of the phyla Proteobacteria (a), Bacteroidetes (c), Planctomycetes (e), Cyanobacteria (g), Verrucomicrobia (i) and of the class Bacteroidia (k) increased in the intervention group during the trial while no changes were observed in the placebo group (b, d, f, h, j, l). Boxplots show medians (thicker line), upper and lower hinges (box), values 1.5 times the interquartile range (whiskers), and values outside hinges (data points, outliers). M. Saarenpaa et al. Environment International 187 (2024) 108705 8 2022, 2023). To conclude, the lack of associations between specific skin bacterial taxa and immune markers does not nullify the importance of the shifts in immune markers as these might have been affected by changes in the gut microbiota. Interestingly, changes in the skin bacterial diversity in the current study occurred mainly at the phylum level, while all the hub OTUs in the co-occurrence networks had a low relative abundance (Table A.2). The implications of this are twofold. First, the common protocol of searching for associations between immune responses and shifts in abundances or diversities at higher taxonomic levels hardly distinguishes the potential Fig. 4. (continued). M. Saarenpaa et al. Environment International 187 (2024) 108705 9 importance of rare taxa, simply because random variation caused by numerous confounding factors in the everyday living environment is likely multiple orders of magnitude higher among rare taxa than major phyla and classes. Second, since variation in phylogeny and niche specialization within classes and phyla can be high, it may not be optimal to use classes and phyla as indicators of health or immune response in studies using soil exposure. Nevertheless, the diversity of the phylum Proteobacteria increased in the intervention but not in the placebo group in the current study, and in an earlier study Proteobac- teria have been identified as possible health indicators of commercially cultivated edible plant species (Koberl et al., 2017). Also, the phylum Bacteroidetes that increased in diversity in the intervention group has an important role in maintaining homeostasis and healthy gastrointestinal functions, especially mucosal immunity (Gibiino et al., 2018; Hutten- hower et al., 2012; Troy & Kasper, 2010). Although we did not study if skin Bacteroidetes were transmitted into the gastrointestinal tract, skin exposure to soil with high microbial diversity is capable of modifying the gut microbiota (Nurminen et al., 2018). Our study did not separate whether skin Bacteroidetes diversity increased via soil contact or via plant contact in the intervention group, while no change occurred in the control group. Interestingly, some studies have found residential green spaces–vegetation patches–to be negatively associated with Bacter- oidetes levels in the gut (Van Pee et al., 2023). In addition to Proteo- bacteria and Bacteroidetes, the Shannon diversity of the phylum Cyanobacteria increased in the intervention group. While not much is known of their role in health and disease, some studies have found them Table 3 Shannon diversities at different time points (0 and 1 month) were compared within the intervention and placebo groups using paired tests. Shannon diversities (mean  standard deviation), p-values, q-values (only intervention group), statistical significance, and test types are given for all tested phyla and classes. Statistical sig- nificance is marked with bold font. Taxon Treatment Shannon diversity index (mean  SD) p-value q-value Significant with FDR 0.1 Test 0 mo 1 mo Phylum Bacteroidetes Intervention 2.76 ± 0.35 3.44 ± 0.73 0.0021 0.0231 Yes t-test Placebo 3.14  0.66 3.22  0.62 0.7671 No U test Class Bacteroidia Intervention 2.75 ± 0.35 3.42 ± 0.71 0.0019 0.0320 Yes t-test Placebo 3.14  0.66 3.22  0.62 0.7639 No t-test Phylum Planctomycetes Intervention 1.02 ± 0.50 1.99 ± 1.40 0.0135 0.0675 Yes U test Placebo 1.42  0.94 1.56  1.19 0.8888 No t-test Class Planctomycetacia Intervention 1.08  0.67 2.16  1.31 0.0214 0.1712 No U test Placebo 1.47  1.13 1.91  0.95 0.2790 No t-test Phylum Proteobacteria Intervention 1.99 ± 0.32 2.51 ± 0.91 0.0184 0.0675 Yes U test Placebo 2.32  0.54 2.38  0.60 0.7981 No U test Class Alphaproteobacteria Intervention 3.43  0.45 3.40  0.91 0.9201 0.9321 No t-test Placebo 3.52  0.48 3.51  0.53 0.9746 No t-test Class Deltaproteobacteria Intervention 2.16  0.40 2.66  0.98 0.2220 0.4364 No U test Placebo 2.24  0.80 2.36  0.91 0.6477 No t-test Class Gammaproteobacteria Intervention 1.69  0.31 1.93  0.68 0.2220 0.4364 No U test Placebo 1.93  0.34 1.96  0.48 0.8511 No t-test Phylum Cyanobacteria Intervention 1.07 ± 0.53 1.50 ± 0.48 0.0307 0.0844 Yes t-test Placebo 1.56  0.68 1.46  0.77 0.7761 No U test Class Oxyphotobacteria Intervention 0.76  0.62 1.00  0.59 0.1091 0.4364 No t-test Placebo 1.28  0.70 1.23  0.68 0.8742 No t-test Phylum Verrucomicrobia Intervention 0.99 ± 0.76 1.73 ± 1.04 0.0426 0.0937 Yes t-test Placebo 1.40  1.07 1.65  1.15 0.5761 No t-test Class Verrucomicrobiae Intervention 0.99  0.76 1.73  1.04 0.0426 0.2272 No t-test Placebo 1.40  1.07 1.65  1.15 0.5761 No U test Phylum Acidobacteria Intervention 2.26  0.51 2.72  0.84 0.1193 0.2187 No t-test Placebo 2.47  0.77 2.68  0.74 0.4629 No t-test Class Acidobacteriia Intervention 1.86  0.50 2.21  0.82 0.2455 0.4364 No t-test Placebo 2.11  0.84 2.34  0.73 0.4761 No t-test Phylum Fusobacteria Intervention 1.73  0.48 1.59  0.25 0.1641 0.2579 No U test Placebo 1.85  0.30 1.73  0.35 0.4017 No t-test Class Fusobacteriia Intervention 1.73  0.48 1.59  0.25 0.1641 0.4364 No U test Placebo 1.85  0.30 1.73  0.35 0.4017 No U test Phylum Deinococcus-Thermus Intervention 1.68  0.42 1.50  0.41 0.2799 0.3849 No t-test Placebo 1.58  0.67 1.61  0.43 0.8888 No U test Class Deinococci Intervention 1.68  0.42 1.50  0.41 0.2799 0.4478 No t-test Placebo 1.58  0.67 1.61  0.43 0.8888 No U test Phylum Chloroflexi Intervention 1.81  0.95 2.05  1.04 0.6062 0.7409 No t-test Placebo 1.66  0.94 1.78  0.84 0.7297 No t-test Phylum Firmicutes Intervention 1.85  0.36 2.07  1.13 0.7120 0.7832 No U test Placebo 2.22  0.40 2.12  0.74 0.9443 No U test Class Negativicutes Intervention 1.03  0.34 1.20  0.34 0.1883 0.4364 No t-test Placebo 1.11  0.27 1.30  0.37 0.2318 No t-test Class Clostridia Intervention 2.82  0.72 3.15  0.72 0.4432 0.5909 No U test Placebo 3.04  0.36 3.19  0.47 0.4264 No t-test Class Bacilli Intervention 1.37  0.33 1.47  0.91 0.9321 0.9321 No U test Placebo 1.66  0.23 1.57  0.53 0.8339 No U test Phylum Actinobacteria Intervention 2.47  0.48 2.50  0.61 0.8266 0.8266 No t-test Placebo 2.68  0.58 2.78  0.66 0.5559 No t-test Class Acidimicrobiia Intervention 1.11  0.73 1.58  1.17 0.3163 0.4601 No t-test Placebo 0.97  0.89 1.35  0.96 0.2240 No U test Class Thermoleophilia Intervention 1.92  0.60 2.10  1.03 0.5815 0.7157 No t-test Placebo 1.98  0.91 2.20  0.79 0.5399 No t-test Class Actinobacteria Intervention 2.43  0.48 2.41  0.51 0.8153 0.9318 No t-test Placebo 2.62  0.55 2.66  0.56 0.8030 No t-test M. Saarenpaa et al. Environment International 187 (2024) 108705 10 to protect the skin against UV-induced damage and pigmentation (Fuentes-Tristan et al., 2019; Li et al., 2020). Co-occurrence network analyses focusing on the microbe-microbe interactions on human skin have not been reported in any previous intervention trials testing the hygiene and biodiversity hypotheses. Despite their limited use, co-occurrence networks can be a convenient tool in revealing positive and negative interactions between microbial taxa. A positive correlation can imply that two species simply favor similar environmental conditions, or that they interact with each other and even co-operate, for example by producing metabolites that benefit one another (Das et al., 2018). In the same way, a negative correlation may indicate that the species require different environmental condi- tions, or that they are competing with each other. Since in our study both groups had daily soil contact, it is not surprising that the number of Fig. 5. The relative abundance of the most common bacterial genera (>1 %) A) on the skin and B) in the soil in the placebo and intervention treatment in the beginning (0 mo) and after one month (1 mo) of the intervention. Table 4 Topological measures of the skin bacterial co-occurrence networks. Taxonomic level Treatment Time point Nodes Node change Edges Edge change Positive Positive % Negative Negative % OTU Placebo 0 mo 1202 66 21171 369 18071 85 % 3100 15 % 1 mo 1268 20802 17720 85 % 3082 15 % Intervention 0 mo 980 502 9630 19365 6890 72 % 2740 28 % 1 mo 1482 28995 23444 81 % 5551 19 % Genus Placebo 0 mo 427 50 3478 677 2983 86 % 495 14 % 1 mo 477 4155 3828 92 % 327 8 % Intervention 0 mo 404 146 1882 6236 1335 71 % 547 29 % 1 mo 550 8118 7858 97 % 260 3 % Family Placebo 0 mo 188 31 763 233 686 90 % 77 10 % 1 mo 219 996 967 97 % 29 3 % Intervention 0 mo 182 60 437 890 325 74 % 112 26 % 1 mo 242 1327 1286 97 % 41 3 % Order Placebo 0 mo 92 22 224 87 187 83 % 37 17 % 1 mo 114 311 301 97 % 10 3 % Intervention 0 mo 76 41 95 237 72 76 % 23 24 % 1 mo 117 332 313 94 % 19 6 % Class Placebo 0 mo 28 12 30 40 26 87 % 4 13 % 1 mo 40 70 66 94 % 4 6 % Intervention 0 mo 31 21 23 58 16 70 % 7 30 % 1 mo 52 81 78 96 % 3 4 % Phylum Placebo 0 mo 5 8 3 7 2 67 % 1 33 % 1 mo 13 10 9 90 % 1 10 % Intervention 0 mo 6 8 3 9 3 100 % 0 0 % 1 mo 14 12 11 92 % 1 8 % M. Saarenpaa et al. Environment International 187 (2024) 108705 11 nodes and edges increased in both groups at almost all taxonomic levels during the trial. As the increases were stronger in the intervention group, the results of the co-occurrence network analyses indicate that microbially diverse soil is superior to microbially poor soil in diversi- fying the skin microbiota. The lack of clinical trials, especially those focusing on skin, that report co-occurrence network results makes forming any estimates on possible health outcomes hard. Comparing co-occurrence results be- tween different types of clinical trials can be particularly challenging, not least due to the obscure nature of positive and negative correlations. Relvas et al. (2021) used co-occurrence network analysis to study the impact of oral health on the salivary microbiome, and found that the health-associated network was characterized by more interconnections between the nodes than the disease-associated. In the current study, the intervention group network contained considerably more in- terconnections than the placebo group network after the trial. Relvas et al. (2021) also found the health-associated network to be better balanced in terms of the proportion of positive correlations. In our study, the proportion of positive correlations increased in both groups, which is understandable as soil bacteria have been found to benefit from inter- specific cooperation (Ren et al., 2015). The hub OTUs identified in our study had a low relative abundance (<1 %)—a finding that Relvas et al. (2021) also reported—which implies that even rare and less-abundant taxa can have an important role in the microbial communities of Fig. 6. Skin bacterial co-occurrence networks at the genus level: intervention group (green) before (a) and after the trial (b), placebo group (yellow) before (c) and after the trial (d). M. Saarenpaa et al. Environment International 187 (2024) 108705 12 human skin. The more complex intervention group co-occurrence networks detected at one month are in accordance with the changes in the skin microbiota diversity. It is logical to assume that there are well- established microbial interaction networks in the intervention growing medium as it consists of multiple natural components with high mi- crobial activity and diversity. As the participants interact with the growing medium, some parts of these complex interconnection networks are transferred onto the skin. The skin swab samples were taken from the back of the hand, not from fingers that the participants mostly used while tending to the crops, which indicates that the detected networks are not solely a result of soil residues being sampled. The nurse visited the participants at different times of the day, and while the participants were given instructions to perform the daily tasks prior to the visit, we, due to ethical reasons, could not monitor the exact time of performing the tasks or handwashing. Due to this limitation, we can only assume that some participants had soil contact and/or washed their hands immediately before sampling while others did not. Another weakness of the current study is its short duration due to the COVID-19 pandemic. Skin and blood samples were not collected at three months of gardening as originally planned. Samples collected at multi- ple timepoints are needed to account for the high temporal variability observed in some individuals (Flores et al., 2014). As the follow-up samples are missing, it is unknown whether the changes in the skin microbiota would have persisted after the trial. Mhuireach et al. (2023) found that soil contact through gardening increases the number of bacterial taxa shared between soil and skin, but this effect largely dis- appears within 12 h. Hence, we assume that no long-lasting microbiota changes would have been detected in the midst of the COVID-19 pandemic. In general, gardening is often a hobby or even a profession for those involved in it, leading to recurring soil contact and microbial exposure, which—based on the current study—might lead to long-term shifts in the immune response (Roslund et al., 2020,2021; Tischer et al., 2022). In addition to being longer-lasting, future studies should be considerably larger and involve people from different geographic re- gions. The current study was heavily skewed towards the female gender, and the results might not be fully generalizable, particularly due to sex differences in immune responsiveness (Beenakker et al., 2019). An optimal study would also consider all possible routes of inoculation. The current study did not differentiate between the role of the consumed plants and aerobiome. Previous studies have found that indoor plants are able to shape the indoor microbial communities (Dockx et al., 2022; Mahnert et al., 2015) and potentially even enhance immune regulation (Soininen et al., 2022). Wider cytokine panels could be utilized to add to the knowledge of the impact of soil contact on immune response. Future studies should also consider using shotgun metagenomic sequencing instead of targeted 16S rRNA as it allows for a more in-depth analysis of the microbial assemblages by enabling the identification of not only bacteria but also fungi, viruses, and other microorganisms and their functional potential. As our study was conducted during the winter season, it demon- strates how beneficial microbial exposure can be obtained year-round. Overall microbial exposure, especially among urbanites, has been found to be low in winter and high in spring and summer (Hui et al., 2019b; Mhuireach et al., 2021). This might also explain why gardening studies executed during the summer months have failed to detect mi- crobial effects (Gascon et al., 2020). In addition to the season, our trial differs from previous gardening studies by taking place indoors and by being double-blinded and more rigorously controlled. In many gardening studies that focused on varying health outcomes the placebo or control group has consisted of wait-list participants not actively gardening (Bail et al., 2018; Davis et al., 2016; Demark-Wahnefried et al., 2018; Gascon et al., 2020), but in our study both the interven- tion and placebo group engaged in gardening using identical equipment, crop species, and instructions. This plausibly is a major advantage of the current study. The majority of the participants in our study reported being satisfied with the trial and that they plan to continue gardening, which might lead to long-lasting immunomodulatory effects. For most, information related to the role of beneficial bacterial exposure in immune regulation was new, but almost everyone was eager to learn more. The findings of this trial show how urban indoor gardening offers a space- and cost- efficient approach with which to increase beneficial microbial expo- sure at all stages of life, for example in kindergartens, schools, offices, and nursing homes. 5. Conclusion The current biodiversity intervention trial demonstrated for the first time that urban indoor gardening has the potential to diversify the microbiota on human skin and to increase anti-inflammatory cytokine levels in plasma. Our findings are in accordance with the hygiene and biodiversity hypotheses, which state that these shifts may ultimately lead to a lower risk or weaker symptoms of certain immune-mediated diseases, such as allergies. Furthermore, the experimental setting illus- trates how beneficial microbial exposure can be obtained indoors and year-round through an activity that is both meaningful and satisfying. Funding This work was supported by The Strategic Research Council (grant numbers 346136 and 346138); by EU Horizon 2020 (grant number 874864); by Business Finland (grant number 9204/31/2019); by Child and Nature Foundation (Lapsi ja Luonto Saatio, personal grant to Mika Saarenpaa); by Onni and Hilja Tuovinen Foundation (Onni ja Hilja Tuovisen Saatio, personal grant to Mika Saarenpaa); by Kone Founda- tion (Koneen Saatio, personal grant to Mika Saarenpaa); and by The Finnish Cultural Foundation, Paijat-Hame Regional Fund (Suomen Kulttuurirahasto, Paijat-Hameen Rahasto, personal grant to Mika Saarenpaa) . CRediT authorship contribution statement Mika Saarenpaa: Methodology, Investigation, Formal analysis, Data curation, Conceptualization, Project administration, Software, Valida- tion, Visualization, Writing – original draft, Writing – review & editing. Marja I. Roslund: Conceptualization, Methodology, Validation, Writing – review & editing. Noora Nurminen: Conceptualization, Formal analysis, Methodology, Validation. Riikka Puhakka: Concep- tualization, Methodology, Supervision, Validation, Writing – review & editing. Laura Kummola: Formal analysis, Methodology, Validation, Writing – review & editing. Olli H. Laitinen: Funding acquisition, Methodology, Resources, Writing – review & editing. Heikki Hyoty: Methodology, Resources, Validation. Aki Sinkkonen: Writing – review & editing, Validation, Supervision, Resources, Project administration, Conceptualization, Data curation, Funding acquisition, Methodology. Declaration of competing interest The authors declare the following financial interests/personal re- lationships which may be considered as potential competing interests: [Aki Sinkkonen reports financial support was provided by The Strategic Research Council of Finland. Olli Laitinen reports financial support was provided by The Strategic Research Council of Finland. Heikki Hyoty reports financial support was provided by Horizon Europe. Aki Sink- konen reports financial support was provided by Business Finland. Mika Saarenpaa reports financial support was provided by Kone Foundation. Mika Saarenpaa reports financial support was provided by Finnish Cultural Foundation. Aki Sinkkonen, Noora Nurminen, Olli Laitinen, Heikki Hyoty has patent #EP3551196 issued to EPO. Aki Sinkkonen, Noora Nurminen, Olli Laitinen, Heikki Hyoty has patent #US- 11786564-B2 issued to USPTO. Aki Sinkkonen, Noora Nurminen, Olli Laitinen, Heikki Hyoty has patent #US-11318173-B2 issued to USPTO. Aki Sinkkonen, Marja Roslund has patent #EP3589300 issued to EPO. M. Saarenpaa et al. Environment International 187 (2024) 108705 13 Aki Sinkkonen, Olli Laitinen and Heikki Hyoty are board members of Uute Scientific Ltd that provides solutions for immune modulation. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper]. Data availability Permission to use personal data can be applied from the ethics committee of the Pirkanmaa Hospital District according to local laws. Sequence data is publicly available. Acknowledgments We would like to thank all the ADELE and the BIWE (biwe.fi) team members for their support, volunteers for participating in the study, CSC–IT Center for Science, Finland, for computational resources, FIMM–Institute for Molecular Medicine Finland for sequencing services, and personnel (most importantly Pekka Keranen) at the Environmental Laboratory at the University of Helsinki for laboratory services. Appendix Table A1 Questionnaire data on additional soil contact and adherence to the trial. Participant Treatment Additional soil contact during trial Break from gardening during trial P1 Placebo No No P2 Placebo No No P3 Placebo No No P4 Placebo No No P5 Placebo No No P6 Placebo Played with kids outside One weekend trip P7 Placebo No No P8 Placebo No One day during a weekend P9 Placebo No No P10 Placebo No One day during a weekend P11 Placebo Installed a flower bed edging No P12 Placebo Planted indoor herbs A long weekend trip P13 Placebo Raked old leaves No I1 Intervention No No I2 Intervention No No I3 Intervention No No I4 Intervention No No I5 Intervention No No I6 Intervention Cut down a yard tree No I7 Intervention Raked old leaves No I8 Intervention No No I9 Intervention Repotted houseplants No I10 Intervention No No I11 Intervention No No I12 Intervention No One day during a weekend I13 Intervention Played with kids outside One weekend trip I14 Intervention No No I15 Intervention No No Table A2 Three main skin bacterial hub OTUs in the intervention and placebo group were identified based on their degree (number of edges). The hub OTUs were unique for both groups and time points, and none of them were among the hundred most abundant OTUs (relative abundances varied between 0.0012 % and 0.0392 %). Treatment Time point Hub OTU Phylum Class Order Family Genus Placebo 0 mo 1. Proteobacteria Alphaproteobacteria Rhizobiales Devosiaceae Devosia 2. Bacteroidetes Bacteroidia Sphingobacteriales Sphingobacteriaceae Mucilaginibacter 3. Bacteroidetes Bacteroidia Chitinophagales Chitinophagaceae Flavitalea 1 mo 1. Acidobacteria Acidobacteriia Acidobacteriales Acidobacteriaceae Unclassified 2. Verrucomicrobia Verrucomicrobiae Verrucomicrobiales Rubritaleaceae Luteolibacter 3. Proteobacteria Gammaproteobacteria Betaproteobacteriales Burkholderiaceae Polaromonas Intervention 0 mo 1. Actinobacteria Actinobacteria Propionibacteriales Propionibacteriaceae Unclassified 2. Proteobacteria Alphaproteobacteria Rhizobiales Hyphomicrobiaceae Filomicrobium 3. Proteobacteria Alphaproteobacteria Rhizobiales Rhizobiaceae Shinella 1 mo 1. Bacteroidetes Bacteroidia Sphingobacteriales Sphingobacteriaceae Mucilaginibacter 2. Proteobacteria Alphaproteobacteria Reyranellales Reyranellaceae Reyranella 3. Gemmatimonadetes Gemmatimonadetes Gemmatimonadales Gemmatimonadaceae Uncultured References Aerts, R., Honnay, O., Van Nieuwenhuyse, A., 2018. Biodiversity and human health: mechanisms and evidence of the positive health effects of diversity in nature and green spaces. Br. Med. Bull. 127 (1), 5–22. https://doi.org/10.1093/bmb/ldy021. Bail, J.R., Fruge, A.D., Cases, M.G., De Los Santos, J.F., Locher, J.L., Smith, K.P., Cantor, A.B., Cohen, H.J., Demark-Wahnefried, W., 2018. A home-based mentored vegetable gardening intervention demonstrates feasibility and improvements in physical activity and performance among breast cancer survivors. Cancer 124 (16), 3427–3435. https://doi.org/10.1002/cncr.31559. Bates, D., Machler, M., Bolker, B., Walker, S., 2015. Fitting linear mixed-effects models using lme4. J. Stat. Softw. 67, 1–48. https://doi.org/10.18637/jss.v067.i01. Beenakker, K.G.M., Westendorp, R.G.J., de Craen, A.J.M., Chen, S., Raz, Y., Ballieux, B.E. P.B., Nelissen, R.G.H.H., Later, A.F.L., Huizinga, T.W., Slagboom, P.E., Boomsma, D. I., Maier, A.B., 2019. Men have a stronger monocyte-derived cytokine production response upon stimulation with the gram-negative stimulus lipopolysaccharide than M. Saarenpaa et al. Environment International 187 (2024) 108705 14 women: a pooled analysis including 15 study populations. J. Innate Immun. 12, 142–153. https://doi.org/10.1159/000499840. Benjamini, Y., Hochberg, Y., 1995. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. Roy. Stat. Soc.: Ser. B (Methodol.) 57 (1), 289–300. https://doi.org/10.1111/j.2517-6161.1995.tb02031.x. Burmeister, A.R., Marriott, I., 2018. The interleukin-10 family of cytokines and their role in the CNS. Front. Cell. Neurosci. 12, 458. https://doi.org/10.3389/ fncel.2018.00458. Civitello, D.J., Cohen, J., Fatima, H., Halstead, N.T., Liriano, J., McMahon, T.A., Ortega, C.N., Sauer, E.L., Sehgal, T., Young, S., Rohr, J.R., 2015. Biodiversity inhibits parasites: Broad evidence for the dilution effect. Proc. Natl. Acad. Sci. 112 (28), 8667–8671. https://doi.org/10.1073/pnas.1506279112. Das, P., Ji, B., Kovatcheva-Datchary, P., Backhed, F., Nielsen, J., 2018. In vitro co- cultures of human gut bacterial species as predicted from co-occurrence network analysis. PLoS One 13 (3), e0195161. Davis, J.N., Martinez, L.C., Spruijt-Metz, D., Gatto, N.M., 2016. LA sprouts: A 12-week gardening, nutrition, and cooking randomized control trial improves determinants of dietary behaviors. J. Nutr. Educ. Behav. 48 (1), 2–11.e1. https://doi.org/10.1016/j. jneb.2015.08.009. Demark-Wahnefried, W., Cases, M.G., Cantor, A.B., Fruge, A.D., Smith, K.P., Locher, J., Cohen, H.J., Tsuruta, Y., Daniel, M., Kala, R., De Los Santos, J.F., 2018. Pilot randomized controlled trial of a home vegetable gardening intervention among older cancer survivors shows feasibility, satisfaction, and promise in improving vegetable and fruit consumption, reassurance of worth, and the trajectory of central adiposity. J. Acad. Nutr. Diet. 118 (4), 689–704. https://doi.org/10.1016/j.jand.2017.11.001. Dinarello, C.A., 2000. Proinflammatory cytokines. Chest 118 (2), 503–508. https://doi. org/10.1378/chest.118.2.503. Dockx, Y., Taubel, M., Bijnens, E.M., Witters, K., Valkonen, M., Jayaprakash, B., Hogervorst, J., Nawrot, T.S., Casas, L., 2022. Indoor green can modify the indoor dust microbial communities. Indoor Air 32 (3), e13011. Donovan, G.H., Gatziolis, D., Longley, I., Douwes, J., 2018. Vegetation diversity protects against childhood asthma: Results from a large New Zealand birth cohort. Nat. Plants 4 (6). https://doi.org/10.1038/s41477-018-0151-8. Ege, M.J., Mayer, M., Schwaiger, K., Mattes, J., Pershagen, G., van Hage, M., Scheynius, A., Bauer, J., von Mutius, E., 2012. Environmental bacteria and childhood asthma. Allergy 67 (12), 1565–1571. https://doi.org/10.1111/all.12028. Flores, G.E., Caporaso, J.G., Henley, J.B., Rideout, J.R., Domogala, D., Chase, J., Leff, J. W., Vazquez-Baeza, Y., Gonzalez, A., Knight, R., Dunn, R.R., Fierer, N., 2014. Temporal variability is a personalized feature of the human microbiome. Genome Biol 15, 531. https://doi.org/10.1186/s13059-014-0531-y. Fuentes-Tristan, S., Parra-Saldivar, R., Iqbal, H.M.N., Carrillo-Nieves, D., 2019. Bioinspired biomolecules: Mycosporine-like amino acids and scytonemin from Lyngbya sp. with UV-protection potentialities. J. Photochem. Photobiol. B Biol. 201, 111684 https://doi.org/10.1016/j.jphotobiol.2019.111684. Fyhrquist, N., Ruokolainen, L., Suomalainen, A., Lehtimaki, S., Veckman, V., Vendelin, J., Karisola, P., Lehto, M., Savinko, T., Jarva, H., Kosunen, T.U., Corander, J., Auvinen, P., Paulin, L., von Hertzen, L., Laatikainen, T., Makela, M., Haahtela, T., Greco, D., Alenius, H., 2014. Acinetobacter species in the skin microbiota protect against allergic sensitization and inflammation. J. Allergy Clin. Immunol. 134 (6), 1301–1309.e11. https://doi.org/10.1016/j.jaci.2014.07.059. Gascon, M., Harrall, K.K., Beavers, A.W., Glueck, D.H., Stanislawski, M.A., Alaimo, K., Villalobos, A., Hebert, J.R., Dexter, K., Li, K., Litt, J., 2020. Feasibility of collection and analysis of microbiome data in a longitudinal randomized trial of community gardening. Future Microbiol. 15 (8), 633–648. https://doi.org/10.2217/fmb-2019- 0195. Gibiino, G., Lopetuso, L.R., Scaldaferri, F., Rizzatti, G., Binda, C., Gasbarrini, A., 2018. Exploring Bacteroidetes: Metabolic key points and immunological tricks of our gut commensals. Dig. Liver Dis. 50 (7), 635–639. https://doi.org/10.1016/j. dld.2018.03.016. Griffith, D.M., Veech, J.A., Marsh, C.J., 2016. cooccur: probabilistic species co- occurrence analysis in R. J. Stat. Softw. 69, 1–17. https://doi.org/10.18637/jss. v069.c02. Gronroos, M., Parajuli, A., Laitinen, O.H., Roslund, M.I., Vari, H.K., Hyoty, H., Puhakka, R., Sinkkonen, A., 2019. Short-term direct contact with soil and plant materials leads to an immediate increase in diversity of skin microbiota. MicrobiologyOpen 8 (3), e00645. https://doi.org/10.1002/mbo3.645. Gupta, S., Hjelmsø, M.H., Lehtimaki, J., Li, X., Mortensen, M.S., Russel, J., Trivedi, U., Rasmussen, M.A., Stokholm, J., Bisgaard, H., Sørensen, S.J., 2020. Environmental shaping of the bacterial and fungal community in infant bed dust and correlations with the airway microbiota. Microbiome 8 (1), 115. https://doi.org/10.1186/ s40168-020-00895-w. Haahtela, T., 2019. A biodiversity hypothesis. Allergy 74 (8), 1445–1456. https://doi. org/10.1111/all.13763. Haahtela, T., Alenius, H., Lehtimaki, J., Sinkkonen, A., Fyhrquist, N., Hyoty, H., Ruokolainen, L., Makela, M.J., 2021. Immunological resilience and biodiversity for prevention of allergic diseases and asthma. Allergy 76 (12), 3613–3626. https://doi. org/10.1111/all.14895. Hanski, I., von Hertzen, L., Fyhrquist, N., Koskinen, K., Torppa, K., Laatikainen, T., Karisola, P., Auvinen, P., Paulin, L., Makela, M.J., Vartiainen, E., Kosunen, T.U., Alenius, H., Haahtela, T., 2012. Environmental biodiversity, human microbiota, and allergy are interrelated. Proc. Natl. Acad. Sci. 109 (21), 8334–8339. https://doi.org/ 10.1073/pnas.1205624109. Honkanen, J., Nieminen, J.K., Gao, R., Luopajarvi, K., Salo, H.M., Ilonen, J., Knip, M., Otonkoski, T., Vaarala, O., 2010. IL-17 immunity in human type 1 diabetes. J. Immunol. 185 (3), 1959–1967. https://doi.org/10.4049/jimmunol.1000788. Hui, N., Gronroos, M., Roslund, M., Parajuli, A., Vari, H.K., Soininen, L., Laitinen, O.H., Sinkkonen, A., 2019a. Diverse environmental microbiota as a tool to augment biodiversity in urban landscaping materials. Front. Microbiol. 10, 536. https://doi. org/10.3389/fmicb.2019.00536. Hui, N., Parajuli, A., Puhakka, R., Gronroos, M., Roslund, M.I., Vari, H.K., Selonen, V.A. O., Yan, G., Siter, N., Nurminen, N., Oikarinen, S., Laitinen, O.H., Rajaniemi, J., Hyoty, H., Sinkkonen, A., 2019b. Temporal variation in indoor transfer of dirt- associated environmental bacteria in agricultural and urban areas. Environ. Int. 132, 105069 https://doi.org/10.1016/j.envint.2019.105069. Huttenhower, C., Gevers, D., Knight, R., Abubucker, S., Badger, J. H., Chinwalla, A. T., Creasy, H. H., Earl, A. M., FitzGerald, M. G., Fulton, R. S., Giglio, M. G., Hallsworth- Pepin, K., Lobos, E. A., Madupu, R., Magrini, V., Martin, J. C., Mitreva, M., Muzny, D. M., Sodergren, E. J., … The Human Microbiome Project Consortium. (2012). Structure, function and diversity of the healthy human microbiome. Nature, 486 (7402), Article 7402. doi: 10.1038/nature11234. Kirjavainen, P.V., Karvonen, A.M., Adams, R.I., Taubel, M., Roponen, M., Tuoresmaki, P., Loss, G., Jayaprakash, B., Depner, M., Ege, M.J., Renz, H., Pfefferle, P.I., Schaub, B., Lauener, R., Hyvarinen, A., Knight, R., Heederik, D.J.J., von Mutius, E., Pekkanen, J., 2019. Farm-like indoor microbiota in non-farm homes protects children from asthma development. Nat. Med. 25 (7), 1089–1095. https://doi.org/ 10.1038/s41591-019-0469-4. Koberl, M., Dita, M., Martinuz, A., Staver, C., Berg, G., 2017. Members of gammaproteobacteria as indicator species of healthy banana plants on fusarium wilt- infested fields in central America. Sci. Rep. 7, 45318. https://doi.org/10.1038/ srep45318. Kohl, M. (2020). MKinfer: Inferential Statistics (0.6) [R]. http://www.stamats.de. Kozich, J.J., Westcott, S.L., Baxter, N.T., Highlander, S.K., Schloss, P.D., 2013. Development of a dual-index sequencing strategy and curation pipeline for analyzing amplicon sequence data on the miseq illumina sequencing platform. Appl. Environ. Microbiol. 79 (17), 5112–5120. https://doi.org/10.1128/AEM.01043-13. Kuwabara, T., Ishikawa, F., Kondo, M., Kakiuchi, T., 2017. The role of IL-17 and related cytokines in inflammatory autoimmune diseases. Mediators Inflamm. 2017, e3908061. Lehtimaki, J., Karkman, A., Laatikainen, T., Paalanen, L., von Hertzen, L., Haahtela, T., Hanski, I., Ruokolainen, L., 2017. Patterns in the skin microbiota differ in children and teenagers between rural and urban environments. Sci. Rep. 7 (1), Article 1. https://doi.org/10.1038/srep45651. Lehtimaki, J., Sinkko, H., Hielm-Bjorkman, A., Salmela, E., Tiira, K., Laatikainen, T., Makelainen, S., Kaukonen, M., Uusitalo, L., Hanski, I., Lohi, H., Ruokolainen, L., 2018. Skin microbiota and allergic symptoms associate with exposure to environmental microbes. Proc. Natl. Acad. Sci. 115 (19), 4897–4902. https://doi. org/10.1073/pnas.1719785115. Lerner, A., Jeremias, P., Matthias, T., 2016. The world incidence and prevalence of autoimmune diseases is increasing. International Journal of Celiac Disease 3 (4), 151–155. https://doi.org/10.12691/ijcd-3-4-8. Li, Z., Bai, X., Peng, T., Yi, X., Luo, L., Yang, J., Liu, J., Wang, Y., He, T., Wang, X., Zhu, H., Wang, H., Tao, K., Zheng, Z., Su, L., Hu, D. (2020). New Insights Into the Skin Microbial Communities and Skin Aging. Frontiers in Microbiology, 11. https://www. frontiersin.org/articles/10.3389/fmicb.2020.565549. Li, M.O., Wan, Y.Y., Sanjabi, S., Robertson, A.-K.-L., Flavell, R.A., 2006. Transforming growth factor-beta regulation of immune responses. Annu. Rev. Immunol. 24, 99–146. https://doi.org/10.1146/annurev.immunol.24.021605.090737. Liddicoat, C., Bi, P., Waycott, M., Glover, J., Lowe, A.J., Weinstein, P., 2018. Landscape biodiversity correlates with respiratory health in Australia. J. Environ. Manage. 206, 113–122. https://doi.org/10.1016/j.jenvman.2017.10.007. Mahnert, A., Moissl-Eichinger, C., Berg, G., 2015. Microbiome interplay: Plants alter microbial abundance and diversity within the built environment. Front. Microbiol. 6 https://doi.org/10.3389/fmicb.2015.00887. McKnight, D.T., Huerlimann, R., Bower, D.S., Schwarzkopf, L., Alford, R.A., Zenger, K.R., 2019. Methods for normalizing microbiome data: an ecological perspective. Methods Ecol. Evol. 10 (3), 389–400. https://doi.org/10.1111/2041-210X.13115. McMurdie, P.J., Holmes, S., 2013. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One 8 (4), e61217. Mehta, D.S., Wurster, A.L., Grusby, M.J., 2004. Biology of IL-21 and the IL-21 receptor. Immunol. Rev. 202 (1), 84–95. https://doi.org/10.1111/j.0105-2896.2004.00201.x. Mhuireach, G.A., Wilson, H., Johnson, B.R., 2021. Urban aerobiomes are influenced by season, vegetation, and individual site characteristics. Ecohealth 18 (3), 331–344. https://doi.org/10.1007/s10393-020-01493-w. Mhuireach, G.A., Van Den Wymelenberg, K.G., Langellotto, G.A., 2023. Garden soil bacteria transiently colonize gardeners’ skin after direct soil contact. Urban Agriculture & Regional Food Systems 8 (1), 1–22. https://doi.org/10.1002/ uar2.20035. Noverr, M.C., Huffnagle, G.B., 2005. The “microflora hypothesis” of allergic diseases. Clinical and Experimental Allergy: Journal of the British Society for Allergy and Clinical Immunology 35 (12), 1511–1520. https://doi.org/10.1111/j.1365- 2222.2005.02379.x. Nurminen, N., Lin, J., Gronroos, M., Puhakka, R., Kramna, L., Vari, H.K., Viskari, H., Oikarinen, S., Roslund, M., Parajuli, A., Tyni, I., Cinek, O., Laitinen, O., Hyoty, H., Sinkkonen, A., 2018. Nature-derived microbiota exposure as a novel immunomodulatory approach. Future Microbiol. 13, 737–744. https://doi.org/ 10.2217/fmb-2017-0286. Nurminen, N., Cerrone, D., Lehtonen, J., Parajuli, A., Roslund, M., Lonnrot, M., Ilonen, J., Toppari, J., Veijola, R., Knip, M., Rajaniemi, J., Laitinen, O.H., Sinkkonen, A., Hyoty, H., 2021. Land cover of early-life environment modulates the risk of type 1 diabetes. Diabetes Care 44 (7), 1506–1514. https://doi.org/10.2337/ dc20-1719. M. Saarenpaa et al. Environment International 187 (2024) 108705 15 Oksanen, J., Blanchet, F. G., Friendly, M., Kindt, R., Legendre, P., McGlinn, D., Minchin, P. R., O’Hara, R. B., Simpson, G. L., Solymos, P., Stevens, M. H. H., Szoecs, E., Wagner, H. (2019). Vegan: Community Ecology Package. R package version 2.5-6. (2.5- 6.) [Computer software]. https://CRAN.R-project.org/packageˆvegan. Opal, S.M., DePalo, V.A., 2000. Anti-Inflammatory Cytokines. Chest 117 (4), 1162–1172. https://doi.org/10.1378/chest.117.4.1162. Ottman, N., Ruokolainen, L., Suomalainen, A., Sinkko, H., Karisola, P., Lehtimaki, J., Lehto, M., Hanski, I., Alenius, H., Fyhrquist, N., 2019. Soil exposure modifies the gut microbiota and supports immune tolerance in a mouse model. J. Allergy Clin. Immunol. 143 (3), 1198–1206.e12. https://doi.org/10.1016/j.jaci.2018.06.024. Parajuli, A., Gronroos, M., Siter, N., Puhakka, R., Vari, H.K., Roslund, M.I., Jumpponen, A., Nurminen, N., Laitinen, O.H., Hyoty, H., Rajaniemi, J., Sinkkonen, A., 2018. Urbanization reduces transfer of diverse environmental microbiota indoors. Front. Microbiol. 9 https://doi.org/10.3389/fmicb.2018.00084. Parajuli, A., Hui, N., Puhakka, R., Oikarinen, S., Gronroos, M., Selonen, V.A.O., Siter, N., Kramna, L., Roslund, M.I., Vari, H.K., Nurminen, N., Honkanen, H., Hintikka, J., Sarkkinen, H., Romantschuk, M., Kauppi, M., Valve, R., Cinek, O., Laitinen, O.H., Sinkkonen, A., 2020. Yard vegetation is associated with gut microbiota composition. Sci. Total Environ. 713, 136707 https://doi.org/10.1016/j.scitotenv.2020.136707. Prud’homme, G.J., Piccirillo, C.A., 2000. The Inhibitory Effects of Transforming Growth Factor-Beta-1 (TGF-β1) in Autoimmune Diseases. J. Autoimmun. 14 (1), 23–42. https://doi.org/10.1006/jaut.1999.0339. Pruesse, E., Quast, C., Knittel, K., Fuchs, B.M., Ludwig, W., Peplies, J., Glockner, F.O., 2007. SILVA: A comprehensive online resource for quality checked and aligned ribosomal RNA sequence data compatible with ARB. Nucleic Acids Res. 35 (21), 7188–7196. https://doi.org/10.1093/nar/gkm864. R Core Team, 2020. R: A language and environment for statistical computing. [Computer software]. R Foundation for Statistical Computing. Relvas, M., Regueira-Iglesias, A., Balsa-Castro, C., Salazar, F., Pacheco, J.J., Cabral, C., Henriques, C., Tomas, I., 2021. Relationship between dental and periodontal health status and the salivary microbiome: bacterial diversity, co-occurrence networks and predictive models. Sci. Rep. 11 (1), Article 1. https://doi.org/10.1038/s41598-020- 79875-x. Ren, D., Madsen, J.S., Sørensen, S.J., Burmølle, M., 2015. High prevalence of biofilm synergy among bacterial soil isolates in cocultures indicates bacterial interspecific cooperation. ISME J. 9 (1), Article 1. https://doi.org/10.1038/ismej.2014.96. Rohr, J.R., Civitello, D.J., Halliday, F.W., Hudson, P.J., Lafferty, K.D., Wood, C.L., Mordecai, E.A., 2020. Towards common ground in the biodiversity–disease debate. Nat. Ecol. Evol. 4 (1), Article 1. https://doi.org/10.1038/s41559-019-1060-6. Rook, G.A.W., 2009. Review series on helminths, immune modulation and the hygiene hypothesis: the broader implications of the hygiene hypothesis. Immunology 126 (1), 3–11. https://doi.org/10.1111/j.1365-2567.2008.03007.x. Rook, G.A., Lowry, C.A., 2022. Evolution, biodiversity and a reassessment of the hygiene hypothesis. Springer. Rose-John, S., 2012. IL-6 trans-signaling via the soluble IL-6 receptor: importance for the pro-inflammatory activities of IL-6. Int. J. Biol. Sci. 8 (9), 1237–1247. https://doi. org/10.7150/ijbs.4989. Roslund, M.I., Puhakka, R., Gronroos, M., Nurminen, N., Oikarinen, S., Gazali, A.M., Cinek, O., Kramna, L., Siter, N., Vari, H.K., Soininen, L., Parajuli, A., Rajaniemi, J., Kinnunen, T., Laitinen, O.H., Hyoty, H., Sinkkonen, A., Adele research, group., 2020. Biodiversity intervention enhances immune regulation and health-associated commensal microbiota among daycare children. Sci. Adv. 6 (42), eaba2578. https:// doi.org/10.1126/sciadv.aba2578. Roslund, M.I., Puhakka, R., Nurminen, N., Oikarinen, S., Siter, N., Gronroos, M., Cinek, O., Kramna, L., Jumpponen, A., Laitinen, O.H., Rajaniemi, J., Hyoty, H., Sinkkonen, A., Cerrone, D., Gronroos, M., Hui, N., Makela, I., Nurminen, N., Oikarinen, S., Sinkkonen, A., 2021. Long-term biodiversity intervention shapes health-associated commensal microbiota among urban day-care children. Environ. Int. 157, 106811 https://doi.org/10.1016/j.envint.2021.106811. Roslund, M.I., Parajuli, A., Hui, N., Puhakka, R., Gronroos, M., Soininen, L., Nurminen, N., Oikarinen, S., Cinek, O., Kramna, L., Schroderus, A.-M., Laitinen, O. H., Kinnunen, T., Hyoty, H., Sinkkonen, A., 2022. A placebo-controlled double- blinded test of the biodiversity hypothesis of immune-mediated diseases: environmental microbial diversity elicits changes in cytokines and increase in T regulatory cells in young children. Ecotoxicol. Environ. Saf. 242, 113900 https://doi. org/10.1016/j.ecoenv.2022.113900. Roslund, M.I., Parajuli, A., Hui, N., Puhakka, R., Gronroos, M., Soininen, L., Nurminen, N., Oikarinen, S., Cinek, O., Kramna, L., Schroderus, A.-M., Laitinen, O. H., Kinnunen, T., Hyoty, H., Sinkkonen, A., 2023. Skin, gut, and sand metagenomic data on placebo-controlled sandbox biodiversity intervention study. Data Brief 47, 109003. https://doi.org/10.1016/j.dib.2023.109003. Ruokolainen, L., von Hertzen, L., Fyhrquist, N., Laatikainen, T., Lehtomaki, J., Auvinen, P., Karvonen, A.M., Hyvarinen, A., Tillmann, V., Niemela, O., Knip, M., Haahtela, T., Pekkanen, J., Hanski, I., 2015. Green areas around homes reduce atopic sensitization in children. Allergy 70 (2), 195–202. https://doi.org/10.1111/ all.12545. Ruokolainen, L., Lehtimaki, J., Karkman, A., Haahtela, T., von Hertzen, L., Fyhrquist, N., 2017. Holistic view on health: two protective layers of biodiversity. Ann. Zool. Fenn. 54 (1–4), 39–49. https://doi.org/10.5735/086.054.0106. Ruokolainen, L., Fyhrquist, N., Laatikainen, T., Auvinen, P., Fortino, V., Scala, G., Jousilahti, P., Karisola, P., Vendelin, J., Karkman, A., Markelova, O., Makela, M.J., Lehtimaki, S., Ndika, J., Ottman, N., Paalanen, L., Paulin, L., Vartiainen, E., von Hertzen, L., Alenius, H., 2020. Immune-microbiota interaction in Finnish and Russian Karelia young people with high and low allergy prevalence. Clin Exp Allergy 50 (10), 1148–1158. https://doi.org/10.1111/cea.13728. Saarenpaa, M., Roslund, M. I., Puhakka, R., Gronroos, M., Parajuli, A., Hui, N., Nurminen, N., Laitinen, O. H., Hyoty, H., Cinek, O., Sinkkonen, A., the ADELE Research Group., 2021. Do rural second homes shape commensal microbiota of urban dwellers? A pilot study among urban elderly in Finland. Int. J. Environ. Res. Public Health, 18(7), Article 7. doi: https://doi.org/10.3390/ijerph18073742. Schloss, P.D., Westcott, S.L., Ryabin, T., Hall, J.R., Hartmann, M., Hollister, E.B., Lesniewski, R.A., Oakley, B.B., Parks, D.H., Robinson, C.J., Sahl, J.W., Stres, B., Thallinger, G.G., Horn, D.J.V., Weber, C.F., 2009. Introducing mothur: open-source, platform-independent, community-supported software for describing and comparing microbial communities. Appl. Environ. Microbiol. 75 (23), 7537–7541. https://doi. org/10.1128/AEM.01541-09. Schloss, P.D., Westcott, S.L., 2011. Assessing and improving methods used in operational taxonomic unit-based approaches for 16S rRNA gene sequence analysis. Appl. Environ. Microbiol. 77 (10), 3219–3226. https://doi.org/10.1128/AEM.02810-10. Selway, C.A., Mills, J.G., Weinstein, P., Skelly, C., Yadav, S., Lowe, A., Breed, M.F., Weyrich, L.S., 2020. Transfer of environmental microbes to the skin and respiratory tract of humans after urban green space exposure. Environ. Int. 145, 106084 https:// doi.org/10.1016/j.envint.2020.106084. Shaffer, M., Lozupone, C., 2018. Prevalence and source of fecal and oral bacteria on infant, child, and adult hands. Msystems 3 (1), e00192–e00217. https://doi.org/ 10.1128/mSystems.00192-17. Shan, Y., Guo, J., Fan, W., Li, H., Wu, H., Song, Y., Jalleh, G., Wu, W., Zhang, G., 2020. Modern urbanization has reshaped the bacterial microbiome profiles of house dust in domestic environments. World Allergy Organ. J. 13 (8), 100452 https://doi.org/ 10.1016/j.waojou.2020.100452. Sobko, T., Liang, S., Cheng, W.H.G., Tun, H.M., 2020. Impact of outdoor nature-related activities on gut microbiota, fecal serotonin, and perceived stress in preschool children: the Play&Grow randomized controlled trial. Sci. Rep. 10, 21993. https:// doi.org/10.1038/s41598-020-78642-2. Soga, M., Gaston, K.J., Yamaura, Y., 2016. Gardening is beneficial for health: a meta- analysis. Prev. Med. Rep. 5, 92–99. https://doi.org/10.1016/j.pmedr.2016.11.007. Soininen, L., Roslund, M.I., Nurminen, N., Puhakka, R., Laitinen, O.H., Hyoty, H., Sinkkonen, A., 2022. Indoor green wall affects health-associated commensal skin microbiota and enhances immune regulation: a randomized trial among urban office workers. Sci. Rep. 12 (1) https://doi.org/10.1038/s41598-022-10432-4. Stein, M.M., Hrusch, C.L., Gozdz, J., Igartua, C., Pivniouk, V., Murray, S.E., Ledford, J.G., Marques dos Santos, M., Anderson, R.L., Metwali, N., Neilson, J.W., Maier, R.M., Gilbert, J.A., Holbreich, M., Thorne, P.S., Martinez, F.D., von Mutius, E., Vercelli, D., Ober, C., Sperling, A.I., 2016. Innate immunity and asthma risk in amish and hutterite farm children. N. Engl. J. Med. 375 (5), 411–421. https://doi.org/10.1056/ NEJMoa1508749. Stiemsma, L.T., Reynolds, L.A., Turvey, S.E., Finlay, B.B., 2015. The hygiene hypothesis: current perspectives and future therapies. Immunotargets and Therapy 4, 143–157. https://doi.org/10.2147/ITT.S61528. Thieurmel, B. (2021). VisNetwork: Network Visualization using “vis.js” Library. R package version 2.1.0. (2.1.0) [R]. Almende B.V. https://CRAN.R-project.org/ packageˆvisNetwork. Tischer, C., Kirjavainen, P., Matterne, U., Tempes, J., Willeke, K., Keil, T., Apfelbacher, C., Taubel, M., 2022. Interplay between natural environment, human microbiota and immune system: a scoping review of interventions and future perspectives towards allergy prevention. Sci. Total Environ. 821, 153422 https:// doi.org/10.1016/j.scitotenv.2022.153422. To, T., Stanojevic, S., Moores, G., Gershon, A.S., Bateman, E.D., Cruz, A.A., Boulet, L.-P., 2012. Global asthma prevalence in adults: findings from the cross-sectional world health survey. BMC Public Health 12 (1), 204. https://doi.org/10.1186/1471-2458- 12-204. Troy, E.B., Kasper, D.L., 2010. Beneficial effects of Bacteroides fragilis polysaccharides on the immune system. Front Biosci 15, 25–34. Turunen, A.W., Halonen, J., Korpela, K., Ojala, A., Pasanen, T., Siponen, T., Tiittanen, P., Tyrvainen, L., Yli-Tuomi, T., Lanki, T., 2023. Cross-sectional associations of different types of nature exposure with psychotropic, antihypertensive and asthma medication. Occup. Environ. Med. 80 (2), 111–118. https://doi.org/10.1136/ oemed-2022-108491. Valkonen, M., Wouters, I.M., Taubel, M., Rintala, H., Lenters, V., Vasara, R., Genuneit, J., Braun-Fahrlander, C., Piarroux, R., von Mutius, E., Heederik, D., Hyvarinen, A., 2015. Bacterial exposures and associations with atopy and asthma in children. PLoS One 10 (6), e0131594. Van Pee, T., Nawrot, T., van Leeuwen, R., Hogervorst, J., 2023. The gut microbiome and residential surrounding greenness: a systematic review of epidemiological evidence. Current Environmental Health Reports 10 (2), 137–153. https://doi.org/10.1007/ s40572-023-00398-4. von Hertzen, L., Haahtela, T., 2006. Disconnection of man and the soil: Reason for the asthma and atopy epidemic? J. Allergy Clin. Immunol. 117 (2), 334–344. https:// doi.org/10.1016/j.jaci.2005.11.013. von Hertzen, L.C., Joensuu, H., Haahtela, T., 2011. Microbial deprivation, inflammation and cancer. Cancer Metastasis Rev. 30 (2), 211–223. https://doi.org/10.1007/ s10555-011-9284-1. von Mutius, E., Radon, K., 2008. Living on a farm: Impact on asthma induction and clinical course. Immunol. Allergy Clin. North Am. 28 (3), 631–647. https://doi.org/ 10.1016/j.iac.2008.03.010. Wicaksono, W.A., Buko, A., Kusstatscher, P., Sinkkonen, A., Laitinen, O.H., Virtanen, S. M., Hyoty, H., Cernava, T., Berg, G., 2022. Modulation of the food microbiome by apple fruit processing. Food Microbiol. 108, 104103 https://doi.org/10.1016/j. fm.2022.104103. Wicaksono, W.A., Buko, A., Kusstatscher, P., Cernava, T., Sinkkonen, A., Laitinen, O.H., Virtanen, S.M., Hyoty, H., Berg, G., 2023. Impact of cultivation and origin on the M. Saarenpaa et al. Environment International 187 (2024) 108705 16 fruit microbiome of apples and blueberries and implications for the exposome. Microb. Ecol. 86 (2), 973–984. https://doi.org/10.1007/s00248-022-02157-8. Wickham, H., 2016. ggplot2: elegant graphics for data analysis. Springer-Verlag. https:// doi.org/10.1007/978-0-387-98141-3. Zhang, D. (2018). Package ‘rsq.’ R-Squared and Related Measures. Available Online: Https://Cran. r-Project. Org/Web/Packages/Rsq/Rsq. Pdf (Accessed on 8 September 2021). https://cloud.r-project.org/web/packages/rsq/rsq.pdf. Zhao, C., Liu, X., Tan, H., Yin, S., Su, L., Du, B., Khalid, M., Sinkkonen, A., Hui, N., 2024. Neighborhood Garden’s age shapes phyllosphere microbiota associated with respiratory diseases in cold seasons. Environmental Science and Ecotechnology 18, 100315. https://doi.org/10.1016/j.ese.2023.100315. M. Saarenpaa et al.