Translational Psychiatry www.nature.com/tp ARTICLE OPEN Gut microbiota signature in treatment-naïve attention-deficit/ hyperactivity disorder Vanesa Richarte1,2,3, Cristina Sánchez-Mora1,2,4, Montserrat Corrales1,2,3, Christian Fadeuilhe1,2,3, Laura Vilar-Ribó 1,2,4, Lorena Arribas1,4, Estela Garcia4, Silvia Karina Rosales-Ortiz4,5, Alejandro Arias-Vasquez6,7, María Soler-Artigas 1,2,4,8, ✉ ✉ Marta Ribasés 1,2,4,8 and Josep Antoni Ramos-Quiroga 1,2,3,4 © The Author(s) 2021 Compelling evidence supports alterations in gut microbial diversity, bacterial composition, and/or relative abundance of several bacterial taxa in attention-deficit/hyperactivity disorder (ADHD). However, findings for ADHD are inconsistent among studies, and specific gut microbiome signatures for the disorder remain unknown. Given that previous studies have mainly focused on the pediatric form of the disorder and involved small sample sizes, we conducted the largest study to date to compare the gastrointestinal microbiome composition in 100 medication-naïve adults with ADHD and 100 sex-matched healthy controls. We found evidence that ADHD subjects have differences in the relative abundance of several microbial taxa. At the family level, our data support a lower relative abundance of Gracilibacteraceae and higher levels of Selenomonadaceae and Veillonellaceae in adults with ADHD. In addition, the ADHD group showed higher levels of Dialister and Megamonas and lower abundance of Anaerotaenia and Gracilibacter at the genus level. All four selected genera explained 15% of the variance of ADHD, and this microbial signature achieved an overall sensitivity of 74% and a specificity of 71% for distinguishing between ADHD patients and healthy controls. We also tested whether the selected genera correlate with age, body mass index (BMI), or scores of the ADHD rating scale but found no evidence of correlation between genera relative abundance and any of the selected traits. These results are in line with recent studies supporting gut microbiome alterations in neurodevelopment disorders, but further studies are needed to elucidate the role of the gut microbiota on the ADHD across the lifespan and its contribution to the persistence of the disorder from childhood to adulthood. Translational Psychiatry (2021)11:382 ; https://doi.org/10.1038/s41398-021-01504-6 INTRODUCTION to be explained, these data also suggest that other mechanisms Attention-deficit/hyperactivity disorder (ADHD) is a neurodevelop- may provide a means for integrating the effects of genetic and mental disorder characterized by a persistent pattern of symp- environmental risk factors and explaining additional phenotypic toms of inattention, hyperactivity, and impulsivity, resulting in variance in ADHD. Among such factors, compelling evidence dysfunction in two or more areas of an individual’s life [1,2,]. ADHD supports a possible role for the gut microbiome in ADHD. is associated with deterioration in the social, family, academic, The gut microbiome is essential for health and plays a role in and/or occupational functioning of the affected subjects and has a the bidirectional regulation of the brain-gut axis. Microorganisms high impact at the socioeconomic level [3]. influence the brain through their ability to produce and modify The prevalence of ADHD in children is approximately 5.3%, and many metabolic, immunological, and neurochemical factors in the of these, 50−70% will still show symptoms in adulthood [4]. The gut that ultimately impact the central nervous system [8–10]; in etiology is complex and multifactorial, with an average heritability turn, brain activity also impacts the gut microbiota composition of 74% [5]. Through the largest meta-analyses of genome-wide [11,12,]. The gut microbiota influences gut barrier integrity and association studies performed so far, the first genome-wide produce neuroactive compounds such as neurotransmitters, significant loci for ADHD were identified [6,7,]. Evidence for a amino acids, and microbial metabolites, including short-chain strong genetic component of common variants in the polygenic fatty acids [10,13,]. These metabolites can interact with the host architecture of ADHD was found, with an SNP-based heritability of immune system, act on the central nervous system by regulating 22% [6]. Given that the large proportion of heritability still needs gene expression, epigenetics and neuroplasticity and affect local 1 Department of Psychiatry, Mental Health and Addictions, Hospital Universitari Vall d’Hebron, Barcelona, Catalonia, Spain. 2Biomedical Network Research Centre on Mental Health (CIBERSAM), Instituto de Salud Carlos III, Madrid, Spain. 3Department of Psychiatry and Legal Medicine, Universitat Autònoma de Barcelona, Barcelona, Catalonia, Spain. 4 Psychiatric Genetics Unit, Group of Psychiatry, Mental Health and Addiction, Vall d’Hebron Research Institute (VHIR), Universitat Autònoma de Barcelona, Barcelona, Catalonia, Spain. 5Department of Biomedicine, University of Barcelona, Barcelona, Spain. 6Department of Human Genetics, Donders Institute for Brain, Cognition and Behaviour, Radboud University Medical Center, 6525 GA Nijmegen, The Netherlands. 7Department of Psychiatry, Donders Institute for Brain, Cognition and Behaviour, Radboud University Medical Center, 6525 GA Nijmegen, The Netherlands. 8Department of Genetics, Microbiology & Statistics, University of Barcelona, Barcelona, Spain. ✉email: marta.ribases@vhir.org ; jaramos@vhebron.net Received: 4 February 2021 Accepted: 24 June 2021 V. Richarte et al. 2 neuronal cells and afferent pathways that signal directly to the Catalonia (Spain) in a specialized outpatient program for adult ADHD and brain [13]. This dynamic bidirectional communication between the by a single clinical group at Hospital Universitari Vall d’Hebron of Barcelona gut microbiota and the central nervous system influences brain (Spain). A description of the sample is provided in Supplementary Table 1. function, cognition, and behavior and highlights the fact that gut The clinical assessment consisted of structured interviews and self- microbiota imbalance may contribute to the pathophysiology of report questionnaires in two different steps: (i) ADHD diagnosis was based on the results of the Structured Diagnostic Interview for Adult ADHD (DIVA neurodevelopmental disorders and mental health outcomes. 2.0) [34] by a psychiatrist; (ii) the severity of ADHD symptoms and levels of Consistently, an increasing number of studies have shown gut impairment and comorbid disorders were assessed by a psychologist. In microbiome alterations in neurodevelopmental disorders [14–19]. this part of the evaluation, the following scales and questionnaires were For ADHD, an increasing number of studies have reported that the used: the ADHD Rating Scale (ADHD-RS), the Clinical Global Impression gut microbial diversity, bacterial composition, and/or relative (CGI), the Wender Utah Rating Scale (WURS), the Sheehan Disability abundance of several bacterial taxa differ between patients and Inventory (SDS), and the Structured Clinical Interview for DSM-IV Axis I and healthy controls [20–28]. Although not confirmed by others II Disorders (SCID-I and SCID-II). Afterward, the psychiatrist and psychol- [20,22,27,], some studies found differences in microbiota alpha ogist integrated the clinical information and self-reports for valid assessment of symptoms and impairments. In case of discordance [23,28,] or beta diversity [23,24,] in ADHD. For example, Prehn- between the different raters for ADHD symptoms or inconsistencies Kristensen et al. observed decreased alpha diversity in ADHD between the reporters in responses to items measuring similar symptoms, patients and differences in beta diversity between patients and clinician-identified symptoms on the DIVA 2.0 prevailed. Clinical informa- controls [23]. Wang et al. [28] also reported differences in alpha tion was reordered at the moment of inclusion, at which time the stool diversity in ADHD, and Szopinska-Tokov found a significant specimen was also collected. Exclusion criteria were as follows: an reduction in beta diversity in patients with ADHD [24,28,]. When intelligence quotient less than 70; lifelong or current history of mood, focusing on specific taxonomic groups, Aarts et al. reported a psychotic, anxiety, substance abuse, and personality disorders; pervasive nominal increase in Bifidobacterium in individuals with ADHD, developmental disorders; a history or the current presence of a condition changes that were associated with a significantly enhanced or illness, including neurologic, metabolic, cardiac, liver, kidney, or predicted synthesis of the dopamine precursor phenylalanine respiratory disease; a chronic medication of any kind; birth weight ≤ 1.5 kg; and other neurological or systemic disorders that might explain [22]. Similarly, Jiang et al reported decreased amounts of the ADHD symptoms. genera Dialister, Lachnoclostridium, Sutterella, and Faecalibacterium The control sample consisted of 100 unrelated healthy donors matched 1234567890();,: in treatment-naïve children with ADHD compared with healthy by sex and ethnicity with the clinical group. The exclusion criteria were controls and a negative association between the abundance of ADHD symptomatology according to the Adult Self-Report Scale A.S.R. S the last taxonomic group and parental reports of ADHD symptoms v1.1. and any prior or current psychiatric comorbidity. [27]. These results were consistent with those from a recent study All subjects reported European ancestry, which was confirmed through by Wan et al., which also detected a reduced relative abundance principal component analysis (PCA) using genetic data. Exclusion criteria of Faecalibacterium, as well as higher amounts of Odoribacter and for all participants included treatment with antibiotics or probiotics up to before stool collection. Enterococcus, in ADHD patients [20]. Moreover, Prehn-Kristensen The study was approved by the Clinical Research Ethics Committee observed distinct abundance in different microbial taxa, including (CREC) of Hospital Universitari Vall d’Hebron. All methods were performed increased Neisseria and decreased Prevotella and Parabacteroides in accordance with the relevant guidelines and regulations, and written in ADHD subjects [23]. Wang et al. compared the fecal microbiota informed consent was obtained from all subjects before inclusion. None of composition between medication-naïve children with ADHD and the participants received any financial compensation. healthy controls and found Fusobacterium genus as a marker for ADHD as well as enrichment of Lactobacillus in the control group Sample collection and DNA isolation [28]. Finally, a recent study conducted by Szopinska-Tokov et al. Human fecal samples were collected at home, stabilized with the revealed an association between the relative abundance of the OMNIgene·GUT (OM-200) (DNA Genotek Inc.) kit, and then transported Ruminococcaceae _UCG_004 genus and ADHD inattention symp- to the laboratory. The samples were aliquoted into 1.5-ml tubes and stored toms [24]. All these previous studies, however, considered small at −80 °C. Microbial DNA was purified from 200 mg of each homogenized sample sizes (from 14 to 51 ADHD patients), mainly focused on the fecal sample using the QIAamp® PowerFecal® DNA extraction kit (QIAgen, childhood/adolescent form of the disorder, and showed no Hilden, Germany). The isolated DNA was quantified using PicoGreen™ overlap or lack of concordance between findings. dsDNA Assay Kit [35]. Additionally, clinical evidence shows that probiotic intervention in early life may improve later outcomes and reduce the risk of Library preparation and Illumina sequencing neuropsychiatric disorders [29], and mice colonized by microbiota The V3−V4 hypervariable region of the bacterial 16S rRNA gene was from subjects with ADHD displayed altered microbial composition amplified for microbiome composition profiling. DNA library construction and behavioral and brain abnormalities compared with mice was performed following the manufacturer’s instructions (Illumina). We used the same workflow as described elsewhere [36] to perform cluster transplanted with the microbiota from individuals without ADHD generation, template hybridization, isothermal amplification, linearization, [30]. These data further support that the gut microbiome blocking and denaturation, and hybridization of the sequencing primers. composition may influence brain function and behavior and play Briefly, the V3−V4 region was amplified using key-tagged eubacterial a role in the disorder [30–33]. primers 5′CCTACGGGNGGCWGCAG3′ and 5′GACTACHVGGGTATC- Considering this background, we performed the largest TAATCC3′, and 300-nt paired-end amplicons were subsequently characterization of the gastrointestinal microbiome composition sequenced in two different rounds using the Illumina MiSeq platform. in 100 medication-naïve adults with ADHD and 100 sex-matched The raw Illumina paired-end reads were merged considering an overlap healthy controls and assessed differences in the microbiota length > 70 bp with the PEAR software v. 0.9.1, providing a single FASTQ composition between both groups and whether such differences file for each of the samples [37]. High-quality reads were extracted by were associated with ADHD clinical symptoms. applying a minimum Phred score of 20 (Q20, 99% based call accuracy). After primer sequences trimming, reads without both primer sequences or with less than 200 bp were discarded with Cutadapt v.1.8.1 [38]. Chimeric sequences were removed using the UCHIME software [39]. After quality MATERIALS AND METHODS control filtering, we obtained 14.7 million high-quality sequences with 45 Participants and clinical assessment 063−216 059 reads per sample from a total of 200 fecal samples. The raw The clinical sample consisted of 100 adult medication-naïve ADHD subjects and clean number of sequences, mean length, total mega bases (DSM-5 criteria) who were referred to an ADHD program from primary care sequenced, and mean quality per sample can be found in Supplementary centers and adult community mental health services. All subjects were Table 2. The remaining reads were clustered into operational taxonomic evaluated and recruited prospectively from a restricted geographic area of units (OTUs), in which unique sequences with a relative abundance above Translational Psychiatry (2021)11:382 V. Richarte et al. 3 0.1% were clustered into OTUs based on 97% sequence similarity [40] groups were observed in the CCA either (ADONIS P-value = 0.31; using the CD-HIT package [41] and the BLAST search against the NCBI 16S Supplementary Fig. 3). rRNA reference database (September 2019) with bastn v.2.10.0+. Compositional analysis of samples revealed that Bacteroidetes, Taxonomic groups (phylum, family, and genus) were assigned with a Firmicutes, Proteobacteria, Actinobacteria, and Verrucomicrobia Python script developed by ADM-BIOPOLIS (Paterna, Valencia, Spain). To were the most abundant phyla in our sample of 200 subjects remove genera with absent or low prevalence, the OTU table was filtered at the genus level. OTUs with nonzero values in less than 10% of the (Supplementary Table 3), with no significant differences in relative samples were removed. OUT counts were normalized by rarefaction with abundance detected for any of them. When we explored the the phyloseq R package according to Weiss et al. [42]. relative abundance of specific microbial taxa, however, we found evidence that several taxa differed significantly between ADHD cases and controls by two different methods, DeSeq2 and/or Statistical analysis Alpha diversity (within-sample diversity) was calculated on rarefied data random forests: 1 phylum, 7 families, and 17 genera showed with the Richness, Simpson and Shannon diversity indices and compared differential abundance (PFDR < 0,05; DESeq2: 1 phylum, 5 families, between individuals with ADHD and controls using the vegan R package and 15 genera; random forests: 5 families and 6 genera; Table 1, (https://github.com/vegandevs/vegan). Beta diversity (between-sample Fig. 1, Supplementary Table 4 and Supplementary Fig. 4). When diversity) was calculated by weighted and unweighted UniFrac and Bray combining the results of both methods, we found overlap for Curtis distances, as represented by two-dimensional principal coordinates three families (Gracilibacteraceae, Selenomonadaceae, and Veillo- analysis (PCoA) plots, and compared between groups by permutation nellaceae) and four genera (Anaerotaenia, Dialister, Gracilibacter, multivariate analysis of variance (PERMANOVA) using the phyloseq R and Megamonas) (Table 1 and Fig. 1). package [43]. The local contribution to beta diversity (LCBD) test was For downstream analysis, we focused on genera that differed in applied to evaluate the contribution of each sample to the diversity between the groups using the adespatial R package (https://github.com/ relative abundance between ADHD and controls with both of the sdray/adespatial). Canonical correspondence analysis (CCA), a multivariate methods described above (Anaerotaenia, Dialister, Gracilibacter, constrained ordination method, on rarefied OTUs was performed and and Megamonas). When we assessed whether they correlated with significance regarding the microbial community composition between each other, we found a moderate correlation between Anaero- groups was assessed by permutational multivariate analysis of variance taenia and Gracilibacter (r = 0.35; P-value = 3.6e−04), a weak (ADONIS) using the vegan R package (https://github.com/vegandevs/ correlation between Anaerotaenia and Megamonas (r = −0.24; P- vegan). value = 0.018), and no correlation between the others (Fig. 2). A Differential abundance comparisons between groups were assessed in model including the four genera and the covariates age, sex, and taxonomic groups showing an average of normalized counts (baseMean) > BMI explained 15% of the variance in ADHD, with significant 10 using the DESeq2 and randomForest R packages for the classification, rfUtilities to estimate the significance of the classification and rfPermute to improvement of the model which included only the covariates (P- evaluate the significance of specific taxa, with 1000 permutations. All value = 8.2e−07), which explained 5.9% of the variance (Supple- comparisons were performed at the phylum, family, and genus levels. Any mentary Table 5). The microbial signature achieved an overall unknown taxonomic level was assigned to the next highest known sensitivity of 74% and a specificity of 71% for the detection of taxonomic rank. individuals with ADHD versus healthy controls. We also assessed Genera showing significant differences in relative abundance between whether the selected genera correlated with age, BMI, or ADHD ADHD cases and controls after multiple comparison corrections in DeSeq2 rating scale scores but found no evidence of correlation between and the random forest comparisons were considered for downstream relative abundance and any of the selected traits (Fig. 2). analyses. Multiple logistic regression models were applied to test the association between ADHD and all selected genera while adjusting for age, sex, and body mass index (BMI). Adjusted Pseudo-R2 was calculated with the McFaddenAdj method and the DescTools R package (https://github. DISCUSSION com/AndriSignorell/DescTools); sensitivity and specificity were calculated To clarify the relationship between ADHD and the gut micro- with the caret R package (https://github.com/topepo/caret/). A likelihood biome, we performed the largest study to date and compared the ratio test with the lmtest R package (https://cran.r-project.org/web/ microbial composition between 100 medication-naïve adults with packages/lmtest/) was employed to assess whether the inclusion of ADHD and 100 sex-matched unrelated healthy subjects. We found selected genera in the multiple logistic regression model fits the data evidence that ADHD subjects exhibit differences in the relative significantly better than the model including only age, sex, and BMI. In the abundance of several microbial taxa. At the family level, our data first model, we considered affectation status as dependent variable and support a lower relative abundance of Gracilibacteraceae and age, sex and BMI as independent variables (ADHD 〜 age + sex + BMI); in the second model, we included selected taxa as independent variables higher levels of Selenomonadaceae and Veillonellaceae in adults (ADHD 〜 age + sex + BMI + Megamonas + Anaerotaenia + Gracilibacter with ADHD. In addition, the ADHD group showed higher levels of + Dialister). Spearman correlation tests were used to assess correlations Dialister and Megamonas and lower abundances of Anaerotaenia between selected genera, age, BMI, and inattention and hyperactive/ and Gracilibacter at the genus level. impulsivity subscale scores or total scores of the ADHD rating scale. These results are in line with recent studies supporting gut microbiome differences in neurodevelopmental disorders. Although the mechanistic explanation for these associations RESULTS remains unknown, a positive correlation between Dialister Bacterial composition based on 16S rRNA sequencing was abundance and activity level has been described in toddlers available for 100 adult ADHD cases and 100 controls. No [44]. Additionally, decreased levels of Dialister were found in differences in intestinal microbial alpha diversity (microbial autism spectrum disorder (ASD) patients [45,46,] or in treatment- community richness and evenness) were found between ADHD naïve children with ADHD [27] compared with healthy controls cases and controls when measured by three different indices and in ADHD individuals on medication compared with non- (Richness, Simpson, or Shannon indices; Supplementary Fig. 1). medicated individuals [24]. Furthermore, multiple taxonomic Beta diversity (between-sample community dissimilarity) accord- groups that differed in relative abundance between ADHD cases ing to weighted and unweighted UniFrac distances as well as the and controls in the present study, including Selenomonadaceae, Bray-Curtis dissimilarity index showed no differences in the Veillonellaceae, and Megamonas, have previously been associated microbial composition between the groups (PERMANOVA P- with other psychiatric conditions that often coexist with ADHD, value>0.05), with no evidence of separate clustering in PCoA such as ASD or depression [27, 47–51]. Given that the ADHD representations (Supplementary Fig. 2). No significant differences subjects in this study displayed no comorbid psychiatric disorders, in the gut microbiota composition between the ADHD and control we cannot discount a possible pleotropic effect of these Translational Psychiatry (2021)11:382 V. Richarte et al. 4 Table 1. Summary of differential abundance results between ADHD patients and controls considering Deseq2 and random forest results. Relative abundance (% mean (SD) Adjusted P-value ADHD Controls DEseq2 Random forests Phylum Candidatus Melainabacteria 0.072 (0.24) 0.22 (0.76) 3.1E−03 0.11 Family Eubacteriaceae 2.105 (1.51) 2.269 (1.35) 0.81 0.02 Gracilibacteraceae 0,503 (0.85) 0,949 (1.49) 0.035 0.05 Lactobacillaceae 0.965 (1.52) 1.077 (1.24) 0.93 0.02 Peptostreptococcaceae 0,327 (0.55) 0,199 (0.23) 0.016 0.27 Selenomonadaceae 0,387 (1.14) 0,071 (0.26) 3.5E−07 0.05 Veillonellaceae 1,658 (1.90) 0,837 (1.43) 0.012 9.9E−03 Verrucomicrobiaceae 0,036 (0.11) 0,063 (0.17) 0.012 0.73 Genus Acetivibrio 0.021 (0.05) 0.056 (0.17) 6.1E−03 0.099 Alloprevotella 0.380 (1.63) 0.182 (0.97) 4.4E−04 0.21 Anaerotaenia 0.072 (0.13) 0.248 (0.49) 2.3E−09 9.9E−03 Dialister 1.377 (1.76) 0.649 (1.26) 0.041 0.02 Flintibacter 1.967 (1.46) 1.588 (1.37) 0.26 0.045 Fucophilus 0.036 (0.11) 0.064 (0.17) 0.012 0.42 Gracilibacter 0.509 (0.86) 0.958 (1.50) 0.040 9.9E−03 Herbinix 0.024 (0.05) 0.042 (0.08) 0.024 0.24 Leclercia 0.084 (0.42) 0.025 (0.12) 9.8E−03 0.30 Megamonas 0.323 (1.04) 0.029 (0.20) 3.2E−29 9.9E−03 Megasphaera 0.209 (0.72) 0.091 (0.42) 7.5E−20 0.80 Odoribacter 0.547 (0.34) 0.751 (0.83) 0.039 0.14 Parasutterella 0.751 (1.30) 1.588 (1.37) 0.70 9.9E−03 Porphyromonas 0.129 (0.52) 0.110 (0.55) 6.1E−03 0.36 Prevotellamassilia 0.356 (1.82) 0.340 (1.69) 6.4E−15 0.31 Romboutsia 0.228 (0.52) 0.126 (0.16) 9.8E−03 0.93 Vampirovibrio 0.073 (0.24) 0.225 (0.77) 2.6E−03 0.38 Differentially abundant taxa identified by both methods, DEseq2 and random forests, are shown in bold. taxonomic groups and that their relative abundance may explain, interpreted with caution given that we selected genera of interest in part, ADHD phenotypic variability. and estimated the variance in ADHD explained by these taxa as Although previous gut microbiome analyses on ADHD have well as the sensitivity and specificity of the regression model using mainly focused on pediatric samples [20,23, 26–28] and there is the same dataset, which may have led to overfitting and further limited research on adults [22,24,], we focused our study on support the use of independent datasets to obtain more accurate adulthood ADHD. Nevertheless, given that the gut microbiome estimates. evolves throughout the lifespan [16,52,53,], whether early-life Microbiome composition is strongly influenced by environ- exposure to environmental risk factors contributes to the gut mental factors such as diet, overall health status, and medication microbiota and impacts neurodevelopment and mental health use [52, 54–56]. The participants in this study were not on outcomes later in life remain to be investigated. Further medication and had not used antibiotics or probiotics in the three longitudinal studies are warranted to provide additional informa- months before sample collection, which may not explain the tion on the role of the microbiome in ADHD symptom trajectories differences detected between ADHD cases and controls. None- from childhood to adulthood as well as mental health outcomes theless, no other environmental exposures, including smoking, and comorbid profiles across the lifespan. stress, dietary habits, or other lifestyle information, that may have We did not detect substantial changes in alpha or beta diversity an effect on microbiota composition were considered. For between ADHD cases and controls. The high heterogeneity in instance, animal models and population-based cross-sectional terms of age, sample size, sex, clinical characteristics, and type of studies support an effect of nicotine or smoking status on the gut controls may explain nonreplicable results and discrepancies microbiome composition and the fecal metabolome [57–59]. In between studies. We sex-matched ADHD cases and controls and addition to environmental factors, consistent evidence suggests restricted the clinical sample to ADHD medication-naïve adult that the host genetic background impacts the composition of gut subjects, which is a major strength of our study design that may microbial communities and that genetic factors influence micro- allow us to identify an imbalance in the gut microbiome biome composition and explain a significant proportion of the composition that might be neglected by broader study designs. variation in the gut microbiome [60–63]. Hence, further integrative In addition, the sample sizes of previous studies on ADHD, were studies considering multiple data sources (i.e., larger sample sizes), relatively small; although our study may also have limited including environmental factors, human genetic variation, and gut statistical power to estimate the magnitude of the differences microbial composition, are warranted to provide deeper insight identified in microbial relative abundance, we assessed the largest into the mechanisms underlying the relationship between the sample size considered thus far. The results, however, need to be microbiota, host genetics, and individual habits, and behavior, as Translational Psychiatry (2021)11:382 V. Richarte et al. 5 well as their roles in ADHD and other neurodevelopmental disorders across the lifespan. DATA AVAILABILITY The datasets supporting the conclusions of this article are included and available online. Raw fastq data will be available upon request to the corresponding author. REFERENCES 1. Faraone, S. V. et al. Attention-deficit/hyperactivity disorder. Nat Rev Dis Prim. 2015;1:15020. https://doi.org/10.1038/nrdp.2015.20. 2. Franke B, Michelini G, Asherson P, Banaschewski T, Bilbow A, Buitelaar JK. et al. Live fast, die young? A review on the developmental trajectories of ADHD across the lifespan. Eur. Neuropsychopharmacol.2018;28:1059–88. https://doi.org/ 10.1016/j.euroneuro.2018.08.001. 3. Thapar A, Cooper M. Attention deficit hyperactivity disorder. Lancet. 2016;387:1240–50. https://doi.org/10.1016/S0140-6736(15)00238-X. 4. Polanczyk G, de Lima MS, Horta BL, Biederman J, Rohde LA. The worldwide prevalence of ADHD: a systematic review and metaregression analysis. Am. J. Psychiatry. 2007;164:942–948. https://doi.org/10.1176/ajp.2007.164.6.942. 5. Faraone, S. V. & Larsson, H. Genetics of attention deficit hyperactivity disorder. Mol Psychiatry. 2019;24:562−575. https://doi.org/10.1038/s41380-018-0070-0. 6. Demontis D, Walters RK, Martin J, Mattheisen M, Als TD, Agerbo E. et al. Discovery of the first genome-wide significant risk loci for attention deficit/hyperactivity disorder. Nat Genet.2019;51:63–75. https://doi.org/10.1038/s41588-018-0269-7. 7. Rovira P, Demontis D, Sánchez-Mora C, Zayats T, Klein M, Mota NR. et al. Shared genetic background between children and adults with attention deficit/hyper- activity disorder. Neuropsychopharmacology. 2020;45:1617–26. https://doi.org/ 10.1038/s41386-020-0664-5. 8. Erny D, Hrabe de Angelis AL, Jaitin D, Wieghofer P, Staszewski O, David E. et al. Host microbiota constantly control maturation and function of microglia in the CNS HHS Public Access Author manuscript. Nat Neurosci.2015;18:965–77. https:// doi.org/10.1038/nn.4030. 9. Clarke G, Grenham S, Scully P, Fitzgerald P, Moloney RD, Shanahan F. et al. The Fig. 1 Differentially abundant taxa between ADHD cases and microbiome-gut-brain axis during early life regulates the hippocampal ser- controls. (A) Differential abundance results according to two otonergic system in a sex-dependent manner. Mol Psychiatry. 2013;18:666–73. different methods, DESeq2 andrandom forests. B Differences in https://doi.org/10.1038/mp.2012.77. relative abundance between ADHD cases and controls for taxo- 10. Morais LH, Schreiber HL, Mazmanian SK. The gut microbiota–brain axis in nomic groups surpassing multiple comparison corrections in behaviour and brain disorders. Nat Rev Microbiol.2020;19:241–55. https://doi.org/ DeSeq2 and/or random forest analyses. 10.1038/s41579-020-00460-0. 11. Kuwahara, A. et al. Microbiota-gut-brain axis: enteroendocrine cells and the enteric nervous system form an interface between the microbiota and the central nervous system. Biomed Res. 2020;41:199−216. 12. Mohajeri MH, La Fata G, Steinert RE, Weber P. Relationship between the gut microbiome and brain function. Nutr Rev.2018;76:481–96. https://doi.org/ 10.1093/nutrit/nuy009. 13. Dalile B, Van Oudenhove L, Vervliet B, Verbeke K. The role of short-chain fatty acids in microbiota–gut–brain communication. Nat Rev Gastroenterol Hepatol.2019;16:461–78. https://doi.org/10.1038/s41575-019-0157-3. 14. Dam SA, Mostert JC, Szopinska-Tokov JW, Bloemendaal M, Amato M, Arias- Vasquez A. The role of the gut-brain axis in attention-deficit/hyperactivity dis- order. Gastroenterol Clin North Am.2019;48:407–31. https://doi.org/10.1016/j. gtc.2019.05.001. 15. Ma Q, Xing C, Long W, Wang HY, Liu Q, Wang RF. Impact of microbiota on central nervous system and neurological diseases: the gut-brain axis. J. Neuroinflammation.2019;16:1–14. https://doi.org/10.1186/s12974-019-1434-3. 16. Borre YE, O'Keeffe GW, Clarke G, Stanton C, Dinan TG, Cryan JF. Microbiota and neurodevelopmental windows: implications for brain disorders. Trends Mol. Med.2014;20:509–18. https://doi.org/10.1016/j.molmed.2014.05.002. 17. Jurek, L. et al. Is there a dysbiosis in individuals with a neurodevelopmental disorder compared to controls over the course of development? A systematic review. Eur Child Adolesc Psychiatry. 2020. https://doi.org/10.1007/s00787-020- 01544-1. 18. Rogers GB, Keating DJ, Young RL, Wong ML, Licinio J, Wesselingh S. From gut Fig. 2 Spearman correlation between the relative abundance of dysbiosis to altered brain function and mental illness: mechanisms and pathways. four bacterial genera (Anaerotaenia, Dialister, Gracilibacter, and Mol. Psychiatry. 2016;21:738–48. https://doi.org/10.1038/mp.2016.50. Megamonas) and age, BMI, and ADHD rating scale scores. Colored 19. Lacorte, E. et al. A systematic review of the microbiome in children with neu- correlations are statistically significant (P-value < 0.05), with positive rodevelopmental disorders. Front Neurol. 2019;10:727 https://doi.org/10.3389/ and negative correlations in blue and red, respectively. Inattention: fneur.2019.00727. score of the inattention subscale of the ADHD rating scale; 20. Wan, L. et al. Case-control study of the effects of gut microbiota composition on hyperactivity_impulsivity: score of the hyperactive/impulsivity sub- neurotransmitter metabolic pathways in children with attention deficit hyper- scale of the ADHD rating scale; total: total scores of the ADHD activity disorder. Front Neurosci. 2020;14:127. https://doi.org/10.3389/ rating scale. fnins.2020.00127. Translational Psychiatry (2021)11:382 V. Richarte et al. 6 21. Boonchooduang N, Louthrenoo O, Chattipakorn N, Chattipakorn SC. Possible 43. Mcmurdie, P. J. & Holmes S. Phyloseq: an R package for reproducible interactive links between gut–microbiota and attention-deficit/hyperactivity disorders in analysis and graphics of microbiome ceus data. PLoS ONE. 8(4):e61217. https:// children and adolescents. Eur J Nutr.2020;59:3391–403. https://doi.org/10.1007/ doi.org/10.1371/journal.pone.0061217. s00394-020-02383-1. 44. Christian LM, Galley JD, Hade EM, Schoppe-Sullivan S, Kamp Dush C, Bailey MT. Gut 22. Aarts E, Ederveen T, Naaijen J, Zwiers MP, Boekhorst J, Timmerman HM. et al. Gut microbiome composition is associated with temperament during early childhood. microbiome in ADHD and its relation to neural reward anticipation. PLoS One. Brain Behav Immun.2015;45:118–27. https://doi.org/10.1016/j.bbi.2014.10.018. 2017;12:e0183509 https://doi.org/10.1371/journal.pone.0183509. 45. Strati F, Cavalieri D, Albanese D, De Felice C, Donati C, Hayek J. et al. New 23. Prehn-Kristensen A, Zimmermann A, Tittmann L, Lieb W, Schreiber S, Baving L. evidences on the altered gut microbiota in autism spectrum disorders. Micro- et al. Reduced microbiome alpha diversity in young patients with ADHD. PLoS biome. 2017;5:1–11. https://doi.org/10.1186/s40168-017-0242-1. One. 2018;13:e0200728 https://doi.org/10.1371/journal.pone.0200728. 46. Finegold SM, Dowd SE, Gontcharova V, Liu C, Henley KE, Wolcott RD. et al. Pyr- 24. Szopinska-Tokov J, Dam S, Naaijen J, Konstanti P, Rommelse N, Belzer C. et al. osequencing study of fecal microflora of autistic and control children. Anaerobe. Investigating the gut microbiota composition of individuals with attention-defi- 2010;16:444–53. https://doi.org/10.1016/j.anaerobe.2010.06.008. cit/hyperactivity disorder and association with symptoms. Microorganisms. 47. Kelly JR, Minuto C, Cryan JF, Clarke G, Dinan TG. Cross talk: the microbiota and 2020;8:406 https://doi.org/10.3390/microorganisms8030406. neurodevelopmental disorders. Front Neurosci.2017;11:490 https://doi.org/ 25. Checa-Ros A, Jer A, Molina-Carballo A, Campoy C, Muñoz-Hoyos A. Current evi- 10.3389/fnins.2017.00490. dence on the role of the gut microbiome in ADHD: pathophysiology and ther- 48. Liu, S. et al. Altered gut microbiota and short chain fatty acids in Chinese children apeutic implications. Nutrients. 2021;13:1–32. https://doi.org/10.3390/ with autism spectrum disorder. Sci Rep. 2019;9:287. https://doi.org/10.1038/ nu13010249. s41598-018-36430-z. 26. Sukmajaya A. C., Lusida, M. I., Soetjipto & Setiawati, Y. Systematic review of gut 49. Jiang H, Ling Z, Zhang Y, Mao H, Ma Z, Yin Y. et al. Altered fecal microbiota microbiota and attention-deficit hyperactivity disorder (ADHD). Ann Gen Psy- composition in patients with major depressive disorder. Brain Behav chiatry. 2021;20:12 https://doi.org/10.1186/s12991-021-00330-w. Immun.2015;48:186–94. https://doi.org/10.1016/j.bbi.2015.03.016. 27. Jiang HY, Zhou YY, Zhou GL, Li YC, Yuan J, Li XH. et al. Gut microbiota profiles in 50. Zheng P, Zeng B, Zhou C, Liu M, Fang Z, Xu X. et al. Gut microbiome remodeling treatment-naïve children with attention deficit hyperactivity disorder. Behav induces depressive-like behaviors through a pathway mediated by the host’s Brain Res.2018;347:408–13. https://doi.org/10.1016/j.bbr.2018.03.036. metabolism. Mol Psychiatry. 2016;21:786–96. https://doi.org/10.1038/mp.2016.44. 28. Wang LJ, Yang CY, Chou WJ, Lee MJ, Chou MC, Kuo HC. et al. Gut microbiota and 51. Zou R, Xu F, Wang Y, Duan M, Guo M, Zhang Q. et al. Changes in the gut dietary patterns in children with attention-deficit/hyperactivity disorder. Eur Child microbiota of children with autism spectrum disorder. Autism Adolesc Psychiatry. 2020;29:287–97. https://doi.org/10.1007/s00787-019-01352-2. Res.2020;13:1614–25. https://doi.org/10.1002/aur.2358. 29. Pärtty A, Kalliomäki M, Wacklin P, Salminen S, Isolauri E. A possible link between 52. Hasan N, Yang H. Factors affecting the composition of the gut microbiota, and its early probiotic intervention and the risk of neuropsychiatric disorders later in modulation. PeerJ. 2019;2019:1–31. https://doi.org/10.7717/peerj.7502. childhood: a randomized trial. Pediatr Res.2015;77:823–8. https://doi.org/10.1038/ 53. O’Toole PW. Changes in the intestinal microbiota from adulthood through to old age. pr.2015.51. Clin Microbiol Infect.2012;18:44–46. https://doi.org/10.1111/j.1469-0691.2012.03867.x. 30. Tengeler AC, Dam SA, Wiesmann M, Naaijen J, van Bodegom M, Belzer C. et al. 54. Flandroy L, Poutahidis T, Berg G, Clarke G, Dao MC, Decaestecker E. et al. The Gut microbiota from persons with attention-deficit/hyperactivity disorder affects impact of human activities and lifestyles on the interlinked microbiota and health the brain in mice. Microbiome. 2020;8:44 https://doi.org/10.1186/s40168-020- of humans and of ecosystems. Sci Total Environ.2018;627:1018–38. https://doi. 00816-x. org/10.1016/j.scitotenv.2018.01.288. 31. De Theije CG, Wopereis H, Ramadan M, van Eijndthoven T, Lambert J, Knol J. et al. 55. Falony G, Joossens M, Vieira-Silva S, Wang J, Darzi Y, Faust K. et al. Population- Altered gut microbiota and activity in a murine model of autism spectrum dis- level analysis of gut microbiome variation. Science. 2016;352:560–4. https://doi. orders. Brain Behav Immun.2014;37:197–206. https://doi.org/10.1016/j. org/10.1126/science.aad3503. bbi.2013.12.005. 56. Zhernakova A, Kurilshikov A, Bonder MJ, Tigchelaar EF, Schirmer M, Vatanen T. 32. Sharon G, Cruz NJ, Kang DW, Gandal MJ, Wang B, Kim YM. et al. Human gut et al. Population-based metagenomics analysis reveals markers for gut micro- microbiota from autism spectrum disorder promote behavioral symptoms in biome composition and diversity. Science. 2016;352:565–9. https://doi.org/ mice. Cell . 2019;177:1600–e17. https://doi.org/10.1016/j.cell.2019.05.004. 10.1126/science.aad3369. 33. Hsiao EY, McBride SW, Hsien S, Sharon G, Hyde ER, McCue T. et al. Microbiota 57. Chi L, Mahbub R, Gao B, Bian X, Tu P, Ru H. et al. Nicotine alters the gut modulate behavioral and physiological abnormalities associated with neurode- microbiome and metabolites of gut-brain interactions in a sex-specific manner. velopmental disorders. Cell. 2013;155:1451–63. https://doi.org/10.1016/j. Chem Res Toxicol.2017;30:2110–9. https://doi.org/10.1021/acs. cell.2013.11.024. chemrestox.7b00162. 34. Ramos-Quiroga, J. A. et al. Criteria and concurrent validity of DIVA 2.0: a semi- 58. Lee SH, Yun Y, Kim SJ, Lee EJ, Chang Y, Ryu S. et al. Clinical medicine association structured diagnostic interview for adult ADHD. J Atten Disord. 2016;23:1−10. between cigarette smoking status and composition of gut microbiota: https://doi.org/10.1177/1087054716646451. population-based cross-sectional study. J Clin Med.2018;7:282 https://doi.org/ 35. Mardis E, McCombie WR. Library quantification using picogreen fluorometry. Cold 10.3390/jcm7090282. Spring Harb Protoc.2017;2017:432–5. https://doi.org/10.1101/pdb.prot094722. 59. Biedermann L, Zeitz J, Mwinyi J, Sutter-Minder E, Rehman A, Ott SJ. et al. Smoking 36. Klindworth A, Pruesse E, Schweer T, Peplies J, Quast C, Horn M. et al. Evaluation of cessation induces profound changes in the composition of the intestinal general 16S ribosomal RNA gene PCR primers for classical and next-generation microbiota in humans. PLoS One. 2013;8:e59260 https://doi.org/10.1371/journal. sequencing-based diversity studies. Nucleic Acids Res.2013;41:1 https://doi.org/ pone.0059260. 10.1093/nar/gks808. 60. Gilbert JA, Blaser MJ, Caporaso JG, Jansson JK, Lynch SV, Knight R. Current 37. Zhang J, Kobert K, Flouri T, Stamatakis A. Genome analysis PEAR: a fast and understanding of the human microbiome. Nat Med.2018;24:392–400. https://doi. accurate illumina paired-end reAd mergeR. Bioinformatics.2014;30:614–20. org/10.1038/nm.4517. https://doi.org/10.1093/bioinformatics/btt593. 61. Sommer F, Bäckhed F. The gut microbiota-masters of host development and 38. Martin M. Cutadapt removes adapter sequences from high-throughput sequen- physiology. Nat Rev Microbiol.2013;11:227–38. https://doi.org/10.1038/nrmicro2974. cing reads. EMBnet J. 2011;17:10–12. 62. Awany, D. et al. Host and microbiome genome-wide association studies: current 39. Edgar RC, Haas BJ, Clemente JC, Quince C, Knight R. UCHIME improves sensitivity state and challenges. Front Genet. 2019;10:637. https://doi.org/10.3389/ and speed of chimera detection. Bioinformatics. 2011;27:2194–2200. https://doi. fgene.2018.00637. org/10.1093/bioinformatics/btr381. 63. Hughes DA, Bacigalupe R, Wang J, Rühlemann MC, Tito RY, Falony G. et al. 40. Goebel BM, Stackebrandt E. Cultural and phylogenetic analysis of mixed micro- Genome-wide associations of human gut microbiome variation and implications bial populations found in natural and commercial bioleaching environments. for causal inference analyses. Nat Microbiol.2020;5:1079–87. https://doi.org/ Appl Environ Microbiol.1994;60:1614–21. https://doi.org/10.1128/aem.60.5.1614- 10.1038/s41564-020-0743-8. 1621.1994. 41. Li W, Godzik A. Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences. Bioinformatics. 2006;22:1658–9. https://doi.org/ 10.1093/bioinformatics/btl158. ACKNOWLEDGEMENTS 42. Weiss, S. et al. Normalization and microbial differential abundance strategies Over the course of this investigation C.S.M. was a recipient of a Sara Borrell contract depend upon data characteristics. Microbiome. 2017;5:27. https://doi.org/ from the Instituto de Salud Carlos III, Ministerio de Economía, Industria y 10.1186/s40168-017-0237-y. Competitividad, Spain (CD15/00199), M.R. was a recipient of a Miguel de Servet Translational Psychiatry (2021)11:382 V. Richarte et al. 7 contract from the Instituto de Salud Carlos III, Spain (CP09/00119 and CPII15/00023), companies in the last 5 years: EliLilly, Lundbeck, Janssen- Cilag, Actelion, Shire, and M.S.A. was a recipient of a contract from the Biomedical Network Research Ferrer, Oryzon, Roche, Psious, and Rubió. Center on Mental Health (CIBERSAM), Madrid, Spain, and a recipient of a Juan de la Cierva Incorporación contract from the Ministry of Science, Innovation and Universities, Spain (IJC2018-035346-I). The research leading to these results received ADDITIONAL INFORMATION funding from the European Union H2020 Program (H2020/2014-20) under grant Supplementary information The online version contains supplementary material agreements no. 667302 (CoCA) and no. 728018 (Eat2beNICE), from the Instituto de available at https://doi.org/10.1038/s41398-021-01504-6. Salud Carlos III (PI15/01789, PI16/01505, PI17/0289, PI18/01788, PI19/00721,P19/ 01224 and PI20/00041), from the Pla estratègic de recerca i innovació en salut (PERIS); Correspondence and requests for materials should be addressed to M.R. or J.A.R-Q. Generalitat de Catalunya (METAL-Cat; SLT006/17/287) and from the Agència de Gestió d’Ajuts Universitaris i de Recerca-AGAUR, Generalitat de Catalunya Reprints and permission information is available at http://www.nature.com/ (2017SGR1461) and cofinanced by the European Regional Development Fund (ERDF) reprints and by “la Marató de TV3” (092330/31). The work was also supported by the ECNP Network ‘ADHD across the Lifespan’ (https://www.ecnp.eu/researchinnovation/ECNP- Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims networks/List-ECNP-Networks/). in published maps and institutional affiliations. COMPETING INTERESTS Dra. Sánchez-Mora, Dra. Soler Artigas, Estela García, and Dra. Ribasés report no biomedical financial interests or potential conflicts of interest. Vanesa Richarte has Open Access This article is licensed under a Creative Commons served as a speaker for Rubió and Shire/Takeda in the last 5 years. She has received Attribution 4.0 International License, which permits use, sharing, travel awards from Shire/Takeda for participating in psychiatric meetings. The ADHD adaptation, distribution and reproduction in any medium or format, as long as you give Program received unrestricted educational and research support from Eli Lilly and appropriate credit to the original author(s) and the source, provide a link to the Creative Co., Janssen-Cilag, Shire/Takeda, Rovi, Psious, and Laboratorios Rubió in the past two Commons license, and indicate if changes were made. The images or other third party years. Dra. Corrales received travel awards from Shire for participating in psychiatric material in this article are included in the article’s Creative Commons license, unless meetings. Christian Fadeuilhe received travel awards from Rubió, Shire/Takeda, and indicated otherwise in a credit line to the material. If material is not included in the Lundbeck for participating in psychiatric meetings. Prof. Ramos-Quiroga was on the article’s Creative Commons license and your intended use is not permitted by statutory speakers’ bureau and/or acted as a consultant for Eli-Lilly, Janssen-Cilag, Novartis, regulation or exceeds the permitted use, you will need to obtain permission directly Shire, Takeda, Bial, Shionogui, Lundbeck, Almirall, Braingaze, Sincrolab, Medice, and from the copyright holder. To view a copy of this license, visit http://creativecommons. Rubió in the last 5 years. He also received travel awards (air tickets + hotel) from org/licenses/by/4.0/. Janssen-Cilag, Rubió, Shire, Takeda, Shionogui, Bial, Medice, and Eli-Lilly for participating in psychiatric meetings. The Department of Psychiatry chaired by him received unrestricted educational and research support from the following © The Author(s) 2021 Translational Psychiatry (2021)11:382