The microbiome-gut-brain axis plays a crucial role in determining behavioral and psychological outcomes during childhood development, a period when mental health issues often first become apparent. The gut microbiome rapidly changes to its nearly adult state from early childhood to around three years of age. Signals from the early microbiome alter the architecture of developing brain circuits. This reflects in the child’s behavior and cognitive performance, including internalizing symptoms: emotions, and behavior directed towards oneself rather than outwardly, such as anxiety, depression and withdrawal.
The early microbiome appears to leave a lasting imprint on the child’s neurobiology, shaping how the brain responds to mental health stressors. This influence is particularly evident in middle school children, a known period of heightened vulnerability. For example, greater microbial diversity has been linked to stronger connectivity in the fronto-parietal network, which supports cognitive control and regulates negative emotions in infancy. Similarly, changes in specific bacterial taxa at one month have been shown to predict brain volumes at one year, suggesting a cascading developmental effect even in the absence of direct associations at a later state.
A new study published in Nature Communications applies machine learning to uncover links between brain network connectivity in six-year-old children, internalizing symptoms observed at 7.5 years, and their gut microbiome composition in early life. It aims to shed light on the links between the microbiome and mental health in this age group. Prior research demonstrated that specific types of microbes are linked to affective outcomes at one year. Increased overall diversity at one month predicted more fearful behavior at one year, but the opposite was true of taxa like Clostridiales and Dialister. This led to the current study, which aimed to help visualize “how early microbial signals may contribute to lifelong trajectories of mental health.
Such studies mainly assess seed-based resting-state functional connectivity (RSFC) where functional magnetic resonance imaging (fMRI) is used to explore functional activity at rest in selected areas of the brain (“seeds”). This limits the ability to view the whole brain function and excludes connectivity between the large brain networks. Yet such networks are crucial to neurocognitive behavior as well as mental ill-health in children. Changes in RFSC in some key pathways are associated with childhood internalizing symptoms. Mechanistically, this reflects changes in connectivity within and between networks, making them more specialized and efficient. Disruptions in this process are linked to mental illness risk in young people.
This study is essential since internalizing symptoms in early and middle childhood predict chronic and recurrent internalizing symptoms in later life. This small, exploratory observational study (n = 55) used data from the GUSTO study, where stool samples were collected at two years and fMRI scans were performed at six years. “Brain signatures” were examined for the highest covariance with internalizing symptoms at 7.5 years. Subsequently, scientists examined whether early gut microbial profiles predicted these brain signatures. The results showed two distinct brain signatures that showed the strongest associations with internalizing symptoms, such as anxiety and depression, at 7.5 years of age.
While no direct links emerged between gut microbiota profiles and these symptoms, one microbial profile was indirectly associated though a specific brain network signature. Notably, the brain signatures were not correlated with self-reported depressive symptoms at 8.5 years. The difference in internalizing symptoms was explained best by the connectivity signatures in two opposing brain networks linked to mental health and cognition. One network comprises functional connections associated with affective disorders in young people, although conflicting results have been reported. Impaired connectivity in this network could partly explain the onset of internalizing symptoms later.
The second pathway involves higher connectivity levels, and disruptions in this area are known to be associated with poor mental health at various stages of development. This includes depression and irritability, but also other mental symptoms across categories, poor emotional regulation and impaired cognitive control. The first brain signature involved the striatal–orbitofrontal–amygdalar (SOFA), medial temporal lobe (MTL), salience (SAL) and parietomedial (PMN) networks. In contrast, the second reflected connectivity between the SOFA and other networks, such as the default mode network (DMN), ventral attention network (VAN), and fronto-parietal network (FPN). These orthogonal network changes were explained best by three microbial abundance profiles at two years.
These are linked to inflammation, which could affect the corresponding brain signature, leading to internalizing symptoms. Notably, one microbial profile was associated with both externalizing and internalizing symptoms at 7.5 years. Microbes of the Clostridiales and Lachnospiraceae order and family, respectively, were linked to internalizing symptoms later in childhood. The mechanism appears to be mediated by changes in functional brain network connectivity, which are regulated by emotion. This agrees with earlier work on young adults with depression. These taxa are linked to inflammation in response to acute stress under laboratory conditions. Such changes often occur in early life adversity, indicating a vulnerable point in the response to such stressors.
The present study suggests that internalizing symptoms are associated with specific brain signatures linked to certain gut microbial abundance profiles. However, the authors emphasize that the findings are preliminary, based on a modest sample, and involve community children with mostly subclinical symptoms rather than diagnosed mental disorders. They caution that the results show association, not causation, and highlight the need for replication in larger and more diverse samples. These findings could help inform appropriate interventions in the future, while demonstrating the durable impact of early childhood microbiome composition on mental health outcomes in childhood.
- Edited by Dr. Gianfrancesco Cormaci, PhD, specialist in Clinical Biochemistry.
Scientific references
Querdasi FR et al. Nature Commun. 2025 Oct; 16(1):9359.
Esfand SM et al. Brain Behav Immun Health. 2024; 42:100880.
Querdasi FR et al. PNAS USA. 2023; 120(30):e2213768120.
