For years, neuroscientists have relied on brain scans to estimate what they call brain age, a metric that can diverge meaningfully from a person’s chronological age. The concept is simple but powerful: machine learning models are trained on large collections of brain imaging data to predict how old a brain looks based on its structural or functional features. When a person’s estimated brain age exceeds their actual age, that gap, often called a brain-predicted age difference, has been associated in prior research with poorer memory performance, diminished thinking skills and mood disturbances. Yet most of those earlier studies focused on older adults or individuals already living with neurological or psychiatric conditions, leaving open a fundamental question about whether this measure carries any meaning for younger, generally healthy people.

The UCLA team set out to answer that question with unusual rigor. The researchers analyzed brain scans from nearly 1,500 adults drawn from three separate study groups, applying a technique that measures how different regions of the brain communicate with one another while the brain is at rest. This resting-state functional connectivity approach captures the coordinated activity patterns that define large-scale brain networks, including those supporting memory and internally directed thought. From these communication patterns, the team built a computer model that estimated each participant’s brain age and then calculated the difference between that estimate and the person’s real age, a value they named the Brain Aging Index, or BAI.

The results were strikingly consistent across all three cohorts. People whose brains looked older than their years tended to score worse on tests of working memory and executive function, the collection of mental skills that govern focus, organization and self-control. They also reported more symptoms of depression. Notably, the patterns consistently converged on brain regions tied to memory and self-referential thought, the circuits most implicated in early cognitive decline and affective disorders. That the same relationships held across independent groups strengthens the case that the Brain Aging Index captures something biologically real rather than a statistical artifact of any single dataset.

But the study’s most provocative findings came from a subset of one study group in which researchers also analyzed stool samples, opening a window onto the gut microbiome, the vast community of microorganisms that inhabits the digestive tract. The team discovered that a higher Brain Aging Index was associated with specific gut bacteria and their metabolic byproducts, including certain fat molecules, a cholesterol-related compound and lower levels of a hormone called estetrol. These were not random associations. When the researchers traced the biological pathways implicated by these molecules, the signals pointed toward the immune system, blood vessel function, communication between brain cells and cellular energy production, all processes with established roles in brain health and neurodegeneration.

The implications reach into one of the most active frontiers of modern neuroscience: the gut-brain axis. Over the past decade, researchers have accumulated evidence that gut microbes can influence the brain through multiple channels, including the vagus nerve, immune signaling, circulating metabolites and effects on the vascular system. What this new work adds is a potential early-warning link. If the microbial and metabolic signature of an older-appearing brain can be detected in stool samples of young and middle-aged adults, it raises the possibility that inexpensive, noninvasive tests might one day flag individuals whose brains are aging faster than they should, well before memory tests or mood questionnaires would detect a problem.

Dr. Arpana Church, the study’s senior author and co-director of the Goodman-Luskin Microbiome Center at UCLA Health, framed the findings as evidence that brain aging is a long process rather than a late-life event. Brain aging does not suddenly begin when we get older, she noted; instead, the biological signals may be detectable decades earlier. By linking these early brain changes with the gut microbiome and its metabolites, she explained, researchers are beginning to identify pathways that could ultimately help clarify who may be at risk and, importantly, where intervention might support healthier brain aging. In her view, the results open the door to exploring whether targeting gut health could one day become a strategy for protecting the brain.

That preventive framing is where the study’s practical significance lies. Current approaches to cognitive decline are largely reactive, identifying problems only after symptoms appear, when the underlying neural damage may be difficult to reverse. If an index derived from routine brain imaging, or even a metabolic profile from a stool sample, could identify higher-risk individuals in their thirties, forties or fifties, interventions ranging from dietary changes to microbiome-targeted therapies could conceivably be deployed far earlier. Church emphasized that the findings could eventually help identify people at elevated risk for cognitive or mood-related decline earlier in life and point to the gut as a potential target for future interventions aimed at supporting brain health and prevention long before old age.

The researchers are careful to note that the study is cross-sectional, meaning it captures a snapshot in time rather than tracking individuals over decades. Association, in other words, is not causation: the data cannot yet tell us whether particular gut bacteria drive accelerated brain aging, whether an aging brain reshapes the gut ecosystem, or whether both are influenced by a third factor such as diet, inflammation or genetics. Longitudinal studies that follow participants over many years, ideally combining repeated brain scans with serial microbiome and metabolite sampling, will be needed to disentangle these possibilities. Interventional trials that test whether modifying the microbiome can shift the Brain Aging Index would represent the decisive next step.

Even so, the study marks a meaningful advance in how scientists think about the trajectory of brain health. It suggests that the seeds of cognitive and emotional decline may be sown in midlife, or earlier, and that the body’s own biology may be broadcasting warnings well before the mind notices. It also adds to the growing recognition that the gut is not merely a digestive organ but an endocrine and immunological hub in constant dialogue with the brain. As the tools for measuring that dialogue become more refined, the prospect of catching brain aging in its earliest stages, and perhaps slowing it through the gut, moves from speculative to genuinely testable. For a field long focused on treating the aftermath of decline, the message of this research is quietly revolutionary: look earlier, and look to the gut.

Subject of Research: Associations between brain aging measured by functional connectivity and gut microbiome metabolites in young and mid-life adults

Article Title: Brain aging may be detected decades before symptoms appear, with links to gut health

Article References: Brain aging may be detected decades before symptoms appear, with links to gut health. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: brain aging, gut microbiome, Brain Aging Index, UCLA Health, resting-state functional connectivity, working memory, executive function, depression, gut-brain axis, metabolites, eBioMedicine, cognitive decline

Transcranial magnetic stimulation, a non-invasive neuromodulation technique, uses magnetic pulses to influence neural activity in specific brain regions. Since receiving FDA approval in 2008 for major depressive disorder, TMS has grown in clinical utility, particularly for patients unresponsive to traditional pharmacological and psychotherapeutic interventions. Historically, determining the target site for TMS has relied on surface anatomical landmarks on the scalp, which serve as proxies for underlying brain structures. While pragmatically sound and easily accessible for widespread clinical use, this traditional approach lacks customization to the patient’s unique brain circuitry, potentially limiting therapeutic gains.

The innovation introduced by this study lies in leveraging functional magnetic resonance imaging (fMRI) to identify individualized treatment targets based on resting-state functional connectivity. This imaging modality measures synchronized activity patterns amongst disparate brain regions while subjects are at rest. By parsing these connectivity networks, the research team pinpointed precise loci within the brain circuits implicated in depression, thereby refining the spatial accuracy of stimulation. This neuroimaging foundation enables a more tailored intervention that accounts for the heterogeneity of depression at the circuit level.

Of particular note is the application of accelerated TMS (aTMS), which compresses multiple treatment sessions into a single day, thereby shortening the overall treatment course from several weeks to a mere week. This intensification not only improves patient convenience but may enhance neurobiological receptivity to stimulation by delivering more frequent pulses within a condensed timetable. The study set out to compare the clinical outcomes of aTMS when targets were defined by fMRI connectivity versus the established scalp-based targeting.

The trial enrolled 40 adult participants with moderate to severe treatment-resistant major depression, spanning a broad age range of 22 to 80 years. Each individual underwent pre-treatment fMRI scanning to delineate functional connectivity profiles. Subsequently, subjects were randomized to receive aTMS directed either at their individualized connectivity-based target or the conventional scalp-based target. Crucially, both patients and clinical raters were blinded to group assignments to mitigate bias.

One month post-treatment assessments revealed that the group receiving connectivity-guided aTMS demonstrated significantly greater alleviation of depressive symptoms compared to their scalp-based counterparts. These improvements were quantified using the Montgomery-Åsberg Depression Rating Scale (MADRS), a gold-standard clinician-administered instrument that sensitively captures changes in depression severity. Furthermore, the response rate — defined by clinically meaningful symptom reduction — was markedly higher in the connectivity group, with 80% responding versus 60% in the traditional targeting group, underscoring the potential clinical advantage of imaging-informed intervention.

This research builds upon prior explorations led by Joseph Taylor and colleagues, including investigations into imaging-based modulation of anxiety circuits within depressive populations, as recently reported in Molecular Psychiatry. These cumulative findings lend prospective support to the concept that precision neuroimaging can transcend theoretical neuroscience and play a direct role in augmenting therapeutic outcomes.

Taylor emphasizes the significance of closing the gap between neuroimaging research and tangible clinical benefit. Historically, the complexity and additional cost associated with imaging have created barriers to its routine clinical adoption for TMS guidance. This study represents a crucial step toward justifying such investment by empirically demonstrating a quantifiable benefit above conventional practice, a vital incentive for healthcare providers and payers considering integration of this technology.

Despite the promising results, the authors acknowledge particular study limitations, including the modest sample size and single-center study design, which may affect generalizability. They advocate for larger, multi-site trials to validate these early findings and explore durability of treatment effects over extended follow-up periods. Broadening the participant demographics will also be essential to ascertain the utility of connectivity-based targeting across diverse patient populations and varying clinical subtypes of depression.

This trial’s implications extend beyond depression, suggesting that functional brain imaging could inform individualized treatment strategies for a range of psychiatric disorders treatable by neuromodulation, including anxiety, obsessive-compulsive disorder, and post-traumatic stress disorder. As aTMS and neuroimaging technologies continue to evolve and become more accessible, the integration of connectivity-guided targeting holds promise for ushering in a new era of personalized psychiatry grounded in neurobiological precision.

In conclusion, the Mass General Brigham team’s randomized controlled trial provides compelling evidence that connectivity-based targeting via functional MRI can substantially enhance the antidepressant impact of accelerated TMS treatment in individuals with refractory depression. This approach offers a paradigm shift toward precision-guided neuromodulation, with the potential to improve patient outcomes and redefine clinical standards for brain stimulation therapies in psychiatric care.

Subject of Research: People

Article Title: Connectivity- versus scalp-based targeting of accelerated TMS for depression: A randomized trial

Web References:

• https://jamanetwork.com/journals/jamapsychiatry/fullarticle/10.1001/jamapsychiatry.2026.1100
• https://www.massgeneralbrigham.org/en/about/neuroscience-institute
• https://www.massgeneralbrigham.org/en/about/complex-psychiatric-care

References:
Taylor, J. et al. “Connectivity- versus scalp-based targeting of accelerated TMS for depression: A randomized trial,” JAMA Psychiatry, DOI: 10.1001/jamapsychiatry.2026.1100

Keywords:
Depression, Transcranial magnetic stimulation, Accelerated TMS, Functional magnetic resonance imaging, Functional connectivity, Neuroimaging-guided neuromodulation, Treatment-resistant depression, Personalized psychiatry, Montgomery-Åsberg Depression Rating Scale

Central to this study is the exploration of resting-state functional connectivity — a method that examines spontaneous brain activity when individuals are not engaged in explicit tasks. Resting-state paradigms have gained immense traction due to their ability to capture the brain’s default mode and intrinsic organization. In patients with mood disorders, disturbances in resting-state networks have been frequently reported, yet their stability and evolution over time in youth remain poorly understood. Here, the authors ambitiously track the same individuals across multiple time points to determine whether the neural substrates of mood dysregulation manifest enduring alterations or if they fluctuate in parallel with symptomatic states.

The research cohort comprises young participants diagnosed with bipolar-I or bipolar-II disorder, conditions marked by alternating episodes of mania or hypomania and depression, often with a complex clinical course. By focusing on symptomatic youths — rather than individuals in remission — the study targets the neural signature corresponding directly to mood instability, thus enhancing ecological validity. Longitudinal monitoring over months or years enabled the differentiation of trait-related neural alterations from transient state-dependent changes, a distinction critical for biomarker development and therapeutic targeting.

Employing functional magnetic resonance imaging (fMRI) as the primary data acquisition modality, the investigators meticulously assessed brain connectivity within canonical mood-related networks, including the default mode network (DMN), salience network (SN), and limbic circuits. Collectively, these systems orchestrate emotional regulation, attention, and reward processing—domains profoundly disrupted in bipolar disorder. The analytical framework incorporated network-based statistics and graph-theoretical models, allowing quantification of network topology, efficiency, and modularity with remarkable precision.

One of the pivotal findings was the demonstration of high test-retest reliability in key mood-related resting-state networks, suggesting that certain aberrations in connectivity are not merely epiphenomena of mood episodes but may represent stable neurobiological traits. Such traits could serve as enduring markers for diagnosis or risk stratification, potentially guiding personalized interventions. Intriguingly, some connectivity measures exhibited subtle modulation corresponding with clinical mood changes, highlighting the dynamic interplay between enduring network architecture and symptomatic expression.

The implications of these observations extend far beyond theoretical neuroscience. Clinically, bipolar disorder in youth poses significant challenges due to diagnostic complexity, heterogeneity, and the risk of poor long-term outcomes when treatment initiation is delayed. Objective neural markers that remain consistent over time could drastically enhance early diagnosis, monitor treatment response, and ultimately improve prognosis. This study’s longitudinal design underscores the feasibility and necessity of integrating repeated neuroimaging assessments in clinical research and practice.

Moreover, the investigation addresses a critical gap in psychiatric research: the underrepresentation of adolescent and young adult populations in longitudinal neuroimaging studies. Most prior research has focused on adult bipolar cohorts, where brain plasticity and illness trajectories differ markedly. By targeting youth with active symptoms, this study captures a developmental window pivotal for intervention, as brain networks are still maturing and may be more amenable to modulation.

Technical execution of the study reflects cutting-edge neurobiological research standards. Rigorous preprocessing steps controlled for potential confounds such as head motion, scanner drift, and physiological noise, thereby enhancing data integrity. Furthermore, the inclusion of multi-echo fMRI sequences improved signal-to-noise ratios, allowing detection of subtle changes within the resting-state networks. Statistical power was bolstered by including a sufficiently large sample and multiple scanning sessions, facilitating robust longitudinal inferences.

The study also explored correlations between network stability and clinical variables, such as symptom severity, medication status, and functional outcomes, although these analyses revealed complex relationships. For instance, certain disruptive patterns in connectivity were linked with greater mood lability and impaired psychosocial functioning, suggesting a neurobiological substrate for clinical heterogeneity. Nevertheless, pharmacological effects could not be entirely disentangled, underscoring the need for further research into medication influences on neural dynamics.

Significantly, these findings contribute to an expanding conceptual framework that views bipolar disorder not simply as an episodic illness but as a disorder with linked persistent network disruptions. This perspective aligns with emerging models of psychiatric conditions as network-based dysfunctions rather than isolated regional abnormalities. Understanding the stability of these neural networks may pave the way for neuromodulatory therapies, such as transcranial magnetic stimulation or neurofeedback, tailored to restore healthy connectivity patterns.

Looking ahead, the authors advocate for extending this line of inquiry by incorporating multimodal imaging techniques, such as diffusion tensor imaging (DTI) to map white matter integrity and electroencephalography (EEG) to capture rapid electrophysiological changes. Multidimensional datasets could unravel mechanistic pathways bridging structural and functional neural alterations in bipolar youth. Additionally, integrating genetic and environmental data may elucidate factors modulating network stability, enabling precision psychiatry.

In sum, this pioneering research advances the neurobiological understanding of bipolar disorder’s developmental trajectory by affirming that key mood-related brain networks demonstrate remarkable longitudinal stability in symptomatic youth. This challenges prevailing assumptions about transient neural disruptions during mood episodes and highlights the potential for trait-like brain network markers to transform clinical practice. The marriage of rigorous longitudinal fMRI methodology with a developmentally focused cohort sets a new standard for future investigations aiming to decode the complex neural fabric of mood disorders.

As bipolar disorder continues to impose significant public health burdens, particularly among young populations navigating critical life transitions, studies like this illuminate pathways toward improved diagnosis and intervention. Through the integration of neuroscience, psychiatry, and data science, the quest to decode mood-related brain networks offers hope for nuanced, biology-informed treatments that can mitigate suffering and foster resilience in youth.

The continuing exploration of resting-state connectivity dynamics exemplifies the power of neuroimaging to capture the brain’s spontaneous functional organization, which remains a frontier in psychiatric research. By anchoring future work on these foundational findings, the field moves closer to unraveling the intricacies of brain network stability and its disruption in major mood disorders. Such knowledge holds promise for ushering in an era of brain-guided precision psychiatry, tailored not only to diagnosis but also to individualized pathways of recovery.

As scientific inquiry refines our understanding of bipolar disorder’s neural underpinnings, collaborative cross-disciplinary efforts will be essential to translate these insights from bench to bedside. Harnessing longitudinal neural metrics may ultimately revolutionize therapeutic paradigms, offering young patients personalized strategies to sustain mood stability and live fulfilling lives.

Subject of Research: Longitudinal stability of mood-related resting-state brain networks in youth with symptomatic bipolar-I/II disorder.

Article Title: Longitudinal stability of mood-related resting-state networks in youth with symptomatic bipolar-I/II disorder.

Article References: Hafeman, D.M., Feldman, J., Mak, J. et al. Longitudinal stability of mood-related resting-state networks in youth with symptomatic bipolar-I/II disorder. Transl Psychiatry 15, 187 (2025). https://doi.org/10.1038/s41398-025-03404-5

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s41398-025-03404-5

Drawing on an unprecedented sample size of over 12,700 children and adolescents aged between 6 and 19 years, this comprehensive mega-analysis leveraged resting-state functional connectivity data to unravel the intricate neural tapestries linked to autism and ADHD. Resting-state connectivity refers to the spontaneous brain activity that occurs when a person is not engaged in any external task, reflecting intrinsic communication within and between large-scale brain networks. By examining the resting brain, researchers bypass potential confounds of task performance variability, honing in on the foundational neurophysiological architecture that may drive behavioral differences.

The research team undertook a multi-layered analytic approach. First, they explored dimensional traits characteristic of autism and ADHD across 10,168 participants to identify brain connectivity patterns associated with varying degrees of symptom severity. Subsequent analyses centered on diagnostic categories, involving over 2,500 participants with confirmed autism or ADHD diagnoses alongside neurotypical controls. This nested design enabled a rigorous dissection of connectivity alterations attributable to symptomatic traits as well as categorical diagnoses, enriching the precision of the findings.

One of the most striking discoveries was the divergent nature of thalamocortical and basal ganglia connectivity alterations linked to autism and ADHD. Specifically, autism traits and diagnostic status correlated with reduced functional connectivity between subcortical structures such as the thalamus and putamen and key cortical networks including the salience/ventral attention and frontoparietal control networks. These networks are crucial for detecting relevant stimuli and exerting executive control over behavior, respectively. In marked contrast, ADHD traits demonstrated an opposing pattern, characterized by increased connectivity in these same circuits. This bidirectional dysregulation suggests that while both disorders implicate shared neural substrates, the directionality of connectivity deviations is disorder-specific.

Moreover, the study revealed that both autism and ADHD groups exhibited hyperconnectivity between the default mode network (DMN) and the dorsal attention network (DAN) compared to neurotypical peers. These large-scale networks typically demonstrate antagonistic activity profiles; the DMN is more active during internally focused states such as mind-wandering, whereas the DAN is engaged during externally directed attention. Hyperconnectivity between these systems may reflect a breakdown in functional segregation, potentially underpinning difficulties with attentional shifting and cognitive flexibility observed in both disorders. Intriguingly, this overconnectivity was more robustly associated with ADHD trait severity, highlighting a nuanced interplay between network dynamics and behavioral manifestations.

Despite uncovering these neural signatures, the authors underscore that the observed effect sizes are modest, reflecting subtle alterations rather than gross disruptions in resting-state brain architecture. This finding aligns with an emerging consensus in neurodevelopmental research: that conditions like autism and ADHD involve complex, distributed, and finely tuned changes in brain connectivity rather than overt lesions or focal abnormalities. Such subtlety underscores the methodological imperative for large-scale datasets and sophisticated analytic frameworks to detect meaningful patterns amidst neural variability.

Technologically, this study capitalized on advances in neuroimaging acquisition harmonization and statistical mega-analytic techniques to integrate data from multiple cohorts and scanners, minimizing site-related confounds. This level of methodological rigor is critical to ensure that detected connectivity differences genuinely reflect neurodevelopmental variation rather than technical artifacts. Further, by parsing trait-level associations from diagnosis-based analyses, the researchers elegantly bridged dimensional and categorical frameworks, fostering a more nuanced understanding of neurodivergence.

Crucially, the differentiation of connectivity alterations linked to autism versus ADHD has direct implications for personalized medicine. Interventions tailored to specific network dysfunctions may enhance therapeutic efficacy. For example, modulating thalamocortical circuitry through neuromodulatory techniques like transcranial magnetic stimulation might yield differential benefits depending on a child’s diagnostic profile. Similarly, understanding the shared DMN-DAN hyperconnectivity could fuel novel cognitive training paradigms designed to improve attentional control across both disorders.

The research also invites a recalibration of conceptual models that emphasize the co-occurrence of autism and ADHD. Rather than conceiving of their overlap as mere symptom comorbidity, the emerging evidence supports a model wherein these conditions constitute distinct but interrelated neural phenotypes. This paradigm shift may help disentangle clinical presentations and diagnostic ambiguities common in pediatric psychiatry, guiding more precise assessments.

Additionally, the findings offer a springboard for exploring developmental trajectories. Resting-state connectivity patterns evolve throughout childhood and adolescence, paralleling cognitive and emotional maturation. Future longitudinal studies building on this mega-analytic framework could elucidate how the identified neural signatures emerge and transform over time, potentially revealing critical windows for intervention.

The study’s magnitude and methodological sophistication also make it a model for future neuropsychiatric research. The use of resting-state functional MRI, large-scale sample amalgamation, and nuanced trait-diagnosis analyses exemplifies best practices for disentangling complex brain-behavior relationships. This approach may prove invaluable for investigating other neurodevelopmental and psychiatric conditions characterized by overlapping phenotypes.

Nevertheless, several limitations warrant consideration. The resting-state paradigm, while powerful, cannot establish causal links between connectivity patterns and behavioral symptoms. The correlational nature of the analyses means that observed connectivity alterations might reflect downstream consequences or compensatory mechanisms rather than primary etiologies. Moreover, the relatively subtle effect sizes highlight that resting-state connectivity constitutes only one facet of the neurobiological landscape of autism and ADHD.

Extending these findings to real-world clinical practice will require additional translational research. Integrating multimodal data such as structural MRI, genetics, and behavioral indices could facilitate the construction of integrative models with enhanced predictive utility. Furthermore, understanding how environmental and developmental factors moderate connectivity patterns could refine individualized treatment approaches.

In summary, this landmark cross-sectional mega-analysis represents a major advance in unraveling the neurobiological complexity of autism and ADHD. By delineating distinct yet overlapping resting-state connectivity alterations in a vast pediatric sample, the research clarifies how these prevalent neurodevelopmental conditions diverge and converge at the neural circuit level. The subtle but consistent findings caution against simplistic categorical assumptions, instead advocating for a dimensional, network-based perspective of neurodivergence. As the field moves forward, these insights pave the way toward precision diagnostics and interventions tailored to the unique neural architectures underlying autism and ADHD, ultimately improving outcomes for millions of affected children and adolescents worldwide.

Subject of Research:
Functional brain connectivity alterations associated with autism spectrum disorder and attention-deficit/hyperactivity disorder traits and diagnoses in children and adolescents.

Article Title:
Cross-sectional mega-analysis of resting-state alterations associated with autism and attention-deficit/hyperactivity disorder in children and adolescents.

Article References:
Norman, L.J., Sudre, G., Bouyssi-Kobar, M. et al. Cross-sectional mega-analysis of resting-state alterations associated with autism and attention-deficit/hyperactivity disorder in children and adolescents. Nat. Mental Health (2025). https://doi.org/10.1038/s44220-025-00431-5

Image Credits:
AI Generated