📄 psychiatry and clinical psychology

Shared host-genetic architecture between gut microbiota and internalizing psychopathology

This study utilizes Mendelian randomization and genetic modeling to demonstrate that specific gut microbial taxa and internalizing psychopathologies (such as depression and neuroticism) share underlying host genetic architecture rather than exhibiting a direct causal relationship.

Original authors: Velez-Pardo, P., Solano, R. J., Quinchia-Figueroa, A. M., Montoya Monsalve, R., Moratto-Vasquez, N. S.

Published 2026-07-15

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Velez-Pardo, P., Solano, R. J., Quinchia-Figueroa, A. M., Montoya Monsalve, R., Moratto-Vasquez, N. S.

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

Imagine your gut is a bustling, noisy city filled with trillions of tiny residents (bacteria). For years, scientists have wondered if these residents are the bosses running the show in your brain, or if they are just fans cheering in the stands, reacting to whatever the brain is doing.

A new study by researchers Pedro Velez-Pardo and his team decided to settle this debate using a clever trick called "Mendelian Randomization." Think of this as using your DNA as a natural, unchangeable lottery ticket. Since your genes are set at birth (long before you get depressed or anxious), they act like a time-traveling referee. If a specific bacterial group is truly causing a mental health issue, the genes that control that bacteria should also predict the mental health issue.

The Big Discovery: The "Internalizing" Neighborhood
The team checked 211 different bacterial groups against nine different mental health conditions, ranging from depression and anxiety to schizophrenia and ADHD. They ran nearly 1,900 tests.

Here is the plot twist: The bacteria didn't seem to be the bosses of all mental health issues. Instead, the signals they found were like a spotlight shining only on one specific neighborhood of the brain: the "Internalizing" district. This district includes depression, neuroticism (worrying a lot), and insomnia.

The Good Guys: They found that a group called Mollicutes/Tenericutes seemed to act like a shield against depression. Another group, Butyrivibrio, looked like a protective friend against neuroticism.
The Trouble-Makers: Conversely, a group called Betaproteobacteria seemed to be linked to higher levels of neuroticism.

Crucially, no bacteria showed a strong, statistically significant link to the "Psychotic" neighborhood (schizophrenia) or the "Externalizing" neighborhood (like ADHD) in the same way. The signals were strictly focused on the internalizing side.

The Plot Twist: It's Not a Direct Causal Chain
This is where the story gets really interesting. You might expect the researchers to say, "Aha! We proved that Butyrivibrio prevents depression!" But they didn't. In fact, they argued against that idea.

When they put the findings through a super-strict "causality filter" (using tools like Bayesian colocalization and CAUSE), the results changed.

The Verdict: The study found zero pairs where a specific bacterium was the direct, sole cause of a disorder.
The Real Explanation: Instead of bacteria causing the illness, the study suggests that you and your bacteria share the same genetic blueprint. It's like finding that a family of musicians and a family of painters often live in the same house. It's not that the music causes the painting; rather, the house (your host genetics) supports both.

The authors found that the genes controlling these specific gut bacteria are the same genes that influence your risk for depression and anxiety. They are "co-conspirators" in your DNA, not a cause-and-effect chain where the bacteria pull the strings.

How Sure Are We?
The researchers are very careful with their language. They didn't "solve" the mystery of mental illness, nor did they prove that eating more of these bacteria will cure depression.

What is measured: They measured genetic correlations and found a strong pattern where 33 out of 45 tested pairs moved in the same direction (protective or harmful) as the hypothesis predicted.
What is ruled out: They explicitly ruled out the idea that these bacteria are the direct cause of the disorders at the level of individual genetic spots.
What is suggested: They suggest that the "internalizing" part of mental health (worry, sadness, sleeplessness) shares a genetic foundation with specific gut microbes.

The Takeaway
Think of your gut and your brain not as a boss and an employee, but as two instruments in the same orchestra, both playing from the same sheet of music (your DNA). The study found that for the "Internalizing" section of the orchestra, the gut bacteria and the brain are playing in harmony because they share the same genetic score.

While this doesn't mean we can pop a probiotic pill to fix depression tomorrow, it gives scientists a new map. It tells them to stop looking for a single "bad bacteria" to blame and start looking at how our own genes shape both our gut and our mood together. The full map of all 211 bacteria and nine disorders is now open for other scientists to explore, ensuring no one misses the quiet signals hidden in the data.

Technical Summary: Shared Host-Genetic Architecture Between Gut Microbiota and Internalizing Psychopathology

Problem Statement
The causal relationship between gut microbial composition and mental illness remains unresolved. While observational studies report altered microbial profiles in conditions such as major depressive disorder (MDD), bipolar disorder (BD), and schizophrenia, these designs cannot distinguish between causation, reverse causation, and confounding factors (e.g., medication, lifestyle). Previous Mendelian randomization (MR) studies have faced three primary limitations: (1) a focus on single outcomes rather than shared latent psychiatric architecture; (2) reliance on GWAS datasets that have since been superseded by larger releases; and (3) a lack of integration between univariable MR, multivariable cross-disorder analyses, and conservative causal tests (e.g., Bayesian colocalization, CAUSE) to rule out horizontal pleiotropy.

Methodology
The authors conducted a two-sample MR study utilizing the largest available genome-wide association studies (GWAS) as of 2026. The study employed a ten-layer analytical framework applied uniformly to 211 gut microbial taxa (instruments derived from the MiBioGen consortium, $N=18,340$ ) and nine psychiatric/psychopathology-related outcomes (MDD, anxiety, BD, insomnia, neuroticism, ADHD, PTSD, schizophrenia, OCD).

Key methodological components included:

Instrument Selection: SNPs at $p < 1 \times 10^{-5}$ were LD-clumped ( $r^2 < 0.001$ ) to create instruments for 211 taxa.
Univariable MR: Primary analysis using inverse-variance weighted (IVW) random-effects models, supplemented by MR-Egger, weighted median, MR-PRESSO, and MR-RAPS to assess pleiotropy and heterogeneity.
Conservative Validation: Application of Bayesian colocalization (to test for shared causal variants), CAUSE (to distinguish causation from correlated pleiotropy), and multivariable MR (restricted to taxon-only instruments to avoid standard-error inflation).
Latent-Factor Modeling: Genomic Structural Equation Modeling (Genomic SEM) was used to fit genetic covariance matrices. The authors tested one-factor, two-factor, and bifactor models. Notably, they re-assigned ADHD from a pre-registered "psychotic" factor to an "internalizing/distress" factor based on genetic correlation data to resolve an improper solution (Heywood case).
Mediation and Functional Annotation: Two-step mediation analysis through tryptophan, kynurenine, CRP, and IL-6, alongside MAGMA gene-set enrichment and exploratory cis-eQTL prioritization.

Key Results

Discovery Signals: Across 1,898 valid tests, seven taxon-outcome associations passed false-discovery-rate (FDR) correction ( $<0.05$ $< 0.05$ ). Crucially, all significant hits were confined to the internalizing spectrum (MDD, neuroticism, insomnia). No associations reached significance for bipolar disorder, schizophrenia, ADHD, PTSD, anxiety, or OCD.
- Protective associations: The Mollicutes/Tenericutes clade with MDD ( $\beta = -0.073$ ) and Butyrivibrio with neuroticism.
- Deleterious associations: Negativicutes/Selenomonadales with insomnia, Betaproteobacteria with neuroticism, and Tyzzerella3 with neuroticism.
Conservative Validation: Strict causal tests tempered the interpretation of direct causation:
- Colocalization: The posterior probability of a shared causal variant (PP.H4) was $<0.05$ at every locus, indicating distinct causal variants for the exposure and outcome.
- CAUSE: Zero of 45 taxon-outcome pairs met the strict criterion for genuine causation.
- Multivariable MR: Most signals were structurally over-parameterized due to weak instruments; only Butyrivibrio retained significant effects in this specification.
Directional Consistency: Despite the rejection of direct per-locus causation, the direction of effect in CAUSE matched the protective-versus-deleterious hypothesis in 33 of 45 pairs ( $p = 1.2 \times 10^{-3}$ ), suggesting a structured directional relationship.
Latent Architecture: Genomic SEM confirmed that the candidate taxa load specifically onto a latent internalizing factor (correlation 0.48 with a separate psychotic factor) and onto internalizing-specific genetic variance in a bifactor model. Betaproteobacteria was the only taxon showing deleterious effects on both internalizing and psychotic factors.
Mediation: No mediation through tryptophan, kynurenine, CRP, or IL-6 exceeded 5% with a bootstrap CI excluding zero.
Functional Annotation: MAGMA identified gene-level hits (e.g., ABCA7, REV3L), but no canonical butyrate/SCFA or FFAR pathways reached cross-taxa significance.

Significance and Claims
The paper claims that the primary finding is not a list of causal microbes, but a re-framing of the relationship between the microbiome and psychopathology. The authors conclude that selected gut microbial taxa and internalizing psychopathology share host-genetic architecture rather than existing in a direct microbe-to-disorder causal chain at the level of individual variants.

The study's contributions are threefold:

Dimensional Specificity: It demonstrates that microbial signals in MR studies concentrate specifically on the internalizing dimension of psychopathology, distinct from psychotic or neurodevelopmental dimensions.
Methodological Rigor: It provides a stringent test of the direct-causation hypothesis using a conservative cascade (colocalization, CAUSE, permutation-null MR-BMA), finding the hypothesis negative while the dimensional-architecture finding remains positive.
Resource Generation: The authors release a full "atlas" of 211 taxa $\times$ 9 outcomes across ten analytical layers, including null results, to prevent selective reporting and serve as a resource for future larger-scale microbiome studies and brain-tissue follow-up.

The authors explicitly state that the observed associations likely reflect shared genetic susceptibility between the host's microbiome composition and internalizing disorders, rather than a direct causal mechanism where specific microbes cause the disease. They position the identified taxa and loci (e.g., ABCA7, REV3L) as candidates for future experimental validation and brain-tissue replication, not as immediate druggable targets.

Technical Summary: Shared Host-Genetic Architecture Between Gut Microbiota and Internalizing Psychopathology

More like this