Genetic analysis reveals two-way link between gut bacteria and insomnia risk

A genetic-microbiome study was published in the open journal General Psychiatry: some groups of intestinal bacteria increase or decrease the likelihood of insomnia, and insomnia itself changes the composition of these bacteria. The authors used the Mendelian randomization method and combined huge data sets - 386,533 people from GWAS on insomnia and 26,548 people from two microbiome consortia. The result: 14 bacterial groups were found associated with a higher risk of insomnia (by 1-4% for each group) and 8 groups with a lower risk (by 1-3%). At the same time, people with insomnia showed significant shifts in the abundance of individual taxa (for example, Odoribacter): in some - a drop of 43-79%, in others - an increase of 65% -> 4 times.

Background

Insomnia is one of the most common sleep disorders (up to 10–20% of adults; even higher in the elderly) and a significant risk factor for depression, CVD, and metabolic disorders. Against the background of limited effectiveness of symptomatic therapy, there is growing interest in targets of the “gut-brain axis”, where microbes and their metabolites affect inflammation, the HPA stress axis, neurotransmitters, and circadian rhythms.

• Biological clues were there even before “genetics.” Microbial products, especially short-chain fatty acids (e.g., butyrate), have been linked to improved sleep in preclinical and early clinical work; tryptophan → serotonin/melatonin metabolism in the microbiota is another plausible pathway for influencing sleep.
• The problem with older studies is causality. Much of the early work was observational: diet, medication, and lifestyle all affect both the microbiota and sleep, so it’s hard to know which is the cause and which is the effect. Hence the shift to tools that are robust to confounding factors, such as Mendelian randomization (MR).
• Why microbiota MR has only recently become possible. Large microbiome-GWAS consortia have emerged with open aggregate data:
  • The international MiBioGen (＞18,000 participants) showed that variations in host genes (e.g. LCT, FUT2) are associated with the abundance of individual taxa;
  • The Dutch Microbiome Project (≈7,738 individuals, Nat Genet, 2022) has clarified the “heritable part” of the microbiota. These kits have become “genetic tools” for MR analyses.
• And on the sleep side, there are also large “genetic maps.” Large GWAS on insomnia cover hundreds of thousands to millions of participants, identifying tens to hundreds of risk loci and providing power for bidirectional MR (“microbe → insomnia risk” and “insomnia → microbiota composition”).
• What interventions have already hinted at. Systematic reviews and meta-analyses on probiotics/prebiotics have shown small improvements in subjective sleep quality, but with high heterogeneity of strains, doses, and populations — i.e. without a firm answer to “why and for whom it works.” Genetic methods help to identify which specific groups of bacteria are potentially causally associated with sleep and are worth clinical testing. Why a new study was needed. To combine “big genetics” on insomnia (≈386 thousand) with the largest microbiome-GWAS to date (MiBioGen + DMP, total ≈26.5 thousand) and test for bidirectional causality: which taxa increase/decrease the risk of insomnia and how genetic predisposition to insomnia restructures the microbiota. Such a design is more resistant to confounding and reverse causality than classical observations.
• Limitations to keep in mind: Microbiota is highly country/ethnic/diet dependent, and most reference GWAS are of European origin; 16S approaches annotate taxa differently; even MR is subject to pleiotropy if genetic tools affect outcome via alternative pathways (hence MR-Egger, heterogeneity tests, etc.). Clinical conclusions require RCTs with verified strains/metabolites and objective sleep metrics.

What exactly did they do?

• We took the largest summary data available today:
  • GWAS for insomnia - 386,533 participants;
  • Genetically indexed microbiome: MiBioGen (18,340 individuals) and Dutch Microbiome Project (8,208 individuals).
    71 common bacterial groups were analyzed together.
• We used bidirectional Mendelian randomization (multiple methods and sensitive tests) to test for causal relationships: “microbe → insomnia” and “insomnia → microbe”. This reduces the risk of confounding with lifestyle factors and reverse causality.

Main results

• Which microbes "push" towards insomnia. Only 14 groups showed a positive causal association with the risk of insomnia (about +1–4% to the odds), and 8 showed a protective association (−1–3%). Among the markers on which the validation sets converged, the genus/class Odoribacter stood out.
• Insomnia “reshapes” the microbiome. Genetically predicted predisposition to insomnia was associated with a sharp decrease in the abundance of 7 groups (−43…−79%) and an increase in 12 groups (+65% to more than 4 times). This is an important argument in favor of a two-way relationship.
• The statistics hold up. The authors found no evidence of strong horizontal pleiotropy — that is, the effect probably occurs through microbial factors, not extraneous pathways.

Why is this important?

So far, we have seen mostly correlations between sleep disorders and gut flora. Here is a big step towards causality: genetic tools show that some microbial groups influence the risk of insomnia, and insomnia changes these groups in response. This opens the way to microbiome-oriented approaches to prevention and therapy – from prebiotics/probiotics to dietary strategies and, potentially, to more targeted interventions.

How it might work (mechanical clues)

The work does not prove specific mechanisms, but fits into the logic of the microbiome-gut-brain axis: microbes and their metabolites (e.g., short-chain fatty acids, neurotransmitter-like molecules) modulate the immune response, inflammation, the HPA stress axis, and neural networks involved in sleep regulation. Recent preclinical and clinical observations have linked, for example, butyrate and the bacteria that produce it to better sleep; this work indirectly confirms that shifts in the microbiome “production lines” can alter sleep.

What does this mean "in practice" now?

• This is not a list of “good” and “bad” bacteria for self-medication: the effects are small in magnitude and depend on the context (diet, medications, comorbidities).
• The smart steps are the same ones that promote a “healthy” microbiome: a variety of plant foods, fiber, fermented foods (unless contraindicated), moderation with alcohol, exercise, stress management.
• For people with chronic insomnia, clinical trials of targeted microbial interventions hold promise—but they are still to come.

Restrictions

• Microbiome composition varies widely across countries/ethnicities; the bulk of data is from European origin, and the generalizability of findings is limited.
• Genetic proxies for microbes (16S/metagenomic consortia data) were used rather than direct measurements in the same individuals as in the insomnia GWAS.
• Diet, lifestyle, and medications (including sleeping pills) that affect the microbiome were clearly not included in the analysis. This means that this is preliminary causal evidence that requires clinical testing.

What's next?

The authors propose testing microbiome strategies as an adjunct to standard insomnia therapy and using microbial signatures as response biomarkers (therapy personalization). A logical route: pilot RCTs of prebiotics/probiotics with objective sleep metrics (actigraphy/polysomnography) and whole-genome metagenomics before/after.

Source: article in General Psychiatry (Investigating bidirectional causal relationships between gut microbiota and insomnia, DOI 10.1136/gpsych-2024-101855 )