It is estimated that each year, 6% of the adult population worldwide experience the debilitating mental health disorder that is major depression (MDD) (Otte et al. 2016). Considerable uncertainty remains about the neurobiological basis for depression and new leads are urgently required to support diagnosis and improve treatment options. What if new therapeutic targets lie below rather than above the neck?

Convincing evidence from preclinical research studies supports a role for the gut microbiome in the regulation of brain function and behaviour (Donoso et al. 2023). Attempts to translate this research from bench to bedside has seen many compositional assessments of the community of microorganisms resident in our gastrointestinal tract, including from these authors (Bastiaanssen et al. 2020; Radjabzadeh et al. 2022). However, the findings from such studies have often been inconsistent, although some common signals are starting to emerge (McGuinness et al. 2022; Nikolova et al. 2021). Making sense of the mechanistic role of the gut microbiome in depression requires a shift in focus from form to function – when it comes to gut microbes in depression, it is not who is there but what they are doing, and missing or depleted microbes might also lead to a loss of key host-microbe interactions.

Amin and colleagues sought to look at some of these open questions by querying the association between blood metabolic signatures of MDD and the gut microbiome (Amin et al. 2023).

Methods

Of the more than 500,000 individuals in the UK Biobank cohort, Amin et al. examined participants aged 37 to 73 who had provided blood samples. Individuals with lifetime and recurrent major depressive disorder (MDD) were included. People with bipolar disorder, schizophrenia, psychosis and other mental health conditions were excluded. Healthy control individuals who had not reported depression at baseline were included for comparison.

Plasma metabolites were assessed using a high-throughput 1H-NMR metabolomics in a random subset of 118,466 individuals. Bidirectional 2-sample Mendelian randomization was used to determine the direction of the association observed between metabolites and MDD. Results from the BBMRI-NL and PREDICT studies were used for replication purposes. You can read more about Mendelian randomization and the role of diet in depression in this recent Mental Elf blog by Crick D (2023).

Results

6,811 individuals with lifetime MDD were compared to 51,446 healthy controls, and 4,370 individuals with recurrent major depressive disorder (MDD) were compared to 62,508 healthy controls. In the fully adjusted model, which incorporated age, sex, several lifestyle factors, alcohol consumption, body mass index, smoking, education, medication use for cardiovascular morbidity, and other technical covariates, 124 metabolites of the 249 measured were found to be significantly (FDR < .05) associated with MDD.

The authors reported that there was a significant increase in total monounsaturated fatty acids and its ratio to total fatty acids in MDD. Meanwhile, the ratios of linoleic acid and polyunsaturated fatty acid to total fatty acids were significantly decreased in individuals with MDD. Apolipoprotein A1, cholesteryl esters, citrate and sphingomyelins were all significantly decreased with alanine and pyruvate significantly increased in MDD. Similar observations were recorded for those with lifetime and recurrent MDD. This indicates that there are alterations in metabolites that are important for either lipid metabolism or the production of energy.

The metabolites identified in Amin et al.’s study were in good agreement with the findings reported in the BBMRI-NL consortium (Bot et al. 2020). The association of MDD with omega 6, polyunsaturated fatty acid, citrate, and pyruvate was replicated in the data from the PREDICT study (Dunlop et al. 2012). Novel associations were reported for 49 metabolites including 2 key metabolites, citrate, and pyruvate, that are known to be involved in lipid metabolism (important for cell migration, apoptosis, autophagy, and cell division and immune system function) or those used to provide energy.

Mendelian randomization analysis suggested that changes in low- and very low-density lipoproteins, intermediate-density lipoproteins, and fatty acids were associated with MDD. There was no evidence of an association between changes observed in high-density lipoproteins, apolipoprotein A1, or metabolites in the tricarboxylic acid cycle and MDD.

To understand microbiome studies, it is important to remember that micro-organisms, like all lifeforms, are often organised into groups or types – a classification system referred to as taxonomical classification. This goes from Phylum to Class to Order to Family to Genus to Species to Strain. Sequencing studies initially view the gut microbiota through a compositional lens and use reference databases to identify the taxa present in a faecal sample, and often report observations from family level onwards. The objective of this approach is to see if there are any important associations between specific taxa (e.g., a microbial family, genus, or species) and for example, a particular symptom or metabolite. You can read more about this in Bastiaanssen et al.’s 2019 review. Like our own families though, all members have their own often unique characteristics and the better the resolution the better!

Amin et al. 2023 found 223 bacterial taxa that were significantly associated with MDD using a proxy association based on correlation between the metabolic signatures of MDD and gut microbial taxa. At the family level, Ruminococcaceae and other families belonging to the order Clostridiales were negatively correlated with MDD. A positive correlation was reported between Lachnospiraceae and Eubacteriaceae, and MDD.  A number of families were negatively correlated with MDD including Methanobacteriaceae, Rhodospirillaceae, Desulfovibrionaceae, Pasteurellaceae, Neisseriaceae, and Oxalobacteraceae, Porphyromonadaceae, Rikenellaceae, and Prevotellaceae. Many of these taxa have previously been associated with stress-related disorders like depression, and are important leads for future research to understand if or how these gut microbes can contribute to biological abnormalities or symptom expression. This information could lead to new therapeutic interventions based around changing the composition of the gut microbiota.

Conclusions

The biomarker associations identified in this research indicate a possible disruption of mitochondrial metabolism in depression. Moreover, the reported changes in lipid metabolism likely reflect microbial regulation of circulating levels of these metabolites. The Mendelian randomization results suggest that some of these changes in lipid metabolism may be associated with disease processes in major depressive disorder (MDD) although those involved in energy metabolism are not.

These observations are consistent with many previous observations and suggest that therapeutic targeting of the gut microbiome might, in the future, be a promising strategy to restore normal lipid metabolism.

Strengths and limitations

Key strengths of this study are the large sample size under evaluation and that the observations recorded were validated in two different replication cohorts. The authors have also taken great care to control for confounding factors. This suggests robust observations that encourage more research in this area.

Many previous studies have relied on compositional assessments of the gut microbiota and the need for more functional readouts is warranted. The metabolomic analysis reported here is thus a very welcome addition. The integration of gut microbiome and metabolic signatures is also an important selling point that further increases the insights from the study.

The gut microbiome is dynamic as are the conversations going on across host-microbe dialogues. Longitudinal studies are needed to understand how interactions vary over time as symptoms wax and wane. The lived experience of depression is different for everyone, and future studies need to take into account different subtypes. This would allow us to understand, for example, how or if the current observations map onto the low-grade inflammation present in some subsets of depression. Detailed participant phenotyping (i.e., observable traits) in clinically diagnosed MDD will be essential in this regard, and it is not clear how well the findings here will generalise due to the reliance on self-reported depression as the participant selection criterion.

A great deal of evidence supports the role of a range of microbial metabolites in stress-related disorders such as short chain fatty acids, indoles produced from tryptophan and bile acids (Caspani et al. 2019). Unfortunately, the metabolomics assay deployed here was not broad enough in scope to interrogate such theories, which will be an important research objective for future studies. Moreover, as the taxonomical resolution of sequencing platforms improves, we can expect to get to the strain-level information which would be more informative than analyses at family or genera level.

Implications for practice

Translating the observations recorded from preclinical and observational studies into healthcare policy and practice is a difficult endeavour. Human studies of the type reported here are needed to distinguish causality from association. While the current study is an important step in the right direction, there are still many unanswered questions remaining and outlined in the limitations section above, before we can safely make new recommendation or transition to clinical implementation.

Assessing the causal role of the gut microbiome in depression remains a challenge. Mendelian randomization is a useful tool to assess how consistent an observational association between a risk factor and an outcome is with a causal effect. Nevertheless, there are a number of possibilities, and the study points us towards several interesting opportunities.

The gut microbiome can be regarded an important biomarker reservoir, the accuracy of which can be increased via the integration of compositional and functional signals. Coupling these approaches with longitudinal assessments and detailed patient phenotyping can improve diagnostic approaches informed more by biological rather than symptom-based classification. Equally important in this regard are the microbiome-independent biomarkers of energy metabolism. Taken together, these observations point towards the need for a broad panel of biomarkers although the implications for the symptoms of depression or the microbiome-dependent vs microbiome-independent observations require further elaboration.

It will also be important to consider the specificity of the findings. There are indications now that there may be transdiagnostic patterns of microbiome alterations across current diagnostic categories (Nikolova et al. 2021). It will be of great interest to see if that is also true in those with a diagnosis of anxiety, bipolar disorder and schizophrenia in the UK Biobank and other cohorts.

The results reported here indicate that manipulation of the gut microbiota could be a useful strategy to fine tune mechanistically-oriented biomarkers such as those linked to lipid metabolism. Obtaining proof will require interventions in randomised controlled trials with dietary modification being an important option that now requires further evaluation.

Assessing causality can also be achieved using faecal microbiota transplantation with back-translation into animal models (Gheorghe et al. 2021; Secombe et al. 2021). This has already been demonstrated for some behavioural features linked to depression (Kelly et al. 2016) and would offer further impetus if this approach could be used to demonstrate microbial regulation of lipid metabolism. Clearly, further efforts at replication across cohorts with a confirmed clinical diagnosis will be critical in addition to the type of approaches taken here to look at cause and effect. The answers lie ahead and within reach, but we are not there yet.

Statement of interests

Gerard has received honoraria from Janssen, Probi, and Apsen as an invited speaker; is in receipt of research funding from Pharmavite, Reckitt, Tate and Lyle, Nestle and Fonterra; and is a paid consultant for Yakult and Zentiva. This support neither influenced nor constrained the contents of this blog.

Links

Primary paper

Amin N, Liu J, Bonnechere B, et al. (2023) Interplay of Metabolome and Gut Microbiome in Individuals With Major Depressive Disorder vs Control Individuals. JAMA Psychiatry. 2023;80(6):597–609. doi:10.1001/jamapsychiatry.2023.0685

Other references

Bastiaanssen, T. F. S., et al. (2019), ‘ Making Sense of … the Microbiome in Psychiatry’, Int J Neuropsychopharmacol, 22 (1), 37-52.

Bastiaanssen, T. F. S., et al. (2020), ‘Gutted! Unraveling the Role of the Microbiome in Major Depressive Disorder’, Harv Rev Psychiatry, 28 (1), 26-39.

Bot, M., et al. (2020), ‘Metabolomics Profile in Depression: A Pooled Analysis of 230 Metabolic Markers in 5283 Cases With Depression and 10,145 Controls’, Biol Psychiatry, 87 (5), 409-18.

Caspani, G., et al. (2019), ‘Gut microbial metabolites in depression: understanding the biochemical mechanisms’, Microb Cell, 6 (10), 454-81.

Crick, D., (2023), ‘Does what you eat affect how you feel?‘, The Mental Elf, 8 Jun 2023.

Donoso, F., et al. (2023), ‘Inflammation, Lifestyle Factors, and the Microbiome-Gut-Brain Axis: Relevance to Depression and Antidepressant Action’, Clin Pharmacol Ther, 113 (2), 246-59.

Dunlop, B. W., et al. (2012), ‘Predictors of remission in depression to individual and combined treatments (PReDICT): study protocol for a randomized controlled trial’, Trials, 13, 106.

Gheorghe, C. E., et al. (2021), ‘Investigating causality with fecal microbiota transplantation in rodents: applications, recommendations and pitfalls’, Gut Microbes, 13 (1), 1941711.

Kelly, J. R., et al. (2016), ‘Transferring the blues: Depression-associated gut microbiota induces neurobehavioural changes in the rat’, J Psychiatr Res, 82, 109-18.

McGuinness, A. J., et al. (2022), ‘A systematic review of gut microbiota composition in observational studies of major depressive disorder, bipolar disorder and schizophrenia’, Mol Psychiatry, 27 (4), 1920-35.

Nikolova, V. L., et al. (2021), ‘Perturbations in Gut Microbiota Composition in Psychiatric Disorders: A Review and Meta-analysis’, JAMA Psychiatry, 78 (12), 1343-54.

Otte, C., et al. (2016), ‘Major depressive disorder’, Nat Rev Dis Primers, 2, 16065.

Radjabzadeh, D., et al. (2022), ‘Gut microbiome-wide association study of depressive symptoms’, Nat Commun, 13 (1), 7128.

Secombe, K. R., et al. (2021), ‘Guidelines for reporting on animal fecal transplantation (GRAFT) studies: recommendations from a systematic review of murine transplantation protocols’, Gut Microbes, 13 (1), 1979878.

Photo credits

• Photo by Nadine Primeau on Unsplash
• Photo by Ryoji Iwata on Unsplash
• Photo by Markus Spiske on Unsplash
• Photo by CHUTTERSNAP on Unsplash
• Photo by John Arano on Unsplash
• Photo by Hadija on Unsplash