The gut microbiome may affect mental health symptoms like anxiety and depression, and possibly even autism

 autism gut microbiome imageSome very interesting work is being conducted to elucidate connections between the gut microbiome and mental health disorders such as anxiety, depression, and even autism. Scientific American has published several reviews on the subject. John Cryan and colleagues call it "a paradigm shift in neuroscience".

The gut is home to the enteric nervous system, which controls digestion and excretion locally. This "second brain" consists of sheaths of neurons, 100 million in all, embedded in the gut walls all the way from the esophagus to the anus. Neural communication between the brain and the enteric nervous system (gut-brain axis) goes in both directions with, surprisingly, as many as 90% of the fibers of the primary visceral nerve (the vagus) carrying information to the brain. More than 30 neurotransmitters are involved in the enteric nervous system (similar to the brain), and 95% of the body's serotonin is found in the gut. The enteric nervous system is thought to influence our state of mind and emotions. For example, electrical stimulation of the vagus nerve, which mimics and/or amplifies gut-to-brain neural activity, has been shown to be a useful treatment for depression.

The gut microbiota comprises about 100 trillion cells (several pounds of body mass) and represents up to 10,000 species of bacteria, eukaryotes, and viruses. The number of bacterial cells in the body outnumbers our own cells by 10 to 1, and the number of bacterial genes outnumbers our genes by 100 to 1.

Scientists now think that the gut microbiota play a key role in the gut-brain axis, using not only neural but also endocrine and immune pathways. In the last decade, gut microbiome influences in various mental health disorders including anxiety, depression, and autism have been reported.

  • In patients with irritable bowel syndrome (IBS), probiotics not only relieved gastrointestinal (GI) symptoms but also reduced anxiety and stress response and improved mood. Probiotics are bacterial strains thought to be beneficial to health; common species include lactic acid bacterial strains (eg, Lactobacillus acidophilus, L. rhamnosus, L. johnsonii, L. delbrueckii subsp. bulgaricus, and Streptococcus thermophilus). Live probiotic cultures are ingested as fermented dairy products, other fermented foods such as tempeh, miso, kimchi, or kefir, or manufactured freeze-dried in tablets or capsules.
  • Another study demonstrated reduced anxiety in rats and beneficial psychological effects and lower serum cortisol in human subjects when treated with L. helveticus and Bifidobacterium longum strains.
  • L. reuteri decreased anxiety in the mouse elevated plus maze model and reduced stress-induced corticosterone release; in this study, vagotomy prevented these effects, suggesting that parasympathetic innervation is involved.
  • B. infantis acted as an antidepressant in the forced swim test and relieved depression symptoms in the maternal separation model, both in rats.


Autism spectrum disorder (ASD) is a heterogeneous set of neurobehavioral diseases that can affect a patient's social interactions, communication (both verbal and nonverbal), behavior, and interests. The incidence of ASD in the US has increased dramatically from 1 in 150 children in 2000 to 1 in 68 in 2010, as reported by CDC, but it's not clear if that increase is due to increased surveillance and reporting, environmental factor(s), or other reasons. Many patients with ASD have GI comorbidities such as diarrhea, abdominal pain, IBS, and constipation, and altered GI motility and increased intestinal permeability have also been reported. Causes of these GI symptoms in patients with ASD are not clear, but Hsiao and colleagues list 8 studies from the literature in which an altered gut microbiome was identified in patients with ASD; elevated levels of Clostridium species were found in 3 of these studies.

In a recent review of this nascent field, Cheryl Rosenfeld summarizes the animal studies, human epidemiologic studies, and work on other microbiomes (eg, oral cavity, placenta), as well as possible mechanisms and therapies. I've selected one interesting animal study to review in detail.

Deep dive: Hsiao et al 2013 "The microbiota modulates gut physiology and behavioral abnormalities associated with autism"

Elaine Hsiao and colleagues at Caltech and Baylor College of Medicine report interesting associations between specific gut microbial species and the behavioral and GI symptoms of autism. Using the mouse maternal immune activation (MIA) model, these researchers induced the cardinal behavioral and neuropathological symptoms of ASD. The MIA mice exhibited increased anxiety in the open field exploration test, impaired pre-pulse inhibition in the sensorimotor gating test, increased stereotyped behavior in the marble burying test, decreased communication in the ultrasonic vocalization test, and deficits in sociability and social preference in the 3-chamber social test.

The authors report a significant deficit in intestinal barrier integrity in the MIA mice, evidenced by increased passage of FITC-dextran in the intestinal permeability assay, along with decreased gene expression of tight junction components in the colon (ZO-1, ZO-2, and occludin) and small intestine (ZO-1) and abnormal intestinal cytokine profiles (increased IL-6 mRNA and protein and decreased IL-12p40/p70, IP-10, MIG, and MIP-1α).

No significant overall differences in fecal bacterial populations were seen in the MIA mice vs. controls with 16S rRNA gene sequencing, but with UniFrac analysis the authors did see significant differences, primarily in the Bacteroidia and Clostridia classes. When key operational taxonomic units (OTUs) were further refined, 67 discriminatory OTUs were identified, the majority of which were again in the Bacteroidia (67.2%) and Clostridia (20.9%) classes. Porphyromonadaceae, Prevotellaceae, and unclassified Bacteriodales (Bacteroidia) and Lachnospiraceae (Clostridia) were more abundant in MIA mice, and Ruminococceae (Clostridia), Erysipelotrichaceae (Erysipelotrichia), and Alcaligenaceae (Betaproteobacteria) were more abundant in the control mice.

Human commensal Bacteroides fragilis has been shown to correct GI pathology in mouse models of colitis and to protect against neuroinflammation in mouse models of multiple sclerosis. In this study, treatment with B. fragilis from weaning to 8 weeks was found to reverse each of the MIA behavioral disabilities except sociability and social preference. MIA mice treated with B. fragilis had significantly improved gut barrier integrity compared with untreated MIA mice. B. fragilis also reversed the MIA-related effects on gene expression for CLDN8 and CLDN15 and their protein products in the colon. Intestinal cytokine analysis showed that B. fragilis treatment corrected only the IL-6 mRNA and protein levels to those of controls; it did not affect other MIA-related cytokine effects. (IL-6 is thought to be related to the development of behavioral symptoms in ASD.) These results indicate that B. fragilis treatment improves defects in GI barrier integrity and corrects alterations in tight junction and cytokine expression.

Of the 67 OTUs determined to be discriminatory in MIA mice, 6 were reversed by B. fragilis treatment, 4 in the Lachnospiraceae family and 2 in the unclassified Bacteroidia class. These changes in the microbiome occurred in the absence of persistent B. fragilis colonization. The authors mention unpublished results that only a single phylotype, very closely related to B. fragilis, was severely depleted in children with ASD, with the greatest depletion in those with most severe GI issues.

Metabolomic profiling using GC/LC-MS demonstrated significant differences in 8% of the 322 serum metabolites identified in the MIA mice vs. controls. Treatment with B. fragilis significantly altered serum concentrations of 34% of the metabolites. The most dramatically affected metabolite was 4 ethylphenylsulfate (4EPS), which derives from a precursor thought to be produced by several Clostridium species; the 46-fold increase in MIA mice was ameliorated by B. fragilis. Serum 4EPS is thought to be derived from or modulated by the commensal microbiota. It is believed to have a role in communication in mice, and it is chemically related to p-cresol, which is a possible urinary biomarker for autism. p-cresol levels were also elevated in the MIA mice relative to controls, but the increase did not reach statistical significance. Other metabolites elevated with MIA and restored to control levels by B. fragilis treatment were serum indolepyruvate (possibly related to ASD), glycolate, imidazole propionate, and N-acetylserine.

In a separate experiment, application of 4EPS to naive mice from 3 to 6 weeks of age resulted in some anxiety indicators (open field exploration assay, center distance, but not distance; pre-pulse inhibition, startle but not PPI%) similar to those observed in MIA mice. However, no significant differences between 4EPS-treated mice and controls were seen in the marble-burying and ultrasonic vocalization tests. Although anxiety is not a core diagnostic criterion of ASD, it is a common comorbidity.

These findings support that autism is at least partially a disease involving the gut that impacts the immune, metabolic, and nervous systems and that microbiome-mediated therapies may offer benefits in the treatment of ASD.

Blogger: Ginny Fleming, Founder, Lucidize Medical & Scientific Editing. Chief capacities: medical, scientific, and technical writing and editing.