Microbes and the Mind: How Bacteria Shape Affect,Neurological Processes,Cognition,Social Relationships,Development,and Pathology

Chronic and acute affective states

At the most general level, maintaining a robust population of beneficial bacteria in the gut can improve mood and well-being, whereas the depletion of these bacteria can negatively affect well-being (Cryan & Dinan, 2012; Cryan & O’Mahony, 2011; Forsythe, Sudo, Dinan, Taylor, & Bienenstock, 2010). For example, in a double-blind placebo-controlled study, medical students consumed a daily drink composed of probiotics, which are supplements containing live, beneficial bacteria, for 3 months leading up to their qualifying exams. The students showed improvements in psychological symptoms related to chronic stress, as indexed by attenuated increases in cortisol, improved sleep quality, and improved parasympathetic activity (Nishida et al., 2017). Recent studies have further demonstrated that male rats that consumed a diet rich in prebiotics, which is fiber that beneficial microbes can digest, exhibited decreases in anxiety-like behaviors and altered gene expression in neural circuits that may support emotion regulation (Mika et al., 2018). By contrast, chemically induced colitis—inflammation of the colon—led to decreases in beneficial intestinal bacteria in rodents and increases in their behaviors associated with negative affect, such as anxiety (Bercik, Park, et al., 2011; Davis et al., 2016) and depression (Slyepchenko et al., 2016).

Bacteria are also related to chronic, trait-like affective states. For example, Martin et al. (2009) reported that human subjects who had chronically low or high levels of anxiety had distinct metabolic profiles indicative of different gut microbial activity. Furthermore, extended dietary interventions that directly influenced gut motility partially normalized these trait-like anxiety-related differences. Recent studies have also shown that individuals who experience chronic negative affect (as indexed via personality traits such as high neuroticism) have gut microbiota that are distinct from individuals whose personalities are characterized by less negative affect and lower levels of neuroticism (H. Kim et al., 2018).

In a similar demonstration from animal research, colonizing phenotypically anxious mice (BALB/c mice) with the microbiota of phenotypically gregarious mice (Swiss mice) resulted in the typically anxious mice becoming more gregarious. The reverse was also true—colonizing gregarious mice with the microbiota of anxious mice induced heightened anxiety-like behaviors (Bercik, Park, et al., 2011). Although conducted with animals, this study and others like it (e.g., Collins, Kassam, & Bercik, 2013) demonstrate that trait-like affective states can be “transferred” from one individual to another via manipulation of the microbiome.

In one of the first studies to examine how bacteria are related to acute affective experiences, Lencner et al. (1984) reported that astronauts experiencing momentary emotional distress during shuttle launches showed distinct alterations in the beneficial bacteria present in their saliva and feces. Experiments with rodents have repeatedly demonstrated similar findings. Acute negative affect brought on by momentary stress (e.g., fear conditioning, forced swim tests, stress-induced hyperthermia) is associated with or directly caused by the overgrowth of pathogenic bacteria or the depletion of beneficial bacteria (Bravo et al., 2011; McLean, Bergonzelli, Collins, & Bercik, 2012; Neufeld, Kang, Bienenstock, & Foster, 2011).

Mood beyond health

Although it may be tempting to simply conclude that the microbiome influences health and that health in turn influences affective states (e.g., “good” bacteria make you healthy, and healthy people are happy), there are also several compelling examples of bacteria influencing emotion in which improvements in physical health were not the primary mediating force. For example, a probiotic yogurt drink developed to reduce constipation was administered to human subjects daily for 3 weeks. Although the probiotic drink did not influence the frequency of defecation (the targeted health outcome), it did influence discrete emotional outcomes. Consumption of the probiotic drink (compared with a placebo drink) improved the self-reported mood of those whose mood was initially poor (Benton, Williams, & Brown, 2007). In another study, healthy human subjects with no symptoms of physical illness who ingested a probiotic supplement for 30 days reported significant decreases in symptoms of psychological distress over time. Specifically, they exhibited decreases in anxiety (as assessed by the Hospital Anxiety and Depression Scale; Zigmond & Snaith, 1983) and fewer symptoms of depression, anger, and hostility (as assessed by the respective subscales of the Hopkins Symptom Checklist; Derogatis, Lipman, Rickels, Uhlenhuth, & Covi, 1974) compared with those taking a placebo (Messaoudi et al., 2011).

Strikingly, the positive effects of bacteria on mood are not limited to how people perceive their own emotions but also how people perceive emotions on the faces of others. To this end, researchers found that when healthy women with no gastrointestinal (GI) or psychiatric symptoms consumed a nonfermented probiotic drink daily for 4 weeks, they showed decreased reactivity to negative facial expressions on other people’s faces (angry and sad faces) compared with subjects who consumed a placebo (Tillisch et al., 2012). Taken together, these results suggest that bacteria do not influence emotions simply by ameliorating poor health, as all of these human subjects were already healthy. Examples from animal research also support this claim. Healthy mice administered beneficial bacteria (compared with healthy mice given a placebo) tend to be more exploratory and outgoing in their behavior and show fewer symptoms of helplessness and depression (Bravo et al., 2011; Davis et al., 2016).

How do bacteria influence emotions, then, if improved mood is not just an outcome of improved health from the presence of beneficial bacteria and negative mood is not just an outcome of poor health induced by pathogenic bacteria? Although health almost certainly mediates the influence of bacteria on emotion in many cases, experimental studies also point to evidence that bacteria can more directly modify a host’s neural functions, affecting a variety of psychological processes and mental states.

Serotonin

It is well established that serotonin is fundamental to physical and psychological health, but less is known about how vitally important it is for GI health (J. J. Chen et al., 2001). In fact, approximately 90% of the body’s serotonin is located along the intestinal tract, where its primary function is regulating gut motility (Berger, Gray, & Roth 2009). The serotonin produced in the gut is also central to healthy nervous system development, and research with rodents indicates that the activation of serotonin receptors in the ENS is linked to its neurogenesis and neuroprotection (De Vadder et al., 2018). Given that many people diagnosed with serotonin-deficit-based mood disorders (e.g., depression, anxiety) have a comorbid gut disorder (Martin-Subero, Anderson, Kanchanatawan, Berk, & Maes, 2016), serotonin-based interventions may be effective in treating both psychological and gut-related symptoms. In fact, depression and anxiety symptoms in humans tend to increase in parallel with GI symptom severity and frequency (Pinto-Sanchez et al., 2015), making this comorbidity even more relevant. Current treatment methods, however, involve administering a selective serotonin-reuptake inhibitor that, although effective in ameliorating behavioral symptoms, can have negative effects from long-term use and does not address the root cause of the issue (Cohen & Baldessarini, 1985). Alternatively, administering beneficial bacteria that are known to increase serotonin levels (Neufeld et al., 2011; Reigstad et al., 2015) in the form of a probiotic or by following a diet that supports bacteria critical for serotonin production may effectively influence mood as well as concomitant GI symptoms. Current research with humans has already shown that an oral probiotic can improve the affective symptoms of those suffering from a poor mood (Benton et al., 2007), and this may be due in part to the influence bacterial populations exert on key neurochemicals.

GABA

γ-Aminobutyric acid (GABA) has a strong connection to depression and anxiety disorders; low levels of GABA correspond to increased depression- and anxiety-related symptoms (Cryan & Kaupmann, 2005; Lydiard, 2003; Nemeroff, 2003). Benzodiazepines are primarily used to treat anxiety disorders because they bind to GABA receptors in the brain, making GABA more efficacious. Antidepressants can also be used to treat GABA imbalances in people who have been diagnosed with depression. It is known that certain gut bacteria naturally produce GABA (E. Barrett, Ross, O’Toole, Fitzgerald, & Stanton, 2012). Furthermore, when probiotics are ingested by rodents, they have been found to alter GABA in the same regionally dependent manner as benzodiazepines and antidepressants (Bravo et al., 2011; Janik et al., 2016). After probiotic treatment with these GABA-producing bacteria, healthy mice demonstrate an increase in exploratory behavior (Collins, Surette, & Bercik, 2012), showing that even a healthy population can benefit from probiotics. Likewise, the administration of commensal bacteria that have been shown to modulate GABA receptor mRNA expression in rodents can reduce anxiety, increase the experience of pleasure, and modulate other affectively laden motivational states (Bravo et al., 2011).

Brain-derived neurotrophic factor

Brain-derived neurotrophic factor (BDNF) has been implicated in an array of psychological disorders—such as major depressive disorder (Kielstein et al., 2015) and schizophrenia (Nieto, Kukuljan, & Silva, 2013). It has been consistently found that low levels of BDNF in humans are associated with poorer psychological outcomes and that certain medications, such as antidepressants, seem to restore BDNF levels for those who have an abnormally low supply (B. Chen, Dowlatshahi, MacQueen, Wang, & Young, 2001; Shimizu et al., 2003). Recent research has found that having an overabundance of pathogenic bacteria results in significantly less hippocampal BDNF in mice, which covaries with more anxious and depressed behaviors. Similar to the effect of antidepressants, these behavioral markers of anxiety and depression were attenuated when a probiotic containing Bifidobacterium longum was administered to the mice (Bercik, Denou, et al., 2011). Other studies have also found that having a healthier microbial profile, either by receiving a probiotic, reducing the abundance of pathogenic bacteria, or supporting communities of beneficial bacteria that are already present, results in more BDNF and exploratory behavior in mice (Collins et al., 2012). These findings suggest that optimizing one’s microbial profile could effectively increase low BDNF levels.

Deficits in bacteria

Studies in germ-free animals—which live in sterile environments and have no exposure to bacteria—have shown that a lack of bacteria results in many detrimental neurological consequences (Al-Asmakh & Zadjali, 2015; Cebra, 1999; Collins et al., 2012). Specifically, germ-free animals lack key neurometabolites (Wikoff et al., 2009), and their brains lack structural integrity (Luczynski et al., 2016). For instance, the blood–brain barrier of germ-free mice is not as tight or well protected as it is in conventional mice (Braniste et al., 2014). Furthermore, the microbiome supports the growth and functioning of microglia cells, which are essential for mounting immune responses to protect the brain from pathogens or internal damage. Even the temporary depletion of a host’s microbiota leads to structurally and functionally defective microglia in rodents (Erny et al., 2015). Altogether, bacteria prove to be not only beneficial but also vital to brain function and development.

Memory

Encoding and storing information is essential to everyday functioning both for humans and animals, and bacteria play a role in these cognitive processes (Benton, Williams, & Brown, 2007; Magnusson et al., 2015; Messaoudi et al., 2011). For example, in a series of studies, Gareau et al. (2011) found that having a bacterial infection that leads to the proliferation of pathogenic bacteria resulted in memory dysfunction in mice that had been exposed to a prolonged water-avoidance stressor, which is a well-established model of psychological stress in rodents (Saunders et al., 2002). In addition, germ-free mice exhibited similar stress-induced memory dysfunction. This memory dysfunction was prevented when mice received a probiotic containing an array of commensal bacteria, suggesting that probiotics designed to normalize gut bacteria can be beneficial for cognitive processing. Likewise, mice with enriched diets display increased working and reference memory (Collins et al., 2013; Li, Dowd, Scurlock, Acosta-Martinez, & Lyte, 2009). Recent research with rodents has demonstrated that probiotics can have strong and significant effects on both long-term object recognition and short-term memory for objects in place (O’Hagan et al., 2017).

These findings may, however, have important moderators. For example, in a recent experimental study with rodents, the administration of probiotics improved some types of memory (spatial) but impaired others (object recognition; Beilharz, Kaakoush, Maniam, & Morris, 2018). Likewise, antibiotics administered to mice to cause dysbiosis—which is an unhealthy imbalance of their internal microbial populations—led to impairments in object recognition but not spatial memory (Frohlich et al., 2016). These findings suggest that the type of memory task being assessed and the cognitive processes it recruits may be differentially affected by bacteria populations. In addition, the health of subjects’ microbiomes before probiotic intervention as well as their routine diets may also be key variables to account for. For example, Beilharz, Kaakoush, Maniam, and Morris (2018) found that administering probiotics to rats with severely dysregulated microbiomes as a result of a high-fat diet did indeed improve some types of memory performance. By comparison, the probiotic intervention led to memory deficits in healthy rats with typical microbiomes. In humans, many diets support a healthy microbiome and, in general, eating more prebiotic fiber and fewer processed foods will confer a health benefit. However, further research will need to examine when and which bacterial strains most directly influence memory-related processes and how this interacts with the host diet.

Learning and attention

Research suggests that the microbiome also influences the ability to process, apply, and ultimately learn from information. For example, when mice were fed commensal bacteria daily for 11 weeks, they showed significant improvements on cognitive tasks. Note that the improvement depended on the exact type of bacteria mice ingested. Those fed B. longum showed the most improvement in object recognition and object discrimination, followed by those fed Bifidobacterium breve, whereas control animals who did not receive a bacteria supplement demonstrated the least improvement and the poorest performance overall (Savignac, Tramullas, Kiely, Dinan, & Cryan, 2015). A study that examined the effects of diet on the microbiome showed similar results. When tested 2 weeks after a controlled diet change, mice fed a high-fat or high-sucrose diet that resulted in the proliferation of pathogenic bacteria exhibited impaired cognitive flexibility compared with control mice that were fed a normal diet (Magnusson et al., 2015).

In a recent study with human subjects, participants who received a 4-week probiotic food supplement experienced significant reductions in cognitive reactivity to sad moods compared with a control group that received a placebo. This effect was driven primarily by decreases in the rumination and aggressive thoughts linked to sad moods and therefore provides some of the first evidence that beneficial bacterial species can influence emotional states via cognitive pathways (O’Mahony et al., 2014). Likewise, human subjects who ingested a prebiotic every day for 3 weeks exhibited less attentional vigilance to negative versus positive stimuli and a subsequent decrease in cortisol levels (Schmidt et al., 2015).

In another study that explored the ways in which cognition and affect may interact to influence learning, mice were fed the ambient bacteria Mycobacterium vaccae before completing increasingly complex maze tasks. The bacteria-fed animals completed the mazes more quickly than the control animals at all levels of complexity. This performance was in part attributed to corresponding decreases in anxiety-related behaviors. Furthermore, the bacteria-fed mice continued to outperform the control mice for a week after the bacteria had been withdrawn from their diets. However, this effect did not persist beyond that time frame, suggesting that the continual rather than intermittent presence of healthy bacteria is necessary to induce positive cognitive and affective outcomes (Matthews & Jenks, 2013).

These findings also must be interpreted with caution, as there exists at least one study with humans wherein the candidate probiotic Lactobacillus rhamnosus failed to modulate cognitive performance in a randomized placebo-controlled trial (Kelly et al., 2017). As discussed above, this highlights the need for further research aimed at successfully identifying what moderators, as well as which strains of bacteria, are most likely to affect cognitive processing in healthy human subjects.

Social niches and interpersonal interactions

The social niche people live in affects their microbiome from birth to death. Being delivered via vaginal birth versus cesarean section sets the stage for the forthcoming bacterial profile of the infant (Dominguez-Bello et al., 2010), and the nutrients present in a mother’s breast milk feed the growing bacterial population in a developing child (Hinde & German, 2012). This means that within the first few days after birth, social networks and social practices are already exerting an influence on the bacteria that humans will share their bodies with for the rest of their lives. In fact, all humans emit a relatively unique “microbial cloud” (Meadow et al., 2015)—a collection of personal bacteria—that is shed into the surrounding air. This cloud influences the microbial composition of humans’ external environment, including shared social spaces and the people in them. Even ethnic affiliation is correlated with microbiota composition (Mason, Nagaraja, Camerlengo, Joshi, & Kumar, 2013; Ravel et al., 2011), presumably because of shared genes, shared diets, shared living spaces, and other shared cultural experiences.

Bacteria respond not only to whom people are sur-rounded by but also to the content of their social interactions—especially social stress (Parashar & Udayabanu, 2016). For instance, mice exposed to socially aggressive cage mates exhibited substantially altered gut microbial communities (Bailey et al., 2011), and mice exposed to repeated experiences of social defeat showed disruptions in their balance of intestinal bacteria (Galley et al., 2014). Furthermore, the depletion of beneficial bacteria in mice has been shown to suppress normative immune responses after social stressors, such as being repeatedly physically attacked by other mice (Galley et al., 2014). This suggests that bacteria not only respond to stress but may also aid individuals in the ability to recover after major social disruptions. Maternal separation, for example, is an early-life relational stressor that can result in lifelong disturbances in stress reactivity (e.g., chronic or prolonged activation of the hypothalamic-pituitary-adrenal, or HPA, axis). After maternal separation, baby mice suffer a significant decrease in beneficial bacteria such as Lactobacillus species (Bailey & Coe, 1999), whereas being treated with that same (and other similar) bacteria before the separation attenuates the stress response, as indexed by reduced corticosterone levels (Desbonnet et al., 2010).

A microbiome abundant in beneficial bacteria species may even promote prosocial behaviors. Social interactions are most likely to occur when exploratory behaviors are enhanced and anxious behaviors are decreased. There are at least a dozen known strains of bacteria that exert anxiolytic—or anxiety-reducing—effects on their hosts and promote exploration of the environment (Wissel, 2018). Consistent with these findings, germ-free mice exhibit significant social impairments, lacking the normative preference for spending time with other mice compared with spending time alone in an empty chamber. Mice with healthy microbiomes, on the other hand, are more likely to explore, interact with novel mice, and demonstrate the normative motivation to socialize with others (Desbonnet, Clarke, Shanahan, Dinan, & Cryan, 2014).

Mate selection and intimate relationships

Once people have established a social niche, instances of interpersonal bacterial exchange become frequent and periodic. Human romantic relationships may have an especially strong influence on the composition of the microbiome, as they offer virtually unparalleled opportunities for bacterial exchange and colonization through multiple channels (e.g., mucus membranes, prolonged skin-to-skin contact).

There is already evidence that people use information from other people’s bodies to assess their reproductive fitness. To this end, sexual arousal engenders a higher tolerance, and even a desire, for biochemical fluids produced by an interaction partner (Borg & de Jong, 2012). Kissing, for example, may contribute to mate selection by offering the opportunity to sample chemical cues from the saliva that indicate health and compatibility (Nicholson, 1984), and smelling or tasting a partner’s sweat may provide similarly rich mating-relevant information (Bhutta, 2007; Gildersleeve, Haselton, Larson, & Pillsworth, 2012; Singh & Bronstad, 2001). Not surprisingly, bacteria are an essential ingredient in each of these mate-selection signals. For example, approximately 80 million bacteria are exchanged in a single mouth-to-mouth kiss, and the more frequently partners kiss, the more similar their oral and tongue microbiota become (Kort et al., 2014). Likewise, sweat itself does not have an odor. Rather, the bacteria colonizing the skin generate the odors that serve as olfactory cues for health, similarity, and social recognition (Archie & Theis, 2011; Kligman & Strauss, 1956; Montiel-Castro, Augusto, Báez-Yáñez Mario, & Pacheco-López, 2014). Additional markers of reproductive fitness that bacteria directly shape include clear skin, lustrous hair, and body mass index (Levkovich et al., 2013; Poutahidis et al., 2013; Poutahidis et al., 2014). This work demonstrates that the microbiome plays an important role in promoting traits that are used in the mate-selection process.

Experimental research has demonstrated that bacteria can also shape mate preferences at the actor level. For example, in an experimental study, administering antibiotics to fruit flies eliminated their normative preference for mates that inhabited environments similar to their own and maintained similar diets (Sharon et al., 2010). In a departure from their normative behavior, the fruit flies given the antibiotics selected mates more indiscriminately, with no preference given to similarity factors. This suggests that the disruption of the microbiome can alter individuals’ mating preferences and behaviors. Likewise, sexual practices can influence the microbiome. For example, non-monogamous female lizards show a more diverse microbiome in their cloaca than do monogamous females (White, Murielle, Massot, & Meylan, 2011). In humans, men who maintain monogamous relationships can be differentiated from those who have had extramarital affairs by the bacterial profile of their reproductive organs (or their community state type; C. M. Liu et al., 2015).

Bacteria also shape mating and relationship processes at the hormonal level. Feeding mice certain types of bacteria elicits the production of oxytocin, a hormone that has been associated with monogamy and pair bonding, in the brain and bloodstream (Varian et al., 2016, 2017). Mice treated with the purified bacteria Lactobacillus reuteri exhibit increases in testosterone, larger testicles, and higher sperm counts and tend to build extra muscle and maintain a competitive edge in their reproductive efforts (Poutahidis et al., 2014).

At the genetic level, microbes can promote bacteria-induced alterations in nuclear genes that code for the production of sex pheromones in animals or that generate molecules that act as sex attractants (Brucker & Bordenstein, 2012). It should be noted that the type of bacteria that are known to possess some of the most robust psychoactive properties are overrepresented in the human reproductive organs (e.g., species of Lactobacillus; Perez-Burgos et al., 2013). This suggests that bacterial transfer via intimate or sexual contact may be especially likely to influence downstream neurological and psychological processes. Thus, in an effort to support their own reproductive success, bacteria have also shaped the traits and behaviors that drive the reproductive success of their hosts.

Gestation, birth, and early infancy

The gut and vaginal microbiomes undergo considerable changes during pregnancy, resulting in a strong maternal signature on the microbial profile of newborns (Palmer, Bik, DiGiulio, Relman, & Brown, 2007). The human uterus was previously thought to be sterile, but recent research suggests that bacteria may actually be transferred to the fetus via the bloodstream and placenta while still in utero (Jiménez et al., 2008; Rautava, Collado, Salminen, & Isolauri, 2012; Satokari et al., 2009). Although this work is still in its early stages, researchers have now catalogued the presence of numerous bacteria isolated in the umbilical cord, suggesting that they were translocated from the mother to the fetus during pregnancy (Matamoros, Gras-Leguen, Le Vacon, Potel, & de La Cochetiere, 2013). More recent attempts to replicate these findings have not always been successful, however, and some researchers have argued that the bacterial presence in the placenta could be caused by contamination of the sample rather than translocation from mother to baby (de Goffau et al., 2018). Regardless, the microbiome of a pregnant woman experiences several pronounced changes throughout gestation that can affect the fetus. Specifically, there is a decrease in proinflammatory bacteria and an increase in anti-inflammatory bacteria from the first to third trimester, as well as a massive proliferation of bacteria that expertly assist in nutrient uptake (Koren et al., 2012; Mueller, Bakacs, Combellick, Grigoryan, & Dominguez-Bello, 2015). This trajectory of microbial development is an example of how bacteria respond to the physiological needs of the host and aid in the safe delivery and development of the baby.

Psychological factors also have an important effect on the microbiome during pregnancy. Maternal stress in particular influences both the microbiome of the mother as well as the fetus. For example, the self-reported stress of mothers during pregnancy—as measured by the Pregnancy-Related Anxiety Questionnaire (PRAQ-R; Van den Bergh, 1990) and the reported frequency of pregnancy-related hassles—predicted differential gut microbiota in infants as well as their susceptibility to GI disease and allergic reactions (Zijlmans, Korpela, Riksen-Walraven, de Vos, & de Weerth, 2015). Maternal stress has also been positively correlated with the prevalence of bacteria associated with respiratory symptoms, HPA reactivity (Golubeva et al., 2015), and the disruption of amino-acid profiles in the developing brain (Jašarević, Howerton, et al., 2015). One possible mechanism of action is that emotional and physiological stress influence the microbial population of the vagina, decreasing commensal and increasing pathogenic bacteria, which then seed the infant’s microbiome as it passes through the birth canal (Jašarević, Howerton, et al., 2015).

A variety of behavioral factors are also known to affect the infant microbiome. For example, the mode of delivery (vaginal vs. cesarean section), the location of delivery (hospital vs. home birth), whether caregivers breastfeed or provide the infant with formula, and whether mothers and babies receive any type of antibiotic treatment early in life all have a lasting impact on the microbiome. Infants born vaginally tend to have microbiomes dominated by Lactobacillus strains from their mothers’ vaginas and fecal matter (Dominguez-Bello et al., 2010; Mueller, Bakacs, et al., 2015; Vaishampayan et al., 2010), whereas the microbiomes of infants born via caesarian section are largely populated by bacteria found on the skin and immediate external environment—such as the room in which the baby was delivered (Dominguez-Bello et al., 2010). Relative to their formula-fed counterparts, infants that are breastfed exhibit a healthy population of Bifidobacterium and Lactobacillus species (Fernández et al., 2013; Jiménez et al., 2012). These bacteria can enhance the ability to uptake and process nutrients (M. S. Kramer et al., 2008; Roger, Costabile, Holland, Hoyles, & McCartney, 2010), which may in turn lead to positive neurodevelopmental outcomes (e.g., improved intelligence and cognitive abilities; M. S. Kramer et al., 2008). Last, the administration of antibiotics during gestation and infancy is robustly associated with decreased bacterial diversity—including and especially the depletion of beneficial bacteria (Mueller, Whyatt, et al., 2015). Although bacteria colonies may repopulate eventually, massive bacterial extinction during these critical periods of development may have long-term deleterious effects on a range of neurological and psychological outcomes.

Adolescence and early adulthood

Adolescence is a transitional period characterized by major neuroanatomical changes that can partly be mapped onto the developmental changes observed in bacterial populations. For example, two of the most critical alterations in the human cortex during late childhood/early adolescence involve axonal myelination and synaptic pruning. Research with germ-free animals has implicated bacteria in both of these processes. Specifically, animals bred without a microbiome show hypermyelination in the prefrontal cortex (Hoban et al., 2016) as well as defective microglia cells (Erny et al., 2015), which are crucial for pruning back neurons during development (Paolicelli et al., 2011; Schafer et al., 2012). Of course, the disruption of neurological development during adolescence as a result of bacterial imbalances will likely have adverse consequences for concomitant cognitive developments, and several experimental studies offer supporting evidence. For example, infecting healthy rodents with harmful bacterial strains can lead to learning deficits (Gareau et al., 2011; Li et al., 2009), whereas administering healthy bacteria to infected mice can ameliorate symptoms of cognitive dysfunction (Bercik, 2011; Gareau, Jury, MacQueen, Sherman, & Perdue, 2007).

As mentioned throughout this review, environmental and psychological stressors alter people’s microbiota, and adolescence is a developmental period rife with these types of assaults. Changes in diet, disrupted sleep patterns, access to drugs and alcohol, and navigating sexual relationships all influence the microbiome, and all typically have onsets during adolescence (De Filippo et al., 2010; Iyer & Vaishnava, 2016; Mutlu et al., 2009). Thus, adolescence is a stage during which neurological, biological, social, and bacterial changes all unfold in concert with one another in a manner that ultimately sets the stage for healthy or pathological development well into adulthood.

Aging populations and older adults

By 2030, people over the age of 60 years will outnumber children under the age of 9 years, and by 2050, there will be more people over the age of 60 years than people between the ages of 10 and 24 years (United Nations, 2015). Although this area of research is in its nascent stages, physiological processes in these older adult populations are especially likely to be connected to the microbiome. For example, inflammation worsens as the body ages (Guigoz, Doré, & Schiffrin, 2008), and bacteria play an important role in regulating immune responses at the systemic and molecular level. In addition, gut motility decreases as individuals age, leading to changes in nutrient uptake, extended digestion times, and irregularity in bowel movements (Camilleri, Lee, Viramontes, Bharucha, & Tangalos, 2000). All of these changes affect the composition of the microbiome and the proportion of bacteria colonizing people’s intestines. Although seemingly trivial, chewing strength may also decline with age, which can change the integrity and type of food the intestines are challenged to digest (Watanabe, 1998). This can lead to different and prolonged functional demands on the bacteria tasked with decomposing the food and extracting appropriate nutrients. Medication regimens as people age may include the prolonged use of antibiotics, which dramatically influence bacteria populations. For all of these reasons, the gut microbiota of human adults approximately 70 years of age and older is significantly different from that of younger adults and infants (O’Toole & Claesson, 2010). Specifically, researchers have observed a decline in the bacterial diversity of older human populations (70–99 years old), and, not surprisingly, this decline was comorbid with decreases in brain weight, brain volume, and cognitive impairments (Biagi et al., 2010, 2013; Claesson et al., 2012).

Critical windows and sensitive periods

Microbial and neurological development unfold together and thus may have overlapping windows of time during which they are more or less responsive to change and intervention. Specifically, emotional and physiological stress during early life have a substantial impact on both the brain and microbiome, and disruptions may be difficult to normalize if interventions are applied too late in development. For example, colonizing mice with commensal bacteria shortly after birth corrected deficits in the development of the HPA axis caused by restraint stress, whereas colonizing mice later in life did not produce the same therapeutic outcomes (Sudo et al., 2004). Likewise, deficits in social-cognitive processes in mice established early in life (because of an absence of key bacterial populations) could not be reversed by bacterial interventions later in life (postweaning), although some behaviors (such as social avoidance) were more amenable to treatment (Desbonnet et al., 2014).

Adolescence may also constitute a sensitive developmental period. For example, bacterial profiles were significantly altered in rats exposed to exercise during their juvenile years, whereas exercise exposure during their adult years showed no such effect. Furthermore, the exercise-induced changes in the microbiota of these adolescent rats were associated with an increase in lean body mass (Mika et al., 2015), suggesting that earlier (but not later) exercise-related interventions may facilitate the development of a microbiome that promotes healthy weight maintenance. Likewise, colonization with commensal bacteria during the adolescent period reversed stress hyperresponsivity in mice (i.e., reversed the prolonged elevation of adrenocorticotropin, or ACTH, after a stressful event). However, bacteria colonization introduced in later adulthood seemed to be ineffective at ameliorating these stress-induced elevations in ACTH (Neufeld, Luczynski, Oriach, Dinan, & Cryan, 2016). These findings suggest that early adolescence is another critical window of development during which gut bacteria can shape the stress response.

It is important to note that despite having sensitive periods of development, the microbiome may be more flexible across the life span and more amenable to change than other physiological systems. Researchers are currently examining a variety of restorative interventions geared toward transforming the microbiome after it has been altered by early-life stress and assault (Cowan, Callaghan, & Richardson, 2016; Mueller, Bakacs, et al., 2015; Salazar et al., 2014). Several preclinical bacteria-based treatments are being investigated for their therapeutic and medical properties (Dinan, Stanton, & Cryan, 2013). Indeed, the microbiome may offer a pathway to correct neurodevelopmental pathologies by acting on elements that influence function and process (e.g., neurochemicals excreted by bacteria) rather than relatively unalterable morphological structures (e.g., brain tissue).

Autism spectrum disorder

The microbiome has recently been placed at the forefront of autism spectrum disorder (ASD) research, and although the role of the microbiome in ASD has enough complexity for its own review article (see Luna, Savidge, & Williams, 2016), some general conclusions can be summarized here. Germ-free mice show similar social deficits and repetitive behaviors as those observed in humans with ASD (Desbonnet et al., 2014), and mouse models of ASD have an altered microbiota (de Theije et al., 2014), suggesting that when beneficial bacteria are missing ASD-like deficits can appear. Because ASD is so diverse and heterogeneous in nature, it is difficult for specific microbiota-related findings to generalize across the entire spectrum (as addressed in Song, Liu, & Finegold, 2004). Despite this limitation, some findings do have more general implications for ASD. For example, rates of GI pathologies and symptoms are disproportionately high in people with ASD compared with neurotypical populations (de Theije et al., 2011). Probiotics, prebiotics, and antibiotics have all shown behavioral benefits for people with ASD (Critchfield, van Hemert, Ash, Mulder, & Ashwood, 2011; Grimaldi et al., 2018), but whether this is due to the bacteria themselves or the natural cycle of ASD is still unclear. All of these findings do, however, suggest that a baseline dysbiotic state may exist in those with ASD.

Clinical depression and anxiety

There has been an abundance of research on the ways in which the microbiome shapes depression and anxiety (for recent and thorough reviews, see Foster & Neufeld, 2013; Huang, Wang, & Hu, 2016). However, most of this work has been conducted with rodents rather than humans. Nonetheless, some recent research in humans has found that depression and anxiety symptoms increase stepwise as GI symptoms and severity increase (Pinto-Sanchez et al., 2015). GI illness is also highly comorbid with clinical depression and anxiety (Maes, Kubera, Leunis, & Berk, 2012; Maes et al., 2013). In addition, probiotics have been found to improve both GI and clinical depression and anxiety symptoms for those with disorders such as irritable bowel syndrome (Pinto-Sanchez et al., 2017; Wallace & Milev, 2017). Although under some circumstances probiotics can have antidepressant or anxiolytic effects robust enough to relieve clinically classified symptoms in humans, how this happens and which combination of probiotics may be most optimal for treatment requires further investigation.

Posttraumatic stress disorder

The microbiome is implicated in both the normative and hyperactivated stress response in humans (R. T. Liu, 2017) and animals (e.g., van de Wouw et al., 2018). Given that severe and prolonged stress is a central feature of posttraumatic stress disorder (PTSD), researchers have begun to explore the ways in which the microbiome is related to the prevalence and time course of this disorder (Malan-Muller et al., 2018). For example, in a recent exploratory study researchers reported that individuals with a PTSD diagnosis had lower proportions of three distinct phyla of bacteria (Actinobacteria, Lentisphaerae, and Verrucomicrobia) compared with controls who had also been exposed to a traumatic event but did not meet the clinical threshold for a PTSD diagnosis. In addition, increased scores on the Clinician-Administered PTSD Scale (CAPS-5; Weathers et al., 2013) were associated with decreases in the three bacterial taxa (Hemmings et al., 2017).

Developing future clinical treatments

In addition to the topics reviewed above, clinical research has also been conducted on how the microbiome may be associated with many other challenges related to physical and mental health, including chronic fatigue syndrome (Jackson, Butt, Ball, Lewis, & Bruck, 2015; Rao et al., 2009), anorexia (Armougom, Henry, Vialettes, Raccah, & Raoult, 2009), alcoholism (Engen, Green, Voigt, Forsyth, & Keshavarzian, 2015; Mutlu et al., 2009; Yan & Schnabl, 2012), schizophrenia (Severance, Yolken, & Eaton, 2016), and manic depression (Dickerson, Severance, & Yolken, 2017; Kanji et al., 2018). It is clear from this wide range of examples that factors that affect the microbiome are relevant to physical and mental health. Exactly which mechanisms are at play, how these mechanisms operate, and what route may be most optimal for treatment is still being debated, but it seems that the administration of probiotics has the strongest impact on behavioral change. Live bacteria that, when administered in a clinical or medical context, have a significant beneficial effect on the host have been termed psychobiotics (Bambury, Sandhu, Cryan, & Dinan, 2018; Dinan et al., 2013; Misra & Mohanty, 2017; Wall et al., 2014). Although there is still a need for more randomized, controlled, clinical trials investigating the impact of psychobiotics on humans (Anglin, Surette, Moayyedi, & Bercik, 2015), their therapeutic potential for a broad range of mental-health experiences is a promising future direction for pharmacologically driven treatments.

Why are microbiologists interested in the vagus nerve?

The vagus nerve plays a central role in conveying information about the state of the GI tract, as well as other organs, to the CNS, with approximately 20% of its fibers dedicated to GI–CNS communication (Forsythe, Bienenstock, & Kunze, 2014; Stakenborg, Di Giovangiulio, Boeckxstaens, & Matteoli, 2013). Note that certain bacteria that have colonized the gut may only be capable of exerting effects on mental and neurological processes if the vagus nerve is functionally intact. That is, the vagus nerve may constitute a direct line of communication between bacteria and the brain. Microbiologists have reached this conclusion after conducting multiple studies demonstrating that surgically severing the vagus nerve eliminates, or in some cases reverses, both the positive psychological effects of commensal bacteria and the negative psychological effects of pathogenic bacteria. For example, severing the vagus nerve of mice eliminated the antianxiety effects of at least two species of commensal bacteria—L. rhamnosus (Bravo et al., 2011) and B. longum (Bercik, Park, et al., 2011)—as well as the antidepressant effects of L. rhamnosus (Bravo et al., 2011). Likewise, when mice were colonized with pathogenic bacteria, severing the vagus nerve prevented maladaptive behavior from emerging, whereas administering pathogenic bacteria to mice with intact vagus nerves led to increased anxiety-like behaviors (Bercik, Park, et al., 2011). Taken together, there is compelling preliminary evidence that one process by which bacteria influence psychological outcomes depends, at least in part, on vagal integrity.

Vagal influence on psychological outcomes

Over the past 20 years, psychological scientists have implicated the vagus nerve in a range of psychological outcomes that are related to those currently being examined by microbiologists. For example, in human subjects, higher heart-rate variability (a measure of healthy vagal functioning) is associated with lower levels of trait anxiety (Svendsen et al., 2016) as well as a more habitually positive mood (e.g., cheerfulness, calmness), higher life satisfaction (Geisler et al., 2010), and more competent social communication (Quintana et al., 2012; Shahrestani et al., 2015). People with greater vagal flexibility (another measure of adaptive vagus nerve functioning) exhibit more accurate perceptions of social-emotional cues and are more sensitive to their social context (Muhtadie et al., 2015). In a longitudinal intervention-based study designed to improve psychological well-being, individuals who had higher baseline vagal tone (an index of both vagal and cardiovascular health) showed significantly larger increases in positive emotions compared with individuals who had low vagal tone (Kok et al., 2013). Furthermore, researchers have found that vagus nerve stimulation can reduce symptoms of depression and anxiety and point to the gut–brain axis as an important mechanism of action (Kong, Fang, Park, Li, & Rong, 2018).

Psychological scientists have thus repeatedly demonstrated that there is a link between the vagus nerve and emotion, cognition, and behavior, whereas microbiologists have demonstrated that the vagus nerve mediates the relationship between bacterial communities and these same outcomes. Psychological scientists have often situated their findings in the context of broader physical health (e.g., HPA reactivity), and given the overlap between psychological and microbial outcomes outlined above, it may be a particularly appropriate and fruitful time to begin considering how the bacteria that people share their bodies with interact with the vagus nerve and, in turn, influence outcomes of great interest to psychological science.

Other pathways of bacteria-brain communication

Researchers are also beginning to identify several additional mechanisms by which the microbiome influences neurological and psychological outcomes (see Fig. 1; for a detailed and accessible review, see Wiley et al., 2017). For example, bacterial communities can interface with the CNS by excreting metabolites with neuromodulatory properties and release hormone-signaling peptides in the intestines that act on both sympathetic and parasympathetic pathways (Schele et al., 2013; Wren & Bloom, 2007). Different species of bacteria are also capable of interacting with the immune system at the molecular level, resulting in changes to the circulation of pro- and anti-inflammatory cytokines that influence CNS functioning (Dantzer, 2009; Dinan & Cryan, 2017; Petra et al., 2015). Furthermore, dysbiosis or an overgrowth of pathogenic bacteria can lead to increased permeability of both the intestinal lining (e.g., “leaky gut”; Hollander, 1999) and the blood-brain barrier (Logsdon, Erickson, Rhea, Salameh, & Banks, 2018), accelerating the rate and proportion of microbial infiltration (de Theije et al., 2011). As the science of the gut–brain axis advances, microbiologists, neuroscientists, immunologists, and medical professionals will continue to uncover more detail concerning these pathways of communication—and psychological scientists motivated to deeply understand the mind-body connection will undoubtedly have important contributions to make.