Cultivating a healthy neuro-immune network: A health psychology approach

Abstract

The field of psychoneuroimmunology (PNI) examines interactions among psychological and behavioral states, the brain, and the immune system. Research in PNI has elegantly documented effects of stress at multiple levels of the neuro-immune network, with profound implications for both physical and mental health. In this review, we consider how the neuro-immune network might be influenced by “positive” psychological and behavioral states, focusing on positive affect, eudaimonic well-being, physical activity, and sleep. There is compelling evidence that these positive states and behaviors are associated with changes in immune activity in the body, including reductions in peripheral inflammatory processes relevant for physical health. Growing evidence from animal models also suggests effects of positive states on immune cells in the brain and the blood-brain barrier, which then impact critical aspects of mood, cognition, and behavior. Tremendous advances are being made in our understanding of neuro-immune dynamics; one of the central goals of this review is to highlight recent preclinical research in this area and consider how we can leverage these findings to investigate and cultivate a healthy neuro-immune network in humans.

The field of psychoneuroimmunology (PNI) examines interactions among psychological and behavioral states, the brain, and the immune system, and the implications of those interactions for physical and mental health. Although historically the brain and the immune system were considered independent systems, research in PNI has documented their complex connections at multiple levels. To date, much of the research in PNI has focused on “negative” psychological states and their links with immunity. In particular, PNI investigators have for many years been interrogating the effects of psychological stress on the peripheral immune system, elucidating a key pathway linking stress and physical health conditions such as cardiovascular disease, cancer, asthma, and viral illnesses (Glaser & Kiecolt-Glaser, 2005). Another major area of research in PNI has examined how the peripheral immune system “talks” to the brain and drives changes in mood, energy, and social behavior, with implications for depression and other psychiatric conditions (Dantzer, O’Connor, Freund, Johnson, & Kelley, 2008). More recent work has focused on immune cells in the brain and their effect on neural function, cognition, and behavior as well as neurological conditions such as Alzheimer’s disease, dementia, and autism (Bilbo & Schwarz, 2012; Filiano, Gadani, & Kipnis, 2017).

Clearly, the complex interactions between the brain and the immune system are highly relevant for our physical and mental health. In this review, we consider how the neuro-immune network might be influenced by “positive” psychological states and behaviors, including positive affect, eudaimonic well-being, physical activity, and sleep. We examine the growing body of evidence that these positive factors can impact the peripheral immune system, focusing on inflammation. In addition, we consider how positive processes may interact with other components of the neuro-immune network that have received less attention in psychological and behavioral science, including immune cells in the brain and the blood-brain barrier (BBB; see Table 1 for key biological terms that will be used throughout this review and their definitions). Finally, we consider emerging work on other cells and systems that regulate the neuro-immune network, including the microbiome and T cells. Tremendous advances are being made in our understanding of neuro-immune dynamics; one of the central goals of this review is to highlight recent preclinical research in this area and consider how we can leverage these findings to investigate and ultimately cultivate a healthy neuro-immune network in humans.

Table 1.

Key biological terms

Term	Definition
Blood-brain barrier (BBB)	A semi-permeable barrier between the circulating blood and the brain. The BBB is formed by tight connections between endothelial cells that restrict the passage of molecules from the bloodstream into the brain.
Brain-derived neurotrophic factor (BDNF)	A protein involved in neuroplasticity and neurogenesis.
Lymphocyte	Type of immune cell that includes natural killer cells, B cells, and T cells.
Meninges	Three membranes (pia mater, arachnoid mater, dura mater) that surround and protect the brain and spinal cord.
Microglia	Resident immune cells of the brain, related to macrophages.
Monocyte/Macrophage	Type of immune cell that is an important source of inflammation. Monocytes circulate in the blood and become macrophages when they migrate into the tissue.
Nuclear factor-κB (NF-κB)	Intracellular transcription factor that is a key mediator of inflammation.
T cells	Type of lymphocyte that plays an important role in adaptive immunity.
Treg (T regulatory) cells	Type of T cell that modulates immunity and inflammation. Tregs are important for suppressing pathological immune responses.

Interactions between the nervous system and the immune system under stress

In this section, we describe the basic pathways linking the brain and immune system with a focus on how these pathways are influenced by stress. Although our primary interest is in positive psychological states and behaviors, we use stress paradigms (which have been much more carefully examined) to describe some of the intricate connections between the immune system and brain. This provides a framework for understanding how neuro-immune processes might be influenced by positive psychological and behavioral processes. In addition, we focus on inflammation as a key immunological process relevant for both mental and physical health.

As shown in Figure 1, the brain sends signals to the immune system through neuroendocrine pathways, including the hypothalamic-pituitary-adrenal (HPA) axis and the autonomic nervous system (ANS). These systems release chemical messengers that are “heard” by immune cells and lead to changes in immune function, including effects on inflammation (Irwin & Cole, 2011). In turn, the immune system sends signals to the brain about infection, injury, and repair processes occurring in the body. One of the key immune-to-brain signaling pathways occurs through release of pro-inflammatory cytokines by peripheral immune cells, which act through neural and humoral (blood borne) pathways to influence neural function and behavior. The effects of pro-inflammatory cytokines have been elegantly demonstrated in preclinical and clinical studies in which exposure to a peripheral immune stimulus leads to changes in activity/fatigue, mood, cognitive function, and social interaction/connection (Dantzer et al., 2008; Eisenberger, Inagaki, Mashal, & Irwin, 2010). Together, these changes are described as “sickness behavior” as they are commonly experienced across different types of illness. Importantly, many of these changes are consistent with those experienced in the context of depression and other psychiatric conditions, and inflammation is increasingly thought to play a role in core dimensions of human behavior and functioning (Dooley et al., 2018; Miller & Raison, 2016).

Figure.

The two panels of the figure illustrate key components of the neuro-immune network in the context of stress (left panel) and in the context of positive psychological processes and behaviors (right panel). The network can be considered at three levels: 1) the peripheral immune system; 2) immune cells in the brain; and 3) signaling pathways between the peripheral immune system and the brain, including the blood brain barrier (BBB). As depicted in the left panel, stress activates the sympathetic nervous system (SNS) and the hypothalamic pituitary adrenal (HPA) axis, which release chemical messengers that influence peripheral immune organs (e.g., bone marrow) and cells (e.g., monocytes). Both acute and chronic stressors are associated with higher numbers of pro-inflammatory monocytes and elevated concentrations of pro-inflammatory cytokines in the periphery. In addition, stress activates immune cells in the brain called microglia and can disrupt the structure and function of the BBB. Together, these changes play a central role in the mental and physical health effects of stress. In contrast, the right panel of the figure depicts a “healthy” neuro-immune network, which is characterized by lower levels of inflammation in the periphery and the brain. In this context, we see decreased activation of the SNS and HPA axis, fewer pro-inflammatory monocytes and lower levels of pro-inflammatory cytokines in the periphery, microglia in their resting state, and BBB integrity. Positive psychological processes and behaviors have demonstrated effects at each level of the network, which may account for their beneficial effects on health and well-being.

The communication pathways between the brain and immune system are an important part of the organism’s defense system and help mobilize appropriate action (immunologically and behaviorally) during times of threat. Indeed, the interactions between the nervous system and immune system have primarily been characterized using stress models. One of the most elegant and detailed neuro-immune models of stress is the repeated social defeat (RSD) paradigm developed by Sheridan and colleagues (Avitsur, Stark, & Sheridan, 2001; Reader et al., 2015). The RSD paradigm involves placing a large, aggressive rat into the resident rats’ cage for 2 hours each day for 1 week. During these 2-hour episodes, the aggressor rat attacks the residents into submission. RSD causes resident rodents to release immature, pro-inflammatory immune cells from the bone marrow, promoting inflammation at the site of injury (Powell et al., 2013). RSD also leads to activation of microglia, the resident immune cells of the brain, which enhance neuroinflammation and facilitate the recruitment of pro-inflammatory monocytes across the BBB and into the brain. Microglial activation and monocyte infiltration in the central nervous system (CNS) underlie behavioral changes observed among animals undergoing RSD, including anxiety-like behavior (Reader et al., 2015). Thus, the RSD stress paradigm influences immune cells in the brain and body as well as the signaling between these systems, with effects on neural function and behavior that may persist for weeks after termination of the stressor and can be re-activated by later stress exposure (Wohleb et al., 2014).

The RSD paradigm is useful as it identifies multiple levels of the neuro-immune network that can be influenced by stress: 1) the peripheral immune system; 2) immune cells in the CNS; and 3) signaling pathways between the CNS and immune system and the BBB. We use these three levels throughout the review to organize research linking psychological and behavioral states to components of the neuro-immune network. Although some of the changes seen in the RSD paradigm are specific to that model (for example, biting-induced injuries play an important role), many of the effects are observed in other stress models, as we describe below.

Beginning with level 1, a robust body of preclinical and clinical work has documented effects of stress on the peripheral immune system. In human studies, there is evidence that acute stressors increase peripheral inflammation, as demonstrated by increased expression of pro-inflammatory cytokine genes in immune cells, increased ability of immune cells to make pro-inflammatory cytokines, and increased concentrations of pro-inflammatory cytokines in circulation (Bierhaus et al., 2003; Marsland, Walsh, Lockwood, & John-Henderson, 2017; McInnis et al., 2015). Similar effects are observed following chronic stress exposure, including stressors such as caregiving, social isolation, low socioeconomic status, and early life adversity (Cole, 2014; Kiecolt-Glaser et al., 2003; Miller, Chen, & Parker, 2011), indicating that stress can leave a lasting imprint on peripheral immune cells. Importantly, cells that are in a more pro-inflammatory state can influence subsequent stress responses; for example, preclinical models have shown that animals whose immune cells have more pro-inflammatory tendencies at baseline show strong inflammatory and depressive-like responses to social stress (Hodes et al., 2014). Thus, both acute and chronic stressors can influence the inflammatory potential of peripheral immune cells.

At the second level, effects of stress on immune cells in the brain have been observed in other animal models of stress and may contribute to stress effects on behavior (Tian et al., 2017; Walker, Nilsson, & Jones, 2013). For example, research by Herkenham and colleagues using a modified model of social defeat stress that does not involve biting has demonstrated increases in microglial activation and associated increases in depressive behavior that are not dependent on monocyte trafficking to the brain (Lehmann, Cooper, Maric, & Herkenham, 2016). Of note, animals exposed to this paradigm do not uniformly develop depressive-like behaviors; those who are “stress susceptible” (the majority of animals) show increases in inflammation-related microglia gene expression, whereas “resilient” animals show activation of gene transcripts implicated in cellular and behavioral plasticity to stress (Lehmann et al., 2018). Chronic unpredictable stress also induces microglial activation which mediates stress effects on anxiety and depressive-like behavior (Wang et al., 2018). Prenatal and early life stress exposure lead to elevated microglial activity (Calcia et al., 2016), which may persist over time and have long-term consequences for neural function, mood, and cognition (Bilbo & Schwarz, 2012). Indeed, this is one pathway through which early life stress may get “under the skin” and, in this case, into the brain with detrimental effects on later-life cognitive and mental health.

At the third level, animal models have demonstrated effects of social defeat on the BBB, which is in turn associated with expression of depression-like behaviors (Menard et al., 2017). Indeed, there is evidence that changes in BBB permeability (Lehmann et al., 2018) and signaling (Reader et al., 2015) play a key role in stress-related changes in behavior. Although there is little human work investigating stress effects on the BBB given challenges of measuring BBB integrity using peripheral markers, there is evidence that early life stress may amplify neuro-immune signaling such that the brain is more sensitive to peripheral inflammation (and vice versa) (Kuhlman, Chiang, Horn, & Bower, 2017; Miller & Cole, 2012). This suggests that the same level of peripheral inflammation can lead to different neural and behavioral effects, which may be related to changes in BBB integrity and function.

Overall, stress has been shown to influence immune cells in the periphery and the brain and the functioning of the BBB, with profound implications for behavior. In the sections below, we consider how positive psychological and behavioral factors may influence these same neuro-immune processes, focusing on positive affect, eudaimonic well-being, physical activity, and sleep. An important question is whether these positive factors influence neuro-immune processes directly (i.e., at rest) and/or serve to buffer the effects of stress. We take a broad view of neuro-immune health that includes decreased inflammation in the periphery and brain and tighter regulation of the BBB both at rest and in response to physical or psychological stress. Thus, we consider studies that examine direct effects of positive psychological states and behaviors on neuro-immune outcomes, as well as studies that use models of stress, injury, or illness to interrogate buffering effects.

Positive psychological states and the neuro-immune network

There is compelling evidence that positive psychological states are associated with level 1 of our model and specifically with reduced peripheral inflammation. Much of the research so far has focused on positive affect, which refers to feelings of pleasurable engagement in response to one’s environment (e.g., happiness, contentment) (Clark, Watson, & Leeka, 1989). Studies suggest that higher levels of positive affect are associated with lower levels of circulating inflammatory markers, including IL-6, TNF-α, and CRP (Brouwers et al., 2013; Moreno, Moskowitz, Ganz, & Bower, 2016; Steptoe, O’Donnell, Badrick, Kumari, & Marmot, 2008) and reduced stimulated cytokine production by immune cells, a measure of inflammatory potential (Prather, Marsland, Muldoon, & Manuck, 2007). Further, positive affect is associated with reduced inflammatory reactivity in response to psychosocial stress both in the laboratory (Aschbacher et al., 2012; Steptoe, Wardle, & Marmot, 2005) and in daily life (Blevins, Sagui, & Bennett, 2017; Sin, Graham-Engeland, Ong, & Almeida, 2015).

Of note, positive affect is a heterogeneous construct, measured in a variety of ways (Pressman, Jenkins, & Moskowitz, 2019). In some studies, for example, positive affect is measured via a 1-item response reflecting “happiness” (Steptoe et al., 2005) whereas in others, positive affect refers to a composite of various mood states, encompassing aspects of enthusiasm, energy level, and enjoyment (Prather et al., 2007). Positive affect can also refer to both low (e.g., contentment) and high (e.g., excitement) arousal states, though few studies separate these dimensions. Importantly, low and high arousal positive affect are differentially associated with neural and neuroendocrine processes (Gerber et al., 2008; Hoyt, Craske, Mineka, & Adam, 2015) and may also have distinct associations with inflammation (Moreno et al., 2016).

In addition to positive affect, a range of positive emotional experiences and resources have been linked with inflammation, including awe (Stellar et al., 2015), self-compassion (Breines et al., 2014), and optimism (Brydon, Walker, Wawrzyniak, Chart, & Steptoe, 2009; Roy et al., 2010). One construct of particular interest is eudaimonic well-being, which encompasses a sense of purpose and meaning in life, social embeddedness, and the potential for personal growth (Keyes, 2002; R. M. Ryan & Deci, 2001; Ryff, 2018). Studies suggest that aspects of eudaimonic well-being, including purpose in life and positive relations with others, are associated with lower circulating markers of inflammation (Friedman, Hayney, Love, Singer, & Ryff, 2007). Of note, these findings are consistent with a large literature linking social support and social integration with lower inflammatory cytokines (Uchino et al., 2018). In addition, eudaimonic well-being has been linked to genomic indicators of immunity, namely the conserved transcriptional response to adversity (CTRA). The CTRA is a gene expression profile that involves up-regulation of pro-inflammatory genes and down-regulation of genes involved in antiviral response and antibody synthesis in immune cells (Cole, 2009). Eudaimonic well-being is associated with lower CTRA gene expression in middle-aged (Fredrickson et al., 2013; Fredrickson et al., 2015; Kitayama, Akutsu, Uchida, & Cole, 2016) and older adults (Cole et al., 2015) that remains when controlling for hedonic well-being (positive affect) and various indices of distress.

The results presented thus far are largely correlational; however, there is preliminary evidence that interventions designed to enhance positive psychological states can influence inflammatory activity. In a recent randomized controlled trial, healthy adults instructed to perform prosocial acts previously linked to increases in eudaimonic well-being exhibited decreased CTRA gene expression compared to those in the control condition (Nelson-Coffey, Fritz, Lyubomirsky, & Cole, 2017). In a small pilot study of patients with heart failure, those assigned to write gratitude journals showed greater reductions in inflammation relative to a treatment-as-usual control (Redwine et al., 2016), though effects of gratitude writing on inflammation are not uniform (Moieni et al., 2018). Further, a study conducted with adolescents found that volunteering led to decreases in the pro-inflammatory cytokine IL-6 as well as positive changes in other cardiovascular risk factors (Schreier, Schonert-Reichl, & Chen, 2013).

Research on mind-body interventions may also be relevant for establishing causal links between positive psychological states and inflammation. In particular, mindfulness-based interventions have been shown to enhance positive affect, eudaimonic well-being, self-kindness, and other positive psychological states (Bower et al., 2015; Labelle, Lawlor-Savage, Campbell, Faris, & Carlson, 2015) and to influence inflammatory biology, including decreases in pro-inflammatory gene expression (Bower & Irwin, 2016; Creswell et al., 2012). However, few studies have directly examined changes in positive processes as mediators of intervention effects. In a recent study of mindfulness meditation for younger breast cancer survivors, we found that increases in eudaimonic well-being were associated with decreases in CTRA gene expression whereas decreases in depressive symptoms were not (Boyle, Cole, Dutcher, Eisenberger, & Bower, 2019), supporting the hypothesis that positive states may play an important role in regulating the peripheral immune system.

Taken together, these studies provide preliminary support for the influential role of positive psychological states on inflammatory activity in the periphery, highlighting effects on level 1 processes. Research has demonstrated both direct and stress-buffering effects of positive states on peripheral inflammation, consistent with the broader literature on positive affect which suggests both main and moderated effects on a variety of health outcomes (Pressman et al., 2019). To date, there has been minimal examination of positive states in relation to the other two levels of our model: inflammatory processes in the CNS and neuro-immune signaling pathways/integrity of the BBB. However, positive psychological states and mindfulness-based interventions can influence neural processes (Creswell et al., 2016; Holzel et al., 2011; Kilpatrick et al., 2011) and could potentially influence activity of immune cells of the brain as well as sensitivity of the CNS to peripheral immune signaling. Indeed, mindfulness-based interventions are thought to work in part by regulating one’s responses to internal and external states, which might operate at the neuro-immune interface (Bower & Irwin, 2016; Holzel et al., 2011).

Healthy behaviors and the neuro-immune network

Two key health behaviors, physical activity and sleep, have been linked to functioning of the neuro-immune network with implications for health and well-being. In this section we review the preclinical literature using experimental animal models to interrogate effects of these behaviors on neuro-immune dynamics as well as studies that have translated those observations to humans.

Physical activity:

Physical activity has been extensively studied as a behavioral factor that can modulate inflammation and its potential consequences. In rodents, access to regular opportunities for exercise (e.g., running wheel) is linked to lower circulating markers of inflammation (Gleeson et al., 2011), demonstrating effects at level 1 of our model. Studies using animal models have also characterized effects of physical activity on inflammation in the CNS. More specifically, exercise appears to shift microglia from an inflammatory to a neuroprotective phenotype (Chirico et al., 2016; Parachikova, Nichol, & Cotman, 2008; Ryan & Kelly, 2016), the latter of which has been linked to better cognitive function and neurogenesis in aged mice (Chirico et al., 2016; Kohman, DeYoung, Bhattacharya, Peterson, & Rhodes, 2012). In addition, exercise following spinal cord injury reduces microglia activation during recovery and mitigates development of pain in a rodent model (Chhaya, Quiros-Molina, Tamashiro-Orrego, Houle, & Detloff, 2018).

Some of the most exciting work on behavioral influences on microglia comes from studies investigating environmental enrichment (EE), which includes physical activity along with intellectual stimulation and social interaction. EE has well-documented effects on neural and behavioral function, including enhancement of hippocampal neurogenesis and synaptic plasticity and improvements in learning and memory (Nithianantharajah & Hannan, 2006). EE also has effects on microglia, which are hypothesized to play an important role in these beneficial effects (Ziv et al., 2006). In addition, EE buffers against neuroinflammation induced by amyloid-β oligomers (involved in Alzheimer’s disease pathology) (Xu, Rajsombath, Weikop, & Selkoe, 2018) and by peripheral LPS and other insults (Williamson, Chao, & Bilbo, 2012). Thus, increasing exercise, cognitive, and social stimulation may have potent effects on microglia and neuroinflammation, with significant implications for age- and insult-related cognitive decline.

At level 3, physical activity can influence the integrity of the BBB in animal models, particularly in the context of illness or injury. In a mouse model of multiple sclerosis (MS), a chronic neuroinflammatory disease, strength and endurance training have been shown to slow disease progression in part by decreasing production of pro-inflammatory cytokines in the spinal cord and reducing BBB permeability (Souza et al., 2017). Exercise in the immediate aftermath of a stroke can prevent subsequent neurocognitive damage via effects on BBB efficiency (Zhang et al., 2013), and exercise interventions may mitigate the consequences of traumatic brain injury in part by fortifying the BBB (Archer, Svensson, & Alricsson, 2012; Mota et al., 2012). Further, voluntary wheel running prevented disruptions to the BBB following methamphetamine-induced oxidative stress (Toborek et al., 2013).

The observations made in animal models appear to be replicated in humans, particularly at level 1 of our model. It is well-established that individuals who exercise regularly have lower circulating and intracellular markers of inflammation (Gjevestad, Holven, & Ulven, 2015; Kasapis & Thompson, 2005; Petersen & Pedersen, 2005). There is also evidence that physical activity interventions, including aerobic exercise and strength training, can reduce markers of peripheral inflammation. One recent review of 83 randomized and non-randomized trials found an overall reduction in CRP following physical activity, albeit with a small effect size (Fedewa, Hathaway, & Ward-Ritacco, 2017). Other reviews have demonstrated similar effects of exercise within specific patient populations such as breast cancer survivors (Meneses-Echavez et al., 2016), patients with type 2 diabetes (Hayashino et al., 2014), and older adults (Sardeli et al., 2018). However, effects are mixed and may depend on the population, exercise intensity and duration, and the inflammatory marker assessed.

Studies looking at the impact of physical activity on integrity of the BBB and neuroinflammatory markers are limited in humans, given challenges in directly assessing these factors in vivo. To date, there are no well-established peripheral markers of BBB integrity, although neuronal-specific enolase (NSE) and S100β have been proposed as markers of BBB permeability. There is preliminary evidence that exercise may modulate these markers in humans, suggesting effects on BBB integrity in the expected direction. In a study of older women, exercise and exercise combined with taurine supplementation (a common amino acid in the body that declines with age) led to reduced peripheral inflammation and increased integrity of the BBB as measured by both S100β and NSE (Chupel et al., 2018). Similarly, exercise led to decreases in S100 (though no changes in NSE) among normal weight patients with MS (Mokhtarzade et al., 2018).

In sum, there is strong evidence from experimental models in rodents that physical activity impacts all levels of the neuro-immune network, with studies showing both direct and buffering effects on peripheral and central inflammation and the BBB. Studies conducted in humans also demonstrate effects of regular physical activity on peripheral markers of inflammation, with preliminary data suggesting effects on BBB integrity. Of note, the type and duration of physical activity may be an important moderator of these effects. For example, acute bouts of exercise have been linked to short-term increases in inflammatory markers (Barcelos et al., 2017; Verheggen et al., 2018) as well as markers of BBB permeability (Koh & Lee, 2014). The degree to which different types and intensity of physical activity influence the neuro-immune network is an important question for future research.

Sleep:

Sleep is an essential biological process with restorative functions that modulate mental and physical health (Underwood, 2013). Indeed, sleep disturbance is associated with increased risk for a variety of diseases as well as earlier mortality (Gallicchio & Kalesan, 2009). One explanation for the link between sleep and health is the impact of sleep disruption on inflammatory processes (Irwin, 2015). Experimental animal models demonstrate that sleep deprivation leads to increases in circulating markers of inflammation in the periphery (Gorczynski et al., 2005; Pandey & Kar, 2011). There is also evidence that sleep deprivation and sleep fragmentation increase the transport of inflammatory cytokines across the BBB, particularly in aged mice (Opp et al., 2015). In addition, prolonged sleep deprivation causes increased microglial activation in rodents which is also associated with the development of neurocognitive decline and anxious behavior (Bellesi et al., 2017; Wadhwa et al., 2018; Wadhwa et al., 2017).

In humans, there are well-documented links between sleep problems and elevated peripheral markers of inflammation (Irwin, Olmstead, & Carroll, 2016). Importantly, experimental sleep deprivation leads to acute increases in genomic markers of inflammation (Irwin, Wang, Campomayor, Collado-Hidalgo, & Cole, 2006), supporting a causal effect of sleep disturbance on inflammatory activity. Sleep loss has also been linked to disruptions in the BBB that lead to greater neuroinflammation (He et al., 2014; Hurtado-Alvarado, Domínguez-Salazar, Pavon, Velázquez-Moctezuma, & Gómez-González, 2016; Hurtado-Alvarado et al., 2013). These observations in both rodents and humans may explain why sleep disruptions are so closely linked to the development of both psychiatric and neurodegenerative disorders, such as depression, anxiety, and dementia.

Overall, work in preclinical and human models suggests an effect of sleep disturbance on all three levels of the neuro-immune network. There is also preliminary evidence that improving sleep can help reduce inflammation, at least in the periphery. Interventions targeting insomnia have demonstrated beneficial effects on sleep, depression, and other sickness behaviors as well as reductions in circulating, cellular, and genomic markers of inflammation (Black, O’Reilly, Olmstead, Breen, & Irwin, 2015; Irwin et al., 2014, 2017). However, the degree to which improvements in sleep mediate intervention effects on inflammatory markers has not yet been determined. Given growing evidence of the role of sleep in mental and physical well-being, the role of sleep interventions and sleep health more generally on the neuro-immune network is a promising focus for future research.

Emerging players in neuro-immune dynamics: gut microbiome and T cells

Thus far, our review has examined how positive states and behaviors modulate levels of the neuro-immune network, focusing on inflammation. Here, we highlight emerging work on other systems and cells that influence neuro-immune interactions: the gut microbiome and T cells. There is growing evidence that these factors influence the neuro-immune network in ways that are relevant for behavior. We review these studies with the goal of identifying new targets that can be used to promote mental and physical health.

Microbiome and the neuro-immune network:

Over the past decade, the role of the gut microbiome in mental and physical health has become an increasing focus of PNI research (Dinan & Cryan, 2017). The “microbiome” refers to the entire microbial population cohabitating our bodies, including microorganisms (bacteria, archaea, eukaryotes, and viruses), their genomes, and the surrounding environment (Marchesi & Ravel, 2015). One of the key pathways through which the gut microbiome may influence health is through effects on the neuro-immune axis (Liu, 2017). At level 1, the gut microbiota are known to impact peripheral immunity and inflammation (Haase, Haghikia, Wilck, Muller, & Linker, 2018), in part through effects on production of short chain fatty acids (SCFAs) and tryptophan metabolites. SCFAs, which are metabolites produced by bacteria from fermentation of dietary fiber, exert anti-inflammatory effects by inhibiting NF-κB signaling within immune cells, downregulating TNF production by peripheral immune cells, and increasing regulatory T (Treg) cell presence in the gut. Gut microbiota also metabolize the essential amino acid tryptophan, resulting in the production of the neurotransmitter serotonin, direct transformation into the aryl hydrocarbon receptor (AhR), or kynurenine production via the IDO1 pathway, all of which have effects on immunity and inflammation (Agus, Planchais, & Sokol, 2018). At level 2 of our model, recent work has highlighted effects of the microbiome on microglia maturation and function (Erny et al., 2015).

What is the evidence that microbiome effects on neuro-inflammatory processes are relevant for behavior? While numerous animal and human studies have linked the microbiome and psychological/behavioral outcomes, few have examined neuro-immune mediators. We highlight here select preclinical studies that document the important role of microbiota as regulators of both neuro-immune and behavioral processes. Of note, all of the studies reviewed interrogated these processes in the context of stress, although the effects of the gut microbiome may not be restricted to these contexts.

In a subordinate colony housing model of chronic stress, mice inoculated with a heat-killed strain of Mycobacterium vaccae (M. vaccae), a probiotic, displayed fewer submissive postures and anxiety-like behaviors following stress exposure; these effects were dependent on peripheral Treg activity (Reber et al., 2016). The probiotic also impacted immune cells in the CNS as indicated by increased microglial density in the prelimbic region of the medial prefrontal cortex. Similar effects were observed in a rodent model of chronic social defeat, in which probiotic treatment with Lactobacillus rhamnosus buffered stress effects on behavior through effects on peripheral Treg cells (Bharwani, Mian, Surette, Bienenstock, & Forsythe, 2017). More specifically, treatment with Lactobacillus rhamnosus increased expression of IL-10 (an anti-inflammatory cytokine) in Treg cells and led to fewer anxiety-related behaviors in defeated mice.

Stress buffering effects of probiotics have also been observed following acute stress. M. vaccae immunization prior to an acute stressor had anxiolytic behavioral effects, induced an anti-inflammatory hippocampal phenotype, and blocked microglial priming to immune challenge (Frank et al., 2018). Importantly, M. vaccae also increased hippocampal gene expression and levels of both IL-4 protein (an anti-inflammatory cytokine) and the antigen Cd200r1, which plays an integral role in microglial immunomodulation.

This initial work suggests that the microbiome can buffer stress-induced changes in behavior via effects on peripheral and central immunity. To our knowledge, the possibility that positive psychological states (positive affective states, eudaimonic well-being) may impact the gut microbiota has not been directly tested, though meditation has been proposed as a mechanism for maintaining a healthy gut microbiota (Househam, Peterson, Mills, & Chopra, 2017). The impact of physical activity and sleep on the gut microbiome has also received little attention, though initial evidence suggests a link between physical activity and gut microbial composition (Monda et al., 2017). Another important health behavior is diet, which can modulate microbial composition and diversity (Xu & Knight, 2015) with implications for health. For example, a recent study found that supplementation with omega-3 fatty acids in pregnant mice and their male offspring led to changes in gut microbial composition, peripheral inflammatory activity, cognition, and behavior (Robertson et al., 2017). Prebiotics and probiotics also directly modulate the microbiome and have demonstrated beneficial effects on peripheral immunity in animal models, though results from clinical trials in humans are more mixed (Quigley, 2019). Overall, despite the recent growth in gut-brain research, greater attention should be directed towards the immune mechanisms through which the microbiome exerts its influences on the brain and behavior, as well as studies on how we can effectively modulate our microbiomes to promote a healthy neuro-immune network.

Beyond inflammation: T cell influences on the brain and behavior

Thus far, our review has focused primarily on inflammation in the periphery and the brain given its relevance for physical and mental health. Here, we expand our focus to include exciting new work on another type of immune cell that influences the brain and behavior: T cells. T cells in the periphery are known to play a critical role in protection against illness; indeed, manipulation of T cells has led to recent breakthroughs in immunotherapy for cancer (Ribas & Wolchok, 2018). Although T cells do not typically enter the brain parenchyma under healthy conditions, they are found within the meninges, the three-layer membrane that surrounds the brain (Filiano et al., 2017). Their most likely route into the brain is through meningeal blood vessels, and recent work has shown that the cerebrospinal fluid that flows within the meninges drains into cervical lymph nodes through meningeal lymphatic vessels (J. Kipnis, 2016). Thus, the meninges serve as an interface between immune cells in the periphery and the CNS (i.e., a level 3 signaling pathway).

Intriguing work in preclinical models has shown that T cell migration to the meninges following stress is associated with reduced anxiety-like behavior and attenuated stress-related loss of brain-derived neurotrophic factor (BDNF) in the hippocampus (Lewitus & Schwartz, 2009). Further, transfer of T cells from chronically stressed animals to naïve animals renders the naïve animals less vulnerable to subsequent stress (Brachman, Lehmann, Maric, & Herkenham, 2015), suggesting that T cells may play a role in promoting stress resilience. Meningeal T cells have also been shown to support healthy cognitive function and social behavior (Filiano et al., 2017; Kipnis, Gadani, & Derecki, 2012). Effects of T cells on the brain are complex, as infiltrating T cells can play a pathological role in the context of neuroinflammatory diseases. However, work on meningeal T cells further supports the importance of the neuro-immune interface as a key regulator of behavior and suggests new pathways through which the immune system may modulate our thoughts, feelings, and behavior.

Conclusions

Our understanding of the complex interactions between the brain and the immune system has grown tremendously over the past several decades, along with recognition of the importance of these neuro-immune dynamics for health. Psychologists have been particularly interested in the effects of stress on the peripheral immune system and downstream consequences for physical health. The goal of this review was to examine the effects of positive psychological processes at multiple levels of the neuro-immune network, including peripheral immunity as well as immune cells in the CNS, and the blood-brain barrier. Because these levels are more difficult to assess in humans, we considered preclinical studies that have examined effects of healthy behaviors (physical activity, sleep) on neuro-immune function. In general, we found compelling evidence that positive psychological states and behaviors are associated with reductions in peripheral inflammation in humans, and that physical activity and sleep have effects at all three levels of the neuro-immune network in animal models. We also identified other systems that modulate neuro-immune dynamics and could potentially be influenced by these (and other) positive states and behaviors, including adaptive immunity (T cells) and the gut microbiome.

One of the implications of the neuro-immune framework is that the immune system can function most effectively when it receives signals from the brain that provide relevant information about the external environment (Irwin & Cole, 2011). Thus, when faced with an actual or perceived threat, it may be adaptive for the immune system to prepare for wounding and associated bacterial infection with a pro-inflammatory response. In contrast, positive psychological states might signal safe environmental conditions, decreasing peripheral stress response signaling and associated mobilization of the pro-inflammatory network. To the extent that positive states involve feelings of social connection or integration, this may shift the immune system away from inflammation and towards an antiviral response program, given elevated risk of viral infection among close others (Eisenberger & Cole, 2012). Of note, positive states may also activate the neural reward system, which has a direct effect on immunity (Ben-Shaanan et al., 2016). These dynamics may help explain why positive states have both direct and stress buffering effects on inflammation, as they may protect against stress-induced activation of inflammatory responses and steer the immune system to prepare for more relevant challenges.

This is an exciting time for psychoneuroimmunology and health psychology, with rapid developments in our understanding of neuro-immune interactions occurring alongside growing awareness of the importance of positive psychological processes as regulators of health and well-being. These advances suggest new models for conceptualizing and investigating psychosocial and behavioral effects on health that incorporate all levels of the neuro-immune network and associated systems. In addition, they broaden the range of psychological processes that may influence this network to include positive psychological states and behavior. There are a number of new therapies that are specifically designed to enhance positive affect and related states, including prosocial and positive psychology interventions, mind-body approaches, and neuroscience-based treatments (Craske, Meuret, Ritz, Treanor, & Dour, 2016). The possibility that these and other therapies may influence neuro-immune dynamics is an exciting prospect for future research. Ultimately, our goal is to understand and harness the power of neuro-immune interactions to enhance long-term mental and physical health.