‘As above, so below’ examining the interplay between emotion and the immune system

Abstract

While the concept of a palpable relationship between our mental and physical well-being is certainly not new, it is only in the light of modern scientific research that we have begun to realize how deeply connected our emotional and immune states may be. We begin this review with a series of studies demonstrating how four fundamental emotional responses: anger, anxiety, mirth and relaxation are able modulate cytokine production and cellular responses to a variety of immune stimuli. These modulations are shown to be either detrimental or beneficial to a patient's health dependent on the context and duration of the emotion. We also discuss the reverse, highlighting research demonstrating how the loss of key immune cells such as T lymphocytes in clinical and animal studies can negatively impact both emotional well-being and cognition. Additionally, to give a more complete picture of the manifold pathways that link emotion and the immune system, we give a brief overview of the influence the digestive system has upon mental and immunological health. Finally, throughout this review we attempt to highlight the therapeutic potential of this burgeoning field of research in both the diagnosis and treatment of immune and disorders. As well as identifying some of the key obstacles the field must address in order to put this potential into practice.

Introduction: defunct medicine

“The mind most effectually works upon the body, producing by his passions and perturbations miraculous alterations … cruel diseases and sometimes death itself”. Robert Burton, The Anatomy of Melancholy, (1621/1893).

Humorism is an ancient medical theory and philosophy centred around the concept that differing combinations of four key bodily fluids (blood, phlegm, black and yellow bile) have distinctive effects on human health and behaviour. Although now widely discredited, for over 2000 years humorism dominated medical practice across much of the civilized world. One of this theory's most famous proponents was the ancient Greek philosopher and paragon of early medicine Hippocrates. His teachings on the subject were still practiced 700 years later during the height of the Roman empire, when the prominent surgeon (and also philosopher) Galen would use humorism as the basis for his hypothesis that a ‘balance of the passions’ was fundamental to good physical health. Indeed up until the late nineteenth century intense emotional states and responses (such as grief, anger and anxiety) were still considered a leading cause of human morbidity and mortality. The precepts of humorism pervaded western culture for such a long time that many of the words we use today to describe our own personalities, such as sanguine, phlegmatic and melancholy have their origins in this ancient medical doctrine.

While undoubtedly erroneous as a form of medicine (though bloodletting appears to be making something of a comeback^1,2) a growing body of scientific evidence does back up a concept lying at the heart of humorist philosophy: that a link exists between emotional, physical and immunological health. In this review we will re-assess the Hippocratic theory of humorism in light of recent discoveries in the field of modern immunology, supporting the idea that changes in immune state can be detected by the nervous system and translated to a corresponding change in emotion. We will also examine the reverse: where distinct emotional states have been demonstrated to bring about alterations in the function of both the immune repertoire and response. Evidence will also be provided of pathophysiological conditions where an impaired emotional response could be considered a biomarker of physical disease. In doing so, we hope to highlight the potential therapeutic value of a patient's emotional state in the treatment of immune diseases, as well as speculate on the possibility of modulating the immune system to support mental well-being, most especially in the field of psychiatry.

Anger, laughter, stress and tai chi: emotion and the immune system

As well as its classical role in protecting against pathogens, the immune system has been established as playing a key part in governing homeostatic conditions linked to immune surveillance against inflammatory disease and tumorigenesis. Over several decades multiple regulatory pathways have been identified linking the immune with nervous and endocrine systems, mediated through the release of cytokines, endocrine hormones and neurotransmitter activity^3,4 (Fig.1). In turn a number of endocrine factors (such as oxytocin, cortisol and noradrenaline) are known to be produced in response to various emotional stimuli.^5–7 Many of these same stimuli have also been shown to elicit a striking effect on the immune response.

Figure 1.

The central nervous system (CNS) mediates the release of various immune influencing glucocorticoids via activation of a series of connected regions within the brain referred to as the hypothalamic–pituitary–adrenal axis. Research suggests the CNS to be primarily immune-suppressive in action, inhibiting production of pro-inflammatory cytokines, chemotactic factors and limiting migration and activation in several immune cell types.^111–114 Additionally the CNS elicits differential immune effects dependent on cell type and stage of development, inducing the expansion and migration of immature dendritic cells while seemingly promoting a tolerogenic phenotype in their mature counterparts.^113,115 Although less studied, several papers have demonstrated that the sympathetic nervous system is able to modulate immune activity through production of epinephrine and norepinephrine promoting a T helper type 2 (Th2) and Th17 phenotype in in T cells and dendritic cells, respectively.^116,117 Produced by the parasympathetic nervous system, acetylcholine has been shown to interact directly with multiple immune cell subsets through expression of acetylcholine receptors, leading to suppression of a number of pro-inflammatory pathways in macrophages and other immune cells.^118–120 Abbreviations: IFN-γ, interferon-γ; IL-17, interleukin-17; NF-κB, nuclear factor-κB; STAT, signal transducer and activator of transcription; Th1, T helper type 1; TNF, tumour necrosis factor.

Anger is a fundamental emotional state triggered reflexively in response to a perceived threat or provocation (this threat could be either physical or emotional). Multiple studies have explored the effects of anger on the immune system, together suggesting that this emotion can induce distinct yet related responses dependent on the context of the triggering event. For example anger associated with a hostile marital interaction has been shown to increase the production of the inflammatory cytokine interleukin-6 (IL-6) and circulating levels of C-reactive protein.⁸ In a different social setting, rugby players underwent psychological and serological evaluation 72 and 2 hr preceding an important match. A positive and independent relationship was identified between feelings of aggression, anger and anxiety (which for most increased between time-points) and circulating levels of IL-1β.⁹ Even the memory of anger-triggering events, elicited using an established technique referred to as the Anger Recall Interview,¹⁰ have been shown to significantly increase peripheral blood monocyte production of tumour necrosis factor-α (TNF-α) and IL-6 following lipopolysaccharide stimulation.¹¹ These same cells were also shown to express significantly higher levels of α₂-integrins in subjects that specifically had an increased anger response, as well as increased norepinephrine levels and accelerated diastolic blood pressure.¹²

One could speculate that in evolutionary terms the seemingly immunostimulatory (or immune-antagonistic) effect of anger is a beneficial one. For our ancestors anger often preceded physical violence, which in turn often resulted in injury. In these circumstances an immune system primed for action would certainly be advantageous, at least in the short term. However, further clinical and experimental studies have shown that in other instances, often when sensations of stress or anger become persistent, our immune response may become notably diminished or imbalanced. Individuals with below average levels of anger control were shown to heal significantly slower than subjects less disposed to this emotion.¹³ Similar observations were made in a study comparing differences in wound healing among students during their summer vacation and the run up to their exams. Recovery during the stressful exam period was on average 3 days slower than during vacation with a reported 68% reduction in IL-β production.¹⁴ In a number of studies Kiecolt-Glaser and colleagues have demonstrated that chronically stressful situations such as those experienced by caregivers can weaken the immune response, significantly diminishing antibody production against influenza and pneumococcal pneumonia vaccines and increasing the chance of latent herpes simplex virus flare ups.^15–17 Further, in a range of clinical and experimental studies Dhabhar and his collaborators have also shown that chronic and acute stress can trigger both a deleterious and a protective T-cell response in a wide variety of immune-related disorders including cancer, allergy and post-operative recovery.^18–25 As you will see throughout this review, there is little that is clear cut within this developing field.

Positive emotional states or emotional well-being (there is some debate in our laboratory as to whether any emotion could be categorized as singularly positive or negative) also have a distinct effect on the immune system. Research studying the effects of relaxing interventions on health has seen enormous growth in recent years (possibly because of the steep increase in recorded stress-related disorders). Seminal studies conducted by Gruzelier^26,27 have shown that hypnosis and guided relaxation cause a significant modulation of the immune response, increasing the number of CD4-positive T cells while buffering the drop in natural killer (NK) and CD8 cells that occurred in human subjects experiencing stress or anxiety. T'ai chi ch'uan, an ancient Chinese martial art form that has recently come into vogue in western society has been subject to multiple studies investigating its psychological benefits and their effects on the immune system. When collated this research suggests that performing T'ai chi exercises leads to an improvement of both cell-mediated immunity and antibody response to infection.^28–36 However, it should be noted that few studies have demonstrated a unified mechanism behind these improvements – instead focusing on specific players in the immune response such as circulating myeloid dendritic cells ^28,37 or pro-inflammatory CD14⁺ CD16⁺ monocytes.³⁸

Laughter has also been reported to have a surprisingly potent modulating effect on the immune system. Studies have shown that laughing therapy (in which at its most simple subjects are made to watch humorous films) up-regulates the expression of genes involved in the NK cell immune response, such as granzymes H and B, perforin, cathepsin and granulysin.^39,40 Similar effects on NK cells have been described in further investigations where laughter was additionally shown to significantly decrease levels of circulating pro-inflammatory cytokines in patients with rheumatoid arthritis.⁴¹ It was further demonstrated that this therapy suppressed the heightened expression of growth hormone and insulin-like growth factor 1 that is often associated with the disease.⁴² In one of the first and more detailed studies on this subject, Berk et al. collected blood samples at regular time-points before and after a subject was exposed to a humorous video. Their results demonstrated that the potentiating effect of laughter on the immune system can last as long as 12 hr with increases in NK cell activity, immunoglobulin levels and functional phenotypic markers for multiple lymphocyte subsets.⁴³ This persisting influence raises the question of the possible genomic effects of ‘happiness’, its molecular mechanisms of action and the possible therapeutic implications it may have. Of course this particular form of ‘mirthful laughter’ as Berk refers to it, is distinct from the laughter that can stem from emotions such as embarrassment and anxiety. It seems feasible that any immune response to this form of laughter may be distinct also.⁴⁴

Leucocytes, lymphocytes, cognition and depression: the immune system and emotion

At the turn of the previous century the once presiding concepts of humorism underwent a dramatic decline. Primarily due to a spate of discoveries that began to reveal the architecture of both brain and body down to the cellular level. Most significantly the development of Magnetic resonance imaging by Nobel laureates Paul C Lauterbur and Peter Mansfield in 2003 allowed us to examine the flashes and pulses of brain activity in both real time and exquisite detail. Neuroscience has now progressed so far that we can now use it to predict the efficacy of psychological therapies. By assessing the occurrence of specific mutations in genes highly relevant to brain function such as nerve growth factor and brain-derived neurotrophic factor one group of researchers have successfully determined the efficacy of patient responses to cognitive behavioural therapy.⁴⁵ This emerging field of research has been recently coined ‘therapygenetics’.^46–48 Accompanying this upsurge of knowledge about the brain and its physical machinations has come an understandable preference to regard this organ as the sole instigator of human thought and feeling, processing events from our external environment into a set of defined emotional states, which in turn activate a corresponding physical response. While there is a wealth of scientific and epidemiological research supporting this top–down perspective of emotional reaction, there is a growing body of evidence to suggest the reverse: a bottom–up response in which events internal to our body yet separate from the brain may have a substantial influence on our mood.

In examining the influence immune cells may hold over the nervous system and brain let us first focus on the influential research brought forward by Kipnis and Schwartz.^49–58 Their studies have demonstrated that depletion of CD4 T cells from adult mice positively correlates with impairment in learning and memory in the Morris's water maze test. Further investigation has established that populations of these cells accumulate in the meningeal space where they release IL-4. While classically a key immunoregulatory cytokine, IL-4 also seems to possess neuroprotective properties, up-regulating production of brain-derived neurotrophic factor, a neural hormone that promotes the growth of new neurons found in areas of the brain that are vital to learning, memory and higher thinking.^54,59 This beneficial and homeostatic effect does not seem to be a unique requisite of CD4 T cells, extending to other cells producing the same cytokine such as M2 anti-inflammatory macrophages.⁶⁰ However, a recent study has suggested that the beneficial effects of IL-4 may be reversed with ageing. Elderly mice displayed an excessive T helper type 2 response coupled with an increase in IL-4 production in their choroid plexus, a region of the brain responsible for the production of cerebrospinal fluid. The presence of this cytokine was shown to activate epithelial production of CCL-11, a chemokine known to be linked with cognitive dysfunction.⁶¹

Clinical studies, experimental evidence and patient accounts often describe the damaging effect of transitory or prolonged absence of T cells on emotional well-being. One of the best examples of T-cell-associated emotional imbalance in a disease condition is HIV/AIDS.^62–65 In a study performed on 96 men without symptoms or previous treatment with retroviral medications, investigators found that stressful life events, dysphoric mood and limited social support were associated with more rapid clinical progression in HIV infection, with serum cortisol also exerting an independent effect on disease progression.⁶⁶ Another comprehensive cross-sectional study expanded these observations, showing that the decreased numbers of T helper memory cell and B cells associated with high distress was mainly dependent on HIV viral load.⁶⁷

Compounding the concept of T-cell-dependent mood modulation, striking alterations in human behaviour can also be induced by a wide range of immunomodulatory drugs. Cyclosporine, a T-cell-directed immunosuppressive peptide widely used in organ transplantation, has been shown to induce a range of neuropsychological problems including depression and anxiety both in patients and in experimental animals.^68–70 Chemotherapy also commonly causes acute psychological complications, too severe to be considered a reaction to the treatments physical effects alone.^71,72

While it seems clear that these cells have an influence on our emotional state, the key question that remains to be addressed is how and by what mechanisms? To this end we began investigating the impact of immunosuppression on emotional behaviour by performing classical behavioural tests on RAG1^−/− mice.⁷³ We additionally assessed the relative contribution of CD8 and CD4 cells to emotional impairment using RAG1^−/−/OT-I and RAG1^−/−/OT-II mice, respectively. Two previous studies, one by Cushman et al.⁷⁴ and the other by McGowan et al.,⁷⁵ have assessed the behavioural defects in these mice and reported a range of phenotypic differences including increased locomotor activity, reduced awareness of threat and impaired social recognition memory. Behaviours we considered analogous to a human state of acute anxiety. Our results complemented these observations and indicated an increased level of anxiety-like and obsessive compulsive behaviour, as demonstrated using the open field and marble burying tests.

A key issue facing thousands of individuals in western societies who currently suffer from anxiety and obsessive compulsive disorders is a diminished ability to perform basic day to day activities (e.g. cooking, cleaning, shopping).^76–78 RAG1^−/− mice present comparable problems, showing a significant decline in their ability to build their own nest and reduced social interaction with other animals. Our results with RAG1^−/−/OT-I and RAG1^−/−/OT-II mice also agreed with previous studies indicating that CD4 rather than CD8 T cells played the leading role in this cross-talk between the immune system and emotional behaviour.^79–81 However, we did not observe comparable effects following a 3-week period of T-cell depletion in wild-type C57BL/6 mice. This led us to hypothesize that while impairment of cognition is a reversible and quickly ameliorated process that manifests following perturbation of the immune repertoire, a longer period of immunodeficiency might lead to a more lasting state level of behavioural dysfunction. Potentially under the influence of genetic control.

To this aim we sought to investigate if the immunodeficiency of RAG1^−/− mice might have a persisting effect on the brain. Using micro-array analysis we were able to identify 111 genes that were specifically altered in these immune-deficient mice. Of these genes many have already been described as being directly involved in the progress or instigation of various neurodegenerative disorders including Huntington's, Alzheimer's and Parkinson's diseases (all of which cause a deficit in cognition and an impaired emotional state). Most strikingly these differences in gene expression were strongly reduced or abolished completely in the brains of RAG1^−/−/OT-II mice, providing what we believe is exciting evidence that these cells may be arbiters of both immunity and emotion.

Gut feelings: immunity, emotion and the digestive system

As a final point let us briefly examine the human digestive system, which has long been known to have a somewhat enigmatic relationship with our emotional health.^82–85 A connection between persistent stress and an increased risk of suffering gastrointestinal disorders has become so established that these maladies are often treated with anti-depressants or behavioural therapies.^83,86–88 This brain–gut axis, as it is often described,^89–91 has experienced a recent upsurge in research focus, with one group going so far as to examine the validity of that commonly given advice to ‘follow your gut’ (the conclusion being ‘yes’ albeit with several provisos).^92,93 Multiple studies published last year have suggested the presence of particular ‘melancholic’ fauna in our bowel⁹⁴ or the absence of particular probiotics^95,96 has been linked to altered states of mental health and cognition. Additionally germ-free mice (which lack an intestinal micro-fauna) display a highly exaggerated response to stressful stimuli when compared with those with a normal compliment of gut bacteria.^82,97 These findings are supported by a recent letter to the journal Nature, which has demonstrated that a functioning microbiota is essential for correct social development in mice.⁹⁸ Mice lacking any gut bacteria from birth have also been reported to possess a substantially underdeveloped immune system.^99,100 An unsurprising phenotype perhaps, if one analogizes the immune system to a muscle that must be trained to develop correctly. Corroborating this work, a study by a team of French scientists has demonstrated that the effects of the immunosuppressive anticancer drug cyclophosphamide are ablated in germ-free mice. The authors suggested that the absence of a gut microbiota in these animals prevented the induction of the specific subset of pathogenic T helper type 17 cells and memory T helper type 1 cells that would target the tumour in normal mice.¹⁰¹ An interesting follow-up study might investigate if the microbiota uses the immune system to influence mood and emotions, such as the way parasitic helminths have been shown to decrease aggression and short-term memory in infected mice.¹⁰² Finally, studies into the ‘hunger hormone’ ghrelin have revealed that this peptide not only plays a significant role in appetite regulation and energy homeostasis but also exerts a significant modulatory effect on both mood and the immune system. Ghrelin-treated mice exhibit reduced inflammation and enhanced lymphocyte development in response to immune stimuli, while ghrelin knockout animals presented heightened anxiety and depressive behaviour when subjected to a variety of emotional stressors.^103–105 Research such as this offers a compelling hint that the nervous, digestive and immune systems are considerably more interconnected and overlapping than previously thought.

Conclusion: above, below and ahead

These studies and numerous others make a strong case for the immune system serving as both a channel and a controller of our emotional state. Regardless, this field remains a very new (or at least very recently renewed) area of research and there is much still to be established.

Foremost, we must consolidate the idea that the immune system can be influenced by the social (not just the physical) environment. A key foundation of this should be a rigorous investigation into how drug-free psychosocial treatments can influence the immune response. There is substantial evidence that the social context of patients suffering psychiatric disorders greatly influences the outcome of any pharmacological treatment they may receive. Seminal work by Priebe et al.¹⁰⁶ has shown how aspects of treatment including how psychiatrists introduce themselves or how many friends the psychiatric patient has¹⁰⁷ are significant factors that influence the outcome of therapy. Hence, if the future of psychiatry is social, as Priebe et al.¹⁰⁸ have suggested, is it possible that the treatment of immune diseases might also include this consideration? Second, a great deal of systematic research (with standardized treatments and techniques) is needed to support these hypotheses and also to address the often conflicting results that this research has generated. Steps in this direction can already be found in the literature. As an example, in a comprehensive study by Marazziti et al.¹⁰⁹ the long-term (12 months) treatment of adult obsessive compulsive disorder patients with selective serotonin reuptake inhibitors significantly reversed the skewed CD4/CD8 ratio (increase of CD8 and decrease of CD4) that was observed at baseline. This reversal occurred in parallel with the patient's clinical improvement.

Recent estimates by the World Health Organization suggest that by 2030 depression and stress-related problems will be the most debilitating and widespread health disorders on the planet, closely followed (rather tellingly) by autoimmune disease and allergy.¹¹⁰ With growing evidence that our emotional and immune states share a complex and bi-directional relationship with one another we believe the time is ripe to begin a serious cross-field examination of the significance of these interdependent states. It is our hope for the future that greater credence may be given to both the physical and emotional wellbeing of a patient when trying give a prognosis. And that the collaboration of clinicians, psychologists and immunologists to understand and ameliorate disease of any kind, be it mental or physical, becomes the norm. Because as that old adage tells us:

Happiness and healthiness go hand in hand.

Disclosures

The authors declare that they have no competing interests.