Insulin Impairment Disrupts Central Serotonin Synthesis: Implications for Stress Resilience

Abstract

This chapter reviews the important neurophysiological mechanisms that drive symptoms characteristic of comorbid depression and metabolic disease. It outlines how insulin impairment in the periphery¹ interferes with central 5-hydroxyindole metabolism and ultimately restricts central² serotonin synthesis. More specifically, peripheral insulin impairment disrupts i) peripheral and central tryptophan stores, ii) tryptophan uptake into the brain, and iii) tryptophan hydroxylase-2 function. Central serotonin availability appears to be increasingly restricted by higher degree and duration of insulin impairment, which can lead to both physiological and behavioral positive feedback loops experienced by individuals as a spiral of deteriorating mental health and tryptophan metabolism. Serotonin and its metabolites are fundamentally homeostatic regulators that serve to enhance adaptive response to stress in all organisms. Considering this essential trait, this review proposes that: disruptions in normal 5-hydroxyindole metabolism of tryptophan during impaired insulin function will disrupt homeostatic adaptive capacity of central serotonin, thereby increasing vulnerability to emotional and energy disturbances, and limiting recovery from such disturbances.

1. Introduction

Serotonin (5-hydroxytryptamine or 5-HT) is an important neurotransmitter and hormone involved in various functions, such as emotion, cognition, learning, metabolism, sleep, platelet function, and gastrointestinal motility [1, 2, 3, 4, 5]. These functions depend on tight regulation of available 5-HT and other indole metabolites of Tryptophan (Trp) throughout the body and brain [1]. Research continues to uncover how imbalances of these molecules is associated with insulin impairment, and a cascade of disruption via complex interactions between the mind and body. This review highlights bidirectional feedbacks between 5-HT and insulin imbalances that are associated with energy and emotion [6, 7, 8, 9]. It attempts to explain some of the underlying limitations to psychological resilience commonly experienced by persons with compromised insulin function, and to explain potential mechanistic links behind highly comorbid depression and metabolic disease.

1.1 Comorbid depression and metabolic disease

Diabetes mellitus is a metabolic condition characterized by chronic high blood sugar levels due to an underlying impairment in insulin secretion (Type 1, T1D), action (Type 2, T2D) or both [10]. Metabolic syndrome – the co-occurrence of insulin resistance, obesity, atherogenic dyslipidemia, and hypertension – is a high-risk factor for the development of T2D [11]. Diabetes is commonly comorbid with mood disorders linked to altered central 5-HT, and most often with disorders involving depression [12, 13, 14]. Diabetics are two times more likely to have depression compared to non-diabetics [15, 16, 17]. This comorbid relationship between mental health disorders and both types of diabetes appears to be bidirectional, and their co-occurrence tends to exacerbate symptom severity [12, 18, 19].

Depression refers to any of several depressive or mood disorders outlined in the DSM³ handbook, broadly involving extended periods of sadness, emptiness and depressed mood, and often accompanied by disturbances in cognition, sleep and energy levels [18, 20, 21]. There is less clarity and consensus on underlying impairments involved in depression. Since the 1960s, depression has most commonly been explained by a deficiency in brain 5-HT activity [22, 23]. This paralleled industry’s need to allocate pharmaceutical prescriptions according to symptomology [24] and thus their marketing of selective serotonergic medications [25, 26]; most commonly, 5-HT reuptake inhibitors (SSRIs) which function to increase synaptic 5-HT and upregulate 5-HT neurotransmission [7].

The neurobiological basis and treatment of depression is still the predominant narrative accepted by the Western public and endorsed by leading research and educational materials [27, 28]. However, the idea that depression is caused strictly by low 5-HT activity is mostly rejected by experts [25, 28, 29]. Some even argue that high 5-HT activity is behind depressive states, and that this is a functional response [7]. What remains largely undisputed is the extensive evidence that changes in brain 5-HT has some role in depressive states and energy regulation.

1.2 Mechanisms behind diagnostic labels

Medical research on diabetes and depression now consider the environmental and sociocultural contexts in which these neurophysiological systems interact, and how these contexts shape our organization of disease and disorder [26, 29, 30, 31]. Sociocultural factors influence how these states are generated, whether their symptoms and syndromes are interpreted as adaptive or maladaptive, and at what point these symptoms become categorized as diseases, syndromes or disorders [24, 30, 31, 32, 33]. Psychological disorders are subjectively categorized by non-universal, Western constructs of self, normality and value [34]. Moreover, our scientific understandings of the neurological basis for cognition, emotion and behavior have been predominantly measured on Western populations, despite the evidence that changes in neurobiology lead to different expressions in different cultures [35].

The RDoC system from the National Institute of Mental Health is a more recent and systemic tool for categorizing mental health by functional constructs that represent a specified functional dimension of behavior [36]. The constructs are systems of response to the environment (e.g., positive valence, or reward system) which are then characterized in aggregate by the genes, molecules, circuits and behaviors involved. The RDoC thus examines the mechanism that “drive psychiatric symptoms” [37].

1.3 Peripheral insulin impairment and central serotonin availability

In light of these evolving explanations, this chapter focuses on mechanistic links between insulin and 5-HT systems that drive common symptoms of metabolic and mood dysfunction, which may shed light on the consequences of abnormal insulin synthesis and sensitivity, regardless of diagnostic labels. In this way, these feedbacks may be interpreted and applied to evolving understandings of disease and disorder in different biological and sociocultural contexts.

We thus use the unconventional term, insulin impairment, to refer to impaired insulin secretion or insulin sensitivity in any condition, and focus on its relationship to 5-HT availability from Trp metabolism via the 5-hydroxyindole pathway; despite the absence of this terminology and this scope of focus in neurophysiological medical research.

Experimental studies over several decades have investigated the mechanisms of Trp metabolism and the impacts of induced-diabetes on the distribution of 5-HT in distinct central and peripheral systems. Yet this literature remains largely fragmented, each representing isolated pieces of complex relationships involved in insulin-5-HT imbalance [38, 39, 40, 41, 42]. Martin and colleagues [43] addressed this gap in a novel review examining the links between central insulin impairment and central serotonergic activity. They highlight several mechanisms potentially linking these systems, including increased oxidative stress, inflammation, and hypothalamic-pituitary–adrenal axis activation. However, the bulk of this research emphasizes the downstream processes of insulin signaling in the brain. Martin and colleagues [43] show that insulin and 5-HT coregulate processes within the brain, including neurogenesis. Thus, the impairment of central insulin function in diabetic states disrupts the effectiveness of 5-HT activity in these pathways, and throughout the brain.

In contrast, the present chapter examines the role of peripheral insulin (that outside of the brain) in regulating availability of central 5-HT and its precursor Trp. We specifically focus on how peripheral insulin impairment reduces peripheral pools of Trp, the rate of Trp transport into the brain, and the rate of central 5-HT synthesis from available Trp. It is important to emphasize that the scope of this chapter focuses on the ultimate outcome of central 5-HT synthesis and thus the amount of 5-HT that could be available for neurotransmission, while Martin and colleagues [43] examine 5-HT neurotransmission (activity in the synapse). Similar to Martin and colleagues’ [43] presentation of altered 5-HT activity, we see a multitude of pathways where the loss of normal insulin function can impair metabolic, mood and energy regulation through positive feedback.

In summary, this chapter aims to highlight the significant role of normal peripheral insulin systems in supporting 5-HT availability to the brain, and the multitude of pathways that impair normal 5-HT availability to the brain when this system is not functioning well. This can occur at varying degrees among individuals across these metabolic diseases and syndromes, and in non-diseased states.

2. Tryptophan metabolism

Tryptophan (Trp) is an essential amino acid that serves as a building block for 5-HT (5-HT), kynurenic acid, melatonin, and quinolinic acid, among other compounds. Imbalances in Trp levels have been associated with various diseases and disorders, including cancer, dementia, diabetes, and depression [44, 45]. Most notably, individuals with diabetes, depression, or both conditions tend to have lower levels of Trp in their blood [46, 47]. Trp is primarily metabolized through three pathways: the indole pathway, the 5-hydroxyindole pathway (which is involved in 5-HT production), and the kynurenine pathway (which converts 90% of Trp into other compounds) [2, 48].

2.1 Central tryptophan regulation

The central and peripheral pools of 5-HT are separated by the blood-brain barrier (BBB): a protective layer between the brain and blood vessels. Unlike 5-HT, Trp can cross the BBB (Figure 1) through a specific transporter called the large neutral amino acid transporter (LAT1) [49]. However, other large neutral amino acids (LNAA), such as tyrosine, phenylalanine, leucine, isoleucine, and valine, also compete for entry into the brain using the same transporter [50, 51, 52]. When there are high levels of these competing LNAA in the blood, the ratio of Trp to LNAA decreases, leading to reduced availability of Trp in the central nervous system [50, 51, 52, 53, 54]. Since Trp and other LNAAs⁴ are essential amino acids, the composition of our diet strongly influences the ratio of Trp to LNAA in the blood and subsequently affects Trp uptake in the brain [53, 55, 56]. Studies have shown that elevated levels of Trp in the blood after a meal can increase peripheral Trp:LNAA [57], which is followed by elevated levels of Trp and 5-HT in the central nervous system [58, 59].

Figure 1.

Tryptophan transport past the blood brain barrier determines tryptophan availability for central serotonin synthesis (5-HT^C = central serotonin, 5-HT^P = peripheral serotonin, Trp = tryptophan, LNAA = large neutral amino acid, BBB = blood brain barrier, TPH = tryptophan hydroxylase).

2.2 Peripheral tryptophan regulation

The ratio of Trp to other large neutral amino acids (LNAA) in our diet is particularly important because Trp is the most limited LNAA, usually comprising only 1–2% of total protein [60, 61, 62].

When high-Trp proteins are consumed alone, they significantly increase the levels of Trp in the blood and the Trp to LNAA ratio compared to low-Trp proteins [39, 63, 64]. However, due to the higher concentrations of other LNAA in most dietary proteins, a high-carbohydrate meal actually increases the Trp to LNAA ratio more than a high-protein meal [65, 66]. This unexpected effect is due to the role of insulin in modulating Trp in the peripheral blood, and the stronger insulinogenic capacity of carbohydrates.

Insulin plays a crucial role in facilitating Trp uptake into the brain (Figure 2). Insulin helps transport LNAA into skeletal tissues, which reduces competing LNAAs in the blood, increases Trp:LNAA, and reduces competition for Trp to enter the brain [40, 67]. Insulin also helps transport Trp into skeletal tissues, but a larger proportion of branched-chain amino acids (BCAA) including valine, isoleucine, and leucine are taken up by muscle tissues in response to insulin. As a result, there is an elevated Trp:LNAA in the blood that is proportional to the circulating insulin levels [57, 68].

Figure 2.

Insulin facilitates tryptophan transport by increasing Trp:LNAA in the blood (5-HT = serotonin, Trp = tryptophan, LNAA = large neutral amino acid, BBB = blood brain barrier, TPH = tryptophan hydroxylase, P-tissue = skeletal tissues, in = insulin).

Insulin does impact central Trp uptake through another mechanism, but this process does not have a significant effect on Trp availability in the central nervous system. To cross the blood-brain barrier via the LAT1 transporter, Trp must be unbound [54]. Within the limited pool of Trp in the blood, most of it is reversibly bound to plasma albumin protein, with only about 10% of Trp being unbound. Insulin comes into play in this pathway by removing non-esterified fatty acids (NEFA) from albumin, creating space for Trp to bind to albumin and reducing the amount of free Trp available for transport [69]. However, the binding of Trp to albumin is transient, so the impact on Trp availability in the central nervous system is negligible [70]. As a result of this constant state of dissociation and binding with albumin, approximately 70–80% of Trp in the blood is available to cross the blood-brain barrier [70].

2.3 Insulin impairment and tryptophan

Research below demonstrates that impaired peripheral insulin function will alter Trp homeostasis and metabolic pathways in two ways: i) by redirecting Trp to regulate immune response in hyperglycemic conditions, reducing peripheral Trp availability and ii) by modifying insulin’s role in regulating amino acid ratios, limiting Trp transport into the brain. Multiple pathways leading to reduced Trp levels and altered Trp metabolism in diabetic states have been extensively studied in rodents and humans; insulin plays a crucial role in modulating these changes.

Chemical impairment of insulin production in rodents⁵ resulted in a significant decrease in total plasma Trp after 7 day [42], 28 days [40], and 35 days [71]. Similar findings have been reported in human studies. Three studies involving children with clinical type 1 diabetes (T1D) found lower plasma Trp and Trp:LNAA levels compared to non-diabetics [38, 72, 73]. Adolescents with metabolic syndrome also showed lower peripheral Trp levels compared to controls [74, 75]. In a study involving healthy controls and diabetic adults individuals with T1D had significantly lower serum Trp levels compared to the control group [76]. Chemically induced acute T1D in rodents also led to a significant increase in plasma branched-chain amino acids (BCAAs) valine, leucine, and isoleucine, resulting in a decrease in the overall Trp:LNAA [71, 77] and subsequently reduced central Trp levels [40, 71, 77, 78, 79]. Human studies have also found low Trp:LNAA in individuals with diabetes [77, 80].

The culmination of this research shows that impaired insulin function and insulin deficiency limit the uptake of Trp competitors into skeletal tissues, maintain low Trp:LNAA levels and LNAA competition for LAT1, and thus result in lower central Trp levels (Figure 3). Importantly, elevated peripheral BCAA concentrations have been associated with an increased risk of future diabetes and may indicate impaired insulin function even before the diagnosis of metabolic disorders [81, 82].

Figure 3.

Insulin impairment lowers serotonin production in the brain by impaired ability to elevate peripheral Trp:LNAA and facilitate central Trp uptake, and by alteraing TPH2 enzyme activity (5-HT = serotonin, Trp = tryptophan, LNAA = large neutral amino acid, TPH2 = tryptophan hydroxylase 2, P-tissue = skeletal tissues, in = insulin, KP = kyneurenine pathway, T1D = type 1 diabetes, T2D = type 2 diabetes, IDO = indoleamine 2,3-dioxygenase).

Studies in rodents and humans have shown that while initial insulin impairment increases unbound peripheral trp chronic stages of insulin impairment eventually led to a significant decrease in plasma Trp, further reducing the Trp:LNAA and central Trp levels [83, 84]. Chronic insulin impairment can affect peripheral Trp levels through changes in Trp metabolism and oxidation. Impaired insulin function disrupts normal glucose uptake, leading to hyperglycemic states that cause inflammation and increased production of reactive oxygen species by mitochondria [85, 86, 87]. Chronic inflammation and oxidative stress trigger an immune response that upregulates Trp metabolism in the kynurenine pathway, ultimately reducing peripheral Trp pools (Figure 3) [72, 88, 89, 90, 91]. This also has direct impacts on Trp levels because the highly reactive indole ring of Trp can be easily damaged by oxidizing species [73, 92].

New research suggests a connection between plasma Trp and peripheral insulin through the GPR142 receptor located on pancreatic islet cells. GPR142 is a G-protein coupled receptor that specifically binds to Trp and phenylalanine [77]. Binding of Trp to GPR142 stimulates the release of insulin, glucagon-like peptide-1 (GLP-1), and glucose-dependent insulinotropic polypeptide (GIP), helping to regulate glucose levels [93]. Administration of Trp in obese mice significantly increased glucose metabolism and insulin secretion [94]. In GPR142 knockout mice, Trp supplementation did not lead to increased insulin release or improved glucose tolerance compared to controls, highlighting the importance of GPR142 in the Trp-insulin link [94]. The action of Trp on GPR142 receptors in the pancreas creates a positive feedback loop where chronically reduced Trp levels in impaired insulin states may further decrease insulin secretion, disrupt glucose homeostasis, and subsequently lower plasma Trp levels.

3. Serotonin synthesis

5-HT is synthesized from tryptophan (Trp) through the 5-hydroxyindole metabolic pathway (Figure 3). This process is catalyzed by either Trp hydroxylase enzyme 1 (TPH1) or Trp hydroxylase enzyme 2 (TPH2). TPH1 is primarily found in peripheral tissues, while TPH2 is responsible for converting Trp to 5-HT in the brain. The rate of 5-HT synthesis is dependent on the activity of the TPH enzyme. In the brain, the synthesis of the intermediate molecule 5-hydroxytryptophan (5-HTP) is the limiting step, and this step is determined by the availability of Trp as a substrate.

3.1 Insulin impairment and serotonin

Insulin impairment not only affects central 5-HT synthesis by altering enzyme activity in the 5-hydroxyindole pathway, but it also impacts the catalytic function of TPH2. In rodents with chemically-induced insulin impairment for 7 days, TPH2 activity was reduced in the cerebral cortex and brainstem [89]. This decrease in activity was previously attributed solely to low central Trp availability as a result of the mechanisms described earlier. However, this same study found that the TPH2 enzyme in diabetic rodents had a significantly lower affinity for Trp and reduced phosphorylating capacity required to stimulate the enzyme. Another study with T1D rodents observed similar catalytic dysfunction of TPH2, as well as reduced expression of the enzyme [42]. These epigenetic and kinetic changes were attributed to hyperglycemic conditions: elevated peripheral glucose levels (hyperglycemia) led to increased brain glucose levels, which can trigger inflammation, reactive oxygen species, and oxidative stress, ultimately causing damage to TPH2 expression and function [95, 96, 97, 98].

Chronic elevation of glucocorticoid levels, commonly seen in diseases associated with insulin impairment, can also explain the reduced expression of TPH2 [99, 100]. Excess glucocorticoids are released in response to chronic stress through activation of the hypothalamic–pituitary–adrenal axis. Glucocorticoid receptors act as transcription factors that inhibit TPH2 expression and can significantly decrease 5-HT levels in the raphe nuclei after one week of excess exposure [101]. Elevated glucocorticoid levels are also independently linked to peripheral tissue insulin resistance, impaired insulin production, and high blood glucose levels [99, 100, 102]. Overall, chronic release of glucocorticoids appears to indirectly impair central 5-HT availability by altering Trp transport and conversion, as well as via hyperglycemic conditions. The correlation between elevated glucocorticoids and depressive symptoms highlights this stress response as a contributing factor in the disrupted regulation of Trp and 5-HT, which underlies psycho-metabolic comorbidities [103]. In summary, both reduced central Trp availability and impaired TPH2 activity disrupt the 5-hydroxyindole pathway in individuals with chronic insulin impairment.

The effects of altered Trp metabolism in response to peripheral insulin impairment, such as lower peripheral Trp, elevated peripheral LNAA, and lower central TPH2 expression and function, are expected to subsequently reduce central 5-HT pools available for neurotransmission. This prediction is supported by several rodent studies, where chemically induced insulin impairment led to a significant decrease in central 5-HT synthesis. This decrease was observed after 4 weeks in the whole brain [78], after 1 week in the whole brain [104], after 2 weeks in the whole brain [41] and the hypothalamus [89], and after 50 days in the striatum and pons medulla [105], as indicated by 5-HTP accumulation. Additionally, 5-HT synthesis measured by TPH2 activity was significantly lower after 7 days in the cerebral cortex and brainstem [89]. While this is beyond the scope of the present chapters, readers may consider that insulin impairment is also expected to compromise other systems dependent on Trp and 5-hydroxyindole metabolism of Trp. For example, the synthesis of the 5-HT metabolite, melatonin (MLT), appears to be significantly reduced in diabetic states [6].

3.2 Insulin treatment

Insulin treatments in diabetic rodents have a limited capacity to restore normal Trp and 5-HT levels in the brain. Prolonged periods of hyperglycemia place greater stress on both insulin secretion and insulin sensitivity, further reducing the potential for recovery [106]. A comprehensive study examined the effects of insulin treatment on rodents with 7 days of insulin impairment by measuring TPH2 activity, central 5-HT levels, central Trp levels, and blood glucose levels [42]. Following insulin treatment, rodents with induced T1D showed normal blood glucose and Trp levels in the brain, but TPH2 activity and 5-HT levels remained depressed after 7 and 14 days of insulin treatment [42]. However, These findings are contradicted by a different study of rodents with 7 days of insulin impairment, where insulin treatment after 14 days was able to restore near-normal central 5-HT levels [104].

A study from 1991 suggested that prolonged insulin impairment may result in irreversible reduction in central 5-HT levels [77]. Rodents with 10–30 days of insulin impairment were treated with insulin at various time intervals ranging from 15 to 135 days. Insulin treatment restored serum Trp levels throughout all time spans but could not restore central Trp levels beyond 15 days. It is important to note that this study induced insulin impairment using alloxan, which is known to cause higher cellular toxicity in non-target cells compared to streptozotocin used in other studies [42, 77, 104, 107].

Human studies also offer insights into the consequences long term insulin impairment, as in the case of clinically diagnosed diabetes. In comparison to normal controls, diabetics had higher levels of plasma BCAA [108], tyrosine and phenylalanine [109], and lower levels of plasma Trp [110]. The subsequently low plasma Trp:LNAA was not reversed by weight loss [111]. Diversion of Trp metabolism to the kynurenine pathway in response to stress persisted after bariatric surgery [112] and plasma Trp and Trp:LNAA were little changed by Trp dosing in obese persons compared to lean [109].

Both rodent and human studies indicate that chronic peripheral insulin impairment may have long-term effects on 5-HT availability that cannot be fully restored by insulin treatment, underscoring the complexity of identifying effective treatments.

4. Behavioral and psychotropic treatments

The dysregulation of insulin and serotonergic systems is further complicated by a positive feedback loop involving common behavioral responses and psychotropic medical treatments. For instance, central 5-HT depletion can lead to cravings for carbohydrates in an attempt to raise peripheral insulin levels [49, 113, 114, 115]. However, a high-glycemic diet increases the risk of insulin resistance, impaired insulin secretion, and depression [106, 116, 117, 118]. Carbohydrate cravings are also common in various disorders associated with mood changes and serotonergic modulation, such as atypical depression, seasonal affective disorder, late luteal phase dysphoric disorder, and binge eating disorder. Serotonergic medications that upregulate 5-HT neurotransmission are often prescribed to treat these disorders [119, 120].

However, these medications may exacerbate Trp and 5-HT imbalances in individuals with peripheral insulin impairment. As serotonergic medications are not specific to the central nervous system, they disrupt 5-hydroxyindole homeostasis in the periphery, where most 5-HT is synthesized [121]. This can lead to apoptosis of pancreatic cells and acute pancreatitis (inflammation and tissue damage), and thereby inhibit insulin secretion [122, 123]. It can also impair insulin receptor function [124]. Several studies have observed an increased risk of type 2 diabetes associated with SSRI use [125, 126, 127, 128], despite the effect of SSRIs on glucose homeostasis [118], and this risk seems to dose, concentration and duration dependent [127, 128]. Moreover, individuals with pre-existing low levels of peripheral Trp are more susceptible to anxious and depressive episodes when taking serotonergic antidepressants (Figure 4) [129, 130].

Figure 4.

Selective serotonin reuptake inhibitors (SSRI) disrupt homeostatic regulation of serotonin in the periphery, which can restrict peripheral insulin secretion and sensitivity, reducing available central 5-HT (5-HT = serotonin, Trp = tryptophan, LNAA = large neutral amino acid, P-tissue = skeletal tissues, in = insulin, KP = kyneurenine pathway, T1D = type 1 diabetes, T2D = type 2 diabetes).

In summary, the regulation of Trp and 5-HT homeostasis is crucial in both the periphery and brain. Insulin impairment and common attempts to regulate emotional changes caused by this impairment can disrupt this delicate balance. The evidence from rodent and human studies presented in this chapter suggests that behavioral and medical treatments that exacerbate insulin impairment also pose a risk of long-term and potentially irreversible impairment of Trp metabolism and transport, as well as central 5-HT availability.

5. Interpretations and applications

How do we interpret the consequences of low Trp levels and restricted central 5-HT in light of positive feedbacks with higher degree and duration of insulin impairment? Returning to ideas presented in the Introduction, what do these interactions and outcomes mean within a new paradigm that dissociates neurobiology from absolute diagnostic labels and questions the 5-HT ‘chemical imbalance’? Low and high 5-HT hypotheses, and the medications targeting neurotransmission upregulation, are not modeled on balance. Rather, they are modeled on extremes and deficits of a single chemical and blind targeting of multiple systems in the body [131], all in an attempt to chronically maintain ideal emotional states.

5.1 Complex adaptive living systems

These mechanisms can instead be viewed with a model of balance that better reflects the reality of functional living systems. Optimization of single brain chemicals is reflective of a linear cause-effect reductionist approach to health [132, 133]. Alternatively, we interpret the patterns of interaction among neurological and metabolic systems with the view that human beings are composed of, and operate within, complex adaptive systems (CAS) [132, 133, 134].

A complex adaptive system is a collection of specialized agents (components or parts) that can be understood within the context of the whole and the interconnecting systems they comprise [133, 134]. Diverse interactions between these systems and environmental stimuli will adapt to give rise to non-linear, unpredictable and ever-changing dynamics [26, 135, 136]. While defined by dynamic change and evolution, complex living systems conserve their conditions for renewability through self-regulating feedback loops that link interacting agents [134, 137]. Collectively, these traits allow CAS to maintain homeostatic equilibrium such that they maintain a dynamic form of stability over time [138] and develop resiliency in the face of disturbance and adversity [133, 134, 135, 137].

This CAS perspective is especially useful in understanding complex systems that are not easily measured, understood and predicted [132]. Applying the CAS perspective to the human brain and body, we view health as homeostatic equilibrium of essential interactions, systems and functions needed for sustaining the human as a self-organizing system. This lens can be applied to neurophysiological systems by viewing them as normally fluctuating between order and disorder, and by focusing on “state change” rather than static maintenance of absolute highs and lows [132]. We interpret positive feedback loops with their “tend[ency] towards chaos and decay” and identify a need to counterbalance by negative feedback [132]. The CAS model thus shifts our view of neurophysiological imbalance from the failure of a single component in a particular place, to the idea that imbalance arises from interaction between multiple systems that are not able to self-regulate and maintain dynamic equilibrium [132, 133]. It also recognizes that these interactions will generate emergent, unpredictable outcomes that manifest differently in unique individuals and sociocultural contexts [133, 137]. We apply these principles to our understanding of the insulin system as it interacts with 5-hydroxyindole system in the body and brain.

5.2 Systems of Insulin and 5-hydroxyindole metabolism

Functional homeostatic regulation occurs when components of the system are maintained within upper and lower bounds, and fluctuate dynamically in response to changes in their environment [24, 26, 135, 139]. Indeed, both an excess and deficiency of 5-HT and its precursor throughout the body can lead to health imbalances [2]: sufficiently low peripheral 5-HT is needed to maintain insulin sensitivity and prevent obesity [140, 141], yet sufficiently high levels of peripheral 5-HT are needed to maintain pancreatic insulin secretion [142] and normal glucose levels [143]. Similarly, a “high” Trp diet in human studies has been labeled both a protective [144] and risk factor [83] for later development of T2D.

A state of acute tryptophan depletion (ATD) can be experimentally induced to explore the behavioral and cognitive responses to low central 5-HT. Reduced central 5-HT from ATD led not only to negative affect bias (sad mood), but also better punishment prediction accuracy and more risk aversive behavior [145, 146, 147]; it also enhanced negative reciprocity, expressed as greater punishment or retaliation in response to perceived unfairness [148]. While these findings tend to be interpreted as explanations of maladaptive depressive states, we can revisit this from a CAS perspective. Low central 5-HT states may be an adaptive, functional response to avoid harm that is learned in environments where punishment is common and then recruited when complex social issues are interpreted as chronic danger. Other scholars predict that hypervigilance is an adaptive response for a child raised in an unpredictable environment [32], and that social avoidance is an adaptive response to volatile social experiences [149].

Evolutionary psychologists tend to view short periods of stress, anxiety or depression as adaptive responses that can increase resiliency in certain contexts if the triggers are addressed [7, 147, 149, 150]. For example, the Analytical Rumination Hypothesis posits that depressive symptoms of cognitive rumination and loss of interest (lack of concentration, appetite, sex drive or socialization) are coordinated as an evolutionary adaptation to solve complex social problems or traumatic experiences with sustained focus [7, 151, 152]. These researchers show overlapping genetic, neurobiological and symptom expression between depressive symptoms and sickness behavior, with the former acting as an emotional fever. Just as our bodies divert energy towards immune function in response to a viral threat at the expense of growth and reproduction, biological trade-offs occur during depression, with a reallocation of resources in response to persistent environmental threats.

Both extremes of lowered central 5-HT availability and heightened 5-HT neurotransmission impair the brain’s ability to accurately interpret facial emotions; ATD inaccurately escalated the interpretation of fearful emotion to anger, while SSRI’s dissolved the ability to distinguish fearful and normal emotions [145]. These results support the idea that functional 5-hydroxyindole metabolism and 5-HT neurotransmission may require dynamic homeostatic regulation.

From a resiliency perspective, ATD alone does not seem to initiate the cascade of positive feedbacks in insulin and 5-hydroxyindole systems; it worsens or triggers them. In fact, Trp-deficiency induced episodes of anxiety and depression in people with a personal or familial history of mental health disorder, yet had little to no emotional effect on healthy controls [153, 154, 155]. Similarly, in comparisons of high-Trp versus low-Trp consumption (where high-Trp levels significantly elevated plasma Trp and Trp:LNAA), high Trp improved memory exclusively in individuals susceptible to high stress [64], and reduced vulnerability to experimental triggers for fatigue, negative affective bias, and diminished well-being [63].

It is important to recognize that these ATD tests represent acutes stress, while stress responses outside of the laboratory are often triggered by more complex, abstract threats and exist in a more pervasive form that requires reflection on previous experiences and learned behavior [146, 147]. The chronic nature of these real stressors can cause damage to the brain and body that correspond with mental unwellness and symptoms of non-communicable diseases [156, 157].

Interpreting these experiments through CAS perspective, we suggest that recovery from fluctuations in mood and depressive episodes may be supported by the adaptive capacity to fluctuate high and low neurotransmission in response to biosocial environmental disturbances. 5-HT availability is a limiting factor in the potential to upregulate neurotransmission at a given time and in response to a given situation. Availability also determines the potential to downregulate in a time-responsive manner relative to upregulation, as needed to re-establish equilibrium.

5.3 Serotonin as resilience molecule

This idea is supported by evolutionary studies of 5-hydroxyindole metabolism arguing that a tightly regulated balance of Trp content and availability is most ideal for all living organisms in order to maintain the adaptive function of its intermediates [158]. The human body has developed several mechanisms for maintaining a narrow range of Trp in different regions [158] and in relation to co-existing amino acids [2]. The 5-hydroxyindole pathway that produces 5-HT from Trp has been highly conserved from unicellular bacteria to mammals, and shares one common function across all organisms and tissues: adaptive response to environmental stress [158, 159, 160]. Despite the body having very low Trp concentrations compared to other amino acids, its metabolite 5-HT serves many crucial biological functions. 5-HT is considered the homeostatic regulator of the central nervous, neuroendocrine, gut and immune systems, and the biochemical connector of mind and body with the environment [8, 158].

The adaptive function of 5-HT as a moderator of stress response can be understood as a trait supporting psychological resilience: the ability to cope with or recover from adversity [161]. We argue that altered Trp metabolism as a result of chronic insulin impairment ultimately impairs psychological resiliency, particularly among already vulnerable individuals. By limiting availability of 5-HT in the brain, chronic insulin impairment disrupts the ability to finely regulate changes in 5-HT neurotransmission in response to changing environmental contexts. Since 5-HT is responsible for adaptive response to stress, this represents a loss of neurological options that will further increase vulnerability to emotional and energy disturbances.

Insulin impairment is one mechanism that impairs the adaptive capacity of Trp metabolites like 5-HT to mitigate impacts of environmental stress. The dynamic state of insulin impairment and associated glucose intolerance throughout the life course of prediabetic and diabetic individuals [10, 162, 163], as well as the diverse experiences of individuals themselves, must be considered in light of experimental studies. Moreover, our understanding of the complex and dynamic state of insulin impairment, diabetes and metabolic disorder, cannot be based on short-term rodent studies alone. In combination with human diabetic studies, these experiments do identify clear limits in physiological function during these complex states; most notably, during early states of insulin impairment that occurs long before changes to blood-glucose, weight, or energy levels are detected at diagnosis of metabolic disorder. Together, these studies demonstrate a strain in adaptive response to stress that will be affected by the severity and duration of insulin impairment and individual human characteristics.

6. Conclusion

Insulin impairment disrupts the homeostatic adaptive capacity to regulate central 5-HT and impairs psychological resilience to stress by altering normal 5-hydroxyindole metabolism of Trp. The small fraction of 5-HT in the brain is more vulnerable to insulin impairment than peripheral 5-HT since, in addition to being limited by low peripheral Trp stores, it is also restricted by impaired TPH2 activity and impaired central Trp uptake. Insulin impairment represents a loss of options for the many roles of central 5-HT, which are increasingly restricted by higher degree and duration of insulin impairment, as well as serotonergic medications and dietary cravings induced by the dysfunction itself.

Neurophysiological studies of high and low 5-HT might be better understood from the perspective homeostatic balance of 5-hydroxyindole metabolism, and how this is shaped by dynamic states of other molecules in the body, beyond insulin impairment. Future studies may elucidate our understanding of a bidirectional relationship between insulin and 5-HT function and explore the 5-hydroxyindole response to glycemic imbalance and oxidative stress under diverse conditions and locations in the brain and body (i.e., how different receptors in different areas of the brain respond to insulin impairment over time). Better integration of experimental studies that focus on the origins of neurophysiological imbalance may help identify treatment that supports adaptive capacity inherent to the 5-HT system. Most significantly, we encourage neurophysiological research to consider experimental design and interpretation with the resiliency model of complex adaptive living systems.

Acknowledgments

Thank you to the Trinity College, University of Toronto Independent Study Course for supporting this research.