A Holistic and Dynamic Network-Level View of the Autonomic Nervous System

Abstract

The autonomic nervous system (ANS) plays a vital role in healthcare for both acute care and chronic diseases. The traditional view of the ANS is to divide it into individual organ systems and study the separate components with a reductionist approach, which has been proven insufficient. Here we argue that a holistic network-level view of the ANS is critical for generating new insights and deepening our understanding of its complex and dynamic functions. In this review, we treat the ANS as such a coordinated and dynamic network. We advocate for studying its interactions with major organ systems and the central nervous system (CNS) using continuous and longitudinal monitoring in ambulatory and at-home settings rather than clinic-based snapshots. We first briefly review ANS physiology, then outline our network perspective, and finally highlight cutting-edge research directions and emerging engineering innovations in ANS monitoring, modeling, and modulation that benefit from this network-level view.

1. INTRODUCTION

The autonomic nervous system (ANS) is a critical component of the nervous system that governs unconscious physiological processes such as cardiovascular regulation, respiratory control, gastrointestinal motility, and thermoregulation [1]. The past few decades have witnessed significant progress in our understanding of the ANS, ranging from functional mapping [2], plasticity and its link to diseases [3], integration with immune and endocrine functions [4, 5], to wearable devices that monitor online autonomic states [6, 7]. This new research has shown that the ANS, in fact, has a far-reaching impact and plays a role in almost all chronic illnesses across body systems and populations [8, 9]. While there is still a long way to go with respect to understanding the ANS fully, the progress that has been made warrants a new look at how we study, diagnose, treat, and design engineering innovations for the ANS.

Specifically, the ANS is currently viewed as several disparate body systems, with what are actually coordinated or connected autonomic changes addressed separately for each organ system. In addition, data about the ANS is collected in discrete and static ‘snapshots’ when the ANS is highly dynamic and varied across individuals. From a biomedical engineering perspective, innovations in ANS space are central to biofeedback systems and physiological monitoring, such as for developing artificial pacemakers, autonomic biosensors, and neural interfaces. Therefore, it is essential to review the emerging insights in the field and the approaches they suggest. The goal of this review is to (i) summarize new insights into the expansive role of the ANS in coordinating across major organ systems, (ii) advocate for a more holistic, network-based approach to the ANS using longitudinal and continuous individual data, and (iii) highlight how this perspective synergizes with recent engineering innovations in ANS monitoring, modeling, and modulation to greatly expand the potential innovation space and subsequent impact on human health.

The remainder of this review is structured as follows. We first review the physiology of the ANS with respect to its role in the major organ systems. Then, we dive into our perspective on the ANS as a coordinated network, why it deserves to be studied through this lens, and what is lost when the current paradigm of healthcare fails to do so. Finally, we highlight various ongoing innovations and opportunities from the engineering perspective to better monitor, model, and modulate the ANS as a coordinated network across the body.

2. A REVIEW OF ANS PHYSIOLOGY

In this section, we first review ANS physiology with respect to individual organ systems in this section before highlighting coordinated functions and network behavior across organ systems in the subsequent section. The ANS is divided into two main functionally distinct branches: the sympathetic and parasympathetic nervous systems. The sympathetic system orchestrates the body’s “fight-or-flight” responses—such as increasing heart rate, dilating bronchioles, and redirecting blood flow—via widespread, fast-acting outputs. Conversely, the parasympathetic system promotes “rest-and-digest” states, slowing heart rate, enhancing digestion, and conserving energy through more targeted and tonic control [1]. ‘Afferent neurons’ take information from the body up to the brain, while ‘efferent’ neurons deliver messages from the brain to the rest of the body [1]. Table 1 summarizes the effects of the sympathetic and parasympathetic systems on each organ system. Figure 1a highlights several organs and systems in the body that are tied to the ANS.

Table 1:

Summary of role of ANS, specifically sympathetic and parasympathetic nervous systems, in each major organ system

	Sympathetic Nervous System	Parasympathetic Nervous System
Cardiovascular	Increases heart rate and blood pressure	Slows heart rate; promotes vasodilation
Respiratory	Bronchodilation; enhances airflow	Bronchoconstriction; increases mucus
Digestive	Slows digestion; reduces secretion and blood flow	Stimulates digestion; increases motility and secretions
Sudomotor	Stimulates sweating via acetylcholine	No direct role
Renal & Urinary	Promotes sodium retention and renin release; relaxes bladder and aids in urine storage	Stimulates bladder contraction and urine flow
Endocrine	Raises stress hormones; inhibits insulin	Stimulates insulin and digestive hormones
Immune	Modulates inflammation and immune cell activity	Reduces inflammation

FIGURE 1.

(a) Illustration of the systemic ANS. (b) Illustration of the brain and its major regions that interact with the ANS.

2.1. Cardiovascular

The ANS is essential in regulating cardiovascular function, particularly in modulating heart rate, vascular tone, and blood pressure. Activation of the sympathetic nervous system increases heart rate and cardiac contractility [10]. Additionally, sympathetic activation induces vasoconstriction in peripheral vessels to uphold blood pressure during times of stress or physical exertion [10]. In contrast, the parasympathetic nervous system, primarily mediated by the vagus nerve, lowers heart rate and promotes vasodilation [11].

These autonomic responses are centrally coordinated by integrating feedback from baroreceptors and chemoreceptors to regulate cardiovascular responses in response to fluctuations in posture, body temperature, and emotional state [10]. A chronic imbalance within the autonomic nervous system, such as prolonged sympathetic hyperactivity or parasympathetic withdrawal, can lead to hypertension, heart failure, and arrhythmias, which are characterized by diminished heart rate variability and impaired reflex control [12, 13].

2.2. Respiratory

The ANS plays a crucial role in regulating respiratory function by controlling airway tone and respiratory rhythm. The sympathetic nervous system promotes bronchodilation [14], thereby increasing airflow during periods of increased oxygen demand. On the other hand, the parasympathetic nervous system induces bronchoconstriction and stimulates mucosal secretion [15]. Central coordination of respiration occurs in the brainstem, which integrates relevant chemosensory inputs, including carbon dioxide, pH, and oxygen levels [14]. Through these mechanisms, the ANS ensures that ventilation aligns with the body’s metabolic requirements, and disruptions in autonomic regulation can result in conditions such as asthma [15], chronic obstructive pulmonary disease [16], and sleep apnea [17].

2.3. Digestive

The ANS is integral to the coordination of essential functions of the gastrointestinal system, encompassing motility, digestion, and blood flow [18]. The parasympathetic division, mediated by the vagus nerve, generally facilitates digestive functions such as stimulating smooth muscle contractions and promoting gastric acid, enzyme, and bile secretion [18]. Additionally, parasympathetic activation relaxes of sphincters and circular muscles to move food through the gastrointestinal tract [18]. Conversely, the sympathetic nervous system exerts inhibitory effects by slowing motility, reducing secretions, causing sphincter contraction, and constricting blood vessels to redirect blood flow away from the GI tract [18]. The autonomic control is fine-tuned throughout the journey of the digestive tract and its different segments. This coordinated autonomic control ensures that each segment of the gastrointestinal tract functions efficiently to regulate motility, secretion, and transit in response to physiological demands.

Autonomic dysfunction primarily arises when nerve damage impairs the balance between parasympathetic and sympathetic activity, resulting in deficient coordination of motility, secretions, and sphincter control [19]. Consequently, patients experience a spectrum of symptoms ranging from gastroparesis (a pause in digestion) and constipation to diarrhea, nausea, and fecal incontinence [19]. Precise coordination between the autonomic branches is essential for maintaining smooth, efficient digestive processes; even small imbalances can lead to significant dysfunction across multiple segments of the gastrointestinal system.

2.4. Sudomotor

The ANS plays a crucial role in regulating sudomotor function, which governs sweating through the activation of eccrine sweat glands [20]. This is an essential process for thermoregulation and the maintenance of homeostasis, particularly in response to heat exposure or during physical exertion [20]. The hypothalamus meticulously regulates this response by monitoring alterations in body temperature [21]. In addition to temperature-driven sweating, psychological stress can also activate sympathetic fibers and result in “emotional sweating” as part of the “fight-or-flight” response, particularly impacting regions such as the palms of the hands and soles of the feet [21]. Sudomotor function is unique in that it is solely regulated by the sympathetic nervous system [21].

Abnormalities in sweat production often reflect autonomic dysfunction and are frequently observed in conditions such as diabetes, Parkinson’s disease, spinal cord injuries, and multiple system atrophy, as well as on their own [22]. For instance, sudomotor dysfunction may manifest as anhidrosis, characterized by reduced or absent sweating, or hyperhidrosis, marked by excessive sweating [23]. These symptoms can present either in a localized or widespread manner, depending upon the degree of nerve involvement.

2.5. Renal and Urinary

Sympathetic fibers innervate the renal vasculature and nephron segments to regulate blood flow, glomerular filtration, tubular function, vasoconstriction, sodium retention, and renin secretion [24]. Within the kidney, these sympathetic fibers closely follow the branching of the renal vasculature, targeting small arteries. Sympathetic stimulation causes vasoconstriction, thereby modulating glomerular filtration rate in response to systemic demands [24]. The same fibers also innervate the proximal tubule and loop of Henle, enhancing sodium reabsorption and contributing to extracellular fluid volume regulation [25]. In a specialized part of the kidney that monitors blood flow and pressure, nerve signals stimulate renin secretion. Renin starts a chain reaction that leads to the narrowing of blood vessels and triggers the release of a hormone that helps the body retain sodium and water [26].

In the lower urinary tract, sympathetic nerves help the bladder store urine by signaling to the bladder wall (detrusor muscle) to remain relaxed and the internal sphincter to stay closed during the filling phase. This coordination prevents early urination and helps maintain continence [27]. The parasympathetic nervous system plays a critical role in urination, where specialized nerves from the sacral spinal cord send signals through the pelvic nerves to the bladder wall. These nerves cause the bladder to contract and the urethral sphincter to relax to allow urine to flow and empty [28].

Autonomic dysregulation, most commonly characterized by sympathetic overactivity and reduced parasympathetic tone, is increasingly recognized in chronic kidney disease and contributes to hypertension. Elevated sympathetic activity promotes inflammation, vasoconstriction, and sodium retention, thereby exacerbating renal injury and cardiovascular risk [28]. Autonomic dysfunction resulting from an imbalance between sympathetic and parasympathetic activity is also linked to conditions such as overactive bladder, which is characterized by symptoms including sudden urges to urinate, increased daytime frequency, and incontinence [29].

2.6. Endocrine

The sympathetic and parasympathetic systems exert modulatory control over the functions of the hypothalamic-endocrine axis and influence peripheral glands such as the pancreas, thyroid, and adrenal glands [30]. Autonomic centers in the hypothalamus regulate the release of oxytocin and vasopressin from the pituitary gland [31]. Sympathetic fibers are critical in stimulating the adrenal glands to release hormones such as epinephrine and norepinephrine into the bloodstream, which act as strong sympathetic signals [32]. This intricate relationship between the endocrine and the sympathetic nervous system enables swift physiological responses in the case of a “fight-or-flight” response [32].

The pancreas receives input from both the sympathetic and the parasympathetic nervous system; sympathetic activation inhibits insulin secretion, while parasympathetic stimulation enhances insulin release [4]. This dual regulatory mechanism ensures that there is sufficient energy and nutrient abundance during times of stress. Similarly, the release of thyroid hormones is modulated by the sympathetic nervous system, which can upregulate iodide uptake and hormone synthesis under stressful conditions [33]. The vagus nerve also contributes to endocrine regulation by transmitting signals that promote the secretion of cholecystokinin (CCK), a hormone integral to digestion and nutrient absorption [34].

Chronic autonomic imbalance, particularly characterized by sustained sympathetic overactivity, is linked to endocrine disorders such as Type 2 diabetes and hyperthyroidism [35]. Individuals exhibiting insulin resistance and metabolic syndrome frequently demonstrate reduced vagal tone [36] and increased sympathetic activity.

2.7. Immune

The parasympathetic and sympathetic nervous systems have varying regulatory effects on immune responses. The vagus nerve drives an anti-inflammatory pathway in which efferent vagus nerve signals suppress proinflammatory cytokines like TNF, IL-1β, and IL-18 from activated macrophages [37]. This direct parasympathetic signaling thus plays a critical role in down regulating inflammation and maintaining immune homeostasis. Conversely, the sympathetic nervous system acts as a pro-inflammatory mechanism when interacting with pathogens in the immune system. In response to a pathogen, the sympathetic neurons express Toll-like receptors that allow them to directly detect pathogen-associated molecular patterns [38]. Sympathetic activation results in the release of pro-inflammatory cytokines such as IL-6, TNF-α, and IL-1β [39] that help in the destruction of pathogens. However, in a surprisingly unique interaction, prolonged sympathetic stimulation actually acts as a negative feedback loop, suppressing the pro-inflammatory immune response [40]. This dual role highlights the complexity of sympathetic-immune interactions, where acute signaling induces a pro-inflammatory immune response while prolonged stimulation leads to immunosuppression.

3. THE ANS AS A CONNECTED AND DYNAMIC NETWORK

Now that we have reviewed the basic physiology of the ANS across organ systems, we will outline how the ANS, with its vast reach, behaves as a coordinated and dynamic network. It interacts both with systemic organ systems as well as with the central nervous system (CNS) to provide feedback and information on the state of the body to the brain and actuate commands from the brain on the body. This network-level perspective, shown schematically in Figure 2a, clearly demonstrates why dysfunction of the ANS can so easily wreak havoc across the entire body in seemingly unrelated or disparate ways. Similarly, this approach helps explain why dysfunction in the CNS or in any single organ system can also lead to similar global consequences.

FIGURE 2.

(a) Schematic of the network perspective ANS showing it as a connected graph of organ systems and the CNS. (b) Schematic illustrating how the monitoring-modeling-modulating paradigm acts on the ANS as a network

3.1. Coordination with CNS circuits

The CNS is made up of the brain and spinal cord. The brain controls most of the functions of the body, such as sensation, awareness, movement, speech, and cognition, in our daily activities. These functions involve interaction and coordination of cortical and subcortical circuits. The spinal cord can be viewed as an extension of the brain and carries messages to and from the brain to the body. Several key brain structures, connectivity, and functions are highlighted in Figure 2b. Network-level coordination within the CNS circuits has been well documented in various animal and human studies through brain recordings [2, 41]. However, the brain, spinal cord, and body also mutually influence each other. The CNS regulates the ANS, and the ANS sends feedback to the CNS to adjust behavior and bodily states. The central autonomic network (CAN) is a sophisticated integrative network within the central nervous system that governs autonomic regulation. It incorporates multiple brain regions and facilitates communication between the central and autonomic nervous systems [42]. Additionally, the peripheral nervous system (PNS) is made up of nerves that connect the brain and spinal cord to the body. The CNS, PNS, and ANS are tightly coordinated to maintain internal balance (homeostasis) while allowing adaptive responses to the environment, varying from running from danger, digesting food, to calming down after stress.

The hypothalamus is a key brain region for autonomic control and regulation across multiple organ systems, including heart rate, blood pressure, digestion, stress response, and body temperature [43]. With direct and indirect connections with other brain regions in the CNS, specific hypothalamic nuclei can integrate inputs from diverse brain regions and peripheral signals and send output signals to both the sympathetic and parasympathetic branches of the ANS. Additionally, hypothalamic activation can trigger the sympathetic nervous system via the brainstem and spinal cord by activating the hypothalamic-pituitary-adrenal (HPA) axis to release cortisol [43].

Coordinated reflexes are involuntary and fast responses to stimuli that involve interplay of multiple muscles, sensory inputs, and organ systems to produce a specific, often protective, response. Table 2 lists several examples. For survival purposes, they are crucial for reacting rapidly to potential dangers and maintaining body balance and posture. Without involving conscious thoughts in higher brain centers, these reflexes are processed through the spinal cord or brainstem and involve the ANS. For instance, the “fight-or-flight” reflex is a multi-system survival response, initiated by threat detection and mediated through central, autonomic, endocrine, and motor systems. In this process, sensory detection is primarily processed by sensory cortices and amygdala (and a small role by the prefrontal cortex), whereas movement is primarily processed by the cerebellum, brainstem, and motor cortex.

Table 2:

Summary of major coordinated reflexes.

Reflex	Primary Output	Organs Involved	Purposes
Fight-or-flight	Sympathetic activation	Heart, lungs, muscles	Immediate survival
Diving	Parasympathetic + vasoconstriction	Heart, blood vessels	Oxygen conservation
Thermoregulation	Sweating/shivering	Skin, muscles, vessels	Heat balance
Baroreceptor	HR/BP adjustment	Heart, vessels	Blood pressure control
HPA axis	Hormonal cascade	Adrenal, brain	Stress adaptation

3.2. Examples of network behavior of the ANS

3.2.1. Pain and injury

Injuries of various types often lead to a pattern of acute or persistent pain that allows safe and effective recovery [44]. Pain is a self-defensive mechanism that serves as a warning signal, alerting the body to injury and prompting protective actions to prevent further harm. Pain involves an active sensing-action loop that is crucial for fitness and survival [45]. It has been well known that pain involves complex interactions and coordination between the CNS, PNS, and ANS. The PNS relays sensory information, including pain signals, to the CNS, while the ANS regulates involuntary functions like heart rate, respiration, and blood pressure, which can be affected by and influence the pain experience. Within the CNS, multiple brain regions (the so-called “pain matrix”) contribute to cortical pain processing [46]. Important regions belonging to the pain matrix, including the insula, anterior cingulate cortex), and prefrontal cortex, interact with the ANS by influencing the body’s physiological response to pain at both sensory and emotional levels. In turn, the ANS modulates pain perception by influencing the brain activity in regions like the anterior cingulate and periaqueductal gray [46].

The behavioral responses to injury and pain, whether being involuntary or voluntary, can be modeled as a control problem, as described in Figure 3. Changes in sensory or motor responses to aversive stimuli reflect adaptive and predictive control and subsequent learning, an ability important for experience-dependent learning, plasticity, and evolution. After injury, whole-body behaviors (such as rest and vigilance) and injury-directed behaviors (such as wound-guarding and avoidance) are influenced by various homeostatic and autonomic systems. Transition from healthy to pathological persistent pain, is referred to as pain chronification. Accordingly, chronic pain can be formulated as a problem of disruption of the network coordination or maladaptivity in learning and information integration [44]. Dysfunction of the ANS is often observed in chronic pain patients [47]. In some chronic pain patients, the experience of pain may not only occur locally at the site of an injury, but also in surrounding and even distant areas of the body—a condition referred to as widespread hyperalgesia. Chronic pain that affects the entire body is called fibromyalgia. Fibromyalgia is linked to both ANS and PNS dysfunction. The ANS dysfunction is associated with increased sympathetic activity and reduced parasympathetic activity, contributing to fatigue and sleep disturbances; the PNS abnormality may involve small fiber neuropathy [47].

FIGURE 3.

Schematic illustrations of feedforward and feedback control systems that involve the ANS. While feedback control can better handle uncertainties of the system and change control signals via error feedback, feedforward control is simpler, faster, and more energy efficient. Unlike feedback control where the system corrects itself only after an error is detected, feedforward control imposes an action in advance based on expectation or a predefined goal. The temperature and heart rate regulation is a simple example of feedforward control. In the example of pain and injury, feedback is used to regulate bottom-up autonomic or endocrine input and compare with top-down reference to generate a prediction error for the feedback controller.

3.2.2. Brain-heart axis

The brain-heart axis refers to the ongoing interaction between cognitive, physiological, and emotional states, facilitated by anatomical and functional links between the nervous and cardiovascular systems. This dynamic relationship operates through three main pathways: neural, mechanical, and biochemical [42]. The neural pathway involves the CAN and the intracardiac nervous system (ICNS), functioning via the sympathetic and parasympathetic (vagal) branches of the ANS and interconnected through ionic currents. The biochemical pathway encompasses neuropeptides, neurohormones, and other biochemical agents (e.g. inflammatory mediators, cardiac steroids, natriuretic peptides) that enable communication between the brain and heart. These signaling molecules facilitate brain-heart communication independently of direct neural or vascular activity. The mechanical pathway relies on mechanoreceptors, particularly Piezo proteins, and blood vessels to transmit signals via blood pressure waves. Each of these pathways, alone or in combination, plays a vital role in regulating key physiological functions, including blood pressure control [42].

In the neural pathway, efferent signals from the CAN are transmitted to the cardiovascular system through autonomic nerves, while afferent input is received in return. As mentioned in Section 2.1, these autonomic branches—sympathetic and parasympathetic—exert direct effects on cardiac rhythm, contractility, and vascular tone. The ICNS is a key element of the brain–heart axis. It serves as a localized neural network within the heart that can function independently of the CNS. The human heart contains over 40,000 ICNS neurons, primarily situated on the posterior atrial surfaces, with smaller groups found in the ventricles and coronary arteries [48–50]. The ICNS is modulated by both sympathetic and parasympathetic inputs. It also includes sensory neurons that monitor mechanical and biochemical states in the heart, sending feedback to local circuits and higher centers in the CAN. Intrinsic autoregulatory mechanisms are crucial to neurocardiac function. Cerebral autoregulation maintains stable cerebral blood flow despite fluctuations in systemic arterial pressure. This mechanical process underscores the tight integration between cardiovascular dynamics and neural control [51].

3.2.3. Brain-gut axis

The brain-gut axis refers to the bidirectional signaling that occurs between the CNS and the gastrointestinal tract, which is mediated by the ANS, endocrine, and immune systems. The brain-gut interaction is a special example of the interaction between the brain and body [2]. The first demonstration of the brain-gut interaction through the brain-gut axis is the cephalic phase of gastric and pancreatic secretion discovered by Ivan Pavlov [52]. Specifically, Pavlov’s seminal findings showed that the sight, smell, or taste of food in dogs with chronic esophagostomy induces a vagal-dependent gastric acid secretion [53]. Therefore, the interaction between the brain and the stomach influences gastric function.

Disturbances in the brain-gut connection can result in widespread functional gastrointestinal disorders [53]. Exploration of brain-gut interactions may enable us to understand their impact on cognitive and emotional processes and their link to dysfunctional brain mechanisms. For instance, glucocorticoids and corticotropin-releasing factors influence the brain-gut interaction, and activation of the HPA axis, the hormonal branch of the brain-gut interaction, is a gastroprotective component in stress [54]. It is known that stress alters the brain-gut interaction and contributes to the development of gastrointestinal disorders such as inflammatory bowel disease (IBD) [55].

More broadly, the brain-gut-microbiota axis expands the brain-gut axis and includes the gut microbiota [56]. Studies on brain-gut-microbiota interactions have shown that gut microbiota not only influence brain functions by modulating neuroinflammatory pathways and neurotransmitter systems, but also contribute to conditions such as anxiety, depression, and neurodegenerative disease through influencing behavior [57, 58].

3.2.4. Sleep and circadian rhythm

Sleep is important for a wide range of functions, such as emotional regulation, memory consolidation, and immune system maintenance [59, 60]. The CNS and ANS coordinate during sleep to regulate unconscious bodily functions, with the ANS shifting between sympathetic (fight-or-flight) and parasympathetic (rest-and-digest) dominance based on CNS signals [61]. This coordination is crucial for sleep’s restorative process according to sleep stage and other physiological factors. During non-rapid eye movement (NREM) sleep, the parasympathetic system generally dominates, promoting relaxation and metabolic recovery, whereas REM sleep involves more fluctuations and periods of increased sympathetic activity [59].

Sleep-wake cycling is also a prime example of how the ANS must coordinate across different body systems to change the state of the body. Sleep is physiologically drastically different from wake as a state, especially NREM sleep. During NREM sleep, strong parasympathetic activity and reduced sympathetic activity leads to the lowest heart rate and blood pressure most people experience throughout the day, in addition to reduced sudomotor and bladder activity. Breathing patterns deepen and slow to promote gas exchange. Despite increased parasympathetic activity, digestion slows to conserve energy for other necessary functions [59–61]. All these changes serve to make sleep an uninterrupted, restorative state focused on tasks such as repairing tissues, consolidating memory, and fighting infection. During REM sleep, sympathetic tone increases relative to NREM sleep, which increases heart rate, blood pressure, and sudomotor activity, and leads to a more irregular pattern of breathing [59–61]. Finally, upon waking, there is a sudden and marked increase in sympathetic tone, which allows for restoration of muscle tone in preparation for movement, significant increases in heart rate and blood pressure to daytime levels, and a sudden surge in sudomotor activity [60]. Dysfunction of this pattern of changing autonomic inputs with circadian rhythms contributes to a variety of sleep disorders such as obstructive sleep apnea, insomnia, restless leg syndrome, REM behavior disorder, narcolepsy, and nocturia, and increases the risk for cardiovascular disease and other chronic illnesses by disrupting the restorative function of sleep [60, 61].

3.2.5. The impact of aging

Aging has significant and widespread effects on coordinated autonomic reflexes, due to gradual changes in nervous, endocrine, and muscular systems. These reflexes become slower, weaker, and less adaptive, increasing vulnerability to stress, illness, and injury [62, 63]. Aging is also accompanied with a series of multi-system changes, including degeneration of cortical and subcortical neurons, reduced synaptic plasticity, slower reaction time in postural reflexes, blunted autonomic reflexes, weakened sensorimotor reflexes (e.g., less effective withdrawal reflexes in response to pain), disrupted circadian and sleep reflexes (e.g., older adults wake up earlier and sleep more lightly, or experience difficulties of staying asleep), as well as hormonal and immune dysregulation [63]. In short, aging degrades the speed, precision, and adaptability of coordinated autonomic reflexes across body systems. This results from cumulative changes in neural integration, sensorimotor transmission, endocrine signaling, and organ responsiveness and results in weaker and more disrupted discharges that are usually precisely coordinated for reflexes [62].

Aging also results in a gradual increase in sympathetic tone throughout the body, which decreases restorative and repair capacity and increases the strain on organs and blood vessels [63]. Aging-induced changes include reduced heart rate variability, increased blood pressure, decreased control of urinary function, diminished gastric motility, and worsened sleep problems, all of which point to increased sympathetic activity and reduced parasympathetic activity [63]. This increasing sympathovagal balance also reduces the natural anti-inflammatory functions of vagal activity, which results in increased inflammation, termed “inflammaging” [64]. It is hypothesized that inflammaging is a heavy contributing to chronic disease progression for all major chronic illnesses [64].

3.2.6. Infection-associated chronic illness

Immune dysfunction and bioenergetic stress play key roles in the symptoms observed in conditions like Chronic Fatigue Syndrome/Myalgic Encephalomyelitis (CFS/ME) and Long Covid [65]. Viral infections, common triggers for both disorders, induce endoplasmic reticulum stress through the release of pro-inflammatory cytokines, which in turn exacerbate immune dysregulation via overexpression of the protein WASF3 [66]. The Vagus Nerve Infection Hypothesis proposes that immune activation within the vagus nerve may amplify fatigue signaling, providing an integrative framework for symptom development [65]. Supporting this, studies of Long Covid patients have detected viral RNA in vagal nerve tissues and observed changes in gene expression related to antiviral responses [5]. These insights highlight how disruptions in autonomic and immune system communication can contribute to a range of chronic conditions characterized by persistent inflammation and dysregulated fatigue signaling, underscoring the importance of this interplay in disease pathogenesis.

3.3. The healthcare system’s view of the ANS

Dysfunction of the ANS, or dysautonomia, can be devastating to basic daily activities and productivity by causing pain, headaches, and sensory disturbances, and interfering with energy level, appetite, digestion, urination, movement, and balance [67]. From a medical perspective, dysautonomia is often non-trivial to identify and diagnose, because it is disguised as the dysfunction of the associated organs or body systems [8, 67]. Many well-known chronic diseases such as diabetes [8, 68, 69], migraines [70, 71], heart failure [72–74], Parkinson’s disease [75–77], lupus [78, 79], and long Covid [9, 80–82] include a strong component of autonomic dysfunction, but what that looks like in practice is a varying, episodic, and patient-specific set of seemingly disconnected symptoms across the body [67, 8]. The varying time course of symptoms in these diseases, compounded by the healthcare system’s practice of dividing treatment of symptoms to different medical specialties by organ system, contributes to a lack of understanding of the role of the ANS in these diseases. This can then contribute to delays in receiving the correct diagnosis or treatment, uncoordinated care across medical specialties, increased medical expenses, and fatigue with the healthcare system [67, 83]. In fact, the average time to receive a diagnosis in patients with dysautonomia in a recent patient-reported study of 672 individuals was 7.7 years [83]. Half of patients must travel over 100 miles to receive care, and over a quarter see over 10 different clinicians before receiving a diagnosis [83]. If diagnosed, dysautonomia can be managed with a combination of medication and lifestyle modification. If left untreated, however, it can wreak havoc on quality of life by interfering with basic daily activities and functions [67, 83].

For example, long Covid affects over 400 million individuals globally and costs $1+ trillion each year in lost productivity [80]. It is characterized by symptoms of dysautonomia spanning the cardiovascular, digestive, and sudomotor systems [81]. However, simply quantifying the prevalence of various symptoms has been challenging due to the lack of consistent diagnostic criteria and involvement of multiple clinical specialties [9]. The first step to identifying and treating dysautonomia in many chronic illnesses like long Covid is for a clinician to identify the possibility of ANS dysfunction and refer the patient to a specialized autonomic clinic for testing. These clinics are scarce, relegated to large academic medical centers, and often have long waitlists to even get an initial appointment [83].

Even at those clinics, the current paradigm of autonomic testing is suboptimal. It consists of a few hours of asking patients to perform a battery of tasks to attempt to mimic naturalistic modulation of the ANS [62]. These tasks can include taking blood pressures in various positions (e.g. lying down, sitting, standing), performing a Valsalva maneuver, doing a paced or deep breathing task, tilt table testing, and occasionally a gastric emptying test or specific sweating tests [8, 68, 84]. However, this process can only capture a “snapshot” of a patient’s reflexes at that day and time, which does not accurately capture the acuity and range of the patient’s symptoms in real life or the naturally varying dynamics of the ANS over time. Due to the episodic nature of many chronic illnesses including long Covid, the patient’s symptoms may be less severe or even absent at the time of clinical testing. In addition, most autonomic clinics are heavily focused on the cardiovascular system [67, 68], with other symptoms relegated to their respective medical specialties and therefore often not included in the evaluation for dysautonomia.

A consequence of the current autonomic clinic paradigm is that for most chronic illnesses, there are wide ranges reported for the prevalence of dysautonomia symptoms across different studies, based on variations in which symptoms are counted, what type of testing is done, and whether symptoms manifest at the time of testing. This precludes mechanistic understanding and awareness of the problem even among clinicians [9]. For example, in long Covid, orthostatic intolerance (improper blood pressure adjustments with postural changes) has been reported in 10–74% of patients, digestive symptoms in 0–92% of patients, and sudomotor symptoms in 36–76% of patients [9, 82]. Similarly in diabetes, dysautonomia is likely underdiagnosed and is estimated to occur in up to 90% of patients [8].

Future directions to improve the evaluation and treatment of dysautonomia should include multi-organ system involvement, including digestive, renal, endocrine, and sudomotor evaluation, instead of an exclusive or disproportionate emphasis on cardiovascular symptoms alone. In addition, given the accelerating adoption of digital health solutions, the use of wearable sensors for at-home evaluation and monitoring can have significant diagnostic value. Not only would this provide a better and more personalized representation of the individual’s dynamic range of autonomic activity, but it would also allow for understanding dynamic patterns over time rather than only a static snapshot in the clinic.

4. EMERGING INNOVATIONS IN THE ANS SPACE

There are many promising ongoing innovations in monitoring technologies, modeling approaches, and modulation techniques for the autonomic nervous system. In this section, we provide several examples of such innovations and highlight opportunities for more. We are choosing to not include purely commercial consumer wearables such as Apple watches and Fitbit-based innovations, since those are easily covered elsewhere. Finally, this is not intended to be a comprehensive list of all ongoing work, just a way to illustrate the breadth of directions being explored. Figure 2b schematically illustrates how these innovations can synergize with the network framework for the ANS.

4.1. Advances in ANS monitoring

The first step to collect data from the ANS involves monitoring, which is contingent upon the available technology (hardware, sensors) for capturing information and storing it. There are many challenges associated with collecting these data, which come from various organs and parts of the body, in a reliable and robust manner [85–87]. In the current age of digital health and personalized medicine, there is a need for solutions that enable continuous monitoring, especially in ambulatory and at-home settings, to enable rich, naturalistic data collection that can allow for individualistic insights [85–87].

The first task is to determine what signal or entity needs to be measured, and where to place a sensor to access that information. What is being measured can largely be divided into (i) electrical signals from the body such as electrocardiogram (ECG) and electrogastrogram (EGG), (ii) motion or activity related signals such as from accelerometers and gyroscopes, (iii) biophysical signals such as temperature and blood pressure, and (iv) biochemical signals such as the presence or quantity of analytes in blood, saliva, sweat, urine, tears, or exhaled breath [85–87]. While epidermal sensors, placed on the surface of the skin, can be used to measure the majority of these signals [86], there are also major innovations happening with respect to implanted and ingestible sensors as well as sensors placed on more sensitive areas, such as in the eye, ear, or mouth [85, 88–90]. Even within epidermal sensors, researchers are expanding beyond sensors simply being placed onto the skin and pushing boundaries in spaces such as textile and footwear-based sensors [85].

Current innovations in autonomic monitoring focus on several key challenges that must be overcome to enable at-scale, continuous monitoring in ambulatory and at-home settings. The first of these is enabling robust and reliable signal quality in such uncontrolled settings. This is being approached in several ways, including new materials, chemistry, and electronics integration such as flexible, stretchable, and skin-like materials with better adherence properties and lower impedance [91–96], as well as novel fabrication techniques [96]. In particular, such techniques have rapidly expanded the area of epidermal biochemical sweat-based sensors, which can simultaneously measure both chemical analytes and electrical signals from the surface of the body [97–104].

A second major challenge is reducing power consumption and prolonging battery life to allow for continuous monitoring without requiring patient compliance for frequent recharging, battery replacement, etc. In this space, there are many creative directions being explored, including low-power wearables [105] and wireless powering [106–108]. A third major challenge involves optimizing ergonomics and ease of use for a variety of patients, who may have motor or cognitive impairments or live alone, to self-administer the sensors themselves. Therefore, directions such as wire-free sensors and single patch array or “peel-and-stick” options for sensors are being developed [108, 109]. Finally, manufacturing and fabrication at scale and for low cost is a challenge for any new sensor to be deployed at a large scale. For this, several interesting fabrication techniques, including AI-assisted optimization [110], 3D-printing and ink-based solutions [111], and chip-less neuromorphic options [6] have been developed.

Such innovations in sensor technology and autonomic monitoring are the crucial first step in the pipeline to reach at-scale, robust, reliable, continuous data collection across many modalities. Without capable hardware up front, the remainder of the pipeline is undoubtedly limited. Continued innovation in this space will surely lead to rapid advancement in the types, quality, and volume of data capable of being collected about the ANS.

4.2. Advances in ANS modeling

The next step after collecting data about the ANS is to develop analysis tools or computational models to capture the most important and relevant information in that data for an individual’s or population’s health. In this section, we will highlight recent advances in ANS modeling, specifically approaches that take into account the multi-system, integrated, holistic nature of the ANS and can be used in the context of continuous monitoring. We will focus on creative attempts to model or hypothesize mechanistically about the nature of the interactions between organ systems in the ANS or between the CNS and ANS, rather than approaches that involve applying a machine learning tool generically or out-of-the-box. We will start with broad examples of multi-system modeling approaches that include at least three organ systems, followed by deep dives into two specific and popular multi-system axes, the brain-heart and brain-gut. While mechanistic multi-system modeling of the ANS is still in its infancy, we believe it has strong potential to uncover previously unseen physiological interactions on a personalized level.

4.2.1. Multi-system modeling.

Multi-system modeling approaches primarily involve combining signals from two or more organ systems with a goal of modeling or tracking a specific autonomic state or function. While there are a range of computational approaches employed, dynamic models like state space models (SSMs), Bayesian filtering, and long short-term memory (LSTM) models are popular because of their versatility and adaptability in structure for specific situations while also preserving some interpretability and mechanistic fidelity [7, 112–113]. While these approaches are typically unsupervised, custom feature definitions using mechanistic models (e.g. point process models) in combination with simpler statistical techniques for supervised learning have also been employed successfully [7, 114]. Approaches typically combine signals such as heart rate variability or heartbeat dynamics measures from ECG, sweat gland dynamics from electrodermal activity (EDA), and brain signals from EEG. States that have been tracked include surgical pain while in the operating room [7], sedation or anesthesia [114], cognitive state during different types of sensory arousal [113], and energy levels or sympathetic arousal [112]. In all such cases, using information from more than one organ system resulted in better performance than using information from a single organ system alone. This suggests that a multi-system approach is not redundant but rather harnesses the synergy of multiple sources of information to yield a more accurate estimate of autonomic state.

Another opportunity for multi-system modeling is computational modeling of vagal nerve stimulation (VNS). Because the vagal nerve acts as a highway connecting many major organs in the body, it is crucial to anticipate potential effects of stimulating both its afferent and efferent fibers as VNS becomes more popular as a therapeutic approach. While these modeling approaches are still in the early stages, it is promising to see attempts to simultaneously model both the nerve conduction properties and the effect on different end organs, such as changes in heart rate or muscle tone [115]. As the volume and diversity of data increases from advances in monitoring, we will require creative approaches to integrate such data into these mechanistic models and resources to make such datasets available to researchers at scale [116].

4.2.2. Brain-heart axis

The brain-heart axis can be understood only considering a multi-system approach. This has led to the development of integrative approaches based on a combination of multiscale, multimodal, and multidimensional variables with imaging techniques to lead to new computational frameworks aimed at revealing the central activation of the ANS within the brain-heart axis paradigm [117–119].

Despite decades of anatomical and functional research, technological advancements continue to uncover new autonomic pathways, particularly involving vagal communication. Gaining a deeper understanding of these complex interactions is crucial for developing effective treatments for disorders linked to brain–heart axis dysfunction, including heart failure, hypertension, and various neurological and neuropsychiatric conditions. Investigating patterns of CAN connectivity and their association with left ventricular ejection fraction (LVEF) in both healthy and diseased states remain essential. Additionally, emerging methods—such as intracranial electroencephalography and the use of animal models—should be further utilized to enhance insights into brain–heart interactions.

4.2.3. Brain-gut axis.

Because of the limited microbiome-based computational approaches intended for autonomic insights, we focus on the brain-gut axis in terms of electrophysiology. While information flow in the brain-gut axis can be modeled biophysically down to the cellular level or using mesh-based approaches combined with 3D imaging [120], these approaches do not lend themselves well to continuous monitoring. Recent work suggests that the primary method of modeling the brain-gut axis computationally involves quantifying gastric motility from EGG signals recorded from the surface of the abdomen in combination with one or more other electrophysiological signals, such as ECG [121, 122], EEG [58, 121], or EDA [121]. A variety of approaches are employed, including Bayesian modeling [120], functional data analysis [121], point process approaches [122], and time-frequency analysis [58, 122, 123] to account for the fact that the various types of information being combined are fundamentally different from each other. Quantifying the spatiotemporal information in the EGG itself requires careful modeling of the propagation of stomach contractions as captured by an array of electrodes on the skin surface [123]. However, when using this kind of multi-system approach to model brain-gut information flow, personalized insights such as the coupling between stomach contractions and brain signals in specific regions [58], individual patterns of post-prandial digestive activity in concert with circadian sleep rhythms [122], and dynamic person-to-person differences in gastrointestinal interoception [121] can be obtained.

4.3. Advances in ANS modulation

The ability to modulate the ANS is a powerful way to improve function in specific body systems or across several of them, alter feedback to the brain, or entrain descending pain modulation pathways. In this section we will highlight several areas in which ANS modulation is already used as a therapeutic approach and others where it has significant untapped potential.

To date, autonomic modulation has been most thoroughly studied in the context of cardiovascular disease. In this domain, autonomic modulation was initially achieved through pharmacologic means, such as using beta blockers, or through invasive surgical means, such as renal denervation and invasive vagal nerve stimulation [124–126]. While effective, both options have many possible risks and side effects that may not be appropriate for all patients. In more recent times, non-invasive stimulation options such as transcutaneous or auricular vagal nerve stimulation have been explored with moderate success [125–129]. So far, the most well-studied and effective disease areas of application for all autonomic modulation approaches have been heart failure [72, 73, 127, 128] and arrhythmias [124–126], in both cases aiming to curb sympathetic overactivity [72, 73, 125]. In future, with the advent of more non-invasive techniques that pose less risk, autonomic modulation may be a key therapeutic approach for a variety of other cardiovascular illnesses or even as a preventive measure for those at high risk of developing cardiovascular disease. Further studies should be done to elucidate the non-cardiac secondary effects of such stimulation as well.

Secondary to the cardiovascular system, there is also expanding interest in developing stimulation-based solutions for the gastrointestinal system to address dysmotility, inflammatory, metabolic, and neuronal disorders [130, 131]. In this case, rather than the purely invasive route, there are creative opportunities to develop ingestible delivery systems that are technically non-invasive and lower risk [132].

Another domain area with rapid expansion of autonomic modulation techniques is chronic pain. The majority of approaches employed have been stimulation-based, but at various anatomic locations in the pain pathway, from the PNS [133], to the dorsal root ganglion and spinal cord [133], vagal nerve stimulation [124, 125], up to the brain with deep brain stimulation (DBS) [133–135]. Most of these approaches are invasive; however, minimally invasive or non-invasive approaches are becoming available for some of these anatomic locations. Initial studies have shown promising results, with larger trials underway.

Finally, there is an entire gamut of behavioral autonomic modulation techniques. These include biofeedback-based approaches, in which subjects are trained to modulate their own autonomic responses [136, 137]; temperature modulation-based approaches, including both cryostimulation [138] and passive heat therapy [139]; and mind-body techniques, such as yoga, meditation, music therapy, altered states of consciousness, and acupuncture [140–142]. Scientific studies on the effects of these approaches are still in preliminary stages, but if successful, these present additional low-risk and low-cost options to achieve potential health benefits.

5. CONCLUSION

The ANS is playing an increasingly important role in both acute care and chronic diseases, including cardiovascular, metabolic, endocrine, gastrointestinal, neurological and psychiatric disorders. Because of its clinical significance, understanding and targeting ANS function has become an important direction for diagnostics, prognostics, and therapy in modern healthcare. In this review, we have outlined a holistic approach to the ANS in terms of its coordination with the CNS and major organ systems in the body. The coordination may occur at multiple levels and multiple timescales, which requires us to examine these interactions through a network-based approach, ranging from multi-system modeling, hybrid (feedforward/feedback) control, to closed-loop modulation. Additionally, the network-level perspective aligns and synergizes with recent engineering innovations in ANS monitoring, modeling, and modulation. We believe that encouraging such innovation through this network lens will maximize the eventual impact on human health. In addition, ongoing innovation will expand our ability to study the ANS as a dynamic network over time, using continuous and longitudinal monitoring approaches that can be employed in any setting rather than being limited to static “snapshots” in a clinic.

The ANS remains understudied and poorly understood, despite its extensive influence across numerous domains of health and disease. This presents a significant untapped opportunity to leverage the ANS for deeper insight into individual health states, improved personalized diagnostics, and even targeted modulation for therapeutic benefit. However, several key questions remain. While it is clear that the ANS is involved in many diseases, its causal role in these conditions has yet to be fully determined. There is also a critical need to develop technologies and methodologies that enable precise modulation of the ANS across different scales—from individual cells or nerve fibers to coordinated activity across multiple organs or systems. Additionally, we are only beginning to demonstrate the potential of leveraging the CNS-ANS connection for both afferent and efferent modulation. Despite these challenges, we remain optimistic that continued research in the coming decades will illuminate the full potential of the ANS to advance human health.

SUMMARY POINTS.

1. The current paradigm of treating the autonomic nervous system (ANS) as separate organ systems that we track through static snapshots in clinical settings is not ideal. It prevents a holistic understanding of ANS dynamics and changes that characterize health and disease, and reduces the potential impact of innovations.

2. A network-level outlook for the ANS is necessary, in which organ systems are thought of as nodes that communicate bidirectionally with each other and with the central nervous system (CNS) to send and receive commands and feedback. This perspective naturally sheds light on the dispersed nature of autonomic symptoms in disease and the link between dysfunction in different organ systems.

3. Personalized, continuous, and longitudinal monitoring of the ANS is necessary in naturalistic settings, such as at home, to fully understand ANS dynamics. Short snapshots in clinic cannot provide the full picture.

4. There are many ongoing innovations and untapped opportunities in ANS monitoring, modeling, and modulation that can benefit from this network-level dynamic ANS perspective. By leveraging the synergy and improved physiologic understanding it can provide, these innovations can maximize their impact on human health.

TERMS AND DEFINITIONS

ANS

Autonomic nervous system

CNS

Central nervous system

PNS

Peripheral nervous system

VNS

Vagal nerve stimulation

CAN

Central autonomic network

Sympathetic

The “fight-or-flight” branch of the ANS

Parasympathetic

The “rest-and-digest” branch of the ANS

Afferent

From the body to the brain

Efferent

From the brain to the body

Sudomotor

Relating to sweating

Dysautonomia

Dysfunction of the ANS

HPA

Hypothalamic-pituitary-adrenal

ICNS

Intracardiac nervous system

ECG

Electrocardiogram, the measurement of electrical signals from the heart

EGG

Electrogastrogram, the measurement of electrical signals from the stomach

EDA

Electrodermal activity, the measurement of skin conductance as a proxy to sweating activity

EEG

Electroencephalogram, the measurement of electrical signals from the brain