Airway reflexes, autonomic function, and cardiovascular responses.

In this article, we review the cardiovascular responses to the inhalation of irritants and pollutants. Many sensory receptors in the respiratory system, from nose to alveoli, respond to these irritants and set up powerful reflex changes, including those in the cardiovascular system. Systemic hypotension or hypertension, pulmonary hypertension, bradycardia, tachycardia, and dysrhythmias have all been described previously. Most of the experiments have been acute and have been performed on anesthetized experimental animals. Experiments on humans suggest we have similar sensory systems and reflex responses. However, we must use caution when applying the animal results to humans. Most animal experiments, unlike those with humans, have been performed using general anesthesia, with irritants administered in high concentrations, and often to a restricted part of the respiratory tract. Species differences in the response to irritants are well established. We must be even more careful when applying the results of acute experiments in animals to the pathophysiologic changes observed in prolonged exposure to environmental pollution in humans.

In this article we review the evidence that activation of airway sensory receptors by pollutants can reflexly influence the cardiovascular system, as well as other systems that may cause secondary cardiovascular changes.
A number of studies in humans show that chronic inhalation of pollutants can induce changes in the electrocardiogram (ECG) (1)(2)(3)(4), but the sensory receptors and neural pathways involved in these changes have not been established. By comparison many studies in experimental animals have allowed identification of sensory receptors and afferent and motor pathways (5)(6)(7)(8)(9)(10).
The mucosae and epithelia of the airways, from nares to bronchioles and alveoli, contain afferent (sensory) nervous receptors that respond to a large variety of pollutant and irritant inhaled substances (Table 1) (5)(6)(7)(8)(9)(10)(11)12), and on activation set up profound reflex changes involving breathing and the autonomic nervous system (Table 2) (5)(6)(7)(8)(9)(10). Although the sensory receptors have been studied mainly with respect to breathing, we concentrate in this review on the reflex cardiovascular changes. Table 1 lists the airway receptor types that have been identified and which are sensitive to pollutants and irritants. There are some disagreements in the literature [for example, on the relative sensitivity of C fiber and of rapidly adapting receptors (RARs) to various endogenous mediators (5,6,(13)(14)(15)], but a majority view is presented in Tables 1 and 2. It seems certain that the list of sensory receptors is an oversimplification, but it may provide a framework for discussion of the role of airway receptors in modulating cardiovascular responses. Functional subgroups of RARs are well established (7); it is likely that subgroups will also be identified for the C-fiber receptors (6,16,17). Single nerve fiber recordings show that many of the groups responding to irritants can be subdivided. Thus, some C-fiber receptors in the nose respond to nicotine but not to ammonia, whereas for others the pattern is vice versa (11). Similar differences in sensitivities to a range of irritants and biologic mediators are seen for tracheobronchial Cfiber receptors and for RARs (5,6,12,16). Although it is known that various RAR subgroups in the lower respiratory tract have different respiratory reflex actions (7,8,17,18), it is uncertain whether this is true for the Cfiber receptors and whether subgroups of RARs and C-fiber receptors also have different cardiovascular reflex actions.
In this review we refer to receptors in their original physiologic sense as nervous or sensory endings which receive stimuli and set up reflex responses. In the pharmacologic sense the nervous endings have chemical receptors on their membranes that when activated set up generator and action potentials in the sensory end organ and nerve fibers. There is an extensive literature on these pharmacologic receptors in airway sensory receptors or end organs (6,17,19,20).

Cardiovascular Reflex Responses to Pollutants
The classical and seminal study was conducted by Kratschmer in 1870 (21,22), who showed that chemical irritants such as ammonia applied to the nose and larynx caused profound hypertension and bradycardia. He also analyzed the afferent and motor pathways. He worked only on experimental animals but realized the significance of his studies for humans. (He later occupied the first Professorial Chair in Public Hygiene in Vienna, Austria.) He wrote We are only repeating the experiments which Nature carries out daily on Man and animals, and thus we can obtain information about processes which, quite apart from their theoretical interest, are of immediate practical significance (22). Since Kratschmer's work, an extensive literature has documented that irritants inhaled into the upper respiratory tract cause cardiovascular reflex changes that are predominantly bradycardiac and include either hyperor hypotension (8-10) ( Table 2). Nasal inhalation of cigarette smoke in rabbits causes hypertension and bradycardia (23) (Figure 1). Because the bradycardia is prevented by administration of atropine (5,8) it is mainly vagally mediated, but studies are limited on the possible additional involvement of the sympathetic innervation of the heart. Receptors in the nasopharynx and pharynx do not seem to be very chemosensitive, possibly because the epithelium in these areas is squamous cell in nature and may be relatively impermeable, but the sites can give rise to powerful cardiovascular reflexes, especially hypertension (8,10,24). Laryngeal irritation can cause a variety of cardiovascular changes, but systemic hypertension has usually been reported, including in humans (25-27) ( Figure 2); in addition, there may be either bradycardia or tachycardia and pulmonary hypertension (27) (Figure 3).
For the lower respiratory tract, RARs and C-fiber receptors occur in and under the epithelium (6,28) ( Figure 4). These latter receptors, as well as C-fiber receptors in the alveolar wall, are very sensitive to irritants and In this article, we review the cardiovascular responses to the inhalation of irritants and pollutants. Many sensory receptors in the respiratory system, from nose to alveoli, respond to these irritants and set up powerful reflex changes, including those in the cardiovascular system. Systemic hypotension or hypertension, pulmonary hypertension, bradycardia, tachycardia, and dysrhythmias have all been described previously. Most of the experiments have been acute and have been performed on anesthetized experimental animals. Experiments on humans suggest we have similar sensory systems and reflex responses. However, we must use caution when applying the animal results to humans. Most animal experiments, unlike those with humans, have been performed using general anesthesia, with irritants administered in high concentrations, and often to a restricted part of the respiratory tract. Species differences in the response to irritants are well established. We must be even more careful when applying the results of acute experiments in animals to the pathophysiologic changes observed in prolonged exposure to environmental pollution in humans.  (29) and reflexly cause pronounced bradycardia and hypotension. As Table 2 indicates, an irritant pollutant distributed throughout the entire respiratory system could activate many reflexes, with diverse actions on blood pressure and heart rate.
Acute cardiovascular reflex responses to inhaled irritants and mechanical stimuli have been reported in humans (25,27) (Figures 2,  3), and in general the patterns are similar to those described for animals, although the stimuli have naturally been much weaker.
Assessing neural mechanisms has seldom been practical.
The receptors for these reflexes have been identified in animals by single nerve fiber recording and in humans and animals by histology. They lie in and under the airway epithelium from nose to alveoli (6,28) ( Figure  4). They are polymodal, responding to many chemical and mechanical irritants, and nociceptive, responding to a variety of pathologic conditions that activate airway reflexes. Their morphology is not well mapped out (6). When stimulated, they release tachykinins such as substance P and neurokinin A, which cause the local axon-reflex effects known as neurogenic inflammation (30) ( Figure 5). This includes mucosal vasodilatation. However, airway neurogenic inflammation has not been convincingly established in humans (30).
Thus, inhaling a chemical irritant will potentially stimulate a large number of airway afferent receptors. Presumably the reflex cardiovascular response will depend on the integration of the reflexes from individual anatomic zones and therefore will in turn depend on the sites of deposition of the pollutant. The integrative mechanisms of such a reflex response have seldom been studied in experimental animals, and not at all in humans.

Dysrhythmias
Human inhalation of pollutants can cause changes in the heart, based on evidence from complex statistical analysis of the pattern of the ECGs (1-4). In animal experiments, inhalation of strong concentrations of irritant vapor can cause dysrhythmias. Kratschmer and colleagues (20,21) noted More recent studies support this possibility; laryngeal irritation in humans can cause cardiac dysrhythmias with depression of the ST complex of the ECG (27) (Figure 3). Before use of local anesthetics in the larynx, cardiac dysrhythmias and even cardiac arrest were not uncommon responses to laryngeal intubation (31). It has long been assumed, but with little evidence, that "restaurant death," the sudden death caused by inhalation of food, is due to dysrhythmia, possibly ventricular fibrillation, and set up by a reflex from the larynx. In humans, the role of the vagus nerves in dysrhythmias caused by inhalation of pollutants has been much discussed (1)(2)(3)(4). It should be   noted that even with strong chemical irritation of the lower respiratory tract (below the larynx) and with pronounced cardiovascular reflex changes in experimental animals, dysrhythmias have not been described.

Sensitization, Inhibition, and Interactions
An important property of the sensory receptors responding to pollutants is their plasticity (17,29,32). This has recently been studied both by recording reflexes and by single nerve fiber recordings of action potentials, in vivo and in vitro. A sensitization to chemical and mechanical stimuli has been shown on a short-term basis-minutes or hours-for RARs with histamine (33), immunologic reactions (34), and by agents such as substance P and lobeline (15,35,36). Longer lasting sensitization-several days to a week-has been shown with ozone (37). Similar sensitization of C-fiber receptors has been established (29) with ozone (38), histamine (39), prostaglandin E 2 (40), and in experimental airways disease (41). Some of the sensitizations occur at the receptor level in the periphery, but interactions between different groups of receptors at the level of the vagal nodose and jugular ganglia have also been established (42). None of these sensitizations has been established for the cardiovascular reflexes but they potentially exist. A sensitized cough reflex after respiratory viral infection, often persisting for months, is a common condition that must depend on sensitized RARs in the airways, as these mediate the cough reflex (14). In addition, there has been much speculation and a number of studies to determine whether the hyperresponsiveness of asthma is due to the sensitization of sensory receptors in the lungs.
Some neural pathways will inhibit the responses from others. A good example is the stimulation of pulmonary C-fiber receptors, which can inhibit cough induced by activation of RARs (14,43). Whether a similar inhibitory mechanism exists for the cardiovascular reflexes is not known. If an irritant at one site (e.g., the nose) causes hypertension, and at the same time at another site (e.g., the lungs) causes hypotension, the result must depend on the interaction of the two afferent pathways. For heart rate, most but not all reflexes from the respiratory tract responding to irritants cause bradycardia, so there may be the possibility of summation or synergy rather then inhibition. Studies are limited on these phenomena but could be important.
A further consideration is that, apart from primary reflexes from irritation of the respiratory tract, the cardiovascular system may be secondarily affected. The changes in breathing, usually apneas, hyperventilation, or cough, will have a mechanical effect on the cardiovascular system, change blood gas tensions with cardiovascular effects via the peripheral chemoreceptors, and alter the discharge of lung receptors (RARs and slowly adapting pulmonary stretch receptors), which in turn may affect the heart and vasculature. For example, in Figure 2 (25) one cannot determine the extent to which the hypertension and tachycardia are influenced by the respiratory changes, cough, and tachycardia, or are primary reflex actions. The hypertensions seen in anesthetized cats with mechanical stimulation of the respiratory tract from nose to trachea are much larger in spontaneously breathing animals than in those paralyzed and artificially ventilated (23). The whole-body mechanisms are very complex.

Application of Animal Results to Humans
Most experiments with inhaled pollutantsnearly all those that analyze nervous pathways and all of those based on single nerve fiber recordings-have been conducted on experimental animals. It has yet to be determined to what extent they are applicable to the mechanisms of cardiovascular responses to inhaled environmental pollutants in humans.

Size of Stimulus
When animals are exposed to chemical irritants, pollutants, and biologic mediators in aerosols, the concentrations are usually far greater than those to which humans are exposed. With some stimulants (water, hyper-and hyposmolar solutions, cold and Environmental Health Perspectives • VOLUME 109 | SUPPLEMENT 4 | August 2001  There have been few or no comparisons of the same stimulus in humans and in unanesthetized animals.

Localization of Stimulus
In the majority of animal experiments, the stimulus has been localized to one site, e.g., nose or lungs. Humans exposed to environmental pollutants inhale them into the entire respiratory system unless the nose, which is the most sensitive reflexogenic zone in most animals, is mechanically blocked. In human experiments inhalation is usually through a mouthpiece. The potential importance of interactions between reflexes arising from different parts of the respiratory system has already been mentioned. Most of the animal species used are rodents-guinea pigs, rats, and mice-which are obligatory nose breathers, unlike humans. The balance of the reflexes from different sites in the respiratory tract may be quite different.

Duration of Stimulus
Most animal experiments are acute, lasting minutes or hours, whether reflexes or afferent nervous activities are being studied. Acute experiments have also been conducted with pollutants in humans, but for natural environmental pollution the period of exposure required for pathologic changes may be months or years.

Anesthesia
Many of the animal experiments have been performed using general anesthesia, and even when the anesthesia is light, it can profoundly change the pattern and size of irritant responses, including those responses in humans. Volatile anesthetics affect, by a peripheral action, the sensitivity and discharge of airway receptors that respond to irritants (5,9), and centrally acting anesthetics act on the brainstem pathways for their reflexes. In humans light general anesthesia enhances the respiratory responses due to stimulation of the larynx (expiratory efforts and cough), whereas deeper anesthesia converts the response to apnea (26,44) ( Figure  6). However, anesthesia seems to have little effect on the reflex laryngospasm due to laryngeal stimulation. In experimental animals, anesthesia may depress the reflex respiratory changes from the larynx (23) ( Figure  1). Blood pressure and heart rate changes have not been studied in this context. It should be emphasized that the brainstem neural pathways for reflexes elicited by inhalation of pollutants have not been mapped out in detail, although the first-order pathways into the nucleus of the solitary tract and adjacent areas have been (45)(46)(47). This lack of detailed knowledge may not be there is much uncertainty as to precisely how, in neuronal terms, general anesthetics affect these reflexes.

Sensitization and Reflex Interactions
These interactions have been mentioned previously. In experimental animals they can occur in the very short term-seconds or minutes. Clinically they can be illustrated by the sensitized cough after respiratory tract infection, and by the hyperresponsiveness observed in subjects with asthma (48). In subjects exposed to pollution for months or years they are certain to exist, but comparison with acute animal and human experiments must be quantitatively uncertain. Cardiac changes in human chronic exposure to pollutants have been well described (1)(2)(3)(4), but it is difficult to establish whether the neural mechanisms identified in the more acute experiments in animals apply to humans.

Conclusion
Abundant evidence exists about reflexes, including those to the cardiovascular system, activated by inhalation of pollutants in experimental animals. Nearly all the studies have been acute or short term. Similar experiments suggest that humans have the same reflexes, but they have not been extensively analyzed, especially with regard to the cardiovascular system. The applicability of this large body of research to the pathophysiologic results of long-term exposure to atmospheric pollutants is at present very tenuous.