The psychophysiology of the sigh: I: The sigh from the physiological perspective

Abstract

Breathing is composed of multiple, distinct behaviors that are bidirectionally regulated through autonomic and voluntary mechanisms. One behavioral component is the sigh, which serves distinct physiological and psychological roles. In two accompanying reviews we will discuss these roles. The present review focuses on the physiological function, where sighs play a critical role in controlling lung compliance by preventing the collapse of alveoli. Implicated in the generation of sighs and normal breathing is the preBötzinger Complex, a rhythmogenic network in the medulla. Although sighs and normal inspiration are generated within the same network, they show distinct temporal characteristics. While sighs occur every few minutes, normal breathing is generated in the range of seconds. Both are differentiated by distinct modulatory and synaptic mechanisms, and recent evidence indicates that these mechanisms are regulated by inputs from different regions of the brain. An important modulator of sighs is hypoxia, implicating sighs in the arousal response.

1. Introduction

The ventilatory role of breathing is critical for homeostatic regulation of oxygen and carbon dioxide in the blood (Forster, Haouzi, & Dempsey, 2012; Guyenet et al., 2019). Yet, there is significantly more to breathing aside from the regulation of blood gases. Breathing is perhaps the most physiologically integrated behavior and is strongly adapted with metabolic, emotional, and cognitive state changes and behaviors (Baertsch, Baertsch, & Ramirez, 2018; Guyenet & Bayliss, 2015; Ramirez & Baertsch, 2018). This state-dependency is bi-directional, since breathing also influences these same activity states. For example, breathing changes in response to pain, anxiety, panic, worry, joy, desire, and relief; conversely, breathing can be a driver of anxiety, panic, and relief, hyperventilation, and can sometimes even induce seizures (Grassmann, Vlemincx, von Leupoldt, Mittelstadt, & Van den Bergh, 2016; Indranada, Mullen, Duncan, Berlowitz, & Kanaan, 2018; Vlemincx et al., 2013; Vlemincx, Van Diest, & Van den Bergh, 2015). The ability of breathing to alter cognitive, emotional, and metabolic states offers important therapeutic opportunities as has long been recognized in Eastern Medicine (Draeger-Muenke & Muenke, 2012). While breathing techniques have become very popular for stress management and relaxation purposes, only recently has research started to investigate the potential of breathing techniques as therapeutic add-ons to standard treatment in Western Medicine (Russo, Santarelli, & O’Rourke, 2017).

However, considering breathing as one behavior is too simplistic. Eupneic breathing (“normal breathing”), gasping, and sighing are distinct breathing behaviors with distinct physiological and psychological roles (Lieske, Thoby-Brisson, Telgkamp, & Ramirez, 2000; Ramirez, 2014). This raises an important question that may seem purely semantic: Do these breathing activities constitute different behaviors or different patterns of the same behavior? Neurophysiologists have debated this question for more than a century, and based on lesion experiments, Lumsden postulated in the early 1900’s that eupneic breathing and gasping are generated by separate centers located in different regions of the brain (Lumsden, 1923). He referred to the pneumotaxic center in the pons and the gasping center in the medulla as being the brainstem regions responsible for the generation of eupnea and gasping, respectively. In vitro approaches combined with contemporary transgenic and electrophysiological experiments in vivo provided an alternative explanation: that the same network can reconfigure to generate distinct motor behaviors (Lieske et al., 2000). The more we learn about the underlying neurobiology, the more it becomes clear that there is no simple answer. Sighs, gasps, and eupnea are three separate motor behaviors that are distinct, yet mechanistically overlapping (Lieske & Ramirez, 2006b; Lieske et al., 2000; Tryba et al., 2008).

In two back-to-back reviews we will focus specifically on one of these breathing behaviors, the sigh, which has long been recognized for its distinct physiological (Part I, this article) and psychological properties (Part II, Vlemincx et al., in press).

2. The “sigh reflex”

Sighs are considered a reflex; these deep breaths can be activated by sensory afferents such as lung stretch receptors (Katagiri, Katagiri, Kieser, & Easton, 1998) that respond to a decrease in lung compliance (Matsumoto, Takeda, Saiki, Takahashi, & Ojima, 1997). Moreover, bilateral vagotomy abolishes sighs in rabbits and dogs (Bowes, Andrey, Kozar, & Phillipson, 1983; Matsumoto et al., 1997), suggesting that afferent input is required for the generation of sighs. Yet, this has not been confirmed by other studies; instead, it has been shown that sighs reappear post-vagotomy (Cherniack, von Euler, Glogowska, & Homma, 1981; Galland, 1984; Marshall & Metcalfe, 1988), and persist following lung transplantation in humans (Shea et al., 1988). Furthermore, sighs respond to chemosensory inputs as hypoxia increases sigh frequency (Cherniack et al., 1981; Glogowska, Richardson, Widdicombe, & Winning, 1972; Marshall, 1987; Schwenke & Cragg, 2000; Widdicombe, 1982). This stimulatory effect depends in part on carotid body inputs, as has been demonstrated by carotid body (CB) denervation studies. However, these studies also show that CB input is not necessary for generating sighs, since sighs persist and continue to respond to hypoxia (Telgkamp & Ramirez, 1999), and hypercapnia (Galland & Cragg, 1985; Schwenke & Cragg, 2000) following denervation.

Still, these afferent inputs are functionally important. For example, combined mechano- and chemosensory inputs likely activate sighs in response to atelectasis (Li & Yackle, 2017; Ramirez, 2014). The tidal volume of a sigh, more than two times larger than a normal “eupneic breath”, allows sighs to reinflate the partially collapsed and under-ventilated regions of the lung, thereby increasing lung compliance (Caro, Butler, & Dubois, 1960; Ferris & Pollard, 1960), and restoring the ventilation/perfusion ratio thereby maintaining normal lung function (Cherniack et al., 1981). Thus, sighs play a critical homeostatic role in maintaining the physiological partial pressure range for arterial O₂ and CO₂. In this functional context, the sigh serves as an important reflex that is activated by several relevant sensory inputs.

3. The origin of the sigh

The importance of the sigh as a reflex suggests that the generation of this large tidal volume is a result of mechanosensory inputs augmenting a normal breath. Indeed, this interpretation inspired the early vagotomy experiments as mentioned above, and this notion continues to persist as indicated by the somewhat misleading, synonymous use of the term “augmented breath” when referring to sighs. The fact that sighs are not simply breaths that are augmented by sensory inputs has clearly been demonstrated in brainstem slice preparations (Lieske et al., 2000; Tryba et al., 2008). The generation of neuronal activity driving sigh activity persists even in small, isolated wedges of brain tissue containing one region of the medulla: the preBötzinger complex (preBötC) (Fig. 1) (Tryba et al., 2008). This approach not only demonstrates that sighs constitute a centrally generated motor behavior but also that sighs possess physiological properties that differentiate them from normal breathing irrespective of the sensory inputs that clearly play a role in vivo. Sighs are periodically generated within the preBötC and occur spontaneously and concurrently with the much faster respiratory drive for rhythmic eupneic activity which, like sigh activity, also persists in the absence of any sensory input. The periodic nature of the sigh is also seen in intact animals and humans. Furthermore, sighs occur more frequently in human infants where they occur every few minutes (Fleming, Goncalves, Levine, & Woollard, 1984; Hoch, Bernhard, & Hinsch, 1998), persisting into adulthood, albeit at a lower frequency (Bell, Azubike, & Haouzi, 2011; Taelman, Vandeput, Vlemincx, Spaepen, & Van Huffel, 2011; Vlemincx et al., 2013, 2009).

Fig. 1.

(A) The transverse medullary brain slice preparation containing the preBötzinger Complex (preBötC) generates both fictive eupnea and sigh activity. (B) Fictive eupnea and sigh activity can further be maintained in a small wedge isolated form the transverse preBötC slice. Scale bars are 1 s, integrated preBötC population recording y-axes are in arbitrary units.

Figure is modified from Tryba et al. (2008).

The possibility to preserve sighs in vitro enabled rigorous cellular investigations aimed at unraveling the neuronal mechanisms that differentiate sighs from normal breathing. Recordings obtained from the preBötC network indicated that the vast majority of rhythmic neurons are activated during both normal, “eupneic” and sigh respiratory activity. A small proportion of sigh-only neurons was identified within the preBötC, but these sigh-only neurons do not possess obvious pacemaker or other regenerative properties that would suggest that they “pace” the sigh rhythm (Tryba et al., 2008). Thus, these insights remain puzzling: how can a shared neuronal network generate two motor activities with distinct temporal properties?

Several neuronal mechanisms distinguish sighs from eupneic activity. One of these is the dependence on the P/Q-type calcium current. Blocking this calcium channel subtype selectively abolishes sighs, but not eupneic activity in vitro (Lieske & Ramirez, 2006a). The genetic knockout of P/Q-type calcium channels in mice confirmed this pharmacological evidence. These mutant mice survive initially, because of their continued ability to generate eupneic breathing, but they eventually succumb to pneumonia and atelectasis due to the failure to sigh (Koch et al., 2013). Sigh generation also depends on a specific metabotropic glutamate receptor (subtype 8, group III) suggesting that the synchronization giving rise to the large amplitude sigh burst depends on glutamatergic mechanisms that differ from those responsible for the eupneic burst (Lieske & Ramirez, 2006b).

Sighs also seem to depend on the persistent sodium current (INap) as suggested by pharmacological experiments, and the intriguing observation that certain INap- dependent pacemaker neurons can concurrently generate small amplitude, eupneic-like bursts occurring at a faster eupneic-like frequency, as well as sigh-like high-amplitude bursts occurring at a slower sigh-like frequency (Toporikova, Chevalier, & Thoby-Brisson, 2015; Viemari, Garcia, Doi, Elsen, & Ramirez, 2013). Even following pharmacological isolation, these two burst types are differentially modulated by muscarinic agonists in a manner that is consistent with the network response: oxotremorine inhibits eupnea while activating large amplitude sigh-like bursts (Tryba et al., 2008). These experiments taken together indicate that distinct neuronal mechanisms contribute to the generation of sighs and eupnea. However, how these mechanisms work together to generate two rhythmic activities with very different temporal, modulatory (Doi & Ramirez, 2010; Peña & Ramirez, 2002, 2004; Reynolds, Vujisic, Davenport, & Hayward, 2008), metabolic (Telgkamp & Ramirez, 1999), and physiological properties (Orem & Trotter, 1993) remains largely unknown.

More recent studies suggest that modulation from areas rostral of the preBötC, or slices that contain more rostral sections of the preBötC favor sigh generation (Li et al., 2016; Ruangkittisakul et al., 2008). Anatomically, preBötC slices that extend further rostral exhibit more sighs, whereas slices which retain more caudal sections reportedly do not (Ruangkittisakul et al., 2008). There are many possible explanations for this rostro-caudal gradient. Differential receptor subtype expression allowing sighs to be modulated in the preBötC could be eliminated in more caudal slice preparations, thereby potentially reducing the probability of the network to reconfigure to generate sigh bursts. Rostral structures such as the peptidergic circuit postulated to control sigh modulation will likely also facilitate the generation of sighs (Li et al., 2016). These modulatory inputs, however, are not essential for rhythmogenesis, since sighs can still be generated in the isolated preBötC as discussed above (Tryba et al., 2008).

There are a number of neuromodulators known to alter sigh frequency including metabotropic glutamate receptors, (Lieske & Ramirez, 2006b), substance P (Ruangkittisakul et al., 2008), norepinephrine (NE) (Viemari et al., 2013), and acetylcholine (Tryba et al., 2008). Thus, the notion that a singular neuromodulator is responsible for sigh generation is likely an oversimplification. Indeed, it is probable that these modulators act together to differentially orchestrate breathing and the generation of sighs, as well as function to coordinate feedback with the rest of the brain. Some modulators only activate sighs, as was described for norepinephrine acting on β-adrenergic receptors (Viemari et al., 2013) and the bombesin-like peptides neuromedin B (NMB) and gastrin releasing peptide (GRP) (Li et al., 2016). Other neuromodulators will facilitate the generation of sighs and simultaneously inhibit eupneic breathing like oxotremorine acting on muscarinic acetylcholine receptors (Tryba et al., 2008). Moreover, various neuromodulators activate both sighs and eupneic breathing synergistically like serotonin (Peña & Ramirez, 2002), noradrenergic modulation acting on alpha-receptors (Viemari, Garcia, Doi, & Ramirez, 2011), or substance P via neurokinin 1 receptors (Ruangkittisakul et al., 2008). These synergistically acting modulators could target developing brain homeobox 1 (Dbx1) or vesicular-glutamate transporter type 2 (Vglut2) populations within the preBötC, since optogenetic activation of these neurons evokes both eupnea and sighs. Somatostatin (SST) has inhibitory effects on both rhythms; however, it appears that SST has a more dramatic effect on sighs since this neuromodulator completely abolishes sighing while reducing eupneic frequency (Ramirez-Jarquin et al., 2012). An important question is how neurons located in different regions differentially modulate sighs within a behavioral context. It has been demonstrated that C1 neurons within the RTN activate arousal and sighs. Interestingly, selectively stimulating NMB neurons in this region failed to stimulate sighs (Souza, Stornetta, Stornetta, Abbott, & Guyenet, 2020), which is inconsistent with the hypothesis that NMB plays a critical role in the generation of sighs (Li et al., 2016). Indeed, it is more likely that NMB neurons in the RTN are critical for the generation of sneezing, given that stimulating NMB neurons in this region evokes active expiration (Souza et al., 2020). NMB neurons also receive input from sneeze-evoking regions that are activated by irritants in the nose and these neurons project to the caudal VRG to activate the sneezing behavior (Li et al., 2021).

During the normal eupneic cycle sighs typically occur every 5–30 min in slice preparations, but rarely occur more frequently without the application of certain modulators (e.g. norepinephrine, βAr1/2 agonist). Even with these neuromodulators, sighs do not occur sequentially. This suggests that the sigh has a refractoriness, similar to normal eupneic activity (Baertsch, Severs, Anderson, & Ramirez, 2019), an idea that remains relatively unexplored. The refractory period of the sigh rhythm could be tied to calcium overactivation in the network, and could also explain why a high amount of variability in the sigh rhythm is observed in transverse preBötC slices, analogous to altering the state of the eupneic network with varying concentrations of potassium (Ruangkittisakul et al., 2008; Smith, Ellenberger, Ballanyi, Richter, & Feldman, 1991).

Differential inhibition has been suggested as a potential mechanism for how eupnea and sighs can both be generated simultaneously from within the preBötC. The emergence of sigh rhythmogenesis during late prenatal to early postnatal stages in the mouse were suggested to be, at least in part, due to the role of glycinergic inhibition (Chapuis, Autran, Fortin, Simmers, & Thoby-Brisson, 2014; Thoby-Brisson, 2018). However, further experimentation has indicated that inhibition during later postnatal stages and into adulthood does not have any effect on the regulation or control of sighs (Borrus, Grover, Conradi Smith, & Del Negro, 2020; Li et al., 2016), a finding that may be due to developmental lineage changes that occur during these earlier time points (Thoby-Brisson & Greer, 2008).

The generation of sighs within the preBötC may be related to the stability of the respiratory rhythm as observed in different slice preparations. The horizontal slice preparation produces a much more stable network state with a reduction in the frequency of sighs. This could be a result of the network being more intact and includes a larger population of both excitatory and inhibitory neurons or could also indicate that other more dorsal areas contribute to sigh generation and are not included in this slice preparation. The horizontal slice preparation with slower rhythms and reduced sigh frequency, albeit complete lack of sighs in some cases, certainly is more similar to the in vivo anesthetized preparation (Baertsch et al., 2018).

There are several recent studies that suggest astrocytes could play a role in sigh modulation. Astrocytes have more recently been found to underly many of the functions of areas endowed with chemosensitive properties, such as the RTN/pFRG region (Funk, 2010; Ramirez, Severs, Ramirez, & Agosto-Marlin, 2018; Sheikhbahaei et al., 2018; Turovsky et al., 2016). Here, astrocytes are known to express pH sensitive currents which are characteristic of Kir4.1-Kir5.1 channels (Mulkey, Wenker, & Kréneisz, 2010; Wenker, Kréneisz, Nishiyama, & Mulkey, 2010) and release different neuromodulators including some vasoactive substances which are able to alter breathing frequency (e.g., prostaglandin E2 (PGE2), ATP, and D-serine (Angelova et al., 2015; Beltrán-Castillo et al., 2017; Eugenin Leon, Olivares, & Beltran-Castillo, 2016; Forsberg et al., 2016; Forsberg, Ringstedt, & Herlenius, 2017; Marina et al., 2017; Sheikhbahaei et al., 2018; Turovsky et al., 2016)). Furthermore, astrocytes express many similar receptor subtypes as neurons, and therefore could modulate neuronal activity through different types of gliotransmitter release in response to activation by similar molecules. Astrocytes are further implicated in regulating breathing during metabolic changes, such as during physical exertion (Sheikhbahaei et al., 2018). Nevertheless, the direct role of astrocytes in modulating sighing is still under direct investigation.

4. Sighs and arousal

A clear indication that sighs are not only linked to the control of lung volume comes from its association with arousal, indicating the role of largely unknown sigh circuitry in connecting to the rest of the brain (Glogowska et al., 1972). Sighs occur simultaneously with arousal (Thach & Lijowska, 1996), as well as during REM and NREM sleep states and the transition between these ‘state switches’ (Anderson, Dick, & Orem, 1996; McNamara, Lijowska, & Thach, 2002; Thach & Lijowska, 1996). Sighs often occur at the initiation of the arousal response, preceding apneas, following increased somatic activity, variable heart rate deceleration, and during sleep state transitions (McGinty et al., 1979). In order to explore whether this arousal response may play a role in sudden infant death syndrome (SIDS), behavioral studies confirmed that in infants, arousal begins with the occurrence of a sigh (i.e., augmented breath), followed by thrashing, eye opening, and repositioning of the head (Lijowska, Reed, Chiodini, & Thach, 1997; McNamara, Wulbrand, & Thach, 1998; Thach & Lijowska, 1996). Indeed, there are reports indicating that some SIDS or near-miss cases report with a reduction in the number of arousals from sleep (Dunne, Fox, O’Regan, & Matthews, 1992; Garcia, Koschnitzky, & Ramirez, 2013; Kahn et al., 1992; Kato et al., 2006; McCulloch, Brouillette, Guzzetta, & Hunt, 1982; Ramirez, Ramirez, & Anderson, 2018; Sawaguchi et al., 2005; Schechtman, Harper, Wilson, & Southall, 1992). However, more studies are required in this area to determine if there is a concrete role for sighing in SIDS victims. As already mentioned, sighs are particularly sensitive to hypoxic challenge (Bartlett, 1971; Bell, Ferguson, Kehoe, & Haouzi, 2009; Bell & Haouzi, 2010; Cherniack et al., 1981; Hill, Garcia, Zenella, & Upadhyaya, 2011; Lieske et al., 2000; Schwenke & Cragg, 2000); moreover, chemical sensitivity to hypoxia exists in the isolated brainstem slice preparation (Fig. 2) (Hill et al., 2011; Lieske et al., 2000). Further mechanistic insights into the association of sighs and arousal come from optogenetic experiments indicating that sighs and the associated cardiorespiratory response are activated by C1 noradrenergic neurons prior to arousal (Burke et al., 2014; Souza, Amorim, Moraes, & Machado, 2017; Souza, Bonagamba, Amorim, Moraes, & Machado, 2016; Souza et al., 2020). These catecholaminergic neurons could activate sighs via β-adrenergic receptors known to specifically activate sighs (Viemari et al., 2013).

Fig. 2.

Hypoxic response in the in vitro preBötC transverse slice and in the in vivo anesthetized preparation. (A-B) The hypoxic response in the preBötC in vitro preparation is comparable to the in vivo anesthetized preparation. This response is characterized by an initial ‘augmenting phase’, resulting in an increased frequency of normal eupneic activity, an onset of sighs, and followed by a decrease in respiratory frequency and onset of gasping. Fictive activity in vitro occurs on a slower timescale than in vivo. In vitro hypoxic response was ended after start of gasping.

The sensitivity to hypoxia and hypercapnia in the context of arousal may be one of the mechanisms that awakens babies sleeping in the prone position. This position leads to a buildup of expired CO₂ and a decrease in inspired O₂ (Bolton, Taylor, Campbell, Galland, & Cresswell, 1993; Chiodini & Thach, 1993; Kemp, Kowalski, Burch, Graham, & Thach, 1993; Kemp & Thach, 1991). The change in blood gases will likely activate C1 neurons and sighs which would contribute to the arousal of the sleeping infant (Lijowska et al., 1997; McNamara et al., 1998). Therefore, in addition to the important role in atelectasis, sighs seem to also play a central role in the events leading to the arousal response and replenishing inspired O₂ (Ayas, Brown, & Shea, 2000; Fewell, 2005; Horne, Parslow, & Harding, 2005;; Lijowska et al., 1997;; Masa et al., 2003; McNamara et al., 1998; Parslow, Harding, Cranage, Adamson, & Horne, 2003; Thach & Taeusch, 1976).

5. Sighing and the control of brain states

The link between C1 neurons and the sigh could provide a mechanistic explanation for the role of sighs in the regulation of different brain states. Sigh generation is facilitated during the transition from wakefulness to NREM sleep, or from sleep to arousal (Burke et al., 2014; Burke, Kanbar, Viar, Stornetta, & Guyenet, 2015; Eckert, Jordan, Merchia, & Malhotra, 2007; Orem & Trotter, 1993). C1 neurons innervate the locus coeruleus (LC), potentially increasing brain-wide noradrenaline by activating populations of neurons in the LC and adrenergic neurons in the pons via glutamatergic drive (Abbott et al., 2012; Guyenet et al., 2013). These C1 neurons are also responsive to hypoxic stimuli (Silva, Takakura, & Moreira, 2016). Direct connections between orexinergic neurons and C1 neurons in the rostral ventrolateral medulla (RVLM) provide further support for the role of this system in regulating sighing behavior at the level of the preBötC (Bochorishvili et al., 2014; Burke et al., 2014; Guyenet et al., 2013; Souza et al., 2020). Orexin, which is released from hypothalamus, will likely contribute to the association between sighing and the regulation of brain states during sleep-wake transitions. This neuropeptide modulates the critical breathing regions, both through direct projections to the preBötC (Young et al., 2005) and through secondary connections from the LC (Nattie & Li, 2012). Orexin modulates chemosensitivity during wakefulness (Gestreau, Bevengut, & Dutschmann, 2008), is further related to stress and anxiety (Johnson, Molosh, Fitz, Truitt, & Shekhar, 2012), and plays a significant role in the modulation of circadian rhythms (Kantor et al., 2009; Tsujino & Sakurai, 2012). Furthermore, studies utilizing the orexin receptor antagonist, almorexant, show that inhibition of orexin signaling reduces the frequency of sighs, as well as reduces the post-sigh apnea (Brisbare-Roch et al., 2007; Li & Nattie, 2010). The involvement or orexin in regulating the sigh is not straightforward, however. Studies in a mouse knockout model of Orexin A/B show that these mice present with a narcoleptic phenotype and have further disruptions in REM sleep. Interestingly, these mice had an increase in the number of central apneas during sleep states, with no difference in sighs or post-sigh apneas. However, the authors did report a reduction in the CO₂- mediated ventilatory response (Nakamura, Zhang, Yanagisawa, Fukuda, & Kuwaki, 2007; Nattie & Li, 2012).

It is relatively well established that the orexin/hypocretin system modulates brain state in part through the LC (Hagan et al., 1999). Projections of orexinergic neurons are most dense in the LC, and Orexin A increases firing of noradrenergic neurons within this area (Hagan et al., 1999). Hypocretin expressing neurons within the lateral hypothalamic area have also recently been identified to activate claustrophobia-related sighing in mice (Li et al., 2020).

Breathing rhythms are understood to contribute to oscillations in other brain regions (Corcoran, Pezzulo, & Hohwy, 2018; Tort, Brankack, & Draguhn, 2018). Respiratory ‘entrained’ oscillations may arise from nasal respiration, initiated in the olfactory bulb and spread to cortical regions and the hippocampus (Heck et al., 2016; Ito et al., 2014; Tort, Brankack, et al., 2018). These rhythms may play a part in synchronizing activity states in different frequency ranges and are suggested to entrain a global rhythm throughout the brain (Tort, Brankack, et al., 2018; Tort, Ponsel, et al., 2018; Zhong et al., 2017). Although there is little evidence as to the functional role of respiratory oscillations, it is suggested that coordination with nasal respiration may be a means of ‘active sensing’ of the environment, thereby organizing the brain’s neural activity in a manner consistent with an organism’s environment (Corcoran et al., 2018). In lieu of this hypothesis, it is reasonable to suggest that the sensitivity of the sigh rhythm to hypoxia is an important component of active sensing, particularly so during infancy. Moreover, as discussed later in detail, the sigh has very important psychological roles (Vlemincx et al., in press); sighs are implicated in several anxiety disorders and are an overt expression of many emotions such as frustration, stress, anxiety and fear, depression, joy, desire and relief (Vlemincx, Meulders, & Abelson, 2017; Vlemincx, Meulders, & Luminet, 2018; Vlemincx, Taelman, Van Diest, & Van den Bergh, 2010; Vlemincx et al., 2009; Vlemincx, Van Diest, & Van den Bergh, 2016). Neural projections from the rostral medulla to the preBötC release the neuropeptides NMB and GRP to stimulate sighing (Li et al., 2016). The same circuit has more recently been identified as an ‘emotional sigh circuit’, as mice placed in confinement conditions experienced ‘emotional sighs’ which are also linked to NMB release from the same rostral area (Li et al., 2020), perhaps indicating the role of bombesin-like peptides stimulates sighs initiated in the cortex, while other mechanisms drive sighing locally at the level of the preBötC. Thus, respiratory entrained oscillations, and sighs specifically, have important implications both physiologically and emotionally. Further research will be necessary to identify whether spontaneously generated versus voluntary sighs incorporate the same neural pathways. Regardless, breathing techniques, such as meditative breathing and slow breathing, have been used as a therapeutic technique for anxiety disorders for thousands of years and are incorporated into many different medical practices all over the world.