Stop and sniff the roses: Insights into the voluntary control of respiration

Respiration, although primarily subserving the fundamental requirement of gas exchange, is a multifaceted process that is intricately linked with other physiological systems. Olfaction is tightly integrated with taste and eating behaviour, and adds to the sensorium required for environmental exploration. Coughing and sneezing form part of our defences against external pathogens, and apnoea combined with cold water immersion can trigger the life-saving diving reflex seen in all air-breathing vertebrates. All of these processes require precise co-ordination and control, carried out in large part in the CNS. Brainstem centres working in tandem with peripheral chemo- and mechano-receptors are responsible for involuntary control of respiration, whereas cortical centres exert voluntary or behavioural control. Pathophysiological disruption of any of these processes at any level (i.e. from the end-organs to peripheral and central control systems) can lead to a multitude of pathological phenomena, including clinically significant human diseases. Although end organ dysfunction and involuntary respiratory control has been relatively well-characterised in many human diseases, CNS dysfunction and its effect on voluntary respiration control is less well understood, particularly in humans. In this issue of The Journal of Physiology, Granget et al. (2025) provide insight into the neurophysiological mechanisms underlying voluntary respiratory control. These findings could in turn lead to a better understanding of the impacts of neurological disease on these systems. Granget et al. (2025) recruited six subjects with drug-resistant epilepsy who were undergoing intracranial EEG evaluation with stereotactically implanted depth electrodes as part of their routine clinical evaluation for epilepsy surgery. This procedure, known as stereo-EEG, involves the placement of multiple (typically 5–15) recording electrodes targeting cortical structures hypothesised to be responsible for seizure generation. Stereo-EEG admissions represent a rare opportunity to explore brain physiology in awake humans. In this study, participants were asked to perform normal tidal breathing, as well as two manoeuvres: sniffing and breath-holding (apnoea). The oscillatory components of the EEG were analysed in typical frequency bands during the preparatory, execution, and recovery phases of these manoeuvres. During sniff manoeuvres, significant changes in activity were seen in the mesial temporal structures (hippocampus and parahippocampal gyrus), mostly in lower frequency bands (alpha and theta), whereas neocortical and insular regions also demonstrated changes, but more in higher frequencies (gamma). During short apnoeas, mostly lower frequency changes were noted in both the mesial temporal structures (particularly the amygdala) and temporal neocortex; however, the dynamics of changes during apnoeas were more complex than those observed during sniffs. The main limitations to the study are shared with all stereo-EEG-based studies: limited spatial sampling and a small number of subjects with varying implantation schemes. The cortex is only sparsely sampled during these procedures, with electrode placements restricted to areas assumed to be involved in a specific individual's seizures. Furthermore, the epileptic brain and networks are inherently pathological and may not represent normal physiology. Granget et al. (2025) attempted to control for the latter issue by excluding contacts with suspected epileptiform activity on EEG, as well as any contacts anatomically located within presumed epileptogenic regions. The results of the study by Granget et al. (2025) help us to understand the potential impacts of human neurological disease on respiratory physiology. Granget et al. (2025) describe some relevant examples of conditions with impaired olfaction and associated neurological dysfunction, such as Parkinson's disease and COVID-19. The study was, however, performed in individuals with epilepsy, and some even more direct clinical implications arise in this patient population. Epilepsy describes the condition of having an enduring predisposition to seizures. These seizures involve uncontrolled electrical activity in often quite specific and localised brain circuits, with seizure symptoms relating to the origin and spread of electrical activity in these networks. The ability to localise the specific site of seizure origin is critical to providing effective diagnoses and therapies for many people. This can be done with careful analysis of seizure symptoms (semiology), but relies heavily on a detailed understanding of the normal and abnormal physiology of the relevant brain circuits. Apnoea is a not uncommon semiologic feature in focal epilepsy, and particularly temporal lobe epilepsy. This is not surprising, as a number of temporal lobe and closely connected structures have long been known to impact respiration. More recently, the amygdala has been more specifically implicated in the genesis of the important phenomenon of ictal central apnoea (Rhone et al., 2020), and a newly-discovered inhibitory circuit between amygdala and brainstem may form part of the underlying mechanism (Gu et al., 2024). The findings reported by Granget et al. (2025) add detail to our understanding of the regional activation and temporal dynamics of EEG changes in these limbic structures, further informing our understanding of focal seizure semiology. Looking towards the future, although useful diagnostically, these findings may also be helpful therapeutically. Deep brain stimulation and responsive cortical stimulation are used to treat many forms of drug-resistant epilepsy. These treatments involve electrical stimulation of cortical and subcortical targets, which may, in turn, affect respiration (Talavera et al., 2023). In designing future neurostimulation therapies and choosing novel targets (Lundstrom et al., 2023), it will be important to understand the neural circuitry underlying apnoea in particular to avoid unintended respiratory consequences. By contrast, these targets and techniques, if able to stimulate respiration effectively, could also be studied as a novel and potentially life-saving intervention for sudden unexpected death in epilepsy (i.e. SUDEP), which is assumed to occur, at least in part, as a result of centrally mediated respiratory arrest following a tonic-clonic seizure. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. No competing interests declared. H.S., S.S. and L.B. were responsible for conceptualization. H.S. was responsible for writing the original draft. S.S. and L.B. were responsible for reviewing and editing No specific funding was received.