Science breaks down Taoist practice into measurable active components, tests their effects against control conditions, and seeks mechanisms compatible with known biology. The concept of qi, understood as a substance or energy circulating through meridians, has never been shown to have a measurable physical counterpart; despite decades of instrumental attempts, none of these research programmes has produced replicable results meeting metrological standards. By contrast, the subjective phenomena that practitioners describe using the vocabulary of qi (diffuse warmth, tingling, a sensation of flow, abdominal fullness, heaviness in the hands) are, in fact, perfectly real as perceptual events, and contemporary science possesses a robust conceptual framework to explain them: peripheral vasodilation of autonomic origin, discharges from cutaneous and deep mechanoreceptors and thermoreceptors, attentional amplification of interoceptive signals, motor and sensory imagery phenomena, and the brain’s predictive inferences regarding the body’s states.
For scientific analysis, a session of Nei Gong or qi gong is not a monolithic entity but a combination of components, each of which has its own experimental literature. Six such components can be distinguished.
The first component is low- to moderate-intensity physical exercise. The dynamic forms of Nei Dan engage the body at a typical intensity of 2 to 4 metabolic equivalents (METs), equivalent to a slow to moderate walk, with a working heart rate of between 40 and 60 per cent of cardiac reserve. This modest but genuine level of exercise is sufficient to trigger some of the classic adaptations associated with physical activity: improved endothelial function, enhanced insulin sensitivity, maintenance of muscle mass and control, and moderate osteogenic stimulation through weight-bearing.
The second component is slow movement under continuous attentional control. Slowness is not merely a reduction in speed: it alters the neurophysiological nature of the movement. A movement performed at a very reduced speed can no longer rely on pre-programmed ballistic motor patterns; it requires closed-loop control, with constant integration of proprioceptive feedback (neuromuscular spindles, Golgi tendon organs, capsular and cutaneous receptors) and continuous corrections via the cerebellocortical loops and the basal ganglia. It is precisely this control mechanism that makes slow movement an intensive proprioceptive training exercise, the effects of which are evident in the documented improvement in balance and postural stability.
The third component is deliberately slowed, deep and abdominal breathing, typically maintained at between four and eight cycles per minute in a trained practitioner, compared with twelve to sixteen during spontaneous rest. As we shall see, this is the component whose physiology is best characterised, with robust and reproducible effects on the baroreflex and heart rate variability.
The fourth component is attentional regulation: sustained attention focused on bodily sensations, the path of the breath, or conventional bodily regions (the lower dantian, the region below the navel, in traditional terminology). This component directly links Nei Dan and nei gong to the family of attentional and meditative practices, the neuroimaging correlates of which are now well described.
The fifth component is mental, motor and sensory attention: the practitioner perceives pathways, flows and expansions. Motor imagery, however, activates a large proportion of the sensorimotor networks involved in actual execution (premotor cortex, supplementary motor area, parietal cortex, cerebellum), and imagery of heat or flow can objectively modulate cutaneous microcirculation via the autonomic nervous system, as demonstrated by studies on thermal biofeedback and research into Tummo meditation.
Finally, the sixth component is contextual and psychosocial: ritualisation, a calm setting, the perceived expertise of the teacher, group practice, and the coherence of the traditional explanatory framework. These factors, far from being mere ‘noise’, are powerful and measurable modulators of therapeutic effects, acting via the now well-mapped pathways of the placebo effect and expectations (activation of endogenous opioid and dopaminergic systems, descending modulation of nociception by the periaqueductal grey matter).
The entire methodological challenge of research into Nei Dan lies in this composite nature: clinical trials most often compare the complete practice with a waiting list, standard care or a conventional exercise, meaning that it is rarely possible to attribute the overall effect to a single component. Mechanistic studies, on the other hand, isolate the components (breathing alone, attention alone, movement alone) and enable the puzzle to be pieced together. It is this dual perspective that structures the remainder of the text.
If one were to identify the physiological core of qigong, it would be slow breathing. The literature on slow-paced breathing constitutes one of the most robust bodies of evidence in all mind-body research, and its conclusions are directly applicable since breath control is explicitly central to the practice (the word ‘gong’ refers to ‘qi’, whose primary meaning is ‘breath’).
In adults at rest, the spontaneous respiratory rate is around twelve to sixteen cycles per minute. When it is deliberately reduced to around six cycles per minute (i.e. a cycle of ten seconds, a frequency of 0.1 Hz), the cardiovascular system enters a specific state known as resonance. Two oscillators then superimpose: respiratory sinus arrhythmia, that is, the acceleration of the heart rate during inhalation and its deceleration during exhalation, mediated by the phase modulation of vagal tone at the sinus node; and the Mayer wave, a spontaneous oscillation in blood pressure with a period of approximately ten seconds, linked to the delay in the sympathetic baroreflex loop. At 0.1 Hz, these two oscillations come into phase and add together: the amplitude of the heart rate oscillations is then at its maximum, heart rate variability (HRV) reaches values far higher than during spontaneous rest, and the baroreflex gain—the sensitivity with which the carotid and aortic baroreceptors correct blood pressure beat by beat—is transiently increased.
The mechanisms underlying this phenomenon are well understood. Deep inspiration lowers intrathoracic pressure, increases venous return and stretches the pulmonary mechanoreceptors; the pulmonary inflation reflex (the Hering–Breuer reflex and afferents from slow-adapting stretch receptors, travelling via the vagus nerve to the solitary tract nucleus) transiently inhibits cardiac vagal tone, resulting in an inspiratory acceleration. During exhalation—particularly if it is prolonged, and Nei Dan practices systematically prolong exhalation—vagal tone is restored and becomes dominant: the heart rate slows. The repetition of this high-amplitude cycle constitutes a genuine exercise of the vagal loop, and longitudinal data show that regular practice increases the resting HRV, particularly its vagal components (RMSSD, high-frequency power), and lowers resting blood pressure in hypertensive patients, with reductions of the order of five to ten millimetres of mercury for systolic pressure in meta-analyses of randomised trials—a clinically significant order of magnitude, comparable to low-intensity antihypertensive monotherapy.
It is important to emphasise why HRV has become the central indicator in this field. High vagally mediated heart rate variability is a cross-sectional marker of autonomic flexibility: it independently predicts cardiovascular mortality and the occurrence of coronary events, and correlates with emotional regulation capacity and executive function, in line with the neurovisceral integration model, which links the prefrontal cortex, the amygdala and the autonomic nuclei of the brainstem into a single regulatory circuit. In this context, the vagus nerve is not merely an efferent pathway for slowing the heart rate: it is a bidirectional cable in which eighty per cent of the fibres are afferent, constantly transmitting information on the mechanical and chemical state of the viscera back to the solitary tract nucleus; this information is then relayed to the parabrachial nucleus, the thalamus, the insula and the cingulate cortex. Slowing down and deepening the breath therefore amounts to a massive alteration of the visceral afferent input received by the brain, which provides a direct mechanistic link between the act of breathing and mental states, and sheds precise light on the traditional intuition that the breath is the lever of the mind.
The abdominal component of breathing, which is practised extensively in neigong, adds a biomechanical layer. Deep diaphragmatic breathing increases the range of movement of the diaphragm, which can range from one to two centimetres during shallow breathing to seven to ten centimetres during trained breathing. This substantial range of movement has several consequences: cyclical mechanical massage of the abdominal viscera and cyclical variation in intra-abdominal pressure, which promotes splanchnic venous return and lymphatic drainage; stimulation of vagal and spinal afferents from the mesentery and the abdominal wall; and coordinated recruitment of the transverse abdominis and the pelvic floor, which makes deep breathing a form of training for the abdominal cage, a key structure in lumbar stabilisation. The so-called ‘reverse’ breathing of Nei Gong (abdominal retraction on inhalation, relaxation on exhalation) is, from this perspective, an exercise in dissociation and fine voluntary control of the diaphragm-transverse abdominis-perineum synergies; its specific physiology has been little studied in its own right, but it can be readily analysed within the framework of respiratory motor control.
Slow breathing alters blood chemistry in a subtle but real way. Unlike hyperventilation, slow, deep breathing, when performed correctly, maintains alveolar ventilation at a more or less constant level: tidal volume increases in proportion to the slowing of the breathing rate, whilst the partial pressure of carbon dioxide remains normal or rises very slightly. This slight relative hypercapnia shifts the haemoglobin dissociation curve (Bohr effect) in a direction that favours the release of oxygen to tissues, increases cerebral blood flow through vasodilation of CO₂-sensitive cerebral arterioles, and reduces neuronal excitability – a pattern consistent with the reported calming effects. The improvement in ventilatory efficiency (ventilation-perfusion ratio, recruitment of the lung bases through diaphragmatic breathing) also explains some of the documented benefits in patients with chronic obstructive pulmonary disease, in whom Nei Dan programmes improve the six-minute walk distance and dyspnoea scores in several randomised trials.
Beyond respiratory mechanics, the most consistent effect of Nei Dan practices relates to the overall balance of the autonomic nervous system and the major neuroendocrine stress axes.
The first axis involved is the hypothalamic-pituitary-adrenal (HPA) axis. Chronic psychological stress maintains elevated secretion of hypothalamic corticotropin-releasing hormone (CRH), pituitary adrenocorticotropic hormone (ACTH) and adrenal cortisol, ultimately leading to a flattening of the circadian rhythm of cortisol, glucocorticoid receptor resistance and the disinhibition of inflammatory cascades normally suppressed by cortisol. Interventional studies on Nei Dan and related practices show, with reasonable consistency though notable heterogeneity, a reduction in salivary or blood cortisol after a few weeks to a few months of regular practice, particularly in subjects who were initially stressed, anxious or depressed. The most parsimonious mechanistic interpretation involves a reduction in repeated amygdala and sympathetic activation: fewer triggers of the HHS axis, improved sensitivity of glucocorticoid feedback, and gradual restoration of the circadian rhythm.
The second axis is the sympathico-adrenergic axis. Regular practice lowers indicators of resting sympathetic tone: circulating catecholamines, skin conductance, the low-to-high-frequency ratio of HRV (with all the interpretative caveats that this ratio entails), and blood pressure. In hypertensive individuals, meta-analyses of randomised trials focusing on Nei Dan conclude that there are clinically significant reductions in blood pressure, of a similar magnitude to those achieved through conventional aerobic exercise, with the practical advantage of often better adherence among elderly or unfit individuals, for whom the low intensity and absence of joint impact remove the main barriers to physical activity.
The third aspect, which has been the most extensively studied over the past fifteen years, concerns the immune and inflammatory systems. Two pathways directly link the nervous system to the immune system. The first is the cholinergic anti-inflammatory pathway described by Kevin Tracey’s team: vagal afferents detect peripheral cytokines; vagal efferents, via the celiac ganglion and the splenic nerve, lead to the release of noradrenaline in the spleen, where a subpopulation of T lymphocytes produces acetylcholine which, by binding to the alpha-7 nicotinic receptors on macrophages, inhibits the production of tumour necrosis factor alpha (TNF-α) and other pro-inflammatory cytokines. Any intervention that sustainably increases vagal tone – and slow breathing is one such intervention – therefore offers a plausible biological pathway towards reducing low-grade systemic inflammation. The second pathway is transcriptional: research into the conserved transcriptional response to adversity (CTRA) has shown that chronic stress shifts gene expression in leukocytes towards a pro-inflammatory profile (overexpression of NF-κB- and AP-1-dependent genes) and a hypo-antiviral profile (underexpression of type I interferon genes). A systematic review of functional genomics studies focusing on mind-body interventions—meditation, yoga, tai chi and Nei Dan—concludes that the most reproducible pattern is precisely a downregulation of NF-κB signalling, that is to say, a partial reversal of the chronic stress signature.
In terms of clinically measured biomarkers, meta-analyses of tai chi and Nei Dan report modest but significant reductions in C-reactive protein and interleukin-6 in subjects with low-grade inflammation at baseline (older adults, cancer patients in remission, those with heart failure, and people with depression), with less pronounced effects in young, healthy individuals, which is to be expected for any intervention acting on a parameter that is already within the normal range. A few randomised trials have investigated markers of cellular ageing: a controlled trial conducted among women exposed to severe chronic stress observed an increase in telomerase activity following a qi gong programme, a finding consistent with the data obtained for meditation and tai chi, but which requires replication on a larger scale before any firm conclusions can be drawn. Overall, the picture that emerges is that of a mild anti-inflammatory intervention whose dominant mechanism is neuro-immune (sympathetic inhibition, vagal tone, HHS axis) rather than metabolic.
Electroencephalography provided the first objective data, from the 1960s to the 1980s on meditators and from the 1990s onwards on qi gong practitioners. The typical electrophysiological pattern associated with the practice combines an increase in alpha power (8–12 Hz), particularly in the posterior and frontal regions, with a frequent shift of the alpha peak towards lower frequencies, and an increase in medial frontal theta (4–7 Hz) in advanced practitioners during states of calm absorption. Medial frontal theta is generated primarily in the anterior cingulate cortex and is observed during states of sustained, relaxed attention; its presence in experienced practitioners, which correlates with hours of practice, indicates effective training of the midline attentional networks. High-density EEG studies with source localisation have also shown that distinct Nei Dan states (for example, concentration meditation on the dantian versus a ‘small celestial circulation’-type attentional circulation) produce differentiated source topographies, suggesting that traditional taxonomies of internal states correspond, at least in part, to distinct neural configurations.
Functional magnetic resonance imaging clarifies the anatomy of these states. Body-centred attentional practices reproducibly engage the insula (particularly the right anterior insula, the primary interoceptive cortex receiving visceral afferents relayed via lamina I of the spinothalamic tract and the nucleus of the solitary tract), the dorsal anterior cingulate cortex, the secondary somatosensory cortex and the frontoparietal networks involved in attentional control. Simultaneously, deactivation or reconfiguration of the default mode network (medial prefrontal cortex, posterior cingulate cortex, precuneus) is observed; hyperactivity and hyperconnectivity in this network are associated with rumination and depressive disorders. Longitudinal and cross-sectional studies focusing specifically on tai chi and Baduanjin—the forms most readily accessible to imaging protocols—report, in regular practitioners, an increase in cortical thickness or grey matter volume in the insula, the hippocampus and certain prefrontal regions, altered functional connectivity between the default mode network and control networks, and correlated improvements in memory and executive function scores among older adults. These findings form part of the broader, very well-established framework of training-dependent neuroplasticity: the adult brain remodels its neural maps and connections in response to attentive repetition, and training that combines fine motor skills, balance, breathing and attention constitutes a multimodal plastic stimulus.
The scientific concept that best unifies the internal dimension of neigong is that of interoception: the brain’s perception of the body’s own physiological state. Interoception has a dedicated anatomical pathway and a computational framework that has become dominant, that of active inference: the brain maintains a predictive model of the body’s states and does not perceive raw signals but rather the discrepancy between prediction and afferent input, weighted by the precision attributed to each source. This framework sheds light precisely on what Nei Gong-style training achieves: by repeatedly directing highly precise attention towards specific bodily regions (the abdomen, palms, spine), it increases the sensitivity of the corresponding interoceptive and proprioceptive signals, refines the cortical maps of these regions, and makes ordinarily subthreshold physiological events perceptible, peripheral pulse, vasomotor waves, peristalsis, and postural micro-adjustments. The sensations of heat, current or tingling described as ‘qi’ find a coherent explanation here: sustained attention directed towards a region objectively increases local cutaneous perfusion (attention-vasomotor coupling mediated by the sympathetic nervous system), lowers perceptual thresholds, and the predictive model—shaped by the cultural expectations of the practice—transforms these amplified afferents into structured percepts of flow. This is not an illusion in the trivial sense: thermal biofeedback studies show that trained subjects actually raise the skin temperature of their hands by several degrees, and research conducted on practitioners of tummo (a Tibetan technique akin to nei gong) has documented measurable increases in peripheral and core body temperature, combining respiratory mechanisms (breath retention and abdominal contraction) with mental imagery. Partial voluntary control of autonomic functions, long considered impossible, is now an established experimental fact, with ‘Nei Gong’-type protocols providing a spectacular demonstration by showing massive voluntary sympathetic activation, the release of adrenaline and an objective attenuation of the inflammatory response to bacterial endotoxin in trained subjects.
Nei Gong differs from simple gentle exercise in its precise motor requirements: axial alignment (suspension of the crown of the skull, elimination of excessive cervical and lumbar curvatures, sacral tilt), relaxation of the superficial muscles in favour of deep support, transmission of forces from the ground to the hands via a continuous chain, and the development of a so-called ‘internal’ force (jin) that is qualitatively distinct from raw muscular strength (li). These traditional principles can be translated fairly faithfully into the language of contemporary biomechanics and motor control.
The first area of translation is postural control. Standing is never static: it is a constant oscillation regulated by the integration of visual, vestibular and proprioceptive inputs. The postures practised in Nei Dan (zhan zhuang, the tree posture) – with slightly bent knees, a wide stance and a lowered centre of mass – constitute training in postural regulation under conditions of increased demand: the sustained flexion of the lower limbs requires prolonged isometric work of the quadriceps and the extensor chains, whilst the instruction to relax the upper body forces a dissociation between stability and stiffness. However, excessive stiffness and generalised co-contraction are precisely the default postural strategy of older adults, anxious individuals and patients in pain, and this is counterproductive: it stiffens the system, impairs the precision of postural corrections and increases energy expenditure. Training in relaxed stability amounts to relearning postural strategies characterised by low co-contraction and a strong contribution from ankle-hip corrections. This is the most likely mechanism underlying the most robust clinical findings in this entire field: the reduction in the risk of falls among older adults, demonstrated for tai chi in large-scale randomised trials and confirmed for forms of Nei Dan in meta-analyses, with reductions in the incidence of falls of between twenty and forty per cent, improvements in balance scores (Berg, Timed Up and Go) and postural confidence.
The second area concerns connective tissues. Research into fasciae—the continuous connective tissue network that envelops and connects muscles, bones and internal organs—has established that these tissues are richly innervated (free nerve endings, Ruffini and Pacini corpuscles, and a majority of fine fibres with interoceptive and nociceptive functions), that they play an active role in the transmission of forces between adjacent muscle groups (lateral myofascial transmission, measured experimentally between synergists and even between segments), and that they remodel their collagen matrix in response to the loads applied to them (fibroblastic mechanotransduction: slow, sustained stretching modulates collagen expression, the activity of metalloproteinases and the production of hyaluronic acid by fasciocytes, which determines the gliding of the layers). The movements of Nei Gong—slow, spiral-like stretches of large amplitude, linked together smoothly—apply to the fascial planes precisely the type of gentle cyclic load which, in experimental models, maintains hydration, inter-layer gliding and the elasticity of the network. The traditional notion of a continuous connection from the foot to the hand finds reasonable anatomical support in the myofascial continuities described by modern dissection, even though the exact functional significance of these chains remains an active and sometimes controversial area of research.
The third axis concerns the abdominal cavity and intra-abdominal pressure. The stability of the lumbar spine depends less on the maximum strength of the erector spinae muscles than on the anticipatory, fine-tuning of the deep muscles—the transverse abdominis, multifidus, diaphragm and pelvic floor—which pressurise the abdominal cylinder a few tens of milliseconds before any movement of the limbs (anticipatory postural adjustments). Nei gong, which explicitly trains the coordination of breathing, the perineum and the abdomen, and the initiation of movements from the centre (the dantian as both the centre of mass and the command centre), can be analysed as systematic training of these deep synergies. Electromyography studies conducted on tai chi practitioners reveal characteristic activation patterns: proximal activation preceding distal activation, brief, phasic bursts in the force output of internal styles, and low levels of antagonist co-contraction during slow movements – all hallmarks of a movement organisation that prioritises the sequential transmission of forces segment by segment (the principle of the proximal-distal kinetic chain, the same principle that governs throwing or a racket swing) over local force production. Internal strength in martial traditions, stripped of its energetic vocabulary, corresponds to an identifiable motor skill: force production through kinetic summation from the points of support, stiffness adjustable in real time, and the utilisation of the elastic pre-stretch of tissues—a skill that can be trained, and whose slow acquisition (taking years) is consistent with the known timescales of expert motor learning.
Because its regular practitioners are predominantly middle-aged adults and older people, Nei Gong has become a prime area for the study of non-pharmacological interventions for ageing, and several lines of evidence merit detailed consideration.
The first line of research concerns cognitive ageing. Randomised trials conducted in healthy older adults or those with mild cognitive impairment show that six months to one year of regular practice improves scores for executive function, working memory and processing speed, with effect sizes comparable to those of aerobic exercise, and that these gains are accompanied, as shown by imaging, by structural changes, increased hippocampal volume or a slowing of its atrophy, prefrontal and insular cortical thickening, as well as an increase in circulating neurotrophic factors, foremost among which is brain-derived neurotrophic factor (BDNF), whose role in adult hippocampal neurogenesis and synaptic plasticity has been established in animals. The overall mechanism is likely to be additive: the cardiovascular component of exercise (cerebral perfusion, angiogenesis, BDNF) is complemented by the component of complex motor learning – memorising and refining sequences of coordinated movements is in itself a cognitive load – and the meditative attentional component, each of which has been shown, in isolation, to have effects on cognition in older adults. This convergence makes Chinese gymnastics a strong candidate for the prevention of cognitive decline, a hypothesis currently being tested in large-scale, multicentre trials.
The second area concerns the ageing musculoskeletal system. Sarcopenia – the age-related loss of muscle mass and function – and osteopenia are primarily addressed by resistance and impact exercises, areas in which qi gong, due to its low intensity, is not the optimal tool; the data nevertheless show significant, albeit modest, effects on lower limb strength (maintaining flexed postures in low positions is equivalent to isometric strengthening), on lumbar and femoral bone mineral density in postmenopausal women, and, above all, on frailty syndrome, a geriatric condition characterised by slowness, weakness, fatigue and vulnerability to physiological stress: recent meta-analyses conclude that there is an improvement in frailty status and overall functional capacity among frail or pre-frail individuals, a population for whom conventional exercises are often inaccessible. This is an important public health issue: the intervention is not the most potent in absolute terms, but it is accessible to precisely those who cannot undertake any other form of exercise, and its long-term adherence is remarkably good; the aesthetic and meditative aspects of the practice, its communal nature and the absence of physical strain all support adherence where medical exercise programmes experience high drop-out rates.
The third area concerns sleep, the deterioration of which with age (reduction in deep slow-wave sleep, fragmentation, phase advance) contributes to cognitive and metabolic decline. Meta-analyses focusing on Nei Dan in older adults conclude that there is a consistent improvement in subjective sleep quality (Pittsburgh Sleep Quality Index), on a par with first-line behavioural therapies, whilst actigraphic and polysomnographic data, though still limited, suggest a reduction in sleep latency and night-time awakenings. The proposed mechanisms are consistent with the overall picture: a reduction in evening sympathetic hyperarousal, improved circadian regularity through morning outdoor practice (light and activity as time cues), and an anxiolytic effect.
A fourth, more nuanced line of research concerns thermoregulation and microcirculation. Laboratory studies conducted on participants during exercise show increases in skin temperature at the extremities of one to three degrees during sessions, accompanied by an increase in microcirculatory flow measured by laser Doppler flowmetry – an objective correlate of the classic sensation of warm, swollen hands. The mechanism involves a reduction in cutaneous sympathetic vasoconstrictor tone combined with local metabolic vasodilation resulting from light muscular work. This phenomenon, which is physiologically commonplace, is interesting for two reasons: it illustrates the attention-autonomy loop already described (vasodilation is more pronounced in regions that are the focus of attention), and it provides the phenomenology of qi with its most immediate basis; the warmth that follows the movement of the hands is not energy circulating, but is precisely what one feels, from within, as an actual redistribution of blood flow.
This is a modern perspective on our Practice.
