Neuroanatomical Foundations of R.O.P.

Labels within the illustrations remain in French. Figure captions and the article text are translated into English.

Authors: Nicolas Boitout, Guy Boitout

R.O.P. does not assume that an organ is physically ‘represented’ at a point on the foot. Its working hypothesis is different: a region of the foot is first a sensory gateway. Pressure deforms the skin and subcutaneous tissues, recruits mechanoreceptors and other somatic afferents, then carries a signal through a peripheral nerve to the spinal cord and higher centres [1-3].

At each of these levels, this activity may interact with somatic, autonomic and visceral circuits. We use the teaching term ‘neuroanatomical bridge’ for each of these possible sites of convergence. A bridge is neither a direct anatomical connection between a point on the foot and an organ nor proof of therapeutic efficacy: it is a plausible mechanism through which somatic input may access a network capable of modulating visceral function.

⌕

Two questions must remain separate. The first concerns the existence of somatovisceral modulation: can stimulation of the foot alter an autonomic or visceral function? The second concerns targeting: do some regions or nerves provide more regional or more reproducible access than others? Confusing these questions is the most common weakness of reflexology texts, and the criticism most justifiably directed at them [16,14,48,49].

Selective stimulation of the plantar branches provides experimental support for investigating differences between inputs within one nerve [15]. In humans, plantar vibration also demonstrates coupling between mechanical stimulation of the skin and sympathetic activity [49]. Findings on the saphenous nerve distinguish segmental proximity from the capacity for modulation; digestive and cephalic reflexes show effects at a distance. These findings support the working hypotheses of R.O.P.; their applicability to manual techniques and their clinical effects still need to be evaluated.

At a glance: essential conclusions

• The foot is a rich sensory interface. The glabrous skin of the sole contains several classes of low-threshold mechanoreceptors, with receptive fields and thresholds that vary by region [2,3].
• Four major somatic gateways can be distinguished at the foot and ankle: plantar/tibial, lateral/sural, dorsal/fibular and medial/saphenous. They are equivalent neither peripherally nor at nerve-root level [1].
• The plantar/tibial gateway is the best documented for mechanical stimulation of the foot: the medial and lateral plantar branches and the calcaneal branches converge into the tibial nerve [2,3,5,6].
• The tibial nerve contains fibres from multiple roots, classically L4–S3. Its potential overlap with the S2–S4 sacral organisation of pelvic function mainly concerns S2–S3. This overlap strengthens a hypothesis of regional pelvic targeting, but does not create a direct pathway to a viscus [1,8-11,14].
• Despite having no sacral overlap, the saphenous nerve modulates the bladder reflex in rats and cats and has yielded an initial clinical signal in humans [17-19,24]. This finding separates segmental proximity from the capacity for modulation: it is the central argument for the two-axis approach.
• Bridge 2 has experimental support: the medial and lateral plantar branches can produce different profiles of bladder modulation under electrical stimulation in rats [15]. In humans, plantar vibration modifies muscle sympathetic nerve activity [49]. These findings support investigating differences between regions and their applicability to manual techniques.
• The lumbosacral spinal cord is a site of integration. Somatic, perineal and visceral afferents may converge on interneurons and sacral circuits; the effect depends on the state of the network and is not a direct command to the viscus [9-11].
• Supraspinal pathways add a second level of integration. The more a mechanism depends on these general relays, the less it can, by itself, establish regional specificity [8,12,13,16,48].
• Two families of arguments must be distinguished. Bridges 1 to 4 build a proximity argument: they explain why the pelvis is the most accessible region from the foot. Bridges 5 and 6 build a network argument: they explain how a function sharing no segment with the foot may nevertheless be modulated.
• Input from the lower limb can modify cranial autonomic output. In rats, hind-paw stimulation recruiting tibial and saphenous afferents alters pupil diameter; the response disappears after section of the oculomotor nerve but persists after section of the cervical sympathetic trunks [43]. This is the only documented pathway in which segmental proximity is absent by definition.
• In the digestive system, somatovisceral reflexes have been documented for the duodenum via vagal pathways, for the jejunum and for the distal colon via sacral parasympathetic pathways [36-42].
• Clinical evidence varies by indication: favourable for overactive bladder [21,22], negative in two controlled trials of faecal incontinence [25,26], and favourable again in certain forms of constipation and dyssynergia [27,28].
• The pelvis remains the best candidate for regional targeting from the foot, and the bladder is the best-documented visceral model. The distal colon and rectum are important extensions [30-35].

1. The foot: a sensory interface and four somatic gateways

1.1. From pressure to afferent input

The glabrous skin of the sole is not a passive surface. Human microneurographic recordings describe four major classes of low-threshold tactile afferents: rapidly adapting type I and II units and slowly adapting type I and II units. Their receptive fields are distributed across the toes, metatarsal heads, arches and heel; their density, size and thresholds are not uniform [2,3].

Manual pressure in R.O.P. may therefore recruit different afferent populations according to the region, contact area, direction, speed, duration and depth of the technique. Neighbouring regions should not be assumed to be equivalent, but differences should be sought in their actual innervation, the tissues stimulated and receptor properties, rather than in a topographical resemblance to an organ.

1.2. Four gateways to compare

These territories overlap and vary between individuals; they should be understood as functional regions rather than precisely bounded areas [1].

All four gateways can generate ascending somatic input. The scientific question is therefore not whether only one region ‘communicates with the brain’, but whether the entry nerve, peripheral branch and recruited afferent population alter the nature, intensity or specificity of the central and visceral response.

1.3. From the foot to the central nervous system

In simplified form, the functional sequence is: local pressure, mechanoreceptors and somatic afferents, peripheral branch, main nerve, spinal roots, spinal cord, ascending pathways and supraspinal networks. This sequence is not a single tract: signals travel through several ascending systems and relays, depending on the sensory modality. The relevant point for R.O.P. is that local pressure can be converted into measurable neural activity that is not confined to the foot [1-3].

Human evidence also shows that cutaneous mechanoreceptor feedback from the hand or foot can modify muscle sympathetic nerve activity. This observation does not demonstrate an effect on a particular organ, but it establishes a possible coupling between somatosensory input and autonomic regulation [49].

2. The visceral side: how internal information reaches the same systems

A bridge cannot be described solely from the foot. Visceral receptors provide information about distension, tension, contractions, the chemical environment and, under certain conditions, inflammation or injury. The cell bodies of sensory neurons lie in the vagal ganglia or dorsal root ganglia [29].

Vagal afferents project to the nucleus of the solitary tract in the brainstem. Spinal visceral afferents enter the dorsal horn after travelling with thoracic or lumbar splanchnic, hypogastric or pelvic pathways. Some viscera, particularly the distal colon, rectum and several pelvic organs, therefore have afferent innervation distributed across several pathways [29,30]. In humans, the distal colon also receives ascending parasympathetic nerves originating in the pelvic territory and travelling over long distances [34], an arrangement that mouse models do not reproduce exactly [35].

The route taken must be distinguished from the nature of the signal. A sensory fibre accompanying a splanchnic or pelvic nerve remains a sensory afferent; it should not be confused with the autonomic efferent fibres that leave the central nervous system to control smooth muscles, glands or blood vessels. We therefore prefer the terms ‘visceral afferents travelling with thoracolumbar splanchnic pathways’ and ‘visceral afferents travelling with pelvic nerves’ [29].

⌕

3. The six neuroanatomical bridges

A bridge is not a direct cable between a point on the foot and an organ: it is an identifiable link in a functional network. The model distinguishes six bridges, from nerve-root overlap to cranial outputs from the brainstem.

These six bridges do not represent the same type of argument, and confusing them has long obscured the discussion. Bridges 1 to 4 build a proximity argument: nerve-root overlap, peripheral branch organisation and convergence in lumbosacral networks. They explain why the pelvis is the most accessible region from the foot. Bridges 5 and 6 build a network argument: supraspinal integration, descending commands and cranial outputs. They explain how a function sharing no segment with the foot may nevertheless be modulated.

These two families are not competing explanations: the tibial nerve combines both. However, they address different questions and require different tests. The proximity model is tested by comparing territories; the network model is tested by looking for an effect where proximity is absent.

Bridge 1 — Segmental overlap: S2–S3 through the tibial nerve

The tibial nerve is classically described as receiving fibres from L4 to S3. The sacral circuits involved in pelvic functions are mainly organised around S2–S4, and somatic control of the sphincters and pelvic floor involves, in particular, the pudendal nerve, also associated with S2–S4 [1,8-11]. In terms of nerve-root overlap, S2 and S3 are possible shared levels between the tibial nerve and part of the pelvic organisation.

This reasoning does not mean that a plantar skin region is an ‘S2’ or ‘S3 dermatome’. Human dermatome maps mainly place L4, L5 and S1 in the distal lower limb and foot [4]. A dermatome describes the dominant nerve-root origin of a skin territory; a mixed peripheral nerve, by contrast, contains axons from several roots supplying skin, muscles, joints and deep tissues.

Descriptions of the S3 contribution to the tibial nerve differ between anatomical sources. This difference should be acknowledged, but does not by itself demonstrate functional variability between individuals. The decisive question remains which fibres a particular stimulus actually recruits; the root range of the whole nerve cannot resolve it.

Bridge 1 therefore requires precise wording: stimulation of a tibial territory may introduce afferent activity into a lumbosacral system that includes S2–S3 and communicates with pelvic circuits. It does not automatically select S2 or S3, reach an organ directly or guarantee a visceral response. It does, however, provide a stronger segmental proximity argument than some gateways arising mainly from lumbar levels.

⌕

Bridge 2 — From plantar region to tibial nerve: which afferents are actually recruited?

The tibial territory is not uniform. The medial plantar nerve, lateral plantar nerve and calcaneal branches carry information from different regions and tissues. Their courses, fascicular separability and territories show anatomical variations [5,6]. Mechanoreceptor properties also vary between plantar regions [2,3].

Kovacevic and Yoo stimulated the medial and lateral plantar nerves separately in anaesthetised rats. Bladder inhibition during and after stimulation was obtained at 5 and 10 Hz for the medial plantar nerve and at 10 and 20 Hz for the lateral plantar nerve. At 50 Hz, stimulation of the lateral plantar nerve or tibial trunk could instead elicit an excitatory response [15]. Two inputs within the same tibial system can therefore produce different functional profiles depending on the branch and stimulation parameters.

Complementary evidence from mechanical stimulation exists in humans. Strzalkowski and colleagues showed that vibration of the sole at 150 Hz reduced the burst frequency of muscle sympathetic nerve activity in healthy adults [49]. This finding establishes coupling between mechanical stimulation of the skin and an autonomic response. It does not measure a bladder response or compare different plantar territories; vibration remains distinct from manual pressure.

These two studies give Bridge 2 a central place in the argument: the first supports functional differentiation between plantar branches; the second documents an autonomic response to mechanical plantar input in humans. Taken together, they support the hypothesis that the choice of region and the parameters of the manual technique may matter. Applying these findings to R.O.P. still requires identification of the afferents actually recruited and measurement of their effects under manual pressure [14,15,49].

The heel may allow preferential recruitment of calcaneal branches and afferents from the superficial and deep tissues of the fat pad. The arch engages the plantar branches more extensively, with possible differences between its medial and lateral aspects. The forefoot has a high sensory density and recruits digital branches of the plantar nerves.

The experimental question now concerns the manual technique: under standardised pressure, do the medial, lateral and calcaneal territories produce different, reproducible autonomic or visceral responses? Comparisons will need to account for the tissues stimulated, depth, mechanoreceptor profile and branch organisation. The available findings justify investigating differences between territories; they do not yet validate a precise map of points or their clinical efficacy under manual stimulation.

⌕

Bridge 3 — Other somatic gateways: sural, fibular, and saphenous

The tibial nerve is not the only possible gateway. The lateral border of the foot is largely supplied by the sural nerve; the dorsum is mainly supplied by the superficial and deep fibular nerves; the medial ankle and part of the proximal foot may receive saphenous innervation [1,7]. The L4–S3 tibial nerve offers the broadest sacral overlap; the S1–S2 sural nerve and L4–S2 fibular system may include an S2 component; the saphenous nerve, mainly L3–L4, has no obvious direct sacral overlap. These distributions are anatomical ranges rather than exclusive codes: local pressure does not select an isolated root.

A tibial > sural/fibular > saphenous gradient can be proposed as a hypothesis if segmental proximity is assumed to govern the response. However, findings on the saphenous nerve show that sacral overlap is not necessary for bladder modulation. By themselves, they do not provide a direct comparison of the effectiveness of all these gateways.

In anaesthetised rats, saphenous nerve stimulation inhibits the bladder reflex in a frequency-dependent manner [17]. This finding is also observed in cats, where saphenous stimulation modifies bladder capacity according to the parameters used [19]. A mechanistic study also showed that section of the hypogastric nerve does not abolish this inhibitory effect, indicating that this peripheral sympathetic pathway alone cannot explain it [18]. In humans, a feasibility pilot investigated percutaneous saphenous stimulation in women with overactive bladder: eighteen were enrolled, sixteen completed the study and fourteen met the authors’ composite criterion [24]. This was a small uncontrolled series without a sham group: a signal, rather than a demonstration.

A further study adds an interesting complication: some forms of percutaneous tibial neuromodulation may also recruit saphenous fibres [20]. The boundary between the ‘tibial gateway’ and the ‘saphenous gateway’ is therefore not necessarily as clear in clinical practice as in anatomical descriptions.

Bridge 3 consequently takes on a different role. The saphenous nerve is not the model’s negative control: conceptually, it is a natural experiment. It separates the properties that the tibial nerve combines—sacral segmental proximity and the capacity for modulation—and isolates the latter. The fact that the saphenous nerve modulates the bladder without sharing S2–S4 is currently the strongest argument for an approach based on two independent axes.

⌕

Bridge 4 — Lumbosacral networks: a spinal integration hub

Afferents entering the lumbosacral spinal cord do not remain confined to an isolated segment. They project to excitatory and inhibitory interneurons, projection neurons and circuits linking several levels. Within pelvic circuits, these networks receive visceral, perineal and somatic information while also being subject to descending commands from the brain [9-11].

In cats, tibial or pudendal stimulation can reduce the activity of sacral neurons responding to bladder distension [11]. This observation localises an effect to the spinal cord without demonstrating that the entire causal pathway is exclusively spinal: descending commands may contribute. Somatic input can therefore modify the processing of visceral information without a direct peripheral pathway to the organ [11,15].

The response depends on the state of the system. Stimulation frequency, intensity and duration, the fibres recruited and the functional state of the viscus can alter the direction of the response. In bladder models, the degree of filling is an important variable [8-11,14]. The same peripheral gateway may therefore be facilitatory, inhibitory or ineffective depending on the conditions.

Interactions between the bladder, distal colon and rectum are documented at several levels, from certain primary afferents to spinal networks and cross-sensitisation [31-33]. They support investigating pelvic functions in their regional context. However, a shared connection may contribute to inhibition, excitation or sensitisation: it does not automatically establish a therapeutic effect common to these organs.

⌕

⌕

Bridge 5 — Supraspinal pathways: descending commands and remote functions

Somatic input from the foot can ascend to higher centres, contribute to brain network activity and then modify descending commands to the spinal cord and autonomic circuits. Experimental work on somatoautonomic reflexes shows, in general, that somatosensory input can influence autonomic functions through spinal and supraspinal mechanisms [48].

This central control is particularly well described for the bladder. Visceral and somatic information is integrated at several levels involving the periaqueductal grey, pontine centres, insula, cingulate cortex and prefrontal regions. These networks contribute to perception of the visceral state, attention, context, decision-making and coordination of autonomic and sphincter responses [8,12,13].

The term ‘limbic brain’ should be used cautiously: naming the structures or networks involved is preferable to suggesting a single limbic centre for R.O.P. Moreover, the brain responses observed during transcutaneous tibial and fibular stimulation are not identical [16]: different peripheral gateways can reach partly shared central networks while retaining distinct response profiles.

The absence of direct nerve-root overlap does not rule out a visceral effect. Nor does a pathway through supraspinal relays necessarily imply a diffuse effect: central circuits can retain functional selectivity, and tibial and fibular inputs produce different brain responses [16]. The relays and effects must therefore be documented for each stimulus. Simply invoking the brain demonstrates neither an effect on a particular organ nor a point-to-organ correspondence.

⌕

⌕

Bridge 6 — Cranial outputs from the brainstem

Bridge 5 describes descending commands to spinal autonomic circuits. Bridge 6 describes a different effector pathway: parasympathetic output leaving the brainstem through cranial nerves without passing through the spinal cord. The two pathways share their ascending route and central integration, but diverge at the output stage.

This sixth bridge has a distinctive feature: segmental proximity cannot be invoked. A plantar afferent and a cranial parasympathetic ganglion share neither a root nor a peripheral pathway. Any observed modulation must therefore be central. Despite the limited data, this makes it the clearest test of the network model.

The oculomotor pathway provides the strongest demonstration. In anaesthetised rats, electrical stimulation of the hind paw produces intensity-dependent pupil dilation. Section of the somatic nerves supplying the stimulated limb abolishes the response, and the authors recorded action potentials in the tibial and saphenous nerves. Bilateral section of the cervical sympathetic trunks does not abolish the response, whereas section of the oculomotor nerves does [43]. Somatic information from the hind limb can therefore, after central integration, modify an ocular function dependent on the parasympathetic pathway of cranial nerve III. In this model, the observed dilation probably reflects modulation—in particular, inhibition of the parasympathetic activity that constricts the pupil—rather than simple parasympathetic activation. However, stimulation was not selectively restricted to the plantar nerves, and no topographical correspondence between a region of the foot and this pathway has been demonstrated.

For the pharynx and larynx, the cited studies describe two distinct observations. Somatic afferents from the lower limb produce responses in the nucleus of the solitary tract [44]. Pharyngeal and laryngeal afferents also converge on neurons in this nucleus [45]. Reaching the same region does not, however, demonstrate that lower-limb and pharyngolaryngeal inputs interact on the same neurons. These studies do not establish an effect of plantar stimulation on pharyngolaryngeal function. Sensory afferents, parasympathetic outputs and branchiomotor control of the striated muscles of the pharynx and larynx must also be distinguished.

For the other cranial ganglia, local physiology is established, but a bridge from the foot is not. Experimental findings establish the local autonomic roles of the pterygopalatine, otic and submandibular ganglia: direct stimulation of the spheno-/pterygopalatine ganglion modifies nasal mucosal blood flow [46], and stimulation of the salivatory nucleus elicits sustained ipsilateral parotid secretion [47]. No known study links selective plantar nerve stimulation to these outputs. A further distinction is needed: the ciliary, pterygopalatine, submandibular and otic ganglia are parasympathetic relays, whereas the trigeminal ganglion and inferior vagal ganglion are primarily sensory ganglia. Grouping them into a single category creates ambiguity.

⌕

4. A two-axis neuroanatomical targeting gradient

The first version of the gradient ranked regions by the density of anatomical arguments linking the foot to a visceral region. The findings presented above require us to separate two dimensions that previously appeared to vary together.

4.1. Horizontal axis: segmental proximity and density of regional bridges

How many anatomical arguments converge on the same region? This axis takes into account the peripheral nerve, probable roots, the presence or absence of overlap with visceral pathways, regional spinal convergence and local functional integration. Tibial territories associated with pelvic networks are the strongest candidates on this axis [8-11,14].

4.2. Vertical axis: evidence for somatovisceral modulation

Is there evidence that somatic stimulation of this pathway actually alters a visceral function? This axis must be qualified by three factors: the species studied, the stimulation site actually used and the study type—an animal mechanistic study, an uncontrolled human series or a randomised trial. A high position on this axis does not mean the same thing when the evidence comes from an animal preparation as when it comes from a controlled trial.

4.3. Four quadrants

Combining the two axes produces four situations, three of which are supported by actual data.

⌕

4.4. Degrees as summary labels

The four-level scale is retained as a summary, provided it is understood as the density of neuroanatomical arguments, never as a scale of clinical effectiveness.

The diaphragm remains a separate case: as a striated muscle under cervical phrenic control, it is not a viscus within this gradient. An influence from the foot would involve central, respiratory or somatic modulation, rather than a sacral segmental visceral bridge [1].

5. State of the clinical evidence

The gradient describes access pathways. It says nothing about therapeutic efficacy. Clinical studies of tibial neuromodulation, the closest to our hypotheses, show markedly different results across indications.

This table is informative in both directions. The same nerve, the same technique and a neighbouring anatomical region produce opposite results depending on the function targeted: a neuroanatomical bridge describes a possible access pathway, never an established therapeutic effect. The converse also holds: failure in one indication does not invalidate the pathway; it indicates that the function, parameters or population studied were not suited to it.

Finally, these trials involved electrical stimulation applied near a nerve trunk. None tested manual plantar pressure.

6. Turning the ROP map into an experimental hypothesis

The value of this framework is not to justify every mapped point retrospectively, but to turn the map into a series of testable hypotheses. Each region can be described through an explicit sequence: stimulated tissue, receptor population, nerve branch, main nerve, probable roots, spinal network, supraspinal relays and measured function.

6.1. Intratibial comparison: testing Bridge 2

Compare several regions within the tibial system—heel, medial arch, lateral arch and forefoot—using a technique that is as standardised as possible. If responses are identical, the peripheral branch would appear to play a limited role. If they differ, afferent distribution, depth, stimulated tissues and fascicular organisation become priority explanatory variables [2,3,5,6].

6.2. Between-nerve comparison: testing Bridge 3

This comparison has already been partly conducted under electrical stimulation in animals, with results pointing towards central mechanisms [17-19]. What remains to be established specifically for R.O.P. is different: does manual pressure on four distinct foot territories produce different response profiles? A tibial > sural/fibular > saphenous gradient would support a segmental contribution; comparable responses from all four gateways would confirm the predominance of central mechanisms [1,16].

6.3. Comparing visceral targets: testing the gradient

The pelvis is the first setting for validation because it combines the highest density of bridges with a functional model that has already been widely studied [8-16]. The distal colon and rectum are a logical second target [30-34]. The targets in quadrant 3—duodenum, jejunum and pupil response—then allow direct measurement of the contribution of intersegmental and supraspinal mechanisms [36-43].

6.4. An accessible first protocol: plantar pressure and pupillometry

Pelvic outcome measures require clinical recruitment, lengthy follow-up and often invasive instrumentation. Pupil diameter, by contrast, provides a non-invasive, continuous, quantifiable autonomic output that can be measured during a session using moderately priced equipment. This is probably the first experiment the R.O.P. programme could feasibly conduct.

The minimal protocol follows from the framework itself. Standardised manual pressure is applied to defined plantar territories—medial plantar, lateral plantar and calcaneal—and these are compared with each other to test Bridge 2. They are then compared with a region belonging to another somatic gateway, such as the medial saphenous territory or dorsum of the foot, to test Bridge 3. Pupillometry is recorded under controlled luminance, using within-subject comparisons, a sham condition and a no-stimulation condition.

What this protocol would establish is precisely the missing link in the framework: that manual plantar pressure, at level T4 of the scale below, produces a measurable autonomic response in humans. It would establish neither a point-to-organ correspondence nor a clinical effect. It would remain a demonstration of coupling, comparable to that already obtained for muscle sympathetic nerve activity [49].

The limitation must be stated at the outset: the available demonstration was obtained in animals using electrical stimulation of the hind paw [43]. There is no guarantee that manual pressure in humans recruits enough fibres to produce a detectable response. That is precisely why the experiment should be conducted rather than its outcome assumed.

6.5. A scale of applicability to manual practice

Any application of a finding to R.O.P. must specify its level. There are four levels, from the most distant to the closest to the manual technique:

The specific question for R.O.P. then becomes: does reproducible manual pressure on an anatomically defined territory generate sufficient afferent activity to modify a neurophysiological, autonomic or visceral marker? Only after this step does testing regional specificity become relevant.

6.6. Variables to document

• Exact anatomical location, side stimulated and probable nerve territory.
• Characteristics of the technique: contact area, direction, depth, intensity, duration and repetition.
• Functional state of the system at the time of testing, particularly important for visceral circuits that depend on the state of the network.
• Comparator: another territory of the same nerve, another nerve of the foot, sham stimulation or no stimulation, according to the protocol.
• Output measure: a sensory, autonomic, neurophysiological, imaging or functional marker appropriate to the research question.
• Explicit distinction between a general effect, a regional effect and an effect attributed to a particular organ.

7. Scope and limitations

This approach does not seek to demonstrate the efficacy of reflexotherapy. It proposes a working framework. An R.O.P. region has a stronger anatomical basis when several concordant elements are present: an identifiable peripheral gateway, a plausible afferent profile, compatible nerve roots, regional spinal convergence and, ideally, experimental evidence of modulation. Its basis is weaker when the link relies mainly on general central relays or a traditional mapping correspondence.

Four limitations must remain explicit. Findings from selective electrical stimulation of plantar branches [15] and human plantar vibration [49] support Bridge 2; their applicability to targeted, reproducible manual pressure remains to be established. Two controlled trials were negative for faecal incontinence [25,26], and this page gives them the same attention as favourable trials. The consistency of the S3 contribution to the tibial nerve is not uniformly established, further weakening predictions of reproducible segmental targeting. An autonomic response is not evidence of visceral targeting: the modulation of muscle sympathetic nerve activity by mechanoreceptor feedback from the foot [49] establishes coupling, rather than a point-to-organ correspondence.

A fifth limitation concerns Bridge 6. Modulation of cranial autonomic output from the lower limb rests on a single, older animal study using electrical stimulation of the hind paw and an outcome—pupil diameter—with no direct relationship to a clinical complaint. It is cited here for its value in demonstrating the mechanism, rather than as evidence of a cephalic effect of R.O.P. For most cranial outputs, this page’s conclusion is explicitly negative: local ganglion physiology is established, but a bridge from the foot is not.

Finally, a clarification of scope: this page describes input from the foot. The R.O.P. protocol also includes occipital input, whose afferents, cervical segments and sites of convergence require a separate analysis that is not undertaken here. Nothing above should be read as a description of this second gateway.

The selected bibliography particularly supports the sole of the foot, the anatomy of the tibial nerve and its branches, bladder neuromodulation, sacral networks, bladder–bowel interactions, digestive somatovisceral reflexes and general somatoautonomic mechanisms. It does not provide the same level of evidence for every viscus mentioned. Proposals concerning the stomach, kidney, spleen or adrenal glands must remain hypotheses of modulation at a distance that require investigation.

8. Synthesis: gateways, bridges, and a gradient

The foot does not contain organs; it contains somatic gateways. These gateways probably differ in sensory density, peripheral organisation and proximity to spinal networks. The plantar/tibial gateway has a particularly rich anatomical basis for the pelvis, but it is neither unique nor exclusive.

The model is organised around six bridges. Bridge 1 examines the tibial nerve’s segmental proximity to sacral networks. Bridge 2 uses functional differences between plantar branches under electrical stimulation and mechanical–autonomic coupling in humans to guide investigation of manual techniques [15,49]. Bridge 3 compares the other somatic gateways and finds, with the saphenous nerve, visceral modulation without sacral overlap. Bridge 4 locates their convergence in lumbosacral networks. Bridge 5 explains how input from the foot accesses supraspinal networks and modifies descending commands. Bridge 6 adds cranial parasympathetic outputs, the only pathway in which a cephalic function has actually been modified from the lower limb.

These six bridges form two families. The first four build a proximity model, explaining why the pelvis is the most accessible region from the foot and remaining the better-supported model clinically. The last two build a network model, explaining how a function sharing no segment with the foot can be modulated; its primary value is mechanistic demonstration.

A simple rule and a qualification follow from this framework. The rule: the more a hypothesis combines an identifiable peripheral branch, nerve-root proximity and regional spinal convergence, the more plausible anatomical targeting becomes. The qualification: this plausibility concerns targeting specificity, rather than the existence of an effect. A gateway without segmental proximity can modulate a viscus—the saphenous nerve demonstrates this for the bladder, the cranial nerve III pathway for the pupil and the vagal pathway for the duodenum. Experimental evidence therefore forms a separate axis that must be evaluated independently.

R.O.P. can thus be reframed as a neuroanatomical research programme: identify the gateways, characterise the recruited afferents, compare nerves, measure network responses and determine how far this organisation explains observed clinical effects. The map is no longer a fixed representation of the body on the foot; it becomes an anatomical and neurofunctional hypothesis to test.

Terminology adopted

Neuroanatomical bridge — A possible site of convergence or relay between somatic input and a network that may contribute to a visceral function.

Somatic gateway — A peripheral territory whose stimulation generates afferent activity through an identifiable nerve.

Visceral afferents travelling with thoracolumbar splanchnic pathways / with pelvic nerves — Preferred terms over ‘sympathetic afferents’ or ‘parasympathetic afferents’ when describing sensory information.

Regional targeting — The anatomical plausibility that a somatic gateway preferentially accesses a regional network; this implies neither clinical efficacy nor point-to-organ specificity.

Visceral modulation — A measurable change in a visceral function or circuit; preferred to ‘direct action on the organ’.

Supraspinal pathway — Involvement of centres above the spinal cord in the integration of somatic input and in descending commands.

Proximity model / network model — The model’s two families of arguments. The former relies on shared spinal segments, Bridges 1 to 4; the latter relies on central integration, Bridges 5 and 6. They require different tests.

Cranial output — A parasympathetic efferent pathway leaving the brainstem through a cranial nerve without a spinal relay, as distinct from the descending commands of Bridge 5.

Dissociation of the axes — The observation that a somatic gateway can modulate a visceral function without sharing a spinal segment with it; this requires segmental proximity and evidence of modulation to be treated as two independent dimensions.

Scale of applicability to manual practice — Four levels, T1 to T4, describing the distance between the protocol of a cited study and the manual R.O.P. technique.

Quadrant — The position of a visceral target at the intersection of the gradient’s two axes; complements, rather than replaces, the degree from 0 to 3.

Bibliography

This bibliography supports the neuroanatomical framework presented here. It is not intended as a systematic review of the clinical efficacy of reflexotherapy, nor does it document every viscus mentioned in the same detail.

A. Peripheral anatomy and afferents of the foot

1. Standring S, ed. Gray's Anatomy: The Anatomical Basis of Clinical Practice. 42nd ed. Elsevier; 2020.

2. Kennedy PM, Inglis JT. Distribution and behaviour of glabrous cutaneous receptors in the human foot sole. J Physiol. 2002;538(Pt 3):995–1002. doi:10.1113/jphysiol.2001.013087.

3. Strzalkowski NDJ, Peters RM, Inglis JT, Bent LR. Cutaneous afferent innervation of the human foot sole: what can we learn from single-unit recordings? J Neurophysiol. 2018;120(3):1233–1246. doi:10.1152/jn.00848.2017.

4. de Souza Faleiros AT, de Lima Resende LA, Zanini MA, de Lima Castro HA, Colichio Gabarra R. L4-L5-S1 human dermatomes. Arq Neuropsiquiatr. 2009;67(2A):265–267. doi:10.1590/S0004-282X2009000200017.

5. Deniel C, Guenoun D, Guillin R, Moraux A, Champsaur P, Le Corroller T. Anatomical study of the medial calcaneal nerve using high-resolution ultrasound. Eur Radiol. 2023;33(10):7330–7337. doi:10.1007/s00330-023-09699-6.

6. Andreasen Struijk LNS, Birn H, Teglbjærg PS, Haase J, Struijk JJ. Size and separability of the calcaneal and the medial and lateral plantar nerves in the distal tibial nerve. Anat Sci Int. 2010;85(1):13–22. doi:10.1007/s12565-009-0045-y.

7. Peng M, Yoo PB, Agur AMR. Distribution, course, and spatial relationships of the saphenous nerve: a 3D neuroanatomical map for nerve stimulation. PLoS ONE. 2024;19(2):e0297680. doi:10.1371/journal.pone.0297680.

B. Neural control of the lower urinary tract

8. Fowler CJ, Griffiths D, de Groat WC. The neural control of micturition. Nat Rev Neurosci. 2008;9(6):453–466. doi:10.1038/nrn2401.

9. Shefchyk SJ. Spinal cord neural organization controlling the urinary bladder and striated sphincter. Prog Brain Res. 2002;137:71–82. doi:10.1016/S0079-6123(02)37008-0.

10. de Groat WC, Vizzard MA, Araki I, Roppolo J. Spinal interneurons and preganglionic neurons in sacral autonomic reflex pathways. Prog Brain Res. 1996;107:97–111. doi:10.1016/S0079-6123(08)61860-9.

11. Yecies T, Li S, Zhang Y, Cai H, Shen B, Wang J, et al. Spinal interneuronal mechanisms underlying pudendal and tibial neuromodulation of bladder function in cats. Exp Neurol. 2018;308:100–110. doi:10.1016/j.expneurol.2018.06.015.

12. Griffiths D. Neural control of micturition in humans: a working model. Nat Rev Urol. 2015;12(12):695–705. doi:10.1038/nrurol.2015.266.

13. de Rijk MM, Fernández Chadily S, Knops A, Schoutens Y, Verstegen AMJ. The periaqueductal gray and its role in the neural control of lower urinary tract function. Auton Neurosci. 2026;265:103413. doi:10.1016/j.autneu.2026.103413.

C. Mechanisms of tibial and saphenous neuromodulation

14. Li X, Li X, Liao L. Mechanism of action of tibial nerve stimulation in the treatment of lower urinary tract dysfunction. Neuromodulation. 2024;27(2):256–266. doi:10.1016/j.neurom.2023.03.017.

15. Kovacevic M, Yoo PB. Reflex neuromodulation of bladder function elicited by posterior tibial nerve stimulation in anesthetized rats. Am J Physiol Renal Physiol. 2015;308(4):F320–F329. doi:10.1152/ajprenal.00212.2014.

16. Krhut J, Tintěra J, Rejchrt M, Skugarevská B, Zachoval R, Zvara P, et al. Differences between brain responses to peroneal electrical transcutaneous neuromodulation and transcutaneous tibial nerve stimulation. Neurourol Urodyn. 2023;42(6):1352–1361. doi:10.1002/nau.25197.

17. Moazzam Z, Yoo PB. Frequency-dependent inhibition of bladder function by saphenous nerve stimulation in anesthetized rats. Neurourol Urodyn. 2018;37(2):592–599. doi:10.1002/nau.23323.

18. Franz KS, Yoo PB. Transecting the hypogastric nerve to uncover the bladder-inhibitory pathways involved with saphenous nerve stimulation in anesthetized rats. Auton Neurosci. 2020;226:102672. doi:10.1016/j.autneu.2020.102672.

19. Li X, Wan X, Wang Z, et al. Frequency-dependent effects on bladder reflex by saphenous nerve stimulation and a possible action mechanism of tibial nerve stimulation in cats. Int Neurourol J. 2021;25(2):128–136. doi:10.5213/inj.2040304.152.

20. Elder CW, Yoo PB. Co-activation of saphenous nerve fibers: a potential therapeutic mechanism of percutaneous tibial nerve stimulation? Conf Proc IEEE Eng Med Biol Soc. 2016:3129–3132. doi:10.1109/EMBC.2016.7591392.

D. Clinical neuromodulation evidence

21. Peters KM, Carrico DJ, Perez-Marrero RA, et al. Randomized trial of percutaneous tibial nerve stimulation versus sham efficacy in the treatment of overactive bladder syndrome: results from the SUmiT trial. J Urol. 2010;183(4):1438–1443. doi:10.1016/j.juro.2009.12.036.

22. Peters KM, MacDiarmid SA, Wooldridge LS, et al. Randomized trial of percutaneous tibial nerve stimulation versus extended-release tolterodine: results from the Overactive Bladder Innovative Therapy trial. J Urol. 2009;182(3):1055–1061. doi:10.1016/j.juro.2009.05.045.

23. MacDiarmid SA, Peters KM, Shobeiri SA, et al. Long-term durability of percutaneous tibial nerve stimulation for the treatment of overactive bladder. J Urol. 2010;183(1):234–240. doi:10.1016/j.juro.2009.08.160.

24. MacDiarmid SA, John MS, Yoo PB. A pilot feasibility study of treating overactive bladder patients with percutaneous saphenous nerve stimulation. Neurourol Urodyn. 2018;37(5):1815–1820. doi:10.1002/nau.23531.

25. Knowles CH, Horrocks EJ, Bremner SA, et al.; CONFIDeNT Study Group. Percutaneous tibial nerve stimulation versus sham electrical stimulation for the treatment of faecal incontinence in adults (CONFIDeNT). Lancet. 2015;386(10004):1640–1648. doi:10.1016/S0140-6736(15)60314-2.

26. Zyczynski HM, Richter HE, Sung VW, et al.; NICHD Pelvic Floor Disorders Network. Percutaneous tibial nerve stimulation vs sham stimulation for fecal incontinence in women (NOTABLe). Am J Gastroenterol. 2022;117(4):654–667. doi:10.14309/ajg.0000000000001605.

27. Ashoush F, Abdelrahim A, Ali O, et al. A randomized controlled trial comparing bilateral percutaneous tibial nerve stimulation versus biofeedback pelvic floor muscle training in obstructed defecation syndrome. Cureus. 2025;17(3):e80885. doi:10.7759/cureus.80885.

28. Wu J, Yuan M, Zhao Y, Yang Y, Yan Z, Wu Y. The efficacy of transcutaneous tibial nerve stimulation as an adjunct to standard rehabilitation for postpartum pelvic floor dyssynergic defecation in primiparous women. Int Urogynecol J. 2026. doi:10.1007/s00192-026-06564-0.

E. Visceral afferents and pelvic networks

29. Jänig W. Neurobiology of visceral afferent neurons: neuroanatomy, functions, organ regulations and sensations. Biol Psychol. 1996;42(1–2):29–51. doi:10.1016/0301-0511(95)05145-7.

30. Brierley SM, Hibberd TJ, Spencer NJ. Spinal afferent innervation of the colon and rectum. Front Cell Neurosci. 2018;12:467. doi:10.3389/fncel.2018.00467.

31. Panicker JN, Marcelissen T, von Gontard A, Vrijens D, Abrams P, Wyndaele M. Bladder-bowel interactions: do we understand pelvic organ cross-sensitization? Neurourol Urodyn. 2019;38(Suppl 5):S25–S34. doi:10.1002/nau.24111.

32. Malykhina AP. Neural mechanisms of pelvic organ cross-sensitization. Neuroscience. 2007;149(3):660–672. doi:10.1016/j.neuroscience.2007.07.053.

33. Christianson JA, Liang R, Ustinova EE, Davis BM, Fraser MO, Pezzone MA. Convergence of bladder and colon sensory innervation occurs at the primary afferent level. Pain. 2007;128(3):235–243. doi:10.1016/j.pain.2006.09.023.

34. Johnson ME, Humenick A, Peterson RA, Costa M, Wattchow DA, Sia TC, Dinning PG, Brookes SJH. Characterisation of parasympathetic ascending nerves in human colon. Front Neurosci. 2022;16:1072002. doi:10.3389/fnins.2022.1072002.

35. Wang L, Taché Y. The parasympathetic and sensory innervation of the proximal and distal colon in male mice. Front Neuroanat. 2024;18:1422403. doi:10.3389/fnana.2024.1422403.

F. Digestive somatovisceral reflexes

36. Noguchi E, Ohsawa H, Tanaka H, Ikeda H, Aikawa Y. Electro-acupuncture stimulation effects on duodenal motility in anesthetized rats. Jpn J Physiol. 2003;53(1):1–7. doi:10.2170/jjphysiol.53.1.

37. Iwa M, Matsushima M, Nakade Y, Pappas TN, Fujimiya M, Takahashi T. Electroacupuncture at ST-36 accelerates colonic motility and transit in freely moving conscious rats. Am J Physiol Gastrointest Liver Physiol. 2006;290:G285–G292. doi:10.1152/ajpgi.00068.2005.

38. Qin QG, Gao XY, Liu K, Yu XC, Li L, Wang HP, Zhu B. Acupuncture at heterotopic acupoints enhances jejunal motility in constipated and diarrheic rats. World J Gastroenterol. 2014;20(48):18271–18283. doi:10.3748/wjg.v20.i48.18271.

39. Gao X, et al. Acupuncture at heterotopic acupoints facilitates distal colonic motility via activating M3 receptors and somatic afferent C-fibers. Neurogastroenterol Motil. 2015. doi:10.1111/nmo.12694.

40. Yuan M, Li Y, Wang Y, Zhang N, Hu X, Yin Y, Zhu B, Yu Z, Xu B. Electroacupuncture at ST37 enhances jejunal motility via excitation of the parasympathetic system in rats and mice. Evid Based Complement Alternat Med. 2016;2016:3840230. doi:10.1155/2016/3840230.

41. Du L, Qin Q, He X, Wang X, Sun G, Zhu B, Liu K, Gao X. Interstitial cells of Cajal are required for different intestinal motility responses induced by acupuncture. Neurogastroenterol Motil. 2025;37(6). doi:10.1111/nmo.14973.

42. Zhang N, Chen H, Gao X, Li S, Li X, Li W, Liu K, Wang S, Zhu B. Multimodal enhancement of colonic motility by acupuncture at ST36 is mediated by TRPV1+ cutaneous sensory fibers. Front Biosci (Landmark). 2026;31(1):46975. doi:10.31083/FBL46975.

G. Cephalic somatoautonomic reflexes

43. Ohsawa H, Yamaguchi S, Ishimaru H, Shimura M, Sato Y. Neural mechanism of pupillary dilation elicited by electro-acupuncture stimulation in anesthetized rats. J Auton Nerv Syst. 1997;64(2–3):101–106. doi:10.1016/S0165-1838(97)00023-4.

44. Toney GM, Mifflin SW. Time-dependent inhibition of hindlimb somatic afferent transmission within nucleus tractus solitarius. Neuroscience. 1995;68(2):445–453. doi:10.1016/0306-4522(95)00156-D.

45. Ootani S, Umezaki T, Shin T, Murata Y. Convergence of afferents from the SLN and GPN in cat medullary swallowing neurons. Brain Res Bull. 1995;37(4):397–404. doi:10.1016/0361-9230(95)00018-6.

46. Zheng C. Control of blood flow in nasal mucosa by electro-stimulation of sphenopalatine ganglion. Zhonghua Er Bi Yan Hou Ke Za Zhi. 1997;32(3):163–166.

47. Donaldson J, Mitchell J, Templeton D. Electrical stimulation of the salivatory nucleus in the rat. J Physiol. 1984;356:1–7. doi:10.1113/jphysiol.1984.sp015448.

H. General somatoautonomic mechanisms

48. Sato A, Sato Y, Schmidt RF. The impact of somatosensory input on autonomic functions. Rev Physiol Biochem Pharmacol. 1997;130:1–328. doi:10.1007/BFb0046598.

49. Strzalkowski NDJ, Incognito AV, Bent LR, Millar PJ. Cutaneous mechanoreceptor feedback from the hand and foot can modulate muscle sympathetic nerve activity. Front Neurosci. 2016;10:568. doi:10.3389/fnins.2016.00568.