Dorsal Foot and Ankle Bone Spurs as a Nerve-Compression Cause

What Works

Two of the three tiers below are about getting the diagnosis right, and that is not padding. On this condition the two tests clinicians reach for both mislead, in opposite directions, and everything downstream depends on not being fooled by either.

Dark cinematic anatomical study of the dorsal foot and its nerve pathway

Tier 1 — Strong evidence

Get the right X-ray, or you are treating blind. A weightbearing lateral view alone detects 40% of anterior tibial and 32% of talar spurs. Adding an oblique anteromedial view takes those to 85% and 73%. HIGH Evidence: STRONG for the diagnostic claim. Prospective accuracy study, N=60 consecutive patients, reference standard combining CT, MRI and arthroscopy (Tol 2004). Technique and its measured 7.3 mm rim-overlap rationale from van Dijk 2002. Corroborated by a 2026 systematic review of 40 studies and 5,217 patients at 93% for oblique views and 92% for MRI.
Do not treat a positive nerve conduction study as proof of cause. Use it as supporting information only. HIGH Evidence: STRONG. 81% of 21 asymptomatic occupationally exposed adults tested positive (Panwar 2019). The AANEM evidence-based practice parameter for the neighbouring tunnel screened 317 articles, found 4 that qualified, could not determine sensitivity or specificity at all, and reached Level C (Patel 2005).

Tier 2 — Moderate evidence

Conservative management of the bony-block presentation: activity modification, footwear modification, range-of-motion work and balance training. MODERATE Evidence: MODERATE. Pooled AOFAS improvement +12.5 with 25% recurrence, against arthroscopic 3-portal debridement at +30.5 with 10% recurrence (Yang 2026). Balance training was identified as reducing recurrence, but its dosing parameters are NOT SPECIFIED in the source and have not been invented here.
Diagnostic anaesthetic block of the deep peroneal nerve, where the genuine question is whether this nerve carries the pain. MODERATE Evidence: MODERATE. Mean 92.1% temporary relief in 37 patients with painful dorsal midfoot osteoarthritis, uncontrolled retrospective review alongside a cadaveric study in which this nerve supplied the second and third tarsometatarsal joints in 11 of 11 specimens (Shi 2019). A medical procedure, outside physical therapy scope in most settings.

Tier 3 — Emerging evidence and clinical reasoning

Remove external compression from the top of the foot. Re-lace to skip the eyelets directly over the bump, pad or replace the boot tongue, change the shoe if the shoe is the problem. EMERGING Honestly, this is mechanism rather than evidence. Tight shoe laces are a named cause in the reference series of 10 patients (Liu 1991), so the mechanism is established by observation, but no study has ever measured what removing them achieves. It goes first in practice because it is free, immediate, reversible and directly targets a documented cause.
Modify sustained loaded plantarflexion where the history contains it. EMERGING Duration of occupational exposure in years was inversely associated with nerve amplitude bilaterally (p<0.001) in 21 palm-tree climbers (Panwar 2019). That is an exposure-response association in a tiny uncontrolled sample. It is not a trial.
Surgical decompression, where a structural lesion is identified and conservative measures have genuinely been tried. MODERATE Moderate for direction, emerging for magnitude. Every series is uncontrolled and the largest has 48 patients.

What Doesn't Work

Strengthening exercise as a treatment for the bony lesion itself. No exercise resorbs a bone spur. Loading and footwear change address the consequences; prescribing strengthening as the treatment implies a mechanism that does not exist.
Aggressive end-range dorsiflexion mobilisation where the block is bony. You are pushing bone into bone. Forcing range is not a stretch here, it is impingement performed deliberately.
Neural mobilisation on its own while the compressing item is still in place. Remove the compression first. No study has tested this in this condition, so this entry is reasoning and is flagged as such.
Reassurance based on a normal lateral X-ray. It misses 60% of anterior tibial and 68% of talar spurs.
Blaming a spur found incidentally on a scan. One in five young adult ankles has one, and midfoot structural findings lost their association with pain once age, sex and BMI were adjusted for.

Exercise Prescription

Read this before the table. There is no research testing exercises for the nerve version of this problem. Not weak research. None. The exercises below keep the ankle and foot working while the pressure comes off, and they are not a treatment for the bump. The sets and reps are clinical reasoning, not evidence-derived, because no study of this condition has ever published an exercise dose.

Calf stretch, back knee straight 3 × 30 seconds · daily Hands on a wall, one foot back, back knee locked, heel down. A stretch in the calf. Stop if it pinches at the front of the ankle.

Calf stretch, back knee bent 3 × 30 seconds · daily Same position, back knee softly bent, which moves the stretch lower. Same front-of-ankle rule.

Single-leg balance 3 × 30 seconds · daily Progress by closing the eyes, then standing on a cushion. Effort, not pain. Balance training was named as recurrence-reducing in the 2026 systematic review, though that review specified no dose.

Toe raises 3 × 15 · daily Heels down, lift the toes and front of the foot, lower slowly. Effort in the front of the shin, no sharp pain.

Heel raises 3 × 12 · every other day Rise onto the balls of both feet, lower over 3 seconds. Progress to one leg when two-leg is easy.

Load management for training: stop deep loaded dorsiflexion where the end-feel is hard and bony, which means full-depth squatting, deep lunges and inclined sled pushes. Raise squat depth above the point where the block appears and switch to a heeled lifting shoe, which is a legitimate tool here rather than a concession. Keep everything else. Acceptable during training is up to 3/10 discomfort that settles within 24 hours; any increase in numbness or burning is a stop, not a threshold, because it is a different signal from joint or muscle pain.

Red Flags

Get these checked rather than working through them. The first one is the reason this section is not optional: an artery runs beside the nerve inside the same tight tunnel, and its blockage produces the same symptoms.

Dark cinematic anatomical study of the anterior ankle and dorsal foot

Refer, do not treat

Absent or diminished pulse on the top of the foot. The dorsalis pedis artery shares this tunnel with the nerve and its thrombosis is a documented cause of the same syndrome. Same-day vascular assessment if the foot is also cold, pale or painful at rest.
Motor weakness that is getting worse, as distinct from weakness that is present and unchanging. Urgent orthopaedic or neurological referral.
Bilateral or non-territorial sensory change, or numbness that does not respect the space between the first two toes. This points to a nerve problem elsewhere rather than a local bump.
An enlarging dorsal lump. A ganglion caused 2 of 13 cases in the only outcome series. Anything growing is imaged, not observed.
A bad ankle sprain that never settled, now with numbness or tingling. A missed talus fracture can produce a delayed neuropathy years later.
Night pain unrelated to position or pressure, or systemic symptoms.

Return to Training

Sensory testing of the first web space is symmetrical with the unaffected side, or the numbness has resolved.
No burning or tingling in the first web space during or within 24 hours of training.
Full previous training depth achieved with no anterior ankle block and no sensory symptoms.
3 sets of 12 single-leg heel raises without anterior ankle pain.
Single-leg balance held 30 seconds, eyes closed, on the affected leg.
Footwear modification in place and confirmed working, meaning you can name what you changed.
Symptom-free for 2 consecutive weeks at full training load.

Touch the skin between your big toe and your second toe, then the same spot on the other foot. Does one side feel duller?

That answer decides everything else. Duller on one side means the nerve is involved. Both the same, with the ankle blocking when you squat, is a different problem with the same bump.

Takes 10 seconds. No equipment.

Do not self-manage if: the pulse on the top of your foot is absent or weak, the foot is cold or pale, weakness is getting worse, or the lump is growing. An artery shares the same tight tunnel as the nerve. Cold or pale means same-day assessment.

The Verdict

One bump on your foot, two problems. Check the feeling between your first two toes, and see a doctor the same day if the foot goes cold or pale.

Think of a garden hose running over a low garden wall, with a tight strap holding it down against the bricks. Left alone the hose is fine. Now imagine a brick slowly pushing up from underneath. The strap will not stretch, so the hose gets pinched between the rising brick and the strap, and the water slows. That is a bone spur growing up into a nerve that is held down by a band of tissue across the front of your ankle. Nothing dissolves the brick. What you can change is the strap, which is your laces and your boots, and that is why the first treatment is a shoelace and not an exercise.

What this actually is: a small bony bump on the front of the ankle or the top of the foot, which about one in five people has and almost nobody notices, sitting in exactly the spot where a nerve crosses over.
What most people get wrong: a normal X-ray does not rule it out, because the standard side-on view finds only about 40% of these spurs, and a positive nerve test does not rule it in, because four out of five people with no symptoms at all test positive.
The one change that matters: take the pressure off the top of your foot, which usually means re-lacing your shoes to skip the eyelets right over the sore spot.

Best for

Adults with pain on the top of the foot or the front of the ankle, especially where footwear, laces or a stiff boot tongue make it worse, and anyone told their foot X-ray was normal but whose symptoms have not settled.

Skip if

The pulse on top of your foot is absent or weak, the foot is cold or pale, muscle weakness is getting worse, or the lump is growing. Those need assessment, not self-management.

Want the full evidence? Keep scrolling

Conviction

MODERATE, endpoint-stratified. The claims on this card are not equally supported, and three of them are not "weak evidence" but no evidence, which is a different statement and is graded as one.

A dorsal bony lesion can compress this nerve	HIGH
A standard lateral X-ray is inadequate to exclude it	HIGH
The lesion is present in about a fifth of young adult ankles	HIGH
A positive nerve conduction study does not establish causation	HIGH
First-web-space sensory change is the discriminator	HIGH
Ultrasound detects dorsal spurs that X-ray misses	MODERATE
Surgical decompression relieves symptoms when a lesion is found	MODERATE
Joint space narrowing, not spur size, predicts surgical outcome	MODERATE
Any specific conservative protocol for the nerve version	NO EVIDENCE
The proportion of dorsal spurs that ever become symptomatic	NO EVIDENCE
Exercise altering the bony lesion itself	NOT APPLICABLE

What would change my mind on the conservative question

A prospective cohort of at least 150 adults presenting to primary-contact physical therapy with dorsal foot or ankle pain, imaged at entry with weightbearing lateral plus oblique views so the lesion is actually detected, stratified at baseline by the presence or absence of first-web-space sensory change, and followed for 12 months on a documented conservative protocol, with the primary endpoint being the proportion whose sensory symptoms resolve. A null result would be as useful as a positive one, and either would replace the largest guess on this card.

What would change my mind on the causation question

A case-control study of at least 200 people, half with dorsal foot pain and half without, all imaged and all electrodiagnosed, reporting how many in each group have a spur and how many have nerve conduction abnormality. The 21 asymptomatic workers suggest the answer will embarrass the diagnosis. If a well-powered version found abnormality in under 10% of pain-free controls, a positive study would carry far more weight and this card's central caution would need softening.

The Full Picture — Anatomy, Diagnosis & Evidence

What's Actually Going On

Dark cinematic anatomical study of the anterior ankle tunnel and its contents

The deep peroneal nerve runs down the front of the ankle and passes beneath a band of tissue called the inferior extensor retinaculum, travelling alongside the dorsalis pedis artery. Cadaveric measurement found that band lying directly over the nerve and vessel in 61.1% of specimens, and found the nerve splitting inside the tunnel in 86.1% (Aktan Ikiz 2007). The roof of this tunnel is already snug before any pathology arrives.

The floor of that tunnel is bone. A dorsal spur at the talonavicular joint grows upward into the tunnel from below, and the band above will not give way. That is the mechanism described in the reference series of 10 patients, where the recorded causes were dorsal contusion, tight shoe laces, talonavicular osteophytosis, ganglion and pes cavus (Liu 1991), and it is what was seen directly at operation in a case where the spurs were found irritating the nerve and the patient was asymptomatic one year after decompression (Huang 1999).

A second mechanism competes with compression and it is not a footnote. One report presented evidence that the syndrome can result from abnormal stretch of the nerve rather than pressure on it (Borges 1981). The two point at different things to change: compression is relieved by taking pressure off the dorsum, stretch by not holding the ankle at end range.

Where the spur itself comes from is contested, and the answer is "both, at different places on the same bone." Medial spurs sit significantly further up the talar neck than the capsule attaches (p<0.01), making them true intra-articular osteophytes, while laterally the difference is not significant (p=0.26), making those traction spurs (Hayeri 2009). The measurement that undercut the traction theory as a general account came first: the capsule attaches proximal to where the spurs originate (Tol 2004). The impact side has direct support, since across 150 kicks by 15 elite soccer players ball contact was made with the anteromedial foot and ankle in the large majority at an average force of 1,025 N, while extreme plantarflexion occurred in only 39% of kicks (Tol 2002).

How to Identify It

Dark cinematic anatomical study of the dorsal foot surface anatomy

Two-point discrimination, first web space, both feet. Sn/Sp DATA UNAVAILABLE Validated against a different cause: in 12 subtle Lisfranc injuries the nerve response was absent with reduced two-point discrimination in all patients (Pourcho 2013).
Standard lateral radiograph. Tibial Sn 40% / Sp 70% Talar Sn 32% / Sp 82% A negative lateral view does not exclude the lesion (Tol 2004).
Lateral plus oblique anteromedial impingement view. Tibial Sn 85% / Sp 45% Talar Sn 73% / Sp 68% Beam tilted 45 degrees craniocaudal, leg in 30 degrees external rotation. Oblique alone for anteromedial osteophytes: tibial 93%, talar 67% (van Dijk 2002; Tol 2004).
MRI. Sn 92% (95% CI 89-95) Pooled across 40 studies and 5,217 patients (Yang 2026).
Brief clinical examination for midfoot osteoarthritis. Sn 29.9% / Sp 87.5% AUC 0.64 in 274 community adults. A positive picture means something; a negative one means very little (Thomas 2015).
Tinel's sign over the prominence. Sn/Sp DATA UNAVAILABLE 12 of 13 in the only surgical series had a bulge with a positive Tinel, but that series has verification bias by design and only about half had any sensory loss (Yassin 2015).

The pattern is worth naming: every imaging test here has published accuracy and every physical test has none. The bedside test this card recommends most strongly is the one with no published accuracy at all, and saying so is more useful than letting a reader infer it from an empty column.

Ultrasound is the outlier worth knowing about. It found midfoot spurs in 344 of 2,445 joints (14.1%) against 13 (0.5%) by radiography, a 26-fold difference, with agreement between the two so weak it was effectively absent (κ 0.029 to 0.035) (Camerer 2017).

The Debate

Does the lateral X-ray settle it?

Standard practice

A weightbearing lateral radiograph rules an anterior spur in or out.

Tol 2004, N=60 prospective; van Dijk 2002

Lateral alone: tibial 40%, talar 32% sensitivity. Adding an oblique view: 85% and 73%. The anterolateral tibial rim projects over the anteromedial rim by an average of 7.3 mm, which is enough to hide a 7 mm spur completely.

Follow the newer evidence. Requesting the second view is the single highest-yield action on this card, and "a normal ankle X-ray" in the notes is not an exclusion.

Does spur size matter?

Intuition, and a lot of clinic conversations

A bigger spur means a worse problem and a worse outcome.

van Dijk 1997, N=62 prospective; Yabiku 2023, N=34 athletes

The degree of joint change predicted outcome far better than spur size or location: 90% good or excellent without joint space narrowing against 50% with it. In athletes, spur protrusion into the joint predicted cartilage damage (p=0.008) while spur size did not (p=0.341).

Measure the joint, not the spur. This is also the number to give a patient deciding about surgery, because it is the one that applies to them.

Does removing the spur fix it for good?

The implicit promise of an operation

Take the spur out and the problem is solved permanently.

Coull 2003, mean 73 months; Tol 2001, mean 6.5 years

Spurs usually recurred and most patients did not feel dorsiflexion returned to normal, yet 23 of 25 without joint space narrowing still had a good or excellent result, and 77% of grade-I patients did well despite recurrence in two-thirds.

Symptom relief and radiographic recurrence are separate outcomes here. The operation buys the first without buying the second, and patients should be told that before they consent.

No clinical practice guideline exists for this condition as of August 2026. The nearest authority is an evidence-based practice parameter on the posterior tarsal tunnel, which screened 317 articles, found four that qualified, could not determine sensitivity or specificity, and reached Level C (Patel 2005). That is a gap statement, not a hedge.

Honest Limitations

1. The conservative evidence is absent, not weak, and the absence hides in the exclusion criteria

Every treatment study in this literature is a surgical study, and every patient in them had already failed conservative care that none of the papers describes. There is no cohort anywhere following a person with a dorsal spur and first-web-space symptoms managed with footwear change and activity modification alone. So treat conservatively first, because the mechanism is external and modifiable, but be explicit with the patient that you are reasoning from mechanism rather than from trial data, and set a review date instead of an open-ended trial.

2. The prevalence numbers come from cadavers and the symptom numbers come from clinics

The two populations never meet, so the number a patient actually wants, the chance that a spur on their scan will ever cause them trouble, has never been measured and does not exist in this literature. Any figure quoted for it has been constructed. Meanwhile midfoot bone marrow lesions, joint space narrowing, cysts and spurs were associated with pain in simple models and stopped being associated once age, sex and BMI were adjusted for (Arnold 2023; Halstead 2025). When an incidental spur is reported, say plainly that about one in five ankles has one and almost none of them hurt.

3. The literature is two different populations wearing one label

The anterior impingement evidence base has a mean age of 32.8 and is predominantly male and athletic. The dorsal midfoot degenerative evidence base has a mean age of 67.8 and is 38 women to 10 men. Evidence generated in a 33-year-old male footballer's ankle does not transfer to a 68-year-old woman's midfoot, and the reverse is equally untrue. Decide which literature your patient belongs to before quoting a number at them, and say which one you are drawing on.

The Nuance

Dark cinematic anatomical study contrasting the bony and neural structures of the dorsal foot

Conservative success, bony-block presentation: pooled AOFAS improvement +12.5 with 25% recurrence across 40 studies and 5,217 patients (Yang 2026). Conservative success, nerve presentation: DATA UNAVAILABLE, because no conservative cohort has ever been published.

Surgical numbers. Anterior tarsal tunnel decompression moved AOFAS from 55 ± 8 to 83 ± 11 at 12 months and 88 ± 10 at 24 months, in 13 patients (Yassin 2015). Dorsal nerve release was excellent in 60%, good in 20% and unimproved in 20% across 20 nerves (Dellon 1990). Nerve division for dorsal midfoot arthritis satisfied 80.8% of 48 patients with 91.7% relieved within six months, but 55.6% still had activity limitations and 10.4% wished they had had a fusion instead (Iturregui 2023). Arthroscopic debridement was good to excellent in 81.04% across 1,506 patients, with a 4.01% complication rate whose commonest entry was mild nerve symptoms (Gianakos 2021).

The honest truth is that those surgical numbers look better than they are, and the reason is design rather than dishonesty. Every series is uncontrolled, the largest has 48 patients, and all of them enrolled people who had already failed something. Nobody knows what happens to the patients who got better with a different boot, because those patients never entered a study.

One bone, several addresses. The same problem occurs further down the foot, where a "dorsal boss" or "tarsal boss" sits at the naviculocuneiform or tarsometatarsal joints, and it occurs with an os intermetatarseum, a congenital accessory bone rather than a degenerative spur, which produces the identical picture of dorsal pain with first-web-space tingling and resolves completely on excision. A plain radiograph shows it if you look in the right place, which is a good reason to say where the symptoms are rather than just "the foot".

Next step

Most foot pain gets treated before it gets identified. The Verdict sends one evidence-graded protocol a week, with the numbers attached and the gaps named out loud.

Join The Verdict — free

Sources

Talbot CE et al., 2018 — 670 ankle specimens from 344 individuals aged 20 to 40. Bony impingement in 21%, bilateral in 8%. PMID 29873114. Large cadaveric series
Tol JL et al., 2004 — N=60 prospective diagnostic accuracy, reference standard CT plus MRI plus arthroscopy. Lateral 40%/32% sensitivity; plus oblique 85%/73%. PMID 14992704.
van Dijk CN et al., 2002 — the oblique technique and the measured 7.3 mm rim overlap that explains why the lateral view fails. PMID 11904689.
Panwar A et al., 2019 — 81% (17/21) of asymptomatic exposed workers had electrodiagnostic anterior tarsal tunnel syndrome. PMID 31602156.
Yassin M et al., 2015 — the only treatment outcome series for the nerve version. N=13, minimum 24 months. Spur was the cause in 4 of 13. PMID 26209470.
Liu Z et al., 1991 — the reference clinical series, N=10 plus 25 cadaveric feet, naming talonavicular osteophytosis and tight shoe laces among the causes. PMID 1670452.
Huang KC et al., 1999 — dorsal talonavicular spurs seen at operation irritating the nerve; asymptomatic at one year after decompression. PMID 10584426.
Yang M et al., 2026 — systematic review and meta-analysis, 40 studies, N=5,217, GRADE and RoB 2 assessed. PMID 42173479. Highest-tier synthesis
van Dijk CN et al., 1997 — N=62 prospective, 2-year follow-up. Joint change predicted outcome better than spur size or location. PMID 9397259.
Hayeri MR et al., 2009 — medial spurs are intra-articular (p<0.01), lateral ones are traction spurs (p=0.26). PMID 19770304.
Patel AT et al., 2005 — AANEM evidence-based practice parameter. 317 articles screened, 4 qualified, sensitivity and specificity could not be determined, Level C. PMID 16003732.
Iturregui JM et al., 2023 — N=48 patients, 55 feet. 80.8% satisfied, 91.7% relieved within 6 months, 10.4% wished for a fusion instead. PMID 36744743.
Shi GG et al., 2019 — nerve supplied the 2nd and 3rd tarsometatarsal joints in 11/11 specimens; mean 92.1% relief after diagnostic block. PMID 31331190.
Coull R et al., 2003 — N=27, mean 73 months. Spurs usually recurred; 23 of 25 without joint space narrowing still did well. PMID 12793562.
Dellon AL, 1990 — 20 nerves in 18 patients. Excellent 60%, good 20%, not improved 20%. PMID 2265812.
Yabiku H et al., 2023 — N=34 athletes. Spur protrusion predicted cartilage damage (p=0.008); spur size did not (p=0.341). PMID 37259105.
Thomas MJ et al., 2015 — N=274 community adults. Clinical examination sensitivity 29.9%, specificity 87.5%, AUC 0.64. PMID 26093213.
Gianakos AL et al., 2021 — 28 articles, 1,506 patients. Good to excellent 81.04%, complications 4.01%. PMID 33400547.
Pourcho AM et al., 2013 — N=12, contralateral control. Nerve response absent with reduced two-point discrimination in all patients. PMID 23539169.
Camerer M et al., 2017 — 2,445 joints. Ultrasound found spurs in 14.1% against 0.5% by radiography. PMID 28478580.
Tol JL et al., 2002 — 150 kicks by 15 elite players, average contact force 1,025 N at the anteromedial ankle. PMID 11798995.
Arnold JB et al., 2023 and Halstead J et al., 2025 — midfoot MRI findings associated with pain in simple models, not after adjustment for age, sex and BMI. PMID 35593411, PMID 39797599.

Full protocol card with 52 sources, evidence grading and clinician-facing assessment detail lives in the Physio Engine library. One retrieved paper carried no abstract and no quantitative claim here rests on it. One retracted publication was found in the corpus and excluded before citation.