Painless Foot Drop from Lumbar Degenerative Disease

What Works

Read the whole section before using any of it. No physical therapy intervention has ever been trialled in this condition in the reviewed evidence. The graded interventions below are diagnostic and surgical. That is the state of the literature, not a gap in this page.

Dark cinematic rendering of lumbar decompression anatomy

Urgent recognition and referral, with the date of onset recorded HIGH

Tier 1 · Evidence: STRONG · Direction unanimous across 6 cohorts and 2 meta-analyses

Refer inside 6 weeks of the weakness starting wherever possible. Surgery within 6 weeks was up to 6 times more likely to produce greater recovery (Baig Mirza 2025, duration p=0.01). Early decompression within 1 month beat late decompression on recovery rate (95% CI 1.59 to 2.57) and neurological improvement (95% CI 0.21 to 1.66) across 312 participants (Song 2022). Duration was significant at OR 3.59 (95% CI 1.09 to 11.8) with a 2 month cutoff (Masuda 2020) and at HR 0.67, p=0.004 (Macki 2016).

The three thresholds disagree at 6 weeks, 1 month and 2 months because each was chosen inside its own cohort. The relationship is continuous and unanimous in direction. Nothing in this literature supports waiting.

Lumbar decompression for a confirmed compressive lesion MODERATE

Tier 1 · Evidence: STRONG for the association, ABSENT for superiority

Across 918 patients, 60% (95% CI 0.44 to 0.75) reached MMT 4 to 5 and 82% (95% CI 0.76 to 0.88) gained at least one grade, with complications at 1% (Than 2025). Every constituent study is observational, heterogeneity ran I² 89 to 97%, and no study compares this with conservative care. Median time to improvement was within 6 weeks of surgery (Macki 2016).

The 3 month re-grade as a prognostic checkpoint MODERATE

Tier 2 · Evidence: MODERATE · One individual patient data meta-analysis, 66 patients

Gaining at least 1 MRC grade within 3 months of surgery carried roughly a 30-fold probability of continuing to improve (Baig Mirza 2025, p=0.006). This is the single most useful post-operative number in the literature. The 3 month mark is not an arbitrary review point, it is where the trajectory becomes informative about the endpoint.

Prognostic stratification at first contact HIGH

Tier 2 · Evidence: MODERATE to STRONG per factor

Preoperative MRC 2 to 3 of 5 versus severe weakness, OR 5.882 (95% CI 4.449 to 7.776). Diabetes, OR 5.657 (95% CI 2.094 to 15.280) for a poorer outcome (Hou 2023). Preoperative tibialis anterior strength OR 12.0 (95% CI 2.41 to 59.9) and age OR 0.93 per year (95% CI 0.87 to 0.98) (Masuda 2020). Age 47 or under was a positive factor (Baig Mirza 2025, p=0.045).

Tier 3 — emerging and untested here

Ankle-foot orthosis as a trip guard EMERGING No reviewed study of this condition evaluated an orthosis. The device evidence in this library comes from foot drop of central neurological origin, which is a different lesion type. The reason to offer one is falls prevention, which is a safety judgement rather than an evidence-based treatment claim, and it should be presented to the patient that way.

Full-endoscopic decompression with minimal root retraction EMERGING One uncontrolled case series, 37 patients across 40 levels, in a mixed "impending deficit" population rather than established foot drop, with 3 revisions (Kim 2022). Mean MRC moved 1.97 to 4.76. No head-to-head against open decompression exists.

Exercise Prescription

There is no exercise programme for this condition NO DATA

That is a finding rather than an omission. No study in the reviewed research tested any exercise, stretch, brace protocol or nerve glide for a foot drop caused by a compressed lumbar nerve root. The muscle is not weak because it is out of condition. It is weak because the nerve supplying it is being squeezed, and training a muscle whose nerve supply is interrupted does not restore the nerve.

What is genuinely useful while the referral goes through

Keep the ankle moving through full range, passively with your hands if you cannot lift it yourself. Several times a day, gently, to a stretch and no further. If the foot sits dropped for weeks the calf shortens and you end up with a stiff ankle on top of a weak one, which is a second and avoidable problem.

Wear a brace whenever walking if you are catching your toe, especially on stairs and uneven ground. This is a falls device, not a treatment.

Keep walking and stay active. Nothing here suggests rest helps.

Check the skin on the foot daily if it is numb or braced. A numb foot can blister without you feeling it.

What Doesn't Work

Treating painlessness as reassurance. No pain suggests no compression, so the spine gets excluded by default and the workup goes below the knee. This is the failure this page exists to prevent.
A watchful waiting period. Six weeks of conservative management is ordinary and defensible for most musculoskeletal presentations. Here it consumes the entire best-supported surgical window.
Functional electrical stimulation offered on the stroke and multiple sclerosis evidence. A compressed nerve root is a lower motor neuron lesion, and stimulation is approved for foot drop of upper motor neuron origin. Stimulating a nerve whose fibres are degenerating does not produce a contraction.
Strengthening the tibialis anterior against a compressed root. The muscle is not weak because it is untrained.
Quoting a recovery percentage without its definition. "60 to 85% recover" merges four different endpoints and two different populations, one of which is painful cases only.

Safety First

Red Flags

Numbness around the groin or between the legs, difficulty passing or controlling urine, loss of bowel control, or weakness in both legs. That is cauda equina syndrome. A and E the same day, not a physical therapy appointment.
A new foot drop, painless or not. This is a motor deficit and the outcome is time dependent. The best supported window is 6 weeks from the day the weakness started (Baig Mirza 2025).
Weakness getting worse while you wait. Progression removes the argument for continuing to wait, because every additional week costs recovery odds.
Weakness spreading beyond the foot, into the calf, the hamstring, or the other leg. That is not this condition.
Fasciculations or wasting that crosses more than one nerve territory, or steady progression with nothing compressive on the scan.

Refer to: spinal surgery or neurosurgery urgently for a new foot drop with a suspected lumbar cause, with the date of symptom onset written on the referral. A and E the same day for suspected cauda equina syndrome. Neurology where the weakness does not fit one root or one nerve.

Dark cinematic rendering of the lumbar spine and descending nerve roots

Return to Training

Concrete and binary. Not "when it feels ready".

Cause identified, and the lesion localised to the L5 root rather than the peroneal nerve or something else
Cauda equina syndrome excluded
Dorsiflexion strength graded and documented at MRC 4 or above against resistance
Full passive ankle dorsiflexion range, with no fixed contracture
Heel walking possible for 10 metres without the foot dropping
Single-leg stance held for 30 seconds without a hip drop, where hip abduction was involved
No toe catch over stairs or uneven ground, without a brace, across a full day
No progression of weakness or sensory change over 2 consecutive weeks

Load management while you wait. Stop anything where a foot that will not lift is a falling hazard: loaded carries, step-ups, box jumps, running, lunges, and standing barbell work. Continue seated and supported training. This is a falls restriction, not a tissue-healing restriction, and saying so out loud is usually what makes it tolerable.

Lie on your side and lift your top leg against someone's hand. If that side is weaker too, the problem is higher up than your ankle.

Hip strength is the one thing that separates a pinched nerve in the back from a pinched nerve at the knee, because the nerve beside the knee cannot reach the hip muscles. If that side is weaker, book an appointment this week.

Takes less than 2 minutes. No equipment needed.

The Verdict

A foot drop that doesn't hurt isn't safer. Get it checked this week, not next month.

Think of the nerve as the cable running from a light switch in your lower back to a bulb in your shin. When the cable is pinched, the bulb goes dim. Changing the bulb does nothing, and exercising the muscle is changing the bulb. The only thing that helps is releasing the pinch, and the longer the cable stays crushed the less of it comes back when you do.

The muscle that lifts your foot is run by a nerve in your lower back, so a disc pressing on that nerve can stop the foot lifting without ever causing back pain.
No pain gets read as no problem, so the search goes to the knee and the ankle while the clock that actually matters keeps running.
Test hip strength on the weak side, write down the exact date the weakness started, and get assessed this week.

Best for

Anyone with a recently weak or dragging foot, with or without back pain, who has not been assessed yet.

Skip if

You have numbness around the groin, bladder or bowel changes, or weakness in both legs. That is an emergency and you should be in A and E today.

Want the full evidence? Keep scrolling

Conviction

MODERATE overall, with the confidence split by claim rather than averaged.

Time-dependence of recovery HIGH
Preoperative motor grade as a prognostic factor HIGH
The 3 month post-operative checkpoint MODERATE
Painlessness as a negative prognostic marker LOW to MODERATE
Surgery superior to conservative care LOW and effectively untested

That the lumbar spine causes painless foot drop, and that this presentation gets attributed elsewhere, is secure and rests on three independent painless-specific sources. That painless cases recover worse is directionally supported by two analyses and statistically significant in only one of them, on 55 patients.

What would change my mind on "painless cases recover worse"

The headline comparison is OR 0.31 with a 95% CI of 0.04 to 2.65, from only two comparing studies (Waseem 2023). That interval spans strong harm to modest benefit, so it is not statistically significant. The direction is carried independently by Tanaka 2021, where absence of radicular leg pain survived multivariate regression in 55 patients. A properly powered comparison of painful against painless cases, with duration and preoperative grade adjusted for, would settle it in either direction. Aono 2007 reported the opposite and did not adjust.

What would change my mind on "surgery works"

A randomised trial of early decompression within 6 weeks against structured conservative management with protocolised escalation for progressive weakness, N of 200 or more, primary endpoint MRC dorsiflexion grade at 12 months, with the painless subgroup pre-specified. No study in this literature has a conservative arm, so nothing separates the effect of the operation from the natural history of a compressed root. A null result would not overturn the timing advice for people who do proceed to surgery, but it would overturn the assumption that decompression is what produces the 60 to 82% recovery.

What would change this protocol overall: a prospective cohort of consecutive adults presenting to primary care or physical therapy with painless isolated dorsiflexion weakness, N of 300 or more, worked up regardless of first impression, reporting what fraction are lumbar rather than peripheral and giving the hip abduction test the sensitivity and specificity it currently lacks. If the lumbar fraction came back under 5%, the clinical priority of this page would drop sharply.

Go Deeper

Don't want to guess whether a new weakness is urgent next time? The Verdict sends one evidence-scored protocol a week, free.

Join The Verdict, free

The Full Picture — Anatomy, Diagnosis & Evidence

What's Actually Going On

The lesion is compression of the L5 nerve root, most commonly at the L4/5 level, by a herniated disc or by stenosis. Across 918 surgically treated patients the cause was disc herniation in 79% (95% CI 0.72 to 0.85) and spinal stenosis in 22% (95% CI 0.15 to 0.30), with overlap (Than 2025).

The L5 root carries the supply to tibialis anterior and extensor hallucis longus, which lift the foot and the big toe. It also contributes to tibialis posterior, which turns the foot inward, and to gluteus medius, which holds the pelvis level when you stand on one leg.

That last detail is the clinically decisive one. The root sits above the point where the sciatic nerve divides, so a root lesion weakens muscles that the common peroneal nerve never reaches. A compression at the fibular head, which is the usual cause of a painless foot drop, cannot weaken hip abduction and cannot weaken inversion. Tanaka's series measured gluteus medius for exactly this reason and reported that all L5-supplied muscles recovered together at final follow-up.

Why the presentation is painless is not established. No reviewed study tested a mechanism for the dissociation between motor loss and pain. That gap is stated here rather than filled with a plausible-sounding account.

Dark cinematic rendering of a lumbar nerve root and its path to the lower leg

How to Identify It

Hip abduction strength (gluteus medius) Sn: NO DATA | Sp: NO DATA
Side-lying resisted abduction, graded 0 to 5, compared side to side. Weakness points above the knee.
Foot inversion strength (tibialis posterior) Sn: NO DATA | Sp: NO DATA
Resisted inversion from a plantarflexed position. Preserved inversion is a hallmark of common peroneal palsy.
Tibialis anterior manual muscle test Prognostic, not diagnostic
Resisted dorsiflexion, graded 0 to 5. Record the number. MRC 2 to 3 carries a far better prognosis than severe weakness, at OR 5.882 (95% CI 4.449 to 7.776).
Tinel sign and palpation at the fibular neck Sn: NO DATA | Sp: NO DATA
A positive finding supports the peripheral lesion. It does not exclude the root.

No published sensitivity, specificity or likelihood ratio exists for any clinical test that separates L5 root foot drop from common peroneal foot drop in the reviewed evidence. The hip abduction discriminator rests on anatomy and on its use for that purpose by Tanaka, not on a diagnostic accuracy study. It is a sound test with no numbers attached, and it should be taught that way.

Dark cinematic rendering of lower limb musculature under clinical lighting

The Debate

No clinical practice guideline for this presentation was identified in the reviewed evidence as of 2 September 2026. No NICE, APTA, BOA, EULAR or ACR guidance was retrieved. The disagreements are between studies, not between a guideline and a trial.

Does painlessness change the prognosis?

Aono 2007, Spine, N=46

Cases without leg pain "were also shown to be treated effectively with surgery". 61% recovered overall.

Tanaka 2021 (N=55) and Waseem 2023 (62 pooled)

Absence of radicular leg pain survived multivariate regression as one of only two independent predictors of non-recovery. Painless versus painful recovery to MRC 3 or above, OR 0.31 (95% CI 0.04 to 2.65).

Analytical, not populational. Aono did not adjust for duration or preoperative grade. Note that the raw proportions never separated: 65% in the painless-only series against 61% in the mixed cohort. Only the adjusted analysis separates them, and the magnitude remains unknown. Follow Tanaka and Waseem on analysis quality, and quote no number to a patient.

Does a complete palsy do better or worse?

Waseem 2023

Patients with a pre-procedure MRC of 0 had a higher median increase in MRC than those above 0.

Baig Mirza 2025 and Hou 2023

MRC 2/5 or above was significantly more likely to achieve greater recovery than MRC 0 (p=0.004). Moderate beat severe weakness at OR 5.882 (95% CI 4.449 to 7.776).

Not a contradiction, a different endpoint. One measures change in grade, which is floor-limited because a patient starting at 0 has more grades available to gain. The other measures the final grade reached. Reading Waseem's line as "complete palsy does better" inverts the clinical advice.

Honest Limitations

The evidence base is surgical-referral data, and the patients this page is about are missing from it

The research finding: 98.8% of the pooled painless cohort were managed operatively, with 74.2% showing symptomatic improvement (Waseem 2023).

The real-world gap: every patient in every one of these studies reached a spine surgeon. The painless foot drop that was labelled a peroneal palsy, braced, and reviewed in six weeks never enters any of these cohorts. The literature describes the outcome of correct identification and is structurally silent on how often identification fails, which is the only part of this a physical therapist actually controls.

Clinical adjustment: treat the 98.8% operative rate as a description of who got written up, never as a treatment recommendation, and never quote it to a patient as "almost everyone has surgery".

Four definitions of "recovery" produce a 38-point spread that is mostly definitional

The research finding: recovery rates run from 46.0% (Masuda 2020) to 84.5% (Saeed 2021), with 60%, 61%, 65%, 73.2% and 74.2% in between.

The real-world gap: these count different things. Reaching MMT 4 to 5, gaining any one MRC grade, a binary "recovered", or symptomatic improvement. Than 2025 contains the proof inside a single cohort, where 60% reached MMT 4 to 5 while 82% gained at least one grade in the same 918 patients.

Clinical adjustment: never quote a range across studies. State one number with its definition attached, or state the uncertainty honestly.

Heterogeneity is extreme, so the percentages are directions and not predictions

The research finding: I² of 89 to 97% on every recovery endpoint in the largest meta-analysis (Than 2025).

The real-world gap: that means the pooled figures summarise populations differing substantially in cause, severity, timing and technique. The point estimate does not describe the individual in front of you.

Clinical adjustment: use the direction and the individual prognostic factors, not the pooled percentage.

The Nuance

Conservative management success rate: no reviewed study reports a conservatively managed comparison group. The pooled painless cohort was 98.8% operative. Wang 2014 states directly that there is insufficient evidence that surgery is superior to conservative therapy.

Surgical success rate: 60% (95% CI 0.44 to 0.75) reach MMT 4 to 5 and 82% (95% CI 0.76 to 0.88) gain at least one grade across 918 patients. In painless cases specifically, 74.2% symptomatic improvement and 66.1% reaching MRC 3 or above, from 62 pooled patients.

Most people who have a lumbar decompression for a foot drop get some motor recovery, and a substantial minority, around 18% (95% CI 0.12 to 0.24), get none at all. What nobody knows is how many of those recoveries the operation caused, because compressed nerve roots also recover on their own and no study has ever compared the two paths. The recommendation to operate early rests entirely on comparing early surgery with late surgery, which is a different question.

So the defensible position is this. The evidence that timing matters is strong and consistent. The evidence that surgery works is suggestive and uncontrolled. A patient is entitled to hear those two separately rather than merged into one confident number. And for a painless presentation there is one extra thing to say out loud, which is that surgery here is being offered for strength alone, because there is no pain to relieve. That is precisely what makes the decision hard, and it was the surgeons who said so first.

Dark cinematic rendering contrasting the lumbar root and the nerve at the outer knee

Sources

Waseem S, et al. 2023, European Spine Journal, PMID 36843147. Systematic review restricted to painless foot drop. 62 pooled patients, 98.8% operative, 74.2% symptomatic improvement, 66.1% reaching MRC 3 or above, painless versus painful OR 0.31 (95% CI 0.04 to 2.65). The entire worldwide evidence base for this presentation.
Tanaka J, et al. 2021, Clinical Neurology and Neurosurgery, PMID 34059400. N=55. Duration before surgery and presence or absence of radicular leg pain were the only two significant multivariate predictors. Measured gluteus medius to separate root from peroneal lesions.
Baig Mirza A, et al. 2025, European Spine Journal, PMID 39961861. Individual patient data meta-analysis, 66 patients. Source of the 6 week window and the 3 month checkpoint.
Than CA, et al. 2025, Journal of Clinical Medicine, PMID 40648844. 20 cohort studies, 918 patients, PROSPERO CRD42024550980. Source of the 18% non-responder fraction and the 60% versus 82% endpoint gap.
Hou Y, et al. 2023, World Neurosurgery: X, PMID 37235062. 9 articles. Preoperative grade OR 5.882 (95% CI 4.449 to 7.776) and diabetes OR 5.657 (95% CI 2.094 to 15.280).
Song SY, et al. 2022, European Spine Journal, PMID 34718868. Timing meta-analysis, 6 studies, 312 participants, 157 early against 155 late.
Aono H, et al. 2014, Journal of Spinal Disorders & Techniques, PMID 24905505. 20 painless cases, 65% recovered, duration had the greatest effect.
Aono H, et al. 2007, Spine, PMID 17426622. N=46, 61% recovered. The source of the opposing position in The Debate.
Masuda S, et al. 2020, Clinical Spine Surgery, PMID 31498277. N=87, 46.0% recovered, 2 month cutoff.
Macki M, et al. 2016, Journal of the Neurological Sciences, PMID 26810531. N=71, 73.2% improved, median time to improvement within 6 weeks of surgery.
Saeed F, et al. 2021, European Spine Journal, PMID 34338874. Painful foot drop comparator, 9 studies, 431 patients, pooled improvement 84.5%.
Wang Y, Nataraj A. 2014, Clinical Neurology and Neurosurgery, PMID 24438801. Narrative review. States that there is insufficient evidence that surgery is superior to conservative therapy.
Kim HS, et al. 2022, Journal of Neurological Surgery Part A, PMID 34237776. 37 patients, full-endoscopic decompression, mean MRC 1.97 to 4.76.

Educational self-management guidance, not personalised treatment. A new or worsening weakness needs assessing in person.