Skier's Thumb (Torn Thumb Ligament)

What Works

Dark cinematic rendering of a thumb immobilised in a supportive cast

Exercise Prescription

Deliberately absent, for the reason above. Your specific movements should come from the clinician who has actually examined your thumb, once they know whether it's stable. The timeline below is real; the exercises inside it are theirs to set.

1. Get it examined properly, with the thumb bent MODERATE-HIGH

The single most useful finding in this literature. One hospital changed the position it tested thumbs in, and the number of severe displaced tears it detected went from 20% to 70%. Same surgeons, same operation. The only thing that changed was the angle the thumb was held at during the test.

2. Protection for a stable tear MODERATE

If the ligament is torn but hasn't displaced, a cast or splint is the actual treatment, not a holding measure. The protocol used in the current trial is 4 weeks in a fixed cast covering the thumb knuckle and its base, then 4 weeks in a removable one.

3. Prompt referral if it's displaced or unstable MODERATE

The 2024 UK guideline puts it simply: no significant looseness means no surgery needed. Significant looseness means you and a hand surgeon decide together, within about two weeks.

4. Ultrasound when the examination is unclear MODERATE

Catches around 9 in 10 of these, and it's quick and cheap. Notably, the 2024 guideline deliberately does not ask for scans on everyone. A good hands-on examination answers most cases.

5. Coming back for a second look if you're too sore to be examined MODERATE

When a thumb is guarded, examination agreed with the reference test about a quarter of the time. Once the swelling settles, nearly always. Booking the follow-up is the cheapest thing on this page.

Weaker options, and one you'll see quoted

Feeling for the displaced ligament end EMERGING — one study of 24 patients. Useful when you find the lump, unreliable when you don't. A negative doesn't rule it out.

Early gentle movement after surgery MODERATE — the only randomised rehab evidence in this literature, 30 patients, and it favoured moving early. That's a decision for your surgeon, because it depends what they did inside.

Returning to sport at 5 to 6 weeks after surgery LOW — you'll see this quoted. It's a proposal in an article with no comparative data and no re-rupture figures, argued against by lab work showing a repaired ligament at roughly a fifth of its original strength on day one. The traditional timeline is 12 weeks.

What Doesn't Work

Clearing it on an x-ray. 40% of these injuries have a bone fragment, and fragments that look identical on film can be attached to the ligament or lying loose beside a completely torn one.
Casting a small chip and hoping. In the only study to test it, all 9 patients still had pain and all 9 ended up having surgery. The chip rotates in a way the x-ray measurement can't see.
Being told it's fine while the thumb is too swollen to examine. 8 of 47 patients cleared that way were actually unstable.
Testing the thumb straight instead of bent. It misses the tear that matters, and it gives a confident wrong answer, which is worse than an uncertain one.
Checking only one side of the thumb. About 1 in 4 of these injuries in one group of athletes had both sides torn, and every one of those needed surgery.
Buddy taping it. No evidence base for this injury, and it doesn't control the sideways force that caused the problem in the first place.

Return to Training

This one is a substitution, not a shutdown. Lower body, trunk and your uninjured arm carry on at full load, so there's no reason for overall training volume to drop. What's out for roughly 8 weeks is anything needing a firm pinch or a wrapped thumb: barbell work with a full grip, pull-ups, kettlebells, racket and stick sports, climbing, and contact sport.

A note on timelines: the widely quoted 132-day return figure comes from professional footballers with daily supervised rehab. There is no published return-to-activity dataset for recreational or sedentary people, so treat milestones as the guide rather than a date.

Cleared by the treating clinician or surgeon, with joint stability specifically confirmed
Protection period completed as prescribed (typically 8 weeks for a stable tear)
Sideways looseness matching the uninjured thumb, or within the range your clinician set
No tenderness over the ligament where it attaches
Pinch strength tested (not before week 8) and close to the other side
Full, pain-free thumb movement
Able to meet your sport or job grip demands without pain or hesitation
No swelling or pain in the 24 hours after a trial session
For contact and racket sports: a taping or splinting plan agreed for the return period

Red Flags — get seen this week

If any of these apply, this needs a hand specialist rather than time.

Cinematic anatomical rendering of the thumb joint and surrounding ligament structures

You can feel a tender lump at the base of your thumb, on the side facing your index finger, that wasn't there before.
The injured thumb bends noticeably further sideways than your other thumb.
You can't pinch or grip properly, or the thumb feels like it gives way.
You had an x-ray showing a small chip of bone. A chip does not mean it's minor. In the one study that tested casting them, all nine patients still had pain and all nine ended up needing surgery.
The thumb hurts on both sides of the knuckle, not just the index-finger side.
Your thumb was too swollen to examine properly and you were told it was fine. That is not the same as it being fine.
Obvious deformity, an open wound, or a possible dislocation. That one is urgent, today.

What to say on the phone: "I forced my thumb sideways. It's swollen on the index-finger side of the thumb knuckle and my pinch grip is weak. I need it stress tested, not just x-rayed."

Hold both thumbs up side by side and have someone gently push each one sideways, away from the index finger. If the injured one bends noticeably further than the good one, or you can feel a tender lump at its base, book an appointment this week.

Comparing against your own uninjured thumb is the comparison the research actually uses, because normal thumb looseness varies enormously between people.

Takes under a minute. No equipment needed.

The Verdict

A jammed thumb can be a torn ligament, and an x-ray won't tell you which.

Think of the ligament as a strap that's been ripped off its anchor point. Normally the loose end stays lying against the bone, and it can knit back down like a plaster re-sticking. But a sheet of tissue runs across the thumb like a seatbelt, and if the strap flicks up over the top of that sheet, the sheet is now wedged between the strap and its anchor. It can't reach the bone any more.

That's why some of these heal in a cast and some never will, and why the difference is invisible from the outside.

What this actually is: A jammed thumb is often a torn ligament, and it's torn in the exact spot your pinch grip depends on.
The one thing that makes it worse: Waiting it out because the x-ray looked fine, when an x-ray shows bone and this is a ligament.
Start here: Stop pinching, gripping and twisting with that thumb, and ask for it to be stress tested rather than just x-rayed.

Best for

Anyone who forced a thumb sideways and now has a weak or painful pinch grip, whether from a fall, sport, or catching it on something.

Skip if

Your thumb pain came on gradually with no injury, or sits at the base of the thumb nearer the wrist. That's a different problem.

Want the full evidence? Keep scrolling. ⌄

How Confident Are We?

Conviction: Moderate

Strong on how to examine this and when to refer. Weak in precisely the place you'd most want certainty: whether the worst version of this injury genuinely needs an operation.

What would change this: a trial called MUSCAT is running right now. It is randomly assigning people with completely torn, displaced ligaments to either surgery or a plaster cast, which nobody has ever done before in the sixty years this rule has existed. When it reports, this page changes either way.

Claim 1: "A displaced tear needs surgery" — why we hold this loosely

This has been an absolute rule since the early 1960s, and it rests on an anatomical argument plus surgical case series rather than a controlled comparison. The 2024 NICE-accredited guideline rates its own recommendation on this as low to very low certainty and moves to shared decision-making.

What would change this: MUSCAT reporting. If a cast performs as well as surgery at 6 months, with roughly 1 in 10 needing surgery later, this becomes a genuine choice rather than a rule. If the cast arm does worse, the sixty-year-old argument finally has evidence behind it.

Claim 2: "Examine it with the thumb bent" — how solid is this?

It rests on a study of 750 normal thumbs plus cadaver work establishing the correct test position, and on one hospital's detection rate tripling after it switched. That second study compared two different time periods rather than randomising, so the exact numbers are soft.

What would change this: a study of around 300 consecutive thumb injuries at first presentation, examined in both positions against a scan, would turn a historical signal into a real figure. Nobody has run it.

Don't want to guess next time you injure something? The Verdict sends one evidence-checked protocol a week, free.

Join The Verdict

The Full Picture — Anatomy, Diagnosis & Evidence

What's Actually Going On

Dark cinematic anatomical view of the thumb metacarpophalangeal joint and its collateral ligament

The ulnar collateral ligament runs along the index-finger side of the thumb knuckle, connecting the thumb's long bone to the first bone of the thumb itself. Its job is to stop the thumb bending sideways away from the hand, which is exactly what your pinch grip loads.

The injury is a forced sideways bend. Classically a ski pole, historically the neck-wringing action that gave gamekeeper's thumb its name, and in practice usually a fall onto an outstretched hand with the thumb sticking out.

Both eponyms are misleading, and that costs diagnoses. In one series of 127 consecutive cases, skiing caused just 2.4% while falls caused 49%. The ski-pole theory even failed its own natural experiment: 65% of injured skiers used traditional strapped poles against only 20% of uninjured skiers, yet the total number of repairs did not fall as pole design changed. A fall onto an outstretched hand simply doesn't prompt anyone to examine the thumb.

The piece of anatomy that decides treatment is the adductor aponeurosis, a sheet of tissue lying over the ligament. When the ligament tears off its lower attachment and the thumb springs back, the torn end can flip up and come to rest on top of that sheet. The sheet is then trapped between ligament and bone. This displaced configuration is the Stener lesion, and the argument that follows is that a ligament separated from its footprint by an interposed sheet cannot heal back down however long you immobilise it.

Associated damage is common: alongside the ligament tear, one series found bone avulsions in 21%, tears of the back of the joint capsule in 57%, and capsule infolding in 29%.

How to Identify It

Cinematic clinical view of a thumb being examined under sideways stress

Typical complaint: "I fell and jammed my thumb. It's swollen and I can't grip anything properly. I can't turn a key or hold a mug without it feeling like it's going to give way."

Sideways stress test, thumb knuckle FULLY BENT, compared against the other thumb. No published accuracy figures exist for this manoeuvre anywhere in the retrieved literature, which is itself notable. But full bending is the correct position: a study of 750 normal thumbs plus cadaver dissection established that it's where cutting the ligament produces obvious looseness, while cutting the surrounding tissue alone barely does.

Examination after numbing the thumb Sn 87.5% | Sp 100% Agreement with the reference test rose from 28% (95% CI 16-43) to 98% (95% CI 88-100) after local anaesthetic, in 47 patients assessed about a week after injury. Eight patients previously recorded as adequately assessed turned out to be unstable.

Ultrasound Sn 88-92% MRI Sn 100% | Sp 100% in a 17-patient study, so read that as clearly better rather than perfect.

Thresholds: bending beyond 35 degrees, or more than 15 degrees further than the uninjured thumb, indicates injury regardless of the x-ray. Surgical thresholds quoted in the operative literature are more than 30 degrees with the joint bent, or more than 20 degrees with it straight. Note: the 35 and 15 degree figures come from a guideline document that could not be independently verified in this review, and it is a paediatric hospital guideline being applied to adults. Treat them as orientation.

The radiograph trap. Of 63 consecutive acute injuries, 25 had a fracture. Only 8 were true ligament avulsions with the fragment attached. Seven looked identical on film but weren't attached at all, and 10 more sat beside a completely torn ligament. The types cannot be told apart on routine films, which is why stress testing is required even when the x-ray looks like a minor undisplaced chip.

What it is not

Dark cinematic comparison view of thumb and wrist anatomical structures

Thumb base arthritis is gradual, has no injury behind it and sits nearer the wrist. A scaphoid fracture is tender in the hollow at the base of the thumb, not at the knuckle. De Quervain's is on the thumbnail side of the wrist and comes on gradually. A radial collateral tear is the same injury on the opposite side of the same knuckle, and is worth checking every single time: 25% of thumb collateral injuries in one group of professional footballers involved both sides, and all of those needed surgery against 63% of the one-sided ones.

The Debate

Does a displaced tear actually need surgery?

Standing position, since the early 1960s

A displaced tear (Stener lesion) is an absolute indication for surgery. The ligament physically cannot reach its footprint, so it cannot heal. Codified in operative indications and taught for sixty years.

VS

MUSCAT trial protocol, 2024

A multicentre trial is randomly assigning complete tears, including displaced ones, to a plaster cast. Its stated rationale: the existing recommendation "is based on expert opinion, anatomic theories, and low-quality observational case series".

The rule was asserted, never tested. The 2024 NICE-accredited guideline rates its own recommendation on this as low to very low certainty and has already moved to shared decision-making. The honest reading is not that surgery is wrong, it's that nobody has ever checked. Keep referring these patients, because a running trial is not a result, and tell people plainly that this is a shared decision rather than a foregone one.

A second, quieter disagreement: practice has tended to treat looseness as the thing that decides surgery, but the one study that allocated by feeling for the displaced ligament end instead, and cast everything non-palpable regardless of how loose it was, found both groups similar at one year. Displacement and looseness are not the same variable.

Honest Limitations

1. This is a surgical literature, and most patients aren't surgical

Almost every number on this page was measured in a thumb that went to theatre. Across 40 retrieved papers there are zero systematic reviews and zero pooled analyses, and all 40 were available only as abstracts. The entire non-surgical evidence base amounts to roughly 60 patients across three small studies. We know a lot about what happens after repair and very little about what happens without it.

2. The accuracy figures come from specialists, not from first contact

That 98% agreement was achieved about a week after injury by a single examiner who knew he was in a study. Real first contact is an acutely swollen, guarded thumb, and the same study's baseline for exactly that situation was 28%. The gap between the two is a process problem, not a skill problem. Nobody can examine through a guarded thumb, however experienced they are. The fix is booking the second look.

3. The return-to-sport data describe professional athletes

95.7% returned to play at an average of 132 days, with daily supervised rehabilitation and a collision-sport injury pattern behind it. Recreational and sedentary people, who are most of the caseload, have no return-to-activity dataset at all. Any confident timeline you're given for ordinary life is extrapolation.

The Nuance

Dark cinematic anatomical rendering of thumb ligament structures in cross-section

For a stable, non-displaced tear, protection is reasonable first-line care and the small amount of available data supports it. For a complete tear with the ligament flipped out of position, the sixty-year answer is surgery, and that answer rests on an anatomical argument plus case series.

Surgery results are consistently good. Complete stability at 3 months in 34 of 34 in one series. Pinch weakness none or mild in 96%, and pain none or mild in 99%, of 69 followed thumbs. 95.7% return to professional sport. For long-standing tears reconstructed with a tendon graft, 24 of 26 were stable and pain-free at an average of 4.5 years, retaining 85% of movement.

The complication rate is real but modest: temporary nerve irritation in 6.5% of 127 operated patients. One rare but serious surgical error is documented in this literature, where the median nerve was taken instead of the intended tendon graft, which is why new numbness after a reconstruction should never be waved off as normal healing.

One long-horizon fact worth knowing. Repair of long-standing tears is durable, but at an average of 24.5 years afterwards, 88% of those patients had arthritis in the joint. Notably, delay to treatment did not correlate with what showed up on the x-rays, so this looks more like the consequence of the original injury than a punishment for waiting.

Sources

MUSCAT Study Group, 2024, Trials (PMID 39468632). Multicentre non-inferiority RCT protocol, NCT05291260. Cast vs immediate surgery for complete ruptures including Stener lesions. Results not yet published.
Dean B, et al., 2024, J Hand Surg Eur Vol 49:1195-1201 (PMID 39315553). BSSH BEST, NICE-accredited clinical guideline. Several recommendations rated low to very low certainty.
Cooper JG, et al., 2005, Emerg Med Australas (PMID 15796727). n=47. Examination agreement 28% before local anaesthetic vs 98% after.
Hergenroeder PT, et al., 1978, J Hand Surg (PMID 722028). 750 normal thumbs plus 25 cadaver thumbs. Established the correct test position.
Heim U, et al., 1986, J Bone Joint Surg Am (PMID 3782203). n=40 operated. Displaced-tear detection 20% vs 70% after switching test position.
Hintermann B, et al., 1993, Am J Sports Med (PMID 8291629). n=63. 40% had a fracture; fragment types indistinguishable on routine films.
Dinowitz M, et al., 1997, J Hand Surg (PMID 9471077). n=9. All nine minimally displaced avulsion fractures failed casting.
Sollerman C, et al., 1990, J Hand Surg (PMID 2348064). n=24. Allocation by palpation rather than by looseness; similar 1-year outcomes.
Colegate-Stone T, et al., 2009, Injury (PMID 19389670). n=127. Ultrasound Sn 92%; falls 49%, skiing 2.4%; nerve irritation 6.5%.
Werner BC, et al., 2017, Am J Sports Med (PMID 27566241). 36 thumbs, 32 NFL players. 25% combined both-sided tears.
Rocchi L, et al., 2014, Eur J Phys Rehabil Med (PMID 24185690). RCT, n=30. The only randomised rehabilitation evidence available.
Lee AT, et al., 2016, Hand (PMID 27698632). Mean 24.5 years. 88% osteoarthritis; delay uncorrelated with radiographic findings.
Werner BC, et al., 2018, Hand (PMID 28836462). Cadaver, 9 paired hands. Repair at 68 N yield vs intact 342 N.
Landsman JC, et al., 1995, Radiology (PMID 7997584). n=17. Ultrasound vs MRI accuracy.

Evidence quality note: no systematic review or meta-analysis exists for this condition. All 40 papers retrieved in this review were available as abstracts only. Treatment recommendations here are guideline-level and case-series-level, not trial-level, and the page says so where it matters.

There's a ligament on the index-finger side of your thumb knuckle.

Skier's Thumb