Back to Blog
Training

Frozen shoulder or shoulder impingement? How to tell them apart at home

Jeffrey Sun

Jeffrey Sun, ACE-CPT

June 28, 2026 · 13 min read

ACE-certified personal trainer specializing in functional movement, mobility, and strength training for busy professionals in San Jose and the Bay Area.

Book a free consultation →
Frozen shoulder or shoulder impingement? How to tell them apart at home

A client came in last month, mid-fifties, told me her shoulder had been bothering her since January. She'd been doing the standard impingement drills she'd read about online, hadn't gotten worse but hadn't gotten better, and now reaching behind her back to clasp her bra had stopped working entirely. She thought she was just being lazy with the exercises.

I had her lie down on her back, relaxed her arm, and slowly lifted it overhead. It stopped at about 130 degrees, hard end-feel. I tried the other arm — that one went to 175, smooth all the way. She'd been treating impingement for four months, and she didn't have impingement. She had frozen shoulder, and it had progressed enough that she needed a clinical referral rather than at-home exercises.

This happens often enough that it's worth being able to tell the two problems apart yourself. Both make overhead reach painful. Both show up in adults over 40. Both can show up gradually without an obvious injury. But they're different problems with different timelines and different fixes, and treating one like the other wastes weeks of training stress on tissue that can't respond to it.

Here's how I work through the difference in a first session. If neither of these fits, no pain at all, just a shoulder that's tighter than it used to be, run the shoulder mobility test instead. It's built for exactly that case.

The two shoulder problems that get confused, and why it matters

Shoulder impingement is a soft-tissue pinch. The rotator cuff tendon, the bursa above it, or both get compressed in the subacromial space between the top of the shoulder blade and the head of the humerus. It hurts when the arm moves into a specific range, usually the middle of overhead reach. At rest the shoulder feels mostly normal. The fix is mobility, scapular control, and rotator cuff strengthening. I covered the whole thing in the shoulder impingement self-checks post last week.

Frozen shoulder, properly called adhesive capsulitis, is a capsular contraction. The connective tissue capsule that surrounds the joint thickens and contracts, physically limiting how far the joint can move. The arm can't go where it used to, full stop. The pain is often dull and constant rather than position-specific, and night pain when lying on the affected side is a hallmark. The fix is mostly clinical: physical therapy, sometimes intra-articular steroid injection, occasionally hydrodilatation or manipulation under anesthesia.

The reason this matters is the early progression. Impingement caught early responds well to a few weeks of band work and t-spine mobility. Frozen shoulder caught early responds much better to a steroid injection plus structured PT than it does to anything you can do alone. People who treat frozen shoulder like impingement for months tend to end up with longer overall recoveries and more residual stiffness.

If you have shoulder pain that's been around for more than a few weeks and isn't responding to the impingement drill protocol, the question to ask isn't "should I push harder," it's "am I treating the right thing."

Side view of a person in low light holding both hands across the front of their chest in a protective shoulder posture

The pain signature: how each one feels different

Beyond the structural difference, the felt experience of these two is different enough that you can sometimes tell from the symptom story alone.

Impingement pain is position-dependent. The shoulder feels fine sitting at your desk, fine carrying a bag, fine pushing a door open. The pain shows up specifically when the arm passes through the painful arc (roughly 60 to 120 degrees of abduction) or when you load overhead. Outside that range, you'd hardly know anything was wrong. The pain is sharp and localized, often on the front or top of the shoulder.

Frozen shoulder pain is range-of-motion-dependent, which sounds the same but plays out differently. The shoulder hurts when you try to go anywhere near the end of available range, but the range itself has gotten smaller, so you hit "the end" much sooner than you used to. Reaching to a high shelf hurts because you can't actually reach the shelf anymore. Reaching behind your back to fasten a bra or pull a wallet out of a back pocket has gotten progressively harder over weeks or months. At rest, the shoulder often aches in a dull, generalized way, especially at night.

The night pain piece is one of the strongest tells. People with impingement sometimes have soreness after an aggravating session but usually sleep fine. People with frozen shoulder describe waking up multiple times a night, unable to lie on the affected side, sometimes needing to sleep in a recliner during the worst stretch. If you can't lie on the affected shoulder without sharp pain, frozen shoulder should be at the top of the differential.

Onset is the other tell. Impingement usually traces to something — started lifting again, started a new sport, ramped up overhead reps. Frozen shoulder typically has no clear trigger. It just shows up gradually, gets steadily worse over weeks to months, and the person remembers no specific event. If someone asks "what did you do to it" and you can't answer, that's a frozen shoulder data point.

Test 1: passive range of motion (the most useful differentiator)

This is the single most useful at-home test for telling these apart. It takes 30 seconds and someone else.

Lie flat on your back on a bed or the floor. Fully relax your arm at your side. Have a friend, partner, or family member slowly lift your arm overhead for you. Their hand should be near your wrist or forearm. You do not help. Stay completely passive.

Impingement pattern: Your friend can usually move the arm overhead through your full normal range with relatively little pain. You can't comfortably do this same motion under your own power because the cuff is irritated and the pinch happens during active loading. The passive motion bypasses the pinch.

Frozen shoulder pattern: Your friend hits a hard stop somewhere short of full overhead range, usually well before 180 degrees. The arm physically won't go further regardless of who's moving it, because the joint capsule itself has contracted. Often the stop comes with significant pain in the deep joint.

Test the unaffected side first to know your baseline. A 30-degree or larger gap between the two sides in passive range strongly suggests frozen shoulder, not impingement.

This test is so useful because it removes the muscular contribution. If your active range is bad but passive range is good, the limit is muscle and tendon, which means impingement-pattern. If passive range is also bad, the limit is the structure of the joint, which means capsular contracture.

Test 2: reach behind your back

A specific frozen shoulder hallmark is loss of internal rotation behind the back. Most people can reach their thumb to somewhere between the bottom of the shoulder blade and the middle of the back. Frozen shoulder usually destroys this range early.

Stand naturally. Reach your affected arm behind your back, palm facing out, and slide your thumb up your spine as high as it will go. Mark the height with your other hand. Then do the same on the unaffected side and mark that.

Pass (impingement-pattern): Both thumbs reach roughly the same height. Maybe the affected side is an inch or two lower with some discomfort, but in the same general range.

Fail (frozen-shoulder-pattern): The affected side reaches significantly lower than the unaffected side — often 4 to 8 inches lower. The thumb might only reach the waistband or even just the hip. Trying to push higher hits a hard end-feel and triggers deep joint pain.

This is the test that catches frozen shoulder the earliest. People often notice they can't fasten a bra or reach a wallet before they notice the overhead loss, because the daily friction is much more constant.

Back view of a person in black and white reaching one arm behind their back to clasp the other at the mid-back, demonstrating shoulder internal rotation reach

Test 3: external rotation at the side

External rotation is the other movement that frozen shoulder restricts early, while impingement usually leaves it alone.

Stand with the affected arm at your side, elbow bent to 90 degrees with the forearm pointing straight forward. Keeping the elbow pressed into your side, slowly rotate the forearm outward, away from your body. Most people can comfortably reach 70 to 90 degrees on each side. Stop when the motion is limited by pain or a hard physical stop. Compare to the unaffected side.

Pass (impingement-pattern): Both sides rotate to roughly the same angle, both within normal range. Maybe slight discomfort on the affected side but no hard stop, and the angles match within 10 to 15 degrees.

Fail (frozen-shoulder-pattern): Significant difference between sides — often 30 degrees or more. The affected arm hits a hard end-feel where the forearm just stops going further, with deep joint pain at the limit.

External rotation loss is one of the clinical signs that physical therapists and orthopedists use specifically to identify frozen shoulder. Catching it on your own is meaningful.

The reason these three tests work together is that they triangulate. One of them coming back positive isn't conclusive. Two or three of them coming back consistent with frozen shoulder is strong evidence for getting a clinical evaluation rather than continuing to self-manage.

What each result means (and the third possibility nobody mentions)

If all three tests look like impingement (good passive range, normal reach behind back, normal external rotation, pain only in the middle of overhead reach), the most likely answer is impingement. Treat it with mobility, scapular control, and gradual loading per the impingement self-check post. Expect change in 4 to 8 weeks.

If two or three of the tests came back consistent with frozen shoulder (limited passive range, limited reach behind back, limited external rotation, especially with night pain or no clear trigger), the most likely answer is frozen shoulder, and the next step is a doctor or orthopedist, not more shoulder exercises.

The third possibility nobody mentions is that you might have both. Frozen shoulder can develop in a shoulder that started with an impingement that wasn't moving, particularly if pain led to weeks of guarded movement. The arm-pinned-to-the-side adaptation can gradually shift into capsular contracture in some people. If the impingement protocol from a few weeks ago made you feel slightly better but then plateaued and is now getting worse, frozen shoulder layering on top is worth considering.

A fourth, less common possibility is a rotator cuff tear. Tears tend to have sudden onset (you can usually point to the moment), significant weakness in specific positions (you can't lift the arm overhead at all against gravity), and don't fit either of the two patterns above cleanly. Cuff tears need imaging and a clinical opinion, often surgical.

The point of these three tests isn't to give yourself a diagnosis. It's to know whether to keep working at home or to book an appointment.

When this is a doctor visit, not a trainer fix

Specific signals that should prompt a clinical visit, not more time waiting it out:

Passive range of motion is meaningfully restricted in any direction — overhead, behind the back, external rotation. The capsule restricts joint motion in ways no exercise outside of a clinical setting can fix early.

Night pain that wakes you up. Frozen shoulder pain at night is severe and consistent enough that most people end up sleep-deprived during the worst stretch. There are specific medical interventions (oral medications, injections) that meaningfully shorten this stage, but they require a prescription.

Progression over weeks to months without injury. A shoulder that started fine and has gotten worse over 6 to 12 weeks without anything happening to it is suspicious for frozen shoulder, especially over 40.

Diabetes, thyroid disease, or recent immobilization. These all raise frozen shoulder risk significantly. A meta-analysis of diabetes and adhesive capsulitis found diabetics were 5 times more likely to develop frozen shoulder than non-diabetics. If you're in any of these buckets and have shoulder pain, frozen shoulder should be ruled in or out by a clinician early rather than late.

Daily tasks have gotten progressively harder. Putting on a coat, washing your hair, fastening a bra, pulling out a wallet. If multiple of these have noticeably worsened over weeks, the conversation is medical, not training.

The reason the early-visit recommendation is so strong here is the natural course of frozen shoulder is bad if untreated. Eighteen months to two years of progressive stiffness, pain, and gradual recovery. Early intervention with a steroid injection plus structured PT can compress that timeline meaningfully, and the longer it's untreated the more residual loss is common.

The cost of seeing a doctor when it turns out to be impingement is low. The cost of self-managing frozen shoulder for six months is high.

If you've already been through PT and the question now is how to get back to training without flaring the shoulder, the post-PT bridge guide covers what comes after discharge.

Where to go from here

If your three tests came back consistent with impingement, the shoulder impingement self-checks post walks through the drills and 4-week protocol that usually moves the pattern.

If you want a baseline read on your shoulder mobility against the other joints in the chain (t-spine, hips, ankles) before deciding what to work on, the movement screen is a five-minute online assessment that scores the relevant axes. It won't replace a clinical eye on a frozen shoulder. It will tell you which parts of your overall movement quality are restricting how the shoulder loads.

And if you're in San Jose or the South Bay and want eyes on whether your shoulder pain is something to train through or something to refer out, come work with me. A first session usually answers the differential question clearly. I've coached more than 12,000 sessions, and the impingement-vs-frozen-shoulder question comes up often enough that I have a clean process for sorting it. The early referral is one of the highest-value things a trainer can do for a client who's heading toward a year of restricted shoulder use, and the right first step is knowing for sure which problem you actually have.

The body usually tells you what's going on if you ask it the right questions. These three tests are the right questions.

Ready to train smarter?

Get a personalized program built around your goals, your body, and your schedule.

Book Your Free Consultation

Related Articles