AC Joint Separation Return-to-Contact Benchmarks: Shoulder Stability Tests Before Tackling

A club rugby prop walks off the field with a shoulder that suddenly has a visible step at the top of it. Two weeks later the acromioclavicular joint doesn't hurt when he reaches overhead, the swelling has settled, and the physio has signed off on full active range of motion. His coach wants him back in the scrum by the weekend. The instinct to let a pain-free, mobile shoulder make that call is understandable, and it's wrong for the same reason it's wrong after a shoulder dislocation: an AC separation is a ligament failure, not a muscle strain, and a ligament that heals back to load-tolerant takes considerably longer than the pain takes to fade. Reaching overhead and absorbing a tackler's body weight through a braced, adducted arm are two entirely different demands on the same joint.

A grade on an X-ray tells you how far the clavicle displaced the day of the injury. It does not tell you whether the coracoclavicular ligaments have remodeled enough to resist that same displacement the next time someone drives a shoulder into the point of his. Clinicians who follow these injuries past the initial diagnosis know that static grade and functional stability under load disagree often enough to matter. This guide sets out the clinical, strength, and contact-simulation benchmarks that predict tackle tolerance more reliably than a grade number or a pain scale, along with what the return-to-sport research on AC joint injuries actually supports and where it falls short for a contact-sport population specifically.

The Rockwood classification runs from Type I, a sprain with no displacement, through Type III, complete disruption of both the AC and coracoclavicular ligaments with the clavicle sitting up to roughly double its normal height above the acromion, up to Types IV through VI, where the clavicle displaces posteriorly, wildly superiorly, or inferiorly enough that surgery is rarely in question. Type I and II injuries almost always get managed conservatively. Type III sits in a genuinely debated middle ground, especially in athletes who need the joint to tolerate repeated collision loading rather than just daily activity. Types IV through VI usually go to surgery regardless of sport.

What that classification doesn't capture is how the joint behaves under the mechanism a tackle reproduces — a sudden compressive and shear force driven along the axis of a flexed, adducted arm, absorbed in a fraction of a second. A single AP or Zanca radiograph measures static displacement under gravity, sometimes with a stress view adding a light hanging weight. Neither test replicates a body checking into a shoulder at speed. Two athletes can carry an identical Type II grade on imaging and differ meaningfully in how much dynamic control their scapular stabilizers and remaining ligamentous structures provide once real contact forces show up, which is exactly the gap the benchmarks below are built to close.

Pallis and colleagues (2012) tracked AC joint injuries across several academic years of intercollegiate and intramural athletics at the U.S. Military Academy, reporting an overall incidence in the range of several injuries per 100,000 athlete-exposures, with male cadets injured at several times the rate of female cadets and wrestling and football producing the highest rates among the contact sports surveyed. The finding is useful for understanding who gets hurt and how, but the limitation is direct: a surveillance study of young military cadets counts injury occurrence, not recovery trajectory, and offers nothing on strength benchmarks or contact-load tolerance.

Mouhsine and colleagues (2003) followed a cohort of patients diagnosed with Type I or II AC dislocations on standard radiographs and managed conservatively, and found that roughly a third of them continued to show functional posterosuperior instability on clinical exam well after the injury, despite normal-looking static imaging. That subgroup accounted for nearly all of the conservative-treatment failures who eventually needed surgery. The limitation is a small, single-center retrospective cohort with instability defined by manual clinical testing rather than a standardized contact-load protocol — but the core finding is precisely why grade-on-X-ray shouldn't be the only input to a return decision. A joint can look mild on film and still fail to control the load a real tackle applies.

Spencer (2007) ran a systematic review comparing operative and nonoperative management specifically for Type III AC separations across the available comparative studies. He found no statistically significant difference in overall long-term outcome scores between the two approaches, but nonoperative patients returned to sport or work considerably sooner — often a matter of weeks rather than the three to four months typical after acute surgical reconstruction — while the operative group carried a distinctly higher complication rate across the pooled studies. The limitation the review itself flags is real: the underlying studies were mostly retrospective, used inconsistent outcome measures, and predate any standardized contact-specific return criteria, so neither treatment arm was tested against a tackle-simulation benchmark before clearance. Taken together, imaging grade and functional stability are separate questions, and neither operative nor nonoperative management alone answers when contact loading is safe — that's a functional-testing question, not a treatment-choice question.

Before any bag work or fall simulation enters the picture, the joint needs to pass a clean clinical exam. These checks catch the residual instability Mouhsine's cohort showed up on exam even when static imaging looked unremarkable.

A clunk during cross-body adduction that wasn't there at the last check deserves more attention than a lingering ache. Pain fades on a predictable timeline; a new mechanical sign usually means the dynamic stabilizers haven't caught up to where the ligaments left off, and a tackle will find that gap first.

Clinical tests confirm the joint isn't actively inflamed. Strength and closed-chain testing confirm the surrounding musculature can actually control the joint once load shows up, which is the entire job description during a tackle's deceleration phase.

The 90% side-to-side threshold on the dynamometer is deliberately conservative for the same reason it is after a dislocation: a shoulder sitting at 80% can feel completely normal in the gym and still fail well before the fourth quarter, once fatigue has eaten into the strength reserve a tackle needs in reserve. Retest weekly rather than trusting a number from two weeks ago — AC joint strength around the scapula tends to plateau below threshold longer than most return timelines assume.

Once the clinical and strength benchmarks above clear, contact itself gets reintroduced in stages, with each stage gated by objective signs rather than how confident the athlete feels that day.

Phase 2 is where most rushed returns unravel. A tackle bag at half speed still delivers real load, and an athlete eager to prove he's fine will sometimes turn his trunk a fraction early to shield the shoulder without realizing he's doing it. That's a compensation pattern, not readiness, and it's worth pulling video specifically to check for it rather than relying on how the session felt.

Most setbacks announce themselves a session or two before they become a real regression. Watch for these rather than waiting for a reported pain spike.

• Point tenderness returning at the joint line the day after contact drills, even if it was absent going into the session — an early sign that session's load exceeded what the joint could absorb that day.
• A visible step-off increasing compared with baseline photos, which can indicate progressive laxity rather than the residual cosmetic bump most AC separations leave behind permanently.
• Trunk rotating early or the shoulder dropping on video during a fall or tackle rep, protecting the joint in a way that wasn't present before the injury.
• Sensor-flagged asymmetric impact absorption during bag drills, particularly a pattern of leaning away from contact on the involved side.
• A new clunk or catching sensation during horizontal adduction that wasn't there at the prior check — a mechanical signal that outranks how the shoulder feels subjectively.