Rotator cuff tears on MRI are nearly universal after 40 - and often irrelevant

If you are over 40 and have no shoulder pain whatsoever, there is still a good chance that an MRI of your shoulder would show a rotator cuff abnormality. A large population-based study from Finland, published in JAMA Internal Medicine, found that rotator cuff findings on magnetic resonance imaging became nearly universal after age 40 - yet showed poor concordance with whether people actually had shoulder symptoms.

The study, led by Thomas Ibounig, MD, of the University of Helsinki, addresses a question that has significant clinical and economic implications: when someone presents with shoulder pain, does ordering an MRI add useful diagnostic information? The answer, according to these data, is often no - at least for atraumatic shoulder pain where no injury has occurred.

What the scans showed - and what they didn't predict

The research team conducted a population-based study examining the prevalence of rotator cuff MRI abnormalities across different age groups and then assessed how well those findings correlated with actual shoulder symptoms among the same individuals. The core finding was stark: abnormalities were common, increasingly common with age, and only weakly linked to whether someone had symptoms.

This pattern is consistent with what is understood about musculoskeletal aging. The rotator cuff - four muscles and their tendons that stabilize the shoulder joint - undergoes age-related wear that produces MRI-visible changes without necessarily causing pain or dysfunction. The imaging is detecting a biological reality of aging, not a pathological process that needs treatment.

The accompanying commentary to the study is co-authored by Brian Feeley, MD, team orthopedist for the San Francisco Giants, lending a sports medicine perspective to findings that have practical implications both in athletic and general clinical contexts.

Why this matters for how shoulders get treated

The problem with ubiquitous MRI abnormalities is that they can prompt unnecessary intervention. A person comes in with shoulder discomfort; an MRI reveals a partial thickness tear; the finding triggers surgical discussion. But if that same partial thickness tear is present in most people the same age who have no symptoms at all, it may be irrelevant to the presenting complaint.

Treatment decisions driven by incidental imaging findings expose patients to the risks of procedures - surgical complications, rehabilitation burden, time off work - without therapeutic benefit. They also drive healthcare costs without improving outcomes. The study supports clinical guidelines that recommend against routine imaging for atraumatic shoulder pain and favor physical examination, activity history, and conservative management as first-line approaches.

Limitations of the study

Population-based studies capture prevalence across a general population, which may not precisely represent patients who present to orthopedic or primary care clinics - who may differ in ways that affect the MRI-symptom relationship. The study also did not track whether participants went on to develop symptoms or functional limitations, so the natural history of asymptomatic findings remains only partially characterized.

Additionally, the population studied was Finnish, and some caution is warranted in generalizing to populations with different physical activity patterns, occupational demands, or genetic backgrounds. The finding that rotator cuff abnormalities are near-universal after 40 is consistent with prior smaller studies, however, and the population-level design provides more statistical confidence than clinical case series.

What good clinical care looks like

The study reinforces a conservative, symptom-led approach to atraumatic shoulder pain. Physical therapy targeting shoulder mechanics, activity modification, and careful history-taking about the onset and character of pain generally provide more actionable information than an MRI scan whose findings are, in this population, nearly universal. When imaging is genuinely warranted - for acute injury, progressive symptoms, or when surgery is being considered - the findings should be interpreted in the full clinical context, not in isolation.