02/17/2026
Imaging can be a hot topic. This is always something to keep in mind: a picture is not good at determining why pan exists!
Taken from
99% of people over 40 have a rotator cuff ‘abnormality’ on MRI.
Do we round up to, everyone?
The FIMAGE study just dropped in JAMA Internal Medicine and it’s important for several reasons.
Here’s a quick overview of the study:
→ 602 people randomly selected from the Finnish general population (not a clinic sample)
→ Aged 41–76
→ Bilateral 3T MRI of both shoulders
→ Blinded reads by experienced musculoskeletal radiologists
→ Standardised clinical exams by shoulder surgeons with 10+ years experience
They found:
→ 98.7% had at least one RC abnormality
→ 7 out of 602 had completely normal tendons
→ 62% had partial-thickness tears
→ 11% had full-thickness tears (rose with age)
→ RC abnormalities were in 96% of pain-free shoulders AND 98% of painful shoulders
78% of all full-thickness tears were found in shoulders with zero symptoms.
Even combining state-of-the-art 3T MRI with experienced surgeon clinical exams could not reliably distinguish a symptomatic shoulder from an asymptomatic one.
The authors argue, and I think the data supports this, that most RC findings after 40 represent normal age-related structural change. Like grey hair. Like wrinkles. Like holes in your socks.
They emphasise a shift in language is warranted: away from ‘tear’ (which implies trauma and repair) toward terms like ‘structural alteration’ or ‘age-related change.’
When the baseline prevalence of an MRI finding approaches 100%, the finding itself has almost no diagnostic value.
How does this change the way you talk to your patients about their MRI results?
