A routine chest CT scan, the kind millions of Americans get every year to check their lungs or heart, may now be able to reveal something doctors never thought to look for: years of hidden stress damage building inside the body.
Story Snapshot
- Johns Hopkins researchers built an AI tool that measures adrenal gland size on chest CT scans to spot chronic stress.
- The new measure, called Adrenal Volume Index, lined up with stress questionnaires and blood cortisol levels in patients.
- Bigger adrenal volume tracked with higher heart disease and heart failure risk in the study group.
- The tool reuses scans already taken for other reasons, so no new tests or extra cost are needed.
What Johns Hopkins Actually Found
Doctors at Johns Hopkins School of Medicine trained a deep learning model to automatically measure the adrenal glands on chest CT scans patients already had done for unrelated reasons. Led by Dr. Elena Ghotbi, the team called this new measurement the Adrenal Volume Index. It was unveiled at the Radiological Society of North America’s annual meeting and described in a press release dated November 25, 2025.
The adrenal glands sit above the kidneys and pump out cortisol, the body’s main stress hormone. A single blood cortisol test only shows stress at one moment in time. The researchers argue that adrenal size builds up slowly, so it works more like a scar tissue record of stress exposure over months or years rather than a snapshot.
Why This Beats a Single Cortisol Test
The team checked their AI measurement against validated stress questionnaires and actual cortisol blood levels in patients, and the numbers lined up. That matters because cortisol tests are expensive, timing-sensitive, and rarely ordered outside of specialized care. A chest CT scan, by contrast, is one of the most common imaging tests done in American hospitals and clinics every single day.
Lead author Dr. Ghotbi explained the appeal plainly: the approach uses imaging data that already exists, opening the door to studying chronic stress across huge numbers of patients without ordering a single new test. For a health system drowning in costs, a free biomarker hiding in old scans is the kind of efficiency conservatives should want more of, not less.
The Heart Failure Connection
Chronic stress has long been suspected of contributing to heart disease, but proving it with hard data has been difficult. This study found that patients with higher Adrenal Volume Index scores also carried greater risk markers for cardiovascular disease and heart failure. That gives doctors a concrete number to track instead of relying on a patient’s self-reported stress level, which can be unreliable or downplayed during a short office visit.
This fits a pattern researchers have seen before. A separate study on “vital exhaustion,” a measure of chronic fatigue and burnout, found that people in the highest exhaustion group had meaningfully elevated levels of heart and inflammation biomarkers, along with higher rates of cardiovascular events. Stress biology consistently shows up as a real, measurable risk factor for heart trouble, not just a vague feeling.
One Marker Among Many, Not a Silver Bullet
No single test should be treated as gospel, and this one is no exception. Cardiologists have spent years building panels of multiple biomarkers, like C-reactive protein and natriuretic peptides, because heart failure risk rarely traces back to one clean cause. The Adrenal Volume Index looks like a strong new addition to that toolbox, not a replacement for careful clinical judgment.
Johns Hopkins has been on a run of AI breakthroughs in cardiac care lately. Earlier this year, a separate university team built a multimodal AI model that outperformed doctors at predicting sudden cardiac arrest by combining heart imaging with medical records. Put together, these efforts point to a hospital system betting heavily on AI to catch deadly heart problems years before symptoms show up.
What Patients Should Take From This
This research is still early. It came from a conference presentation and press release rather than a fully published, peer-reviewed clinical trial, and larger studies will need to confirm the findings hold up across different patient populations. Even so, the underlying idea is common sense: stress leaves fingerprints on the body, and modern imaging combined with smart software may finally let doctors read those fingerprints without ordering a single extra test.
For the average American already getting chest scans for other health reasons, that is genuinely good news. It means valuable information may already be sitting unused in medical records nationwide, waiting for AI tools like this one to unlock it and give doctors an early warning before chronic stress turns into a heart attack or heart failure diagnosis.
Sources:
docs.google.com, psychologytoday.com, facebook.com, pmc.ncbi.nlm.nih.gov
