Blood antibodies may signal bowel disease up to a decade early

Those elevated responses were most noticeable in people who later developed complicated Crohn’s disease or disease affecting the ileum, the final section of the small intestine.

Antibody responses to encapsulated bacteria gradually decreased as diagnosis approached. These bacteria carry protective outer coatings that can help them evade immune defenses.

Future ulcerative colitis cases showed a somewhat different combination. Responses to herpesviruses and autoantigens increased, while responses to encapsulated bacteria declined.

Autoantigens are components of the body’s own tissues that the immune system can mistakenly target.

Together, these patterns show that measurable immune differences can precede an IBD diagnosis, although the study does not establish whether the antibody responses cause the disease.

How accurately could the models identify future disease?

Researchers applied machine learning, a form of artificial intelligence that identifies patterns in complex data, to the antibody profiles.

They tested whether those profiles could distinguish people who would later develop IBD from healthy controls using samples collected up to 10 years before diagnosis.

The area under the curve (AUC) measures how well a model separates two groups. A score of 0.50 is no better than a coin flip, while 1.00 represents perfect discrimination.

• Future Crohn’s disease versus healthy controls: AUC 0.90.
• Future ulcerative colitis versus healthy controls: AUC 0.84.
• Complicated versus less complicated Crohn’s disease: AUC 0.85, approximately 10 years before diagnosis.

The models also separated people who later developed complicated Crohn’s disease from those whose disease would remain less complicated, roughly a decade before diagnosis.

Those scores describe group-level discrimination in the studied sample. They do not mean that an individual has a 90% or 84% probability of receiving a future diagnosis.

Lead author Dr. Arno Bourgonje emphasized the timing of the signal.

“It is fascinating to see that complicated CD at diagnosis is associated with a distinct antibody profile years before the first clinical symptoms and is not simply the consequence of disease progression after diagnosis.”

He said the early profile could support earlier risk detection, distinguish disease subtypes and help identify patients who may benefit from more intensive therapy.

Those findings could therefore guide research into earlier risk detection and more intensive care, but they do not show that acting on a risk signal improves outcomes.

Could antibody profiling become a routine screening test?

Not yet. The researchers described the study as a proof of principle, meaning preliminary evidence that antibody profiling contains useful information about future IBD risk.

The team must test the signatures in independent, more diverse populations that researchers follow prospectively, meaning they observe participants over time rather than only reviewing stored samples.

They must also determine whether the test adds meaningful information beyond genetic information, clinical assessments and existing blood-based measurements.

An eventual laboratory test would need to be accurate, affordable and reproducible, and trials would need to show that using it improves patient outcomes.

Dr. Bourgonje named validation as the next priority.

“The next priority is to validate these signatures in independent, more diverse and prospectively followed populations.”

He said the team must “determine their added value alongside genetic, clinical and other blood-based markers.”

Researchers also need to “translate the findings into a reproducible and affordable targeted assay” and ultimately “demonstrate that using such a test actually leads to better clinical outcomes.”

For now, the signals remain research findings. If later studies confirm them, immune profiling could give clinicians years of advance notice and help distinguish disease subtypes, but its clinical value remains unproven.