A place for everything and everything in its place.
Except for the appendix. That can go.
For most of the last century, medicine classified it as vestigial, a biological relic of an earlier digestive system, present but purposeless. Removing it became one of the most routine surgeries in the world. More than 300,000 people in North America have the procedure every year.1
A new study suggests we may have been removing more than we thought.
Researchers at the University of Technology Sydney and Massachusetts General Hospital/Harvard Medical School trained an AI model on data from nearly 10,000 people, analyzing more than 120 everyday factors (diet, gut bacteria, medical history, lifestyle) to identify which were most strongly associated with Alzheimer's risk.2 What it surfaced was unexpected: among the strongest predictors of Alzheimer's risk in the entire dataset was a history of appendix removal.
The researchers' explanation is careful but striking. The appendix, they suggest, may function as a reservoir for beneficial gut bacteria, a backup system the microbiome draws on to replenish itself after illness, infection, or a course of antibiotics.
🧠The mechanism runs through the gut-brain axis, a communication network between the digestive system and the brain that has become one of the more active areas of neuroscience research in recent years.
Remove it in your twenties, and that recovery mechanism is gone. Over decades, the disruption may compound, leaving the gut progressively less able to maintain the conditions the brain depends on.
Alzheimer's may not begin in the brain at all, but in the gut.
Why This Matters
The appendix finding is associative, not causal. The researchers are explicit: the mechanism they describe is speculative, and the study requires validation in larger cohorts. A history of appendectomy may also serve as a proxy for other gut-disrupting factors, repeated antibiotic use or infection history, that compound over time. The direction of the evidence is clear; the precise mechanism is still being mapped.
The AI model points toward a different kind of screening. Achieving over 80% accuracy using only non-invasive everyday factors, comparable to expensive biomarker-based approaches.2 The framework points toward a future where community-level screening for Alzheimer's, and potentially other diseases, could be delivered at a fraction of current cost and before symptoms appear. The model requires longitudinal validation, but the approach itself represents a template: existing data, accessible inputs, no lab required.
The gut-brain axis connects microbiome disruption directly to neurodegeneration. Beneficial bacteria in the gut produce short-chain fatty acids that maintain the gut's protective barrier and suppress neuroinflammation.2 In participants with Alzheimer's disease, those bacteria were significantly depleted, replaced by a more inflammatory environment capable of sending damaging signals to the brain over time.2
There are answers to questions we haven’t thought to ask yet, sitting in data we already have.
This finding came not from a new clinical trial, but from data collected more than a decade ago,3 because AI doesn't generate new evidence; it reads what we already collected, at a scale and resolution we couldn't manage before. But the same tool that uncovers something new also places what we thought we knew back into question. The appendix finding does both at once: a potential Alzheimer's connection nobody had seen, and decades of confident dismissal undone by the same model. That combination will play out beyond medicine. Nutrition science, criminal justice, economics — every field that drew its conclusions before this capability existed is now sitting on findings that may not be final. The data hasn't changed. What's still buried in it might.
Sources
1 Appendectomy incidence in North America: Sauerland et al., Annals of Surgery, 2017. https://journals.lww.com/annalsofsurgery/fulltext/2017/08000/the_global_incidence_of_appendicitis__a_systematic.8.aspx
2 Primary study: Jabeen, Karimi, Zomorrodi, Khalilpour. "Multi-modal machine learning and gut microbiome pathway analysis for Alzheimer's risk prediction." Alzheimer's & Dementia: Diagnosis, Assessment & Disease Monitoring, 2026. https://dx.doi.org/10.1002/dad2.70340
3 Cohort data source: American Gut Project, launched 2012. McDonald et al., "American Gut: an Open Platform for Citizen Science Microbiome Research." mSystems, 2018. https://pmc.ncbi.nlm.nih.gov/articles/PMC5954204/ (EBI accession PRJEB11419)