Korean national data keeps finding new diseases after gallbladder removal. When those diseases appear tells a different story

South Korea removes a great many gallbladders — the age-standardised rate more than tripled between 2003 and 2017 — and it records what happens to every one of those patients afterwards in a single national insurance database. That combination has produced a steady output of studies with a common shape: people who had a cholecystectomy went on to be diagnosed with something else more often than people who did not.

The list now runs to cancer, type 2 diabetes, chronic kidney disease, metabolic syndrome, fatty liver disease and depression. Read as a sequence of headlines it suggests that removing a gallbladder is quietly dangerous. Read with attention to when the extra diagnoses appear, it suggests something considerably less alarming.

What the studies report#

Each of these is a large, competently conducted matched cohort study, and each was published in a peer-reviewed journal. Taken individually, none is obviously wrong. Taken together, they raise the question of why one operation should independently cause conditions as unrelated as bile duct cancer, kidney failure and depression.

The timing is the tell#

The largest of these studies was run by Korea's National Cancer Center and published in Scientific Reports in January 2025. It compared 715,872 people who underwent cholecystectomy between 2004 and 2020 against 1,431,728 matched controls, and it did something the others largely did not: it reported when the excess cancers were diagnosed.

A cause that operates through the body's physiology — bile draining continuously into the intestine instead of in pulses, say — would be expected to do its damage slowly, so that risk accumulates the longer a person lives without a gallbladder. What the data show is the opposite: a spike in the years immediately after surgery, then convergence with everyone else.

That is the signature of looking rather than of causing. A person who has an operation is scanned, has blood drawn, is followed up in a clinic and is under the eye of a hepatobiliary service. A matched control who never had gallstones is not. Cancers that were already present and would have surfaced later surface sooner, and cancers of the bile ducts — the two largest effects in the study, at rate ratios of 1.92 and 1.78 — are precisely the ones that imaging of the biliary tree would find.

The depression cohort shows the same pattern in a different organ system. Risk was clearly raised in the first three years after surgery and was not statistically distinguishable from the control group after that, and there was no association with suicide in any period. Recovering from an abdominal operation is a plausible reason to be low for a year. Not having a gallbladder for a decade is a less plausible one.

The problem no adjustment fixes#

There is a second reason to be cautious that applies to all of these studies at once, and it is not a defect in any of them. It is a limit of the design.

People do not receive a cholecystectomy at random. They receive one because they have gallstone disease, and gallstone disease keeps company with obesity, insulin resistance and metabolic syndrome — the same conditions that precede type 2 diabetes and chronic kidney disease. Matching on age and sex does not make two such groups alike. Adjusting for recorded comorbidities and health behaviours, as the diabetes and kidney studies did, narrows the gap without closing it, because what is adjusted for is what happens to be coded in an insurance claim.

1. Confounding by indication: the surgical group was metabolically different before anyone operated, and the reason for the operation is itself a risk factor for the outcomes being counted.
2. Detection bias: the surgical group receives imaging, laboratory tests and specialist follow-up that the control group does not, which produces more diagnoses of everything without producing more disease.
3. Shared data and shared design: these are largely the same database analysed repeatedly with the same matched-cohort approach, so agreement between them is weaker evidence than it looks.

What this does and does not license#

None of the above shows that the reported associations are spurious. Cholecystectomy does change bile physiology, and a genuine long-term effect on some of these outcomes remains possible; the diabetes and kidney findings in particular have not been explained away and their authors call for work on the mechanism.

What the evidence does not support is the reading these studies most often receive in the retelling: that having a gallbladder removed makes a person meaningfully more likely to get cancer. The largest study of that question found the excess concentrated in the years when a newly operated patient is being watched most closely, and gone by the time they are not.