Gallstones on a GLP-1: The Numbers, and Why They Matter More for Women

Why this one is different

Most of what women ask about these medicines runs into thin evidence. This is the exception. Gallbladder risk is quantified, it comes from a large body of randomised trials, it replicates outside those trials, there is a plausible mechanism with human data behind it — and it happens to land on a group already more prone to gallstones.

So this page is mostly arithmetic, then a short list of symptoms worth memorising, then the one lever that is more useful than "lose weight slower".

The numbers from the trials

A systematic review and meta-analysis in JAMA Internal Medicine pooled 76 randomised controlled trials covering 103,371 patients¹. Randomisation to a GLP-1 receptor agonist was associated with increased risk of gallbladder or biliary disease overall — a relative risk of 1.37 (95% CI 1.23–1.52). Broken down:

And the finding that matters most for anyone reading this site: in the trials conducted for weight loss specifically, the risk was substantially higher — a relative risk of 2.29 (95% CI 1.64–3.18)¹. Roughly double.

That pattern points at dose and at the amount of weight lost rather than at something peculiar to one molecule. Work on tirzepatide has examined the same gallbladder and biliary safety question².

It appears on the label too, though the number is smaller than the headlines suggest. In Wegovy's adult weight-reduction trials, cholelithiasis was reported by 1.6% of patients against 0.7% on placebo³. (The 4%-versus-0% figure that circulates comes from the trial in adolescents aged 12 and over, where it was 3.8% against 0% — a different population, and not the one this page is about.)

Does it hold up outside a trial?

This is the question that usually kills a promising signal, and here it survives it.

A 2026 retrospective cohort using the TriNetX network compared adults with type 2 diabetes who received a GLP-1 receptor agonist against matched controls — 39,140 patients in each group after propensity-score matching. At two years, the treated group had higher rates of gallstones, with an adjusted odds ratio of 1.44 (95% CI 1.24–1.65)⁴.

Trial data and real-world data agreeing on roughly the same effect size is about as good as this subject gets. It is why the rest of this page is written with more confidence than most.

The mechanism, which is unusually satisfying

Most side-effect pages have to say "we don't know why". Here we mostly do.

Your gallbladder empties when a hormone called cholecystokinin (CCK) tells it to. CCK is released when food — particularly fat — arrives in the small intestine, and it drives both bile secretion and gallbladder contraction.

Researchers infused GLP-1 or saline into ten healthy people and ten people with type 1 diabetes during and after a meal, and measured CCK directly. During the saline infusion, CCK rose after eating as it should. During the GLP-1 infusion, CCK secretion was significantly suppressed in both groups⁵.

A gallbladder that contracts less often is a gallbladder where bile sits longer. Bile that sits concentrates, and concentrated bile is where cholesterol crystals come out of solution. A broader review of how these drugs affect gastric, biliary and intestinal motility reaches the same place: the effects on the gallbladder are real, and they cut both ways — some benefit to postprandial triglycerides, some increase in biliary risk⁶.

So this is not a mysterious drug injury. It is the predictable consequence of turning down the signal that empties an organ.

How much of this is the drug, and how much is the losing?

Here is the context that reframes every number above, and it comes from a study run decades before any of these drugs existed.

In 1992, researchers followed 457 people entering a 520-calorie-a-day weight control programme, confirming by ultrasound that none of them had gallstones at the start. Among the 248 still enrolled at 16 weeks, 10.9% had formed gallstones⁷.

Read that against the 1.6% in Wegovy's adult trials. Losing weight fast — by any means, with no GLP-1 anywhere near it — is one of the most reliable ways to make a gallstone, and a 520-calorie diet did it to more than one in ten people in four months.

The same study found something else worth knowing. The risk factors that predict gallstones in the general population, including age, female sex and number of pregnancies, did not predict who formed stones during rapid weight loss. What predicted it was starting BMI, the amount of BMI lost, and triglyceride levels⁷.

That is the single most useful sentence on this page. During active weight loss, the strongest thing you control is not your history. It is how fast you are going.

But the weight loss is not the whole story, and the label says so outright. Wegovy's prescribing information puts it directly: substantial or rapid weight loss can increase the risk of cholelithiasis — however, the incidence of acute gallbladder disease was greater in Wegovy-treated patients than in placebo-treated patients even after accounting for the degree of weight loss³.

That sentence rules out the comfortable version of this story. There is a drug effect on top of the weight-loss effect — which is exactly what the mechanism above predicts, because a placebo arm that loses weight still has a gallbladder emptying normally, and a treated arm does not.

Why women, particularly

Gallstones are already more common in women. A review of gallbladder disease epidemiology lists the non-modifiable risks as ethnic background, increasing age, female sex and family history — and the modifiable ones as obesity, rapid weight loss and a sedentary lifestyle⁸. Oestrogen raises the cholesterol saturation of bile, which is the mechanistic reason pregnancy and oral contraceptives show up on the same lists.

So the question is not "does this drug cause gallstones" in the abstract. It is what a 1.27 to 2.29 relative risk does on top of a baseline that is already raised — and then what rapid loss adds on top of that. Three things stacking in the same direction, two of which you can influence.

The symptoms worth memorising

This is the practical core. Gallbladder pain is specific enough to recognise.

• Pain in the upper right abdomen, often after a fatty meal, that builds over minutes and then sits there — classically radiating to the right shoulder blade
• Pain lasting more than a few hours
• Fever or chills alongside that pain
• Nausea and vomiting with it
• Yellowing of the eyes or skin, or dark urine with pale stools — that suggests a blocked duct and is urgent

Upper-right pain plus fever, or any yellowing, is an emergency-department symptom rather than a message-your-provider one.

The reason to know this list is that nausea and abdominal discomfort are ordinary on these drugs, so there is a real risk of filing gallbladder pain under "the usual side effects" and waiting it out. The distinguishing features are the location, the duration and the fever.

What to actually do

Most advice here reduces to "lose weight more slowly", which is true and unsatisfying. There is a second lever, and it follows directly from the mechanism.

• Eat enough fat in a meal to make the gallbladder contract. Reviews of gallbladder motor function during very-low-calorie dieting identify gallbladder stasis as a driver of stone formation, caused by too little stimulation from low-fat intake — and put the threshold for efficient emptying at roughly 10 g of fat in a meal⁹. On a drug that suppresses appetite and a diet culture that treats fat as the enemy, three near-fat-free meals a day is an easy accident. It is also, mechanistically, a gallbladder left un-squeezed all day.
• Do not skip meals for long stretches. Fasting is on the same list of biliary risks⁸, and appetite suppression makes accidental fasting effortless.
• Ask about titration pace if you are losing very fast. The amount of BMI lost is the modifiable predictor with the best evidence behind it⁷.
• Say so if you have had gallstones before. It changes the conversation about pace rather than ruling anything out.

Where this sits on the evidence scale

High, and this is the strongest page on this site.

Seventy-six randomised trials pooled, consistent direction, confidence intervals that exclude no effect, a dose-response pattern that makes mechanistic sense, an independent real-world cohort of 39,140 matched pairs landing on a similar effect size, and a measured human mechanism explaining why. Very little in women's GLP-1 coverage is supported this well.

What would refine it is per-molecule and per-dose data in weight-management populations specifically rather than pooled across indications, and a trial testing whether deliberate fat intake at meals reduces stone formation during GLP-1 treatment — which, as far as the published research goes, nobody has run.