GLP-1 and Gallstones: SUSTAIN-6 Signal + Cholecystectomy Risk Evidence

Every GLP-1 prescribing information section 5 lists cholelithiasis as a known adverse reaction, and every major outcomes trial has captured the events. The 2022 JAMA Internal Medicine meta-analysis of 76 trials and 103,371 patients (He 2022^[3]) pooled the risk at a relative risk of 1.37 (95% CI 1.23–1.52) for any gallbladder or biliary disease, rising to roughly 2-fold for cholecystitis specifically. In SUSTAIN-6 (Marso 2016^[1]) and LEADER (Marso 2016^[2]), the absolute event rates were 1–2% over a median 2–3 years of follow-up. In STEP-1 (Wilding 2021^[5]) the cholelithiasis rate was 2.6% on semaglutide 2.4 mg vs 1.2% on placebo over 68 weeks. The risk is real but the absolute numbers are small, and the mechanism — rapid weight loss plus reduced gallbladder motility — is well-characterized, which is why a published prophylaxis protocol (ursodeoxycholic acid 600 mg/day, Sugerman 1995^[8]) already exists from the bariatric-surgery literature.

The honest summary

• The pooled GLP-1 signal is real. He 2022 (JAMA Internal Medicine^[3]) meta-analyzed 76 RCTs of 103,371 patients and found a 37% relative increase in any gallbladder or biliary disease (RR 1.37, 95% CI 1.23–1.52). Cholecystitis specifically ran about 2-fold higher.
• Absolute rates are still small. In STEP-1^[5] cholelithiasis was 2.6% vs 1.2% over 68 weeks; in SURMOUNT-1^[7] the rates were on the order of 1% per arm. Over a 2- to 3-year window in SUSTAIN-6^[1] and LEADER^[2], gallstone-related events accumulated to a few percentage points.
• Rapid weight loss is the dominant driver. Shiffman 1991^[10] documented 36% new gallstone formation in the first 6 months after gastric bypass — the GLP-1 numbers sit well below that, even on the highest doses, but the mechanism is the same.
• Ursodeoxycholic acid prophylaxis works in the bariatric setting. Sugerman 1995^[8] RCT showed UDCA 600 mg/day reduced new stone formation from 32% (placebo) to 2% over 6 months after gastric bypass; the Wudel 2002 pilot^[9] reproduced the direction. UDCA has not been formally trialed in GLP-1 patients, but the mechanism transfers.
• Symptom education is mandatory. Patients should know that post-prandial right-upper-quadrant pain radiating to the right shoulder blade, especially after a fatty meal, is the classic biliary-colic presentation and warrants a same-week ultrasound.

What the CV-outcomes trials actually captured

Two cardiovascular outcomes trials are the cleanest long-term gallstone data we have on GLP-1 agonists because both followed large cohorts for years on background diabetes care.

SUSTAIN-6 (Marso 2016 NEJM^[1]) randomized 3,297 patients with type 2 diabetes and high cardiovascular risk to semaglutide 0.5 mg or 1.0 mg weekly or placebo for a median 2.1 years. The published safety table listed cholelithiasis among the prespecified adverse events; the rate ran higher on semaglutide arms than placebo, in line with the broader meta-analytic signal.

LEADER (Marso 2016 NEJM^[2]) randomized 9,340 patients to liraglutide 1.8 mg daily or placebo with a median 3.8 years of follow-up. Acute gallstone disease events were captured and ran higher on liraglutide; the cholecystitis hospitalization rate was about 1.5x placebo across the He 2022 meta-analytic pooling^[3].

The STEP and SURMOUNT cholelithiasis percentages

STEP-1 (Wilding 2021 NEJM^[5]) randomized 1,961 adults with obesity to semaglutide 2.4 mg weekly or placebo for 68 weeks. The published safety table reported cholelithiasis in 2.6% of the semaglutide arm vs 1.2% of placebo — a doubling at the absolute number level. STEP-2 (Davies 2021 Lancet^[6]), which enrolled patients with type 2 diabetes, reported lower absolute rates but the directional signal was consistent.

SURMOUNT-1 (Jastreboff 2022 NEJM^[7]) randomized 2,539 adults to tirzepatide 5, 10, or 15 mg or placebo for 72 weeks. Cholelithiasis ran on the order of 1% across arms, with a slight numerical bump on active vs placebo — smaller in absolute terms than STEP-1, but the populations and protocols differ enough that direct cross-trial comparison is hazardous.

The mechanism: rapid weight loss plus reduced gallbladder motility

Two mechanisms compound during GLP-1 therapy. First, the rapid-weight-loss pathway: hepatic cholesterol mobilization outpaces bile-acid synthesis, bile becomes supersaturated with cholesterol, and cholesterol monohydrate crystals nucleate in the gallbladder. Shiffman 1991^[10] documented this in the post-gastric-bypass cohort with a 36% new-stone rate at 6 months. The faster the weight loss, the higher the rate — loss greater than 1.5 kg/week is the conventional threshold for high gallstone risk in the bariatric literature.

Second, the GLP-1-specific mechanism: GLP-1 receptor agonism delays gastric emptying and reduces gallbladder motility, producing biliary stasis and sludge formation that accelerates the rapid-weight-loss substrate. The He 2022 JAMA Internal Medicine meta-analysis^[3] specifically tested whether the gallbladder signal was fully accounted for by weight loss vs an independent drug effect; the pooled signal was stronger than weight-loss alone would predict, consistent with a GLP-1-specific motility contribution layered on top of the rapid-loss substrate. The Yang 2024 Obesity Reviews meta-analysis^[4] reached a similar conclusion across a broader weight-loss-intervention panel.

Who is at highest risk

• Female sex — the classic 4F mnemonic (female, forty, fertile, fat) still applies. Women carry roughly 2- to 3-fold the lifetime gallstone risk of men.
• Age over 40 — stone prevalence rises with age in both sexes.
• Higher baseline BMI — the higher the starting weight, the larger the absolute drop on a GLP-1 and the bigger the cholesterol-mobilization load.
• Native American or Hispanic ancestry— well-documented genetic susceptibility to cholesterol gallstones (Pima population data are the archetype).
• Rapid weight loss — greater than 1.5 kg/week sustained is the threshold above which gallstone formation rates climb steeply in the bariatric literature.
• Family history of gallstones in a first-degree relative.
• Concurrent estrogen exposure (oral contraceptives, HRT) and pregnancy increase cholesterol saturation.

Magnitude: annual incidence of biliary events

Cholecystitis vs cholelithiasis vs biliary colic vs gallstone pancreatitis

The four terms come up interchangeably in patient conversation but mean different things clinically. Cholelithiasis is the presence of gallstones, often asymptomatic; the majority of stones never cause a problem. Biliary colic is intermittent right-upper- quadrant pain caused by a stone transiently obstructing the cystic duct, classically after a fatty meal, lasting minutes to a few hours. Acute cholecystitis is sustained gallbladder inflammation from a stone impacted at the cystic duct — constant pain, fever, leukocytosis, positive Murphy sign, requiring antibiotics and (usually) cholecystectomy within 24–72 hours. Gallstone pancreatitis is the most dangerous variant: a stone migrates into the common bile duct, obstructs the pancreatic outflow, and triggers acute pancreatitis — managed with ERCP plus subsequent cholecystectomy.

Prevention: ursodeoxycholic acid and the rapid-loss threshold

The strongest prevention evidence comes from the bariatric literature. Sugerman 1995 (Am J Surg^[8]) randomized 233 post-gastric-bypass patients to ursodiol 300, 600, or 1,200 mg/day vs placebo. At 6 months, new gallstones formed in 32% of placebo, 13% of UDCA 300 mg, 2% of UDCA 600 mg, and 6% of UDCA 1,200 mg — identifying 600 mg/day as the apparent sweet spot. Wudel 2002 (J Surg Res^[9]) reproduced the direction in a smaller pilot. UDCA has not been formally trialed in GLP-1 patients, but the mechanism transfers: bile-acid pool shifting toward ursodiol reduces cholesterol saturation, which is the proximal biochemical trigger common to both rapid-loss settings.

The other lever is dose-ladder pacing. Loss greater than 1.5 kg/week is the bariatric-literature threshold above which gallstone formation rates climb. On Wegovy or Zepbound, that rate can be exceeded during the steepest weeks of the titration. Stretching the dose ladder by 4–8 weeks for high-risk patients (older women with higher baseline BMI, Hispanic or Native American ancestry, family history) is a reasonable risk-mitigation step. Maintaining some dietary fat intake (not a strict low-fat diet) preserves the physiological CCK signal that keeps the gallbladder emptying.

Workup and management

Right-upper-quadrant ultrasound is the first-line test — sensitivity and specificity both greater than 95% for cholelithiasis, no radiation, widely available. HIDA scan (cholescintigraphy) is reserved for functional questions when ultrasound is equivocal or when biliary dyskinesia is suspected. CT is generally not the right first test because non-calcified cholesterol stones are radiolucent.

Management depends on the clinical syndrome:

• Asymptomatic stones found incidentally: observe; cholecystectomy is not routinely indicated.
• Symptomatic biliary colic: elective laparoscopic cholecystectomy. Surgical cost typically runs $10,000–$25,000 depending on facility and insurance.
• Acute cholecystitis: admission, IV antibiotics, and cholecystectomy within 24–72 hours per Tokyo Guidelines.
• Gallstone pancreatitis: admission, ERCP if the stone has not passed, then cholecystectomy during the same hospitalization (or shortly after). The rechallenge calculus for GLP-1 patients with prior pancreatitis is covered separately.

The practical protocol

1. Screen high-risk patients at baseline. Consider a baseline right-upper-quadrant ultrasound for women over 40 with BMI over 35, Hispanic or Native American ancestry, or family history. Asymptomatic stones found at baseline change the conversation but do not automatically disqualify a patient from GLP-1 therapy.
2. Educate every patient on biliary-colic symptoms. Post-prandial right-upper-quadrant pain radiating to the right shoulder blade, especially after a fatty meal, lasting minutes to hours, warrants a same-week ultrasound.
3. Consider ursodiol prophylaxis at 600 mg/day during the rapid-loss phase (months 1–9) for high-risk patients, off-label by extrapolation from Sugerman 1995^[8]. There is no GLP-1-specific RCT, so this is a clinician judgment call.
4. Maintain some dietary fat intake. Strict low-fat diets reduce CCK release and worsen gallbladder stasis. A normal mediterranean-style fat intake (25–35% of calories) preserves the emptying signal.
5. Slow the dose ladder if loss exceeds 1.5 kg/week sustained. Stretching titration by 4–8 weeks reduces the cholesterol-mobilization load.
6. Hold the GLP-1 before any planned cholecystectomy per the ASA periprocedural guidance — covered in our perioperative GLP-1 hold guide.

Related research and tools

• GLP-1 with a history of pancreatitis — the rechallenge calculus when prior pancreatitis (including gallstone pancreatitis) is in the history.
• First 30 days on a GLP-1 — the early-titration symptoms (nausea, biliary discomfort, RUQ aches) and when to escalate.
• Perioperative GLP-1 hold — ASA guidance on stopping GLP-1 ahead of cholecystectomy or any elective surgery to manage aspiration risk from delayed gastric emptying.
• GLP-1 muscle-loss prevention protocol — the protein-and-resistance-training side of the rapid-weight-loss countermeasure package.
• GLP-1 gallbladder disease: FDA-label rates by drug — the companion piece with verbatim Section 5 cholelithiasis percentages from every FDA prescribing information.

Important disclaimer. This article is educational and does not constitute medical advice. Decisions about ursodiol prophylaxis, baseline imaging, and dose-ladder pacing should be made with the prescribing clinician. UDCA is not FDA-approved for GLP-1-associated gallstone prevention — the supporting evidence is extrapolated from post-bariatric-surgery RCTs. Patients with new or worsening right-upper-quadrant pain, fever, jaundice, or post-prandial symptoms should seek same-week evaluation; signs of acute cholecystitis or gallstone pancreatitis (constant severe pain, fever, vomiting) warrant emergency-department assessment. PMIDs were verified live against the PubMed E-utilities API on 2026-05-29.

Last verified: 2026-05-29. Next review: every 12 months, or sooner if a GLP-1-specific UDCA prophylaxis RCT or new pooled CV-outcomes safety analysis is published.