Bile has a job that is chemically awkward. It has to carry cholesterol — a molecule essentially insoluble in water — dissolved in a watery fluid, and it does so by packaging it with bile salts and phospholipids into mixed micelles and vesicles. That solubilisation has a limit. Exceed it, and cholesterol comes out of solution as crystals, which aggregate into stones.
This makes gallstones the third member of a trio. Gout is urate exceeding its solubility in joint fluid; kidney stones are urinary salts exceeding theirs in urine; gallstones are cholesterol exceeding its solubility in bile. Three diseases, three fluids, one physical principle — and, as this review will show, the same twist in all three: the driving force is present in healthy people too.
Gallstones are common. Prevalence is best established by ultrasound screening of populations [1] [2], with substantial ethnic variation in the United States [3]. Most are cholesterol stones in Western populations. Pigment stones form by a different route — through unconjugated bilirubin, whose enterohepatic cycling has been proposed as the mechanism [4] and whose generation by non-enzymic hydrolysis of bilirubin glucuronides is well described [5] [6] — and chronic haemolysis, cirrhosis and ileal Crohn's disease are the risks for black pigment stones [1]. The two types are not entirely separate: unconjugated bilirubin has also been proposed as a nucleating factor in cholesterol cholelithiasis [7], and whether stone type can be predicted from risk factors has been tested directly [8].
And the single most important clinical fact is that most gallstones do nothing. In a population screening study, 739 people aged 35–85 were examined by ultrasound; gallstones, sludge or cholesterolosis were found in 123, of whom 120 were followed for a median of 87 months. Fourteen were admitted and treated for gallstone-related symptoms or complications over that time [9]. The great majority needed nothing. That is the fact that decides management, and it is why an incidentally discovered gallstone is usually a reason to do nothing rather than a reason to operate.
The classic risk factors — the mnemonic is "fat, female, forty, fertile" — are real but shallow as an explanation. The mechanistic versions are: obesity and the metabolic syndrome, female sex and pregnancy, increasing age, rapid weight loss, and a sedentary lifestyle, with ethnicity and family history among the non-modifiable ones [1] [10]. Each has been examined on its own terms — the mechanisms by which exogenous oestrogen promotes stone formation in women [11], recreational physical activity against the risk of cholecystectomy [12], and diet and activity together in a population-based case-control study [13] — and the genetic epidemiology has been reviewed [14].
Centerpiece: supersaturation is necessary, and not sufficient
The chemistry is quantified as the cholesterol saturation index — the amount of cholesterol in bile divided by the maximum that bile's particular bile-salt and phospholipid content can dissolve. Below 1, cholesterol stays in solution. Above 1, it does not have to.
Several independent lines confirm that the saturation index is the operative quantity, and one of them is a natural experiment. ABCB4 encodes MDR3, the canalicular translocator that pumps phosphatidylcholine into bile. Mutations produce low biliary phospholipid and consequently a high cholesterol saturation index [15] — removing one of the two solubilising agents raises saturation and causes stone disease, which is what the model predicts and about as direct a test as human genetics offers. It is not confined to the severe paediatric syndrome: MDR3 defects have been found in adults with symptomatic intrahepatic and gallbladder cholesterol lithiasis [16] [17], and the resulting entity is named for the mechanism — low-phospholipid-associated cholelithiasis [18]. Pregnancy provides another: the lithogenic index of fasting hepatic and gallbladder bile rose during the second and third trimesters in healthy women, while the phase of the ovulatory cycle had no effect [19] — so "fertile" in the mnemonic resolves to a measurable shift in bile chemistry. Rapid weight loss raises saturation further because the liver secretes more cholesterol into bile [20] [10] [21].
But here is the finding that makes this disease interesting, and it is the same shape as the one in the kidney-stones review. Normal human gallbladder bile is routinely supersaturated with cholesterol and does not form stones. What distinguishes people who make stones is not that their bile crosses the solubility limit — most people's does — but how long their bile can hold cholesterol in suspension above it.
That was established by an experiment with an unusually clean logical structure [22]:
1. The onset time for cholesterol crystal nucleation in supersaturated normal gallbladder bile is consistently longer than in bile from patients with gallstone disease. 2. Model bile made up to the identical cholesterol saturation index and molar lipid composition nucleated much faster — so the metastability is not explained by how much of each lipid is present. 3. Normal biles supplemented to a standard supersaturated composition still held out three- to fifteen-fold longer than the matched model bile. 4. The total lipid extract of those same normal biles, reconstituted to the same composition, did not differ from the model solution at all — so whatever protects is not in the lipid fraction. 5. Gel chromatography then localised inhibitory activity to a protein-containing fraction, which raised nucleation time about 1.5-fold.
Step 4 is the one that closes the argument. Take the protective bile, extract its lipids, put them back at the same concentrations, and the protection is gone — which rules out the lipids as the explanation and forces the inhibitor to be something else.
The check the arithmetic was never given. The paper reports the intact-bile effect (three- to fifteen-fold) and the isolated-fraction effect (about 1.5-fold) and never divides them. Working with the excess over the model-bile reference, one pooled chromatography fraction recovers (1.5 − 1)/(3 − 1) = 25 percent of the intact effect at best, and (1.5 − 1)/(15 − 1) = 3.6 percent at worst. A single fraction reproduces at most a quarter of what whole bile does. The inhibitory activity is therefore not one protein — it is a system, and the paper's own numbers say so without saying so.
The five-fold span is itself informative. Normal biles standardised to the same composition still varied three- to fifteen-fold in nucleation time. With the driving force held constant by design, that variation has to be in the inhibitors. Which means susceptibility to gallstones is substantially a matter of how good your bile is at suspending cholesterol it has already dissolved too much of.
The teaching point, and the trio. Gallstones are a supersaturation problem — but so is everyone's bile. Gout, kidney stones and gallstones all share this: the thermodynamics say crystals can form in most people, and what decides whether they do is kinetics and inhibitors. Urine is routinely supersaturated with calcium oxalate and most people never form a stone; bile is routinely supersaturated with cholesterol and most people never form one either.
Three honest limits. The foundational phase diagram is not in this substrate — the tool's search floor is 1980 and Admirand and Small's triangular cholesterol/bile-salt/lecithin diagram dates from 1968, so this review describes the solubility limit from later sources that use it rather than citing the original. The width of the metastable zone is a kinetic property rather than a fixed saturation value, and the boundary on the figure is schematic for that reason. And nucleation in vitro is not stone formation in vivo: gallbladder emptying, mucus and time all intervene, and mucus hypersecretion has its own role in the evolution of cholesterol gallstones [23].
Pillar 1: measurement and diagnosis
Ultrasound
Ultrasound is the imaging standard and the basis of every prevalence estimate [1]. It is quick, radiation-free, and good at what it is asked to do: see stones in the gallbladder.
The syndromes, which matter more than the stones
Biliary colic is the uncomplicated symptom: a stone transiently obstructing the gallbladder outlet, producing severe constant right-upper-quadrant or epigastric pain that builds and then settles over hours. Despite the name it is not colicky.
Acute cholecystitis is sustained obstruction with inflammation and infection of the gallbladder wall — persistent pain, fever, and tenderness. The Tokyo Guidelines set out its diagnostic criteria and severity grading and the management flowchart [24] [25].
Choledocholithiasis — a stone in the common bile duct — obstructs the drainage of the liver, producing jaundice and abnormal liver enzymes, and risks cholangitis, infection of an obstructed biliary tree, which is a medical emergency. Guidelines exist for its management [26] [27] [28]. Liver function tests are the usual first signal and have been assessed for exactly that job — against ultrasound for diagnosing duct stones [29] and as predictors of duct stones in acute calculous cholecystitis [30] — which is why the bloods are ordered even though the stones are found by imaging.
Gallstone pancreatitis occurs when a stone obstructs the pancreatic outflow at the ampulla. It is one of the two commonest causes of acute pancreatitis, and its management sits inside the general acute-pancreatitis guidelines [31] [32] [33].
Gallbladder cancer is uncommon in developed countries but its risk factors overlap heavily with gallstone disease [34] [35] [36] [37]. It is not a reason to remove silent stones, and the epidemiology reviews are explicit that the numbers are small in most populations.
Pillar 2: treatment
Silent stones: do nothing
Given that 14 of 120 people with incidentally found stones needed treatment over a median of seven years [9], and given that cholecystectomy has real if low complication rates, observation is the correct default for asymptomatic gallstones [38] [39]. The exceptions are specific, not general.
Symptomatic stones: take the gallbladder out
Laparoscopic cholecystectomy is the definitive treatment and among the commonest operations performed. It has been compared with the open operation in symptomatic cholecystolithiasis [40], and the reasons it displaced open surgery — shorter stay, faster recovery — are why gallstone surgery became as accessible as it is.
Removing the gallbladder rather than the stones is worth a word of explanation to patients: the gallbladder concentrates bile and is where stones form, so leaving it behind means leaving the factory. Roughly one operation in twenty or so is converted to open surgery, and the factors determining conversion have been characterised [41]. Chronic post-surgical pain is a recognised outcome of surgery generally and is worth counting honestly in the consent conversation [42].
The complication that defines this operation is bile duct injury. It is uncommon and it is serious, and it became a defined problem of the laparoscopic era [43]. The response has been systematic: intraoperative cholangiography and early detection have been studied for their effect on outcome [44], the World Society of Emergency Surgery has issued guidelines for detection and management [45], and a multi-society consensus exists on how to do the operation safely [46]. A very common operation with a rare catastrophic complication is precisely the setting in which technique standardisation earns its keep.
Duct stones: ERCP
Stones in the bile duct are removed endoscopically, by cannulating the ampulla and performing sphincterotomy [47] [28] [27]. The procedure has real risks and they are well quantified — complications of endoscopic biliary sphincterotomy were characterised in a landmark prospective series [48], and ERCP-related adverse events have European guidance devoted to them [49] [50]. Pancreatitis is the commonest.
For severe gallstone pancreatitis and cholangitis, urgent ERCP with sphincterotomy was tested in a controlled trial as far back as 1988 [51] — one of the earlier demonstrations that relieving the obstruction promptly changes outcome.
Dissolution therapy
Ursodeoxycholic acid lowers the cholesterol saturation of bile, which is the model above run backwards. As a treatment for existing stones it is slow, works only for small radiolucent cholesterol stones in a functioning gallbladder, and stones recur when it stops — so it is reserved for people who cannot have surgery.
Its clearest role is prevention in the rapid-weight-loss setting, and the trial is instructive. Sixty-eight obese subjects without gallstones on a 520 kcal/day programme were randomised double-blind to ursodeoxycholic acid, aspirin or placebo for up to 16 weeks, with ultrasound and duodenal bile sampling. No gallstones and no cholesterol crystals formed in the ursodeoxycholic acid group; in the placebo group gallstones formed in five and cholesterol crystals in six; and the bile saturation index rose in the placebo group by week four [52].
Note what that trial measured: stones, crystals, and the saturation index — the outcome, the intermediate and the mechanism, in the same subjects. It is the model of this review tested end to end.
The scale of the problem it addresses is substantial: new gallstones reach 10–12 percent after 8–16 weeks of a low-calorie diet and more than 30 percent within 12–18 months after gastric bypass, with about one-third of those stones symptomatic [20]. Identified risk factors during weight loss include losing more than 24 percent of initial body weight, losing faster than 1.5 kg per week, a very-low-calorie diet with no fat, a long overnight fast, and high serum triglycerides — and saturation returns toward normal once weight stabilises, sometimes allowing spontaneous dissolution [20].
That list is the supersaturation model as a set of practical instructions. Some fat in the diet keeps the gallbladder emptying; a shorter overnight fast does the same; slower weight loss limits the cholesterol flux into bile.
Pillar 3: what is unresolved
Who among the asymptomatic will become symptomatic. The population data give a rate but not a prediction [9] [38]. Identifying the minority who will develop complications is the stated goal of that literature and remains unmet.
The inhibitor system. Whatever suspends supersaturated bile is multiple, protein-containing and incompletely characterised [22], and mucus hypersecretion pulls in the other direction [23]. A disease whose susceptibility is governed by an uncharacterised inhibitor system is a disease with no preventive drug beyond bile-acid manipulation.
Metabolic and microbial links. Gallstones travel with obesity and the metabolic syndrome, and the rising prevalence of both predicts rising gallstone frequency [1] [10]. Gut microbiota have been implicated in promoting cholesterol gallstone formation by modulating bile acid composition and biliary cholesterol secretion [21] — mechanistically plausible and not yet actionable.
Gallbladder-preserving approaches. Whether stones can be removed while keeping a functioning gallbladder is periodically revisited; the objection has always been that the organ that made stones once will make them again.
Dig deeper in lmmol
Kidney stones is the direct sibling and should be read alongside this one: the same physical principle in a different fluid, and — strikingly — the same structural finding, that the biological fluid is routinely supersaturated in healthy people and what differs in disease is the inhibitors and the kinetics rather than the driving force. Gout completes the trio with urate, and adds the crystal-inflammation half that gallstones largely lack. Obesity is the dominant modifiable risk, and unusually it is risky in both directions here — carrying the weight raises stone risk, and losing it rapidly raises it further [20] [1]. Fatty liver disease shares the metabolic-syndrome substrate and the organ [10]. And pancreatic cancer sits at the far end of the same anatomical corridor: gallstone pancreatitis and pancreatic malignancy both present through obstruction of the shared biliary-pancreatic outlet, which is why jaundice in this territory always demands an explanation [31]. The full collection is at health.