A millionfold gradient
held by one pump, behind one layer of mucus.

1. Histamine is the main stimulus to acid, not gastrin. Gastrin’s dominant action is indirect: it releases histamine from enterochromaffin-like cells, which acts on the parietal H₂ receptor. Three stimulatory receptors, one inhibitory.
2. The H⁺/K⁺ ATPase is the final common pathway. Every stimulus, and every drug, converges on it. Bicarbonate leaves the basolateral membrane in exchange for chloride, which is the alkaline tide.
3. 30, 60, 10. Cephalic, gastric and intestinal phases contribute roughly those percentages of the acid secreted for a meal.
4. Solids and liquids empty by different shapes. Solids have a 30-minute lag then fall approximately linearly, half-time about 2 hours. Clear liquids fall exponentially from time zero, half-time about 20 minutes. Drawing both as exponentials is a common and invisible error.
5. Receptive relaxation is vagovagal; accommodation is local. Both are mediated by nitric oxide and vasoactive intestinal peptide from non-adrenergic non-cholinergic neurones, and both are why a litre of food raises intragastric pressure very little.
6. The duodenum controls gastric emptying, not the stomach. Four duodenal signals slow it: distension, acid, fat and protein breakdown products, and hyperosmolarity. Fat is the most powerful, acting through cholecystokinin.

The stomach has four regions — cardia, fundus, body and antrum — and three functions: to store a large meal, to mix and break it down into chyme, and to release it slowly into the duodenum. It stores up to 1.5 to 2 litres. Functionally it divides in two rather than four: a proximal stomach (fundus and body) that is the reservoir and produces smooth tonic contractions, and a distal stomach (antrum) that grinds, mixes and sieves with high-amplitude contractions. Particles must be reduced to 1 to 2 mm before the pylorus will let them through.

View full size

Two defects in the supplied artwork, neither affecting the mechanism. The secretory canaliculus is drawn but not labelled — it is the deep apical invagination lined with microvilli into which the pump secretes. And two labels render the potassium ion as “K*” rather than K⁺, in “K* channel” and “H⁺/K* ATPase”, although the hydrogen superscript in the same label is correct.

A single parietal cell is drawn in section, with the gastric lumen above and a capillary below. Its apical membrane is thrown into microvilli around a secretory canaliculus. In that apical membrane sit an H+/K+ ATPase, which exports hydrogen into the lumen in exchange for potassium, a potassium channel that returns potassium to the lumen to supply the pump, and a chloride channel through which chloride diffuses out, so that hydrochloric acid is the net luminal product. Tubulovesicles carrying more pump molecules are shown fusing with the apical membrane. In the cytoplasm, carbonic anhydrase converts carbon dioxide and water into hydrogen and bicarbonate ions. At the basolateral membrane, a chloride-bicarbonate exchanger sends bicarbonate into the capillary, labelled the alkaline tide, and takes up chloride; a sodium-potassium ATPase maintains the sodium and potassium gradients. Three stimulatory receptors sit in the basolateral membrane: a gastrin CCK2 receptor reached by gastrin arriving in the capillary, a histamine H2 receptor reached from an adjacent enterochromaffin-like cell, and a muscarinic M3 receptor reached from a vagal nerve terminal. A D cell at the left releases somatostatin, drawn as an inhibitory bar onto the cell.

Close

The mechanism runs in four steps. Carbon dioxide diffuses into the cell from the blood. Carbonic anhydrase converts it with water to carbonic acid, which dissociates into H⁺ and HCO₃⁻. At the apical membrane the H+/K+ ATPase — the proton pump — exports hydrogen into the secretory canaliculus in exchange for potassium, against a concentration gradient of roughly three million to one, which is why the cell is packed with mitochondria. At the basolateral membrane bicarbonate is exchanged for chloride; the bicarbonate enters the blood and the chloride crosses the cell and leaves through an apical chloride channel, so the net luminal product is hydrochloric acid and the net blood product is sodium bicarbonate.

The negative feedback matters as much as the stimulation. Once acidic chyme reaches the duodenum, secretin is released and inhibits gastrin release from the G cells; gastric inhibitory peptide promotes somatostatin release, which inhibits both the G cell and the parietal cell directly; and the enterogastric reflex, carried by the enteric nervous system and by the extrinsic sympathetic and vagal supply, reduces both secretion and motility. That matching of delivery to the duodenum’s handling capacity is the single most important control in the section, and it is the same set of signals that will slow emptying in section 08.

The gastric mucosal barrier is not one thing. It has a pre-epithelial, an epithelial and a subepithelial component, and clinically useful damage requires more than one of them to fail.

About 2000 mL of gastric juice is produced a day. Its composition varies with the rate of secretion, because it is really two secretions mixed: a non-parietal component resembling plasma, and a parietal component of essentially pure hydrochloric acid at about 150 mmol/L. At high secretion rates the mixture approaches pure parietal secretion. That is why prolonged vomiting or nasogastric aspiration produces a hypochloraemic, hypokalaemic metabolic alkalosis with hypovolaemia: the fluid lost is rich in hydrogen, chloride and potassium, and its bicarbonate counterpart has already gone into the blood.

The distal stomach does the opposite work. Its basal electrical rhythm is set by interstitial cells of Cajal on the greater curvature at about three cycles per minute, and that pacemaker drives the whole distal stomach at its own rate. Slow waves alone do not reach threshold: they determine when a contraction may occur, and vagal activity, gastrin and acetylcholine determine whether one does, by raising the plateau of the slow wave until it does reach threshold.

At two hours — the clear-fluid fasting interval — about 1.6% of a clear liquid remains, against about 50% of a solid meal. That difference, not a difference in risk of reflux, is why liquids may be taken so much closer to induction than solids.

Fasting gastric juice has a pH of about 1 to 1.5 and the stomach secretes about 2000 mL a day. A fasted stomach is not an empty one: it always contains some residual volume of secretion, swallowed saliva and refluxed duodenal content, and that residual volume is what is available to be aspirated at induction.

The injury itself is Mendelson's syndrome: a chemical pneumonitis from the aspiration of acidic gastric contents. The mechanism is direct acid injury to the alveolar epithelium and capillary endothelium, producing an inflammatory exudate, surfactant loss, atelectasis and a large intrapulmonary shunt. Particulate material, even at a neutral pH, causes obstruction and a foreign-body reaction instead — which is why particulate antacids are avoided before induction and non-particulate sodium citrate is used: raising the pH is no help if the agent that raises it is itself damaging when aspirated.

A full stomach is a physiological state rather than a history. Treat the stomach as full, whatever the fasting interval, when emptying has been arrested or the volume is being added to: pain, trauma, opioids, an acute abdomen, bowel obstruction — in which the stomach is filling from below rather than emptying — labour, autonomic neuropathy, gastroparesis, raised intracranial pressure, and critical illness. In each of these the curves in section 08 simply do not apply, because the mechanism that generates them has been interrupted.