Compound

L-citrulline

L-citrulline, the non-protein amino acid and arginine precursor. FOUR BOUNDARIES. (1) CITRULLINE MALATE IS THIS NODE, WITH THE FORM RECORDED. It is the commonest supplement form, usually 2:1 citrulline to malate, and the citrulline is what the trials are testing. The same decision as glucosamine sulfate versus hydrochloride and as the EPA ester forms: splitting would divide one literature across nodes that tie on every paper. Write the form and ratio into dose_regimen as the paper words it, because whether the malate contributes is a live dispute in this field and recording it is what lets that be read off the graph rather than hidden inside it. Do NOT decline citrulline malate as a combination on the malate alone. (2) WATERMELON IS NOT CITRULLINE. Watermelon juice, rind extract and flesh carry citrulline alongside lycopene, sugars, potassium and other amino acids, and a watermelon arm against a comparator lacking it is a combination exposure — the intervention contains more than the compound and no analysis separates them. Record what was given in co_exposure. MIXED POOLS: A REVIEW THAT POOLS FORMS THIS NODE COVERS WITH FORMS IT DOES NOT IS NOT AN ESTIMATE OF THIS COMPOUND. This is the rule the first citrulline batch needed and did not have. PMID 33308806 pooled seven trials of which three delivered citrulline as watermelon juice, and its dose_regimen was recorded as "free L-citrulline 3-6 g" — those exact figures, 3.3 g and 6 g, ARE the watermelon arms, and not one pooled trial gave free citrulline at all. The review says so only in its table; its Results prose does not. Two verification layers found it independently and both claims were refused. So: count the forms from the review TABLE, never from its prose. Where every pooled arm is a form this node covers (free citrulline, citrulline malate), extract and record the mix in dose_regimen. Where any pooled arm is watermelon or another food matrix, the pooled estimate is a combination exposure — decline it and name the split in co_exposure — UNLESS the review reports a subgroup restricted to the covered forms, which is then the estimate to take. And if you extract from a mixed review on any other ground, dose_regimen must state what the pool ACTUALLY contained. Writing the covered form alone is a false statement about what was administered, and that field renders verbatim to readers. (3) PLASMA CITRULLINE MEASURED IS NOT CITRULLINE GIVEN. Citrulline is an established biomarker of enterocyte mass and absorptive capacity, used in short bowel syndrome, mucositis and coeliac disease. Where the exposure is a diet, a drug or a disease and citrulline is the readout, the claim runs backwards through the graph: subject_is_the_outcome. (4) ARGININE IS A DIFFERENT NODE. L-arginine is CHEBI:16467. Citrulline given WITH arginine only in the treatment arm is a combination exposure; citrulline against arginine head-to-head is comparator_not_absence; and a trial dosing arginine while measuring citrulline is a subject mismatch. Note that citrulline raising plasma arginine is a mechanism, not a licence to merge the nodes.