Narrower indications

• Postprandial hyperglycaemia — 3 graded rows from 3 compounds checked

Compounds with a graded row

Ordered by rung, strongest first. The rung rates the study design; the chip beside it says what the study measured. Both are needed.

1. Cagrilintide

   human RCT Direct outcome In registered trials

   Who was studied

   adults with type 2 diabetes and BMI 27 kg/m2 or higher on metformin with or without an SGLT2 inhibitor, n=92 randomised (cagrilintide 2.4 mg monotherapy arm n=30), 32 weeks; 59 (64%) male, mean age 58 years

   No single-dose or meal-test study of cagrilintide is on file: the only glucose data located come from 32 weeks of weekly dosing in a phase 2 trial (NCT04982575) whose primary endpoint was HbA1c, with fasting plasma glucose and continuous glucose monitoring as secondary measures. The abstract reports the fasting glucose and time-in-range figures for the cagrilintide-plus-semaglutide arm rather than for cagrilintide alone; the cagrilintide monotherapy arm changed HbA1c by -0.9 percentage points. No level 2 or 3 hypoglycaemia was reported in the trial. Cagrilintide is not approved by any regulator.

   Source PubMed 1 primary source read for this row

2. Lixisenatide

   human RCT Direct outcome Approved for this condition

   Who was studied

   adults with type 2 diabetes on optimised insulin glargine with or without metformin, n=142, multicentre randomised open-label three-arm, 8 weeks; primary measure was the 4-hour postprandial plasma glucose AUC after a standardised solid breakfast

   Lixisenatide 20 mcg once daily reduced the 4-hour postprandial glucose AUC more than liraglutide 1.2 mg (-108.3 mg/dL.h difference) or 1.8 mg (-83.0 mg/dL.h), with greater slowing of gastric emptying, while HbA1c ended similar across arms (about 6.1-6.2%). Symptomatic hypoglycaemia was more frequent with lixisenatide than with liraglutide in this insulin-treated population — the abstract does not quantify the rates. The measurement is the meal excursion in hours after 8 weeks of dosing, not long-term control.

   Source PubMed 1 primary source read for this row

3. Pramlintide

   human RCT Direct outcome Approved for this condition

   Who was studied

   adults with type 2 diabetes, n=24 in a single-blind randomised crossover (12 on exogenous insulin, 12 managed with diet and/or oral agents); 5-hour intravenous pramlintide or placebo infusion with a test meal one hour in, glucose measured over the following 4 hours

   Postprandial glucose, insulin, C-peptide and lactate were measured over four hours after a test meal. Glucose fell significantly only in the insulin-treated group; in the diet/oral-agent group insulin, C-peptide and lactate fell but mean glucose did not, with individual glucose reductions tracking baseline glycated haemoglobin. This is an acute infusion experiment on the analogue pramlintide, not native amylin; the FDA approval is as a mealtime adjunct to insulin, and the label carries a boxed warning for severe insulin-induced hypoglycaemia.

   Source PubMed 1 primary source read for this row

4. Semaglutide

   human RCT Direct outcome Approved for this condition

   Who was studied

   adults with type 2 diabetes, n=15 enrolled (13 completed), mean age 58.2 years, 86.7% male, mean HbA1c 6.9%, diabetes duration 3.1 years, BMI 30.8 kg/m2; randomised double-blind placebo-controlled crossover, two 12-week periods, glucose measured across a 5-hour standardised breakfast and an 8-hour fat-rich breakfast

   Oral semaglutide escalated to 14 mg once daily reduced the postprandial glucose area under the curve over five hours by 29% against placebo (treatment ratio 0.71), with a similar reduction after a fat-rich meal and delayed gastric emptying. The measurement here is the meal glucose excursion in hours, not HbA1c over months; the compound was dosed for 12 weeks beforehand, so this is not a single-dose result. Semaglutide is FDA-approved for glycaemic control in type 2 diabetes; the trial did not report hypoglycaemia as an outcome.

   Source PubMed 1 primary source read for this row

5. Tirzepatide

   human RCT Surrogate marker Approved for this condition

   Who was studied

   adults with type 2 diabetes on metformin at two German centres, n=117 (tirzepatide 15 mg n=45, semaglutide 1 mg n=44, placebo n=28), 28 weeks of weekly subcutaneous dosing; hyperinsulinaemic-euglycaemic and hyperglycaemic clamps plus mixed-meal testing

   A phase 1 clamp trial measuring insulin secretion and insulin sensitivity rather than glucose itself: clamp disposition index improved against placebo (treatment difference 1.92) and against semaglutide (0.84), with reduced glucose excursions on meal testing. It is graded surrogate_marker because the clamp fixes glucose by design and reports secretion and sensitivity, which stand in for the capacity to control glucose rather than being a glucose result. Nausea occurred in 24% of tirzepatide recipients; tirzepatide is FDA-approved for glycaemic control in type 2 diabetes.

   Source PubMed 1 primary source read for this row

6. Exendin-4

   human case series Direct outcome Not approved for this condition

   Who was studied

   subjects with type 2 diabetes; study A n=24 dosed twice daily with meals for 5 days, study B n=13 dosed once after an overnight fast with 8 hours of observation; sex, HbA1c range and background therapy not reported in the abstract

   Synthetic exendin-4 (AC2993) lowered postprandial and fasting plasma glucose over hours to days in two early clinical studies. Both had placebo arms but the abstract does not state that allocation was randomised, so the rung is set below human_rct. The measurements are acute — plasma glucose, insulin and glucagon over a single 8-hour observation and across 5 days of mealtime dosing; nothing on file measured a durable glycaemic outcome for the native venom peptide. The marketed synthetic form of this same molecule is exenatide.

   Source PubMed 1 primary source read for this row

7. GLP-1 (7-36)

   human case series Direct outcome Not approved for this condition

   Who was studied

   patients with type 2 diabetes in poor metabolic control on diet and sulphonylurea, some with added metformin or acarbose, n=10, HbA1c 11.6 +/- 1.7%; single 4-hour intravenous infusion, within-subject placebo comparison

   Native GLP-1 (7-36 amide) infused intravenously at 1.2 pmol/kg/min brought fasting plasma glucose down to normal fasting concentrations within four hours, with insulin and C-peptide rising and glucagon falling. This is an acute physiology experiment on the native hormone over a single session, not a treatment: nothing durable was measured, and the native peptide is not administered as therapy in any regulator-approved form.

   Source PubMed 1 primary source read for this row

8. Amylin

   human case series Mechanistic only Not approved for this condition

   Who was studied

   six volunteers, native islet amyloid polypeptide (amylin) infused intravenously at 25 and 50 pmol/kg/min against saline, each session followed by an intravenous glucose bolus of 0.5 g/kg; health status of the volunteers not further characterised in the abstract

   Recorded, and not evidence for this condition

   No study on file has given native amylin to lower a raised blood glucose, and this row is mechanistic only. In the one acute human infusion study located, 25 pmol/kg/min changed neither glucose disposal nor the insulin response to intravenous glucose, and only 50 pmol/kg/min — reaching circulating concentrations above 90 times normal postprandial peaks — reduced the insulin response; the authors concluded that circulating amylin is unlikely to be a hormone influencing carbohydrate metabolism in man at physiological concentrations. Almost all human amylin literature uses the analogue pramlintide rather than the native peptide, which is graded on its own row.

   Source PubMed 1 primary source read for this row

9. GIP (Gastric Inhibitory Peptide)

   human case series Mechanistic only Not approved for this condition

   Who was studied

   patients with mild type 2 diabetes (fasting plasma glucose 7.8 mmol/L, HbA1c 6.3 +/- 0.6%), n=9, versus 9 age- and weight-matched normal subjects; single hyperglycaemic clamp sessions

   Recorded, and not evidence for this condition

   No study on file has given GIP to lower a raised blood glucose; this row is mechanistic only. What was measured is the insulin secretory response to synthetic human GIP infused while glucose was held raised by a hyperglycaemic clamp, which was 54% lower in the type 2 diabetes group than in matched normal subjects while GLP-1 kept most of its insulinotropic effect. That characterises the incretin defect in the disease — glucose was clamped by the investigators, not moved by the peptide — and GIP receptor agonism reaches patients only inside the dual and triple agonists graded on their own rows.

   Source PubMed 1 primary source read for this row

10. MOTS-c

    animal in vivo Mechanistic only Not approved for this condition

    Who was studied

    mice, including high-fat-diet-fed and aged animals, intraperitoneal administration; group sizes and durations as reported in the source. No human participants have received MOTS-c in any study on file

    Recorded, and not evidence for this condition

    No study has measured blood glucose in people given MOTS-c — no acute infusion, no glucose tolerance test, no meal test, no fasting glucose — and this row is mechanistic only. What exists is intraperitoneal administration in mice, where glucose infusion rate during clamp was the readout and diet-induced insulin resistance was reduced; those figures are murine and by a route not used in humans, so the link to raised blood glucose in a person is an argument from rodent physiology rather than a result.

    Source PubMed 1 primary source read for this row