C-peptide protocols in the published literature

What amounts of C-peptide did the replacement studies give?

review

The published summaries behind this page describe the infusions by their aim and their length, and never by a figure. A 2004 review reports that C-peptide infusion restores red blood cell deformability and microvascular blood flow concomitantly with Na+,K(+)-ATPase activity, which is a pump in the cell wall that moves sodium out and potassium in.4 Deformability is how easily a red cell bends to squeeze through the smallest vessels. The same review closes by arguing that physiological C-peptide infusion could be beneficial for the prevention of diabetic complications.4 Physiological is the word doing the work in that sentence. The aim was to put back roughly the level a working pancreas would supply, rather than to push the level as high as it would go. A 2023 review shows the target was still open, saying future studies will need to pay more attention to post-treatment C-peptide levels to explore the optimal range of fasting C-peptide and postprandial C-peptide maintenance.2 Postprandial means after a meal. So what these sources record is an aim, a normal blood level, and not an amount that reached it.

Who received C-peptide in those studies?

review

People with type 1 diabetes, whose pancreas makes little or no insulin and therefore little or no C-peptide, and diabetic animals. A 2017 review argues that the available information suggests that type 1 diabetes should be regarded as a dual hormone deficiency disease, meaning a body short of C-peptide as well as of insulin.1 That framing explains the choice of patient. Replacement only makes sense where something is missing. A 2004 review draws the same line through red blood cells, reporting that the pump activity is decreased in the red blood cell membranes of type 1 diabetic individuals, irrespective of the degree of diabetic control.4 In type 2 diabetes, where a pancreas usually still makes C-peptide, the review says the same activity is less impaired or even normal in those of type 2 diabetic patients.4 The animal arm used diabetic rats, and the review notes that C-peptide infusion to diabetic rats increases endoneural ATPase activity in rat, endoneural meaning inside the nerve.4 Every group described is short of C-peptide to begin with, which is not the same as a person whose own supply is normal.

How was C-peptide given, and in what form?

human pilot / early trial

By infusion in people, meaning a slow drip of fluid, and in three other forms that should not be confused with it. In the laboratory, a 2004 review reports that incubation of diabetic red blood cells with C-peptide at physiological concentration leads to an increase of the pump's activity.4 Incubation means the cells sat in a dish with the peptide around them. In pieces of rat kidney, the same review says C-peptide stimulates in a dose-dependent manner Na+,K(+)-ATPase activity, which means more peptide in the dish produced more pump activity.4 A different use altogether runs through the immune system, and a 2021 editorial records that C-peptide fragments are successfully administered in immunotherapy of type I diabetes.5 Immunotherapy here means teaching the immune system to tolerate the cells it attacks, which is a different goal from topping up a missing hormone, and a fragment is not the whole peptide. A concentration in a dish, a fragment given to the immune system and an infusion into a person are three different things, and a figure from one cannot be read as a figure for another.

How long did the courses run, and was there a cycle?

human pilot / early trial

Every human course in these summaries is described as short-term, and a 2017 review ends by calling clinical trials of C-peptide in diabetic nephropathy both justified and urgently required.1 Diabetic nephropathy is kidney damage caused by diabetes. Asking for trials in 2017 is a statement that the long ones had not yet settled the question. A 2021 editorial gives the later verdict, writing that the peptide turned out to matter little for managing diabetes, while every phase of the research changed how insulin and other hormones are understood.5 No cycle, break or taper appears anywhere in the replacement work described in these sources, because none of it ran long enough to need one. That matters for anybody comparing this with a compound sold with an on-and-off schedule. A short course that moved a kidney marker and a long course that changes a kidney outcome are different questions, and only the first was answered in people.

Is a C-peptide blood result a dose?

human RCT

No, and this is the confusion most likely to bring somebody to a page about C-peptide dosage. A result is a concentration, an amount per litre of blood at one moment. One large trial shows the difference inside a single paper: its 5,047 participants were taking 1,000-2,000 mg metformin daily, which is a dose, while their fasting C-peptide was something the investigators measured.6 Changes are reported the same way, so a 2013 trial wrote that patients given teplizumab had a reduced decline in C-peptide at 2 years of mean -0.28 nmol/L, against -0.46 nmol/L in the control group.7 Those figures describe how much C-peptide was found in the blood, made by the patients' own cells. Nobody was given C-peptide in that trial. Reading a result in nanomoles per litre as an amount to take would be a category error, the way a fuel gauge reading is not an amount of fuel to buy. What a reading means for diabetes type and insulin decisions is covered on the C-peptide overview.

What is given during a stimulated C-peptide test?

human RCT

This is where most of the real amounts in the C-peptide literature sit. A stimulated test gives the pancreas something to respond to, then measures the C-peptide it releases. In a 2026 report on low-dose antithymocyte globulin, people were asked to obtain C-peptide via a commercial laboratory 90 min after drinking 6 mL per kilogram of a liquid meal.3 A 2021 trial in recent-onset type 1 diabetes measured the response through the area under the concentration-time curve over a 4 h period, a total of everything released across four hours.8 A 2017 study in eighteen healthy men compared an equivalent dose of branched chain amino acids, 30.7 ± 1.1 g, given by mouth and into a vein.9 Branched chain amino acids are three of the building blocks of protein. Every one of those amounts is a challenge given to provoke a response, measured against a fixed clock. None of them is an amount of C-peptide, and a drink sized per kilogram for a test says nothing about anything a person would take.

How can someone increase their C-peptide?

human RCT

Most of what the trials behind this page tried was slowing the fall of C-peptide in new type 1 diabetes, or replacing the cells that make it, and each came with a schedule. In a 2013 trial, 52 people with new-onset type 1 diabetes received teplizumab, an antibody that calms part of the immune system, for 2 weeks at diagnosis and after 1 year.7 A 2026 report describes people with stage 2 or stage 3 type 1 diabetes treated clinically with low-dose ATG at 2.5 mg/kg, another immune therapy.3 In a 2025 cell therapy trial, participants in the later parts received a full dose of zimislecel, about 0.8 billion lab-grown islet cells, as a single infusion.10 Those are prescription immune therapies and cell transplants given to people whose own production was failing, under close monitoring. The one rise in people without diabetes came from a 2026 first-in-human trial of KN069, an experimental weight-loss compound given in single doses of 12-120 mg to men with overweight or obesity, where the reported changes included elevated insulin/C-peptide.11 A higher reading there is a pancreas being pushed to release more, which is not the same as a pancreas being repaired, and slowing a decline is a different question again.

Which amounts were tried and did not hold C-peptide up?

human RCT

The failures are as informative as the successes, because they show an amount can be stated precisely and still do nothing. A 1990 double-blind trial gave nicotinamide, a form of vitamin B3, in a dose of 100 mg/year of age up to a maximum of 1.5 g/day to children and adolescents with new type 1 diabetes.12 Its conclusion was flat: nicotinamide, at this dosage, does not preserve residual insulin secretion in subjects with newly diagnosed type 1 diabetes.12 In the 2021 trial of 308 adults, the decline was not significantly smaller with liraglutide alone, a ratio of 1·12 against placebo with p=0·38, even though the combination arm did better.8 A precise dose is a fact about a protocol, and it is not evidence that the protocol worked. Read that carefully whenever an amount is quoted alongside C-peptide as if the number itself were the result.

How can someone bring a high C-peptide down?

human RCT

The studies behind this page lowered the reading in a few ways, and none of them is a fix in the sense a searcher means. In a 1997 rabbit experiment, animals were injected subcutaneously with 6 i.u. of human insulin daily for 24 weeks, and their plasma C-peptide ran lower than that of rabbits given placebo.13 That is a rabbit amount, it does not convert to a person, and the reading fell because insulin arrived from outside and the pancreas made less of its own. In a 2021 trial in 316 people with type 2 diabetes, tirzepatide 10 and 15 mg significantly decreased fasting insulin compared with placebo and dulaglutide.14 Those are amounts of an approved diabetes medicine, and they belong to its label rather than to C-peptide. A lower reading can mean a body needs less insulin, or that its pancreas is producing less. Those are different things, and a falling number on its own does not tell you which.

Does an approved dose apply to C-peptide sold as a research compound?

human pilot / early trial

There is no approved dose to borrow. None of the research behind this page describes C-peptide approved as a medicine anywhere, so no label exists to quote, and no schedule from a regulator can be carried across to a vial. What has lasted is the measurement, and a 2021 editorial notes that plasma C-peptide levels remain a standard for measurement of beta cell activity in patients.5 The 1982 review in Diabetes Care says much the same about where its value lies, in clinical research, where it offers a unique opportunity to follow the B-cell secretion in diabetic subjects.15 The amounts of approved medicines that appear on this page, such as tirzepatide or metformin, belong to those medicines. None of them describes C-peptide, and none of them would carry to C-peptide sold as a research compound.

Is Lipo-C the same as C-peptide?

Many searches for a C-peptide dose are really searches for a product sold as Lipo-C. The two share a letter, and nothing else that we can document. None of the research behind this page concerns Lipo-C. Every paper cited on this page concerns C-peptide, the short piece cut from proinsulin when the pancreas makes insulin. An amount printed for one says nothing about the other, and a chart for one should never be read as a chart for the other.

Why do none of these figures add up to a protocol?

human RCT

Because every figure on this page was attached to a diagnosis, a setting and a team watching the result, and the response still varied. In the zimislecel trial, all the participants also received glucocorticoid-free immunosuppressive therapy, which is a second set of medicines that came with the cells.10 The teplizumab trial found that the same schedule worked unevenly, concluding that metabolic and immunologic features at baseline can identify a subgroup with robust responses to immune therapy.7 So even where a schedule is written down, it belongs to a particular patient group under particular monitoring. Lifting it out of that context is not the same as following it. The C-peptide infusions themselves left no figure in these summaries at all, and what they were measured to do, in kidneys, nerves and red cells, is laid out outcome by outcome on the C-peptide benefits page.