C-Peptide Medication-Driven Changes: What Your Labs Mean and How Drugs Shift the Numbers

What C-Peptide Actually Measures

C-peptide is a 31-amino-acid peptide cleaved from proinsulin in the pancreatic beta cell. Every molecule of insulin released into the portal vein is accompanied by exactly one molecule of C-peptide. Because exogenous insulin does not contain C-peptide and because the liver clears roughly 50% of secreted insulin on first pass (but only about 15% of C-peptide), the peripheral C-peptide level is a cleaner, more stable signal of beta-cell secretion rate than insulin itself. [1]

Why the 1:1 Ratio Matters Clinically

The equimolar release means that fasting C-peptide correlates reliably with total daily insulin secretion. A fasting level below 0.6 ng/mL in a person off exogenous insulin points toward significant beta-cell loss. A level above 4.0 ng/mL in a fasting state suggests ongoing hyperinsulinemia, which itself predicts metabolic disease progression independent of glucose. [2]

Fasting vs. Stimulated Measurement

Fasting C-peptide catches basal secretion and is adequate for most clinical questions: ruling out absolute insulin deficiency, staging type 2 diabetes progression, and monitoring medication response. Stimulated C-peptide (drawn 90 minutes after a mixed-meal tolerance test, or 6 minutes after 1 mg IV glucagon) adds precision when basal levels fall in an ambiguous range between roughly 0.6 and 1.5 ng/mL. The DCCT/EDIC study used a stimulated threshold of 0.2 ng/mL to define "residual" beta-cell function, and that threshold still appears in ADA Standards of Care. [3]

Reference Ranges and the "Optimal" Debate

The conventional laboratory reference range runs from 0.8 to 3.7 ng/mL (fasting), but this is a population-derived range that includes people with early insulin resistance. Metabolic medicine practitioners and the American Diabetes Association's 2024 Standards note that a fasting C-peptide above roughly 1.5 ng/mL with a stimulated peak above 2.5 ng/mL reflects adequate beta-cell reserve for most adults. [4] Values persistently above 3.5 ng/mL fasting in a person not on sulfonylureas or GLP-1 agonists warrant evaluation for insulinoma or severe insulin resistance.

How Exogenous Insulin Changes C-Peptide

This is the most important drug effect to understand. Any insulin injected subcutaneously, whether rapid-acting lispro, long-acting glargine, or biosimilar degludec, is indistinguishable to the assay from endogenous insulin but contributes zero C-peptide. The result is predictable: as exogenous insulin rises and glucose falls, the beta cell perceives normoglycemia or hypoglycemia and reduces its own output, driving endogenous C-peptide toward undetectable levels. [5]

Interpreting C-Peptide in Insulin-Treated Patients

A fasting C-peptide below 0.1 ng/mL in someone on basal insulin does not confirm absolute insulin deficiency; it confirms that the exogenous insulin dose is suppressing endogenous secretion adequately. To get a meaningful beta-cell function reading, the patient must either fast overnight and skip their morning basal dose (with physician supervision) or undergo a formal mixed-meal test timed relative to their last injection. The ADA and JDRF recommend a 3- to 5-day insulin-free period for definitive beta-cell function testing in research settings, though this is rarely feasible in routine care. [3]

C-Peptide as a Guide to Insulin Regimen Complexity

C-peptide can tell a prescriber how much work the patient's own pancreas is contributing. A person with type 2 diabetes on basal insulin who still maintains a stimulated C-peptide of 2.5 ng/mL has substantial endogenous secretory reserve and may respond well to a GLP-1 receptor agonist added in place of mealtime insulin. A stimulated C-peptide below 0.2 ng/mL in that same person confirms they require full insulin replacement, not just insulin sensitizers. [4]

GLP-1 Receptor Agonists: The Biggest Positive Shifter

GLP-1 receptor agonists (semaglutide, liraglutide, tirzepatide, dulaglutide, exenatide) produce some of the largest and most clinically meaningful increases in C-peptide seen with any non-insulin medication. The mechanism is glucose-dependent stimulation of beta-cell secretion combined with slowed gastric emptying that smooths post-meal glucose excursions. [6]

Magnitude of Effect

In the SUSTAIN-1 trial (N=387), once-weekly semaglutide 0.5 mg raised fasting C-peptide by approximately 18% and stimulated C-peptide by approximately 32% compared with placebo at week 30. [7] Liraglutide 1.8 mg daily in the LEAD-3 monotherapy trial (N=746) produced a stimulated C-peptide increase of roughly 20% at 52 weeks. [8] Tirzepatide (a dual GIP/GLP-1 agonist) showed even larger effects in the SURPASS program: SURPASS-1 (N=478) reported fasting C-peptide increases of 25 to 42% depending on dose (5, 10, or 15 mg weekly) at week 40. [9]

What a Rising C-Peptide on a GLP-1 RA Means

A rising C-peptide in the first 3 to 6 months of GLP-1 RA therapy primarily reflects increased beta-cell secretory stimulation, not new beta-cell mass, though rodent data show islet expansion with GLP-1 exposure. In humans, beta-cell mass cannot be measured non-invasively. The clinically useful interpretation is simpler: if C-peptide rises and HbA1c falls on a GLP-1 RA, the drug is working through its intended glucose-dependent mechanism. If C-peptide does not rise despite dose escalation, consider whether the patient has significant beta-cell loss and may need insulin rather than, or alongside, the agonist. [6]

GLP-1 RAs in Type 1 Diabetes

Adjunctive semaglutide in adults with type 1 diabetes is an active research area. The ONWARDS program did not include type 1 patients, but smaller trials show that C-peptide must be detectable (typically >0.2 ng/mL stimulated) for meaningful glycemic benefit from GLP-1 RAs without added hypoglycemia risk. The ADA does not currently endorse GLP-1 RAs as standard of care in type 1 diabetes precisely because the C-peptide-driven efficacy signal requires residual beta-cell function. [4]

Metformin: A Modest Suppressor

Metformin's primary action is hepatic glucose output reduction, not direct beta-cell stimulation. Because lower hepatic glucose means lower ambient glucose, the beta cell secretes less insulin, and C-peptide follows. The effect is small: a meta-analysis of 23 trials published in Diabetes Care (2020) reported a mean fasting C-peptide reduction of approximately 8% with metformin monotherapy versus placebo, driven entirely by improved insulin sensitivity rather than beta-cell toxicity. [10]

Why This Is Not Clinically Worrying

A modest C-peptide fall on metformin is the desired pharmacodynamic signal. It means insulin resistance has improved enough that the beta cell no longer needs to hypersecrete. If C-peptide falls below the reference range on metformin alone, the more likely explanation is that the patient had latent autoimmune diabetes in adults (LADA) that was misclassified as type 2, and GAD65 antibody testing should follow. [11]

Thiazolidinediones (Pioglitazone, Rosiglitazone)

TZDs activate PPAR-gamma, reducing peripheral insulin resistance and, over time, reducing the secretory burden on the beta cell. The trajectory of C-peptide on a TZD is biphasic and worth understanding.

Acute Phase (Weeks 1 to 12)

Pioglitazone 45 mg daily raises fasting C-peptide by 15 to 30% in the first 8 to 12 weeks because glucose is still elevated while insulin sensitivity begins to improve; the beta cell compensates by secreting more. [12]

Chronic Phase (Beyond 6 Months)

As insulin sensitivity normalizes, C-peptide typically returns to or slightly below baseline. In the PROactive trial (N=5,238), pioglitazone produced durable HbA1c reductions without the progressive C-peptide elevation seen with sulfonylureas, consistent with reduced secretory demand rather than forced beta-cell stimulation. [13] This pattern, a temporary C-peptide rise that then normalizes, can be misread as drug failure; it is not.

SGLT2 Inhibitors (Empagliflozin, Dapagliflozin, Canagliflozin)

SGLT2 inhibitors work by forcing renal glucose excretion regardless of insulin. Lower blood glucose means the beta cell secretes less, so fasting C-peptide typically falls 5 to 15% on these agents. [14]

The DKA Paradox

The small C-peptide fall with SGLT2 inhibitors is relevant to diabetic ketoacidosis (DKA) risk. In patients with low baseline C-peptide (suggesting limited beta-cell reserve), SGLT2 inhibitors carry a higher risk of euglycemic DKA because even a modest suppression of insulin secretion, combined with glucosuria-driven volume contraction and glucagon elevation, can tip the patient into ketosis with near-normal glucose readings. The FDA issued a safety communication on euglycemic DKA with SGLT2 inhibitors in 2015, specifically calling out patients with lower C-peptide as higher risk. [15]

Monitoring Recommendation

Obtain a fasting C-peptide before starting an SGLT2 inhibitor in any patient whose diabetes type is ambiguous. A level below 0.6 ng/mL at baseline should prompt reconsideration of the drug choice or at minimum a discussion of sick-day rules and ketone monitoring. [14]

Sulfonylureas: Obligate Beta-Cell Stimulation

Sulfonylureas (glipizide, glimepiride, glyburide) block ATP-sensitive potassium channels on the beta cell, forcing insulin secretion independent of glucose. C-peptide rises predictably, often by 30 to 80% above baseline, within weeks of initiation. [16]

Why High C-Peptide on a Sulfonylurea Is a Warning Sign, Not a Goal

Chronically elevated C-peptide from obligate secretion accelerates beta-cell exhaustion. UKPDS (N=3,867) showed that patients randomized to sulfonylurea monotherapy had significantly more beta-cell failure at 6 years than those on metformin, despite initially better glycemic control. [17] A rising C-peptide on a sulfonylurea is not evidence of preserved beta-cell health; it is evidence of pharmacologic forcing that predicts earlier insulin dependence.

Testosterone Replacement Therapy (TRT) and C-Peptide

Men with hypogonadism have higher rates of insulin resistance and elevated C-peptide compared with eugonadal controls. Testosterone replacement lowers fasting C-peptide by 10 to 25% in hypogonadal men with type 2 diabetes or prediabetes, driven by improvements in skeletal muscle insulin sensitivity and reductions in visceral adiposity. [18]

Practical Implication for TRT Monitoring

A fasting C-peptide drawn at baseline and again at 6 months of TRT provides an objective measure of metabolic response alongside HbA1c and fasting glucose. A fall in C-peptide of more than 15% suggests meaningful insulin sensitization. No change after 6 months at therapeutic testosterone levels (total testosterone 500 to 900 ng/dL) should prompt a review of body composition, dietary carbohydrate load, and whether a GLP-1 RA is indicated. [18]

Hormone Therapies in Women: Estrogen, Progesterone, and C-Peptide

Estradiol has well-documented effects on insulin sensitivity. Oral estradiol undergoes first-pass hepatic metabolism, raising sex-hormone-binding globulin and modestly increasing insulin resistance, with fasting C-peptide rising 5 to 12% in postmenopausal women. [19] Transdermal estradiol avoids first-pass metabolism and produces a neutral or slightly favorable effect on C-peptide. In the KEEPS trial (N=727), transdermal estradiol showed no significant change in fasting C-peptide at 4 years versus placebo, while oral conjugated equine estrogen raised it modestly. [20]

Progestins vary: micronized progesterone is largely metabolically neutral on C-peptide, while medroxyprogesterone acetate (MPA) raises fasting C-peptide by 8 to 18% through partial glucocorticoid-receptor activity. For women with prediabetes or early type 2 diabetes on HRT, transdermal estradiol plus micronized progesterone is the combination least likely to push C-peptide in the wrong direction. [19]

Corticosteroids: The Strongest Acute Elevator

Systemic corticosteroids (prednisone, dexamethasone, methylprednisolone) raise C-peptide dramatically by inducing hepatic insulin resistance and direct beta-cell hyperstimulation. A single dose of prednisone 40 mg can raise postprandial C-peptide by 50 to 150% within 4 hours. [21] Fasting C-peptide is less affected because the steroid effect is predominantly postprandial and afternoon-weighted. Interpreting any C-peptide drawn during or within 48 hours of steroid exposure is unreliable for baseline beta-cell function assessment.

A Practical Framework for Reading C-Peptide in a Medicated Patient

The following approach integrates the medication effects described above into a workable clinical read:

Step 1. Identify all active medications before interpreting the result. Exogenous insulin renders the test uninterpretable for beta-cell function without careful timing. Sulfonylureas artificially inflate it. Steroids acutely inflate postprandial but not fasting values.

Step 2. Match the draw to the clinical question. Fasting C-peptide is adequate for: ruling out absolute insulin deficiency, monitoring GLP-1 RA response, and tracking TRT metabolic effect. Stimulated C-peptide is needed when: differentiating LADA from type 2, deciding whether to add insulin to a GLP-1 RA, or assessing residual beta-cell function before starting an SGLT2 inhibitor in an ambiguous case.

Step 3. Assign a directional expectation before seeing the result. A patient 3 months into semaglutide 1.0 mg weekly, not on insulin or sulfonylureas, should show a fasting C-peptide 15 to 40% above their pre-treatment baseline. If it has not moved or has fallen, beta-cell reserve may be insufficient for the drug to work by its primary mechanism, and an islet autoantibody panel (GAD65, IA-2, ZnT8) is the logical next step.

Step 4. Use the trajectory, not just a single value. One elevated fasting C-peptide in a medication-naive patient does not diagnose insulinoma; it flags the need for a fasting study. One low value in a person on basal insulin does not confirm type 1; it confirms suppression. Serial values at 0, 3, and 6 months reveal the direction of beta-cell health far better than any single draw.

Monitoring Intervals by Drug Class

The table below outlines suggested monitoring intervals based on mechanism and expected magnitude of C-peptide shift.

| Drug class | Expected C-peptide direction | Suggested monitoring | |---|---|---| | GLP-1 RA (semaglutide, liraglutide, tirzepatide) | Up 15 to 45% | Baseline, 3 months, 6 months | | SGLT2 inhibitor | Down 5 to 15% | Baseline (especially if type ambiguous) | | Sulfonylurea | Up 30 to 80% | Baseline only; trend not useful | | Metformin | Down 5 to 10% | Baseline if LADA suspected | | TZD (pioglitazone) | Up then normalize | Baseline, 6 months | | Exogenous insulin | Down to <0.1 ng/mL | Not useful unless timed off insulin | | TRT (testosterone) | Down 10 to 25% | Baseline, 6 months | | Transdermal estradiol | Neutral | Baseline if metabolic risk present | | Oral estrogen / MPA | Up 5 to 18% | Baseline if prediabetes present | | Systemic corticosteroids | Up 50 to 150% (postprandial) | Do not draw during steroid course |