Research report: Case Report: GADA-positive type 1 diabetes with diabetic ketoacidosis in a young adult with obesity temporally associated with unsupervised tirzepatide use for weight reduction

A 28-year-old man with obesity was diagnosed with glutamic acid decarboxylase antibody-positive type 1 diabetes and diabetic ketoacidosis about four weeks after discontinuing self-administered tirzepatide, according to a single-patient case report from Zhongnan Hospital of Wuhan University. The report documents a temporal association between the unsupervised weight-loss treatment and the recognition of autoimmune diabetes, but it cannot establish whether tirzepatide exposure or withdrawal caused the condition, or even whether autoimmune diabetes was present before treatment began. The patient had no previously documented hyperglycemia.

A routine fasting blood glucose measurement two years before admission was 5.2 mmol/L, and a random fingerstick reading immediately before he began tirzepatide was 5.8 mmol/L. However, no recent fasting glucose, glycated hemoglobin, islet autoantibody or C-peptide measurements were available, so the investigators described his pretreatment status as having no documented hyperglycemia rather than confirmed normoglycemia. He self-purchased the drug and injected it subcutaneously without medical supervision, beginning with 2.5 mg once weekly and increasing to 5 mg and then 7.5 mg once weekly over approximately 12 weeks. He lost 15 kilograms during the initial weekly dosing schedule, falling from 94 kilograms to 79 kilograms and reducing his body-mass index from 28.9 to 24.3 kilograms per square metre.

He then used 5 mg every two to three weeks for approximately four weeks as an intermittent maintenance or tapering schedule. After reaching his target weight, he took his final dose and completely discontinued tirzepatide. Weight fell by another 2 kilograms during tapering and after cessation, while his appetite rebounded.

Roughly four weeks after the last dose, he developed polydipsia, polyuria and dry mouth. His symptoms continued for five days before he sought medical care, bringing the total weight loss to 19 kilograms from his starting point. At admission, his weight was 75 kilograms, with a body-mass index of 23.15 kilograms per square metre, and his vital signs were normal, with no Kussmaul breathing.

Laboratory testing showed a fasting glucose concentration of 13.99 mmol/L and a two-hour postprandial glucose concentration of 25.62 mmol/L after a standardized steamed-bun meal test. His glycated hemoglobin was 10.8%, equivalent to 95 mmol/mol. The investigators interpreted this value as indicating that substantial hyperglycemia was likely present for weeks to months rather than being confined to the five-day symptomatic period, meaning it may have begun before the drug was completely withdrawn.

His triglyceride concentration was also elevated at 3.95 mmol/L, while his estimated glomerular filtration rate and urine albumin-to-creatinine ratio were within the reported reference ranges. Evidence of ketoacidosis included a serum beta-hydroxybutyrate concentration of 4.44 mmol/L, urine ketones reported as +++ and serum bicarbonate of 16.9 mmol/L, producing an anion gap of 19.7 mmol/L. The authors classified the presentation as diabetic ketoacidosis using the 2024 hyperglycemic-crises consensus criteria of glucose at or above 11.1 mmol/L, ketosis and metabolic acidosis. Arterial blood gas was not obtained on admission, so the initial arterial pH remains unknown; the diagnosis therefore rested on the laboratory measurements available at presentation.

On hospital day three, after fluids and insulin had normalized the ketones, arterial testing showed a pH of 7.45 and an actual bicarbonate concentration of 28.5 mmol/L, consistent with recovery. The results supported autoimmune type 1 diabetes rather than type 2 diabetes, despite the patient’s obesity and a family history of type 2 diabetes. His glutamic acid decarboxylase antibody concentration exceeded 2,000 IU/mL, far above the laboratory reference limit of 10 IU/mL, and his islet cell antibody level was 129 U/mL, above the cited upper reference value of 29.79 U/mL.

Insulin autoantibody was reported as normal. His fasting C-peptide concentration was 0.798 ng/mL, below the cited range of 1.06–4.81 ng/mL, despite marked hyperglycemia; two-hour postprandial C-peptide was 1.076 ng/mL, also below the cited range of 2–5 times fasting. Together, the strong antibody results, limited beta-cell response, ketoacidosis and immediate need for insulin supported the investigators’ classification of adult-onset autoimmune type 1 diabetes rather than latent autoimmune diabetes in adults, which they describe as usually having a more insidious course and delayed insulin requirement.

Treatment included intravenous fluids, continuous subcutaneous insulin infusion, continuous glucose monitoring and one week of structured type 1 diabetes education. Ketones normalized by hospital day three, after which continuous insulin infusion was transitioned to multiple daily injections. The patient was discharged with his symptoms resolved and stable glucose control.

At discharge, continuous-glucose-monitor time in the target range of 3.9–10.0 mmol/L was 80%. At one month, his glycated hemoglobin was 7.4%, or 57 mmol/mol, and time in range was 90%. At three months, glycated hemoglobin had improved to 6.7%, or 50 mmol/mol, with 88% time in range and a stable insulin regimen.

These outcomes describe his course after recognition and treatment of diabetes; they do not establish an effect of tirzepatide on the autoimmune process. The temporal pattern leaves several possible explanations unresolved. Tirzepatide is a dual agonist of glucose-dependent insulinotropic polypeptide and glucagon-like peptide-1 receptors that promotes insulin secretion, suppresses glucagon release, delays gastric emptying and reduces appetite, resulting in weight reduction.

One hypothesis offered by the investigators is that substantial weight loss and improved insulin sensitivity during treatment temporarily masked progressive beta-cell failure, with autoimmune diabetes becoming clinically apparent as he tapered and stopped the medication. They also acknowledge that type 1 diabetes may have progressed independently and that the timing could be coincidental. Because pretreatment antibody and C-peptide measurements were absent, the report cannot establish when autoimmunity began, whether limited beta-cell reserve already existed before the first injection, or whether discontinuation precipitated decompensation.

The case is also diagnostically important because weight loss, reduced appetite and ketoacidosis can have several causes during incretin-based treatment. Previous reports summarized by the authors describe tirzepatide-associated hypoglycemic or starvation ketoacidosis in people without diabetes, often accompanied by vomiting, low food intake or low glucose. New-onset autoimmune type 1 diabetes has also been reported after compounded semaglutide and after online-prescribed semaglutide without pretreatment screening.

Those accounts provide context rather than results from a systematic review, and the authors explicitly distinguish starvation or hypoglycemic ketoacidosis without autoimmunity from new-onset autoimmune type 1 diabetes. The present patient had severe hyperglycemia, strongly positive glutamic acid decarboxylase antibody, an impaired C-peptide response and a clear requirement for insulin, features not explained by starvation ketosis alone. For science and medicine, the principal lesson is diagnostic vigilance rather than a verdict against tirzepatide.

The report does not show that the drug caused type 1 diabetes, nor does it justify avoiding the treatment in appropriate candidates. It also does not establish a need for routine antibody or C-peptide testing for every user or for routine post-discontinuation monitoring. Instead, it illustrates that unexplained hyperglycemia, continued weight loss or ketoacidosis during or after unsupervised incretin-based weight management may warrant prompt evaluation for autoimmune diabetes, even in a young adult with obesity and no documented history of high glucose.

The patient said he wished he had consulted a doctor before starting tirzepatide and monitored his blood glucose during treatment and after stopping it. Even that experience belongs to one person, however, and the larger uncertainty remains: the case raises a clinical question about recognition and timing without resolving whether tirzepatide played any causal role in his autoimmune diabetes.

Subject of Research: Single-patient case report of autoimmune type 1 diabetes associated temporally with unsupervised tirzepatide use

Article Title: Case Report: GADA-positive type 1 diabetes with diabetic ketoacidosis in a young adult with obesity temporally associated with unsupervised tirzepatide use for weight reduction

Article References: Lei, M., Liu, F., Sun, L., Li, S., Liu, J., Wang, Y., & Tang, J. (2026). Case Report: GADA-positive type 1 diabetes with diabetic ketoacidosis in a young adult with obesity temporally associated with unsupervised tirzepatide use for weight reduction. Frontiers in Endocrinology, 17, Article 1995854. https://doi.org/10.3389/fendo.2026.1995854

Image Credits: AI Generated

DOI: 10.3389/fendo.2026.1995854

Keywords: tirzepatide, type 1 diabetes, diabetic ketoacidosis, autoimmunity, GADA, obesity, weight loss, incretin therapy, C-peptide, case report, Case, Report