Tirzepatide in Type 1 Diabetes

A retrospective UK cohort study associated tirzepatide use with lower weight, HbA1c and insulin requirements over about one year in adults with type 1 diabetes and obesity. Compared with 50 matched controls, the 142 people who started tirzepatide had a 13.4 kg greater mean reduction in weight, a 0.6 percentage-point greater reduction in HbA1c and a 29.9-unit greater reduction in total daily insulin dose^[1]. These are substantial differences. They are observational findings, not randomized estimates of what the drug caused.

Safety findings were encouraging but limited by sample size and record-based collection. The study found no statistically significant between-group difference in diabetic ketoacidosis, severe hypoglycaemia or hospitalization rates. Forty tirzepatide participants reported an adverse effect, most often nausea or vomiting, and 89.4% remained on treatment at the end of follow-up^[2]. A phase 3 randomized trial is underway to answer the efficacy and safety question with stronger controls^[3].

What the study actually was

Researchers at Sheffield Teaching Hospitals NHS Foundation Trust reviewed electronic health and pharmacy records for every adult with type 1 diabetes who started tirzepatide there between March 2024 and June 2025. All 142 treated participants had obesity. The team selected 50 controls from the same records, matched on age, sex and body mass index. Median follow-up was 365 days for the tirzepatide group, while controls had a full year of data^[2].

Treatment was not assigned by researchers. Individual clinicians prescribed tirzepatide in routine practice, with no fixed protocol. Thirty treated participants switched directly from another GLP-1 receptor agonist. Controls could also use another GLP-1 drug, and three did. The groups looked similar on age, sex, BMI and HbA1c, but the tirzepatide group used more insulin at baseline and more often used hybrid closed-loop pumps^[2]. Those imbalances leave room for selection effects and regression to the mean.

This is why the article belongs in a trial-readout format only as an explicit evidence-boundary exercise. It reports a comparative cohort, not a clinical trial. Matching can make groups more comparable on selected characteristics, but it cannot balance every measured and unmeasured factor the way random assignment aims to do.

Weight and glycaemia moved together

Among participants with usable weight data, mean weight fell by 11.5 kg with tirzepatide and rose by 1.9 kg in controls. The between-group difference was 13.4 kg. Mean percentage weight loss in the tirzepatide cohort was 10.8%. At least 5% weight loss was recorded in 71.5% of treated participants versus 8.3% of controls, while at least 10% loss occurred in 44.7% versus 2.1%^[2].

For HbA1c, the analysis excluded people whose insulin-delivery mode changed and those with missing paired data. That left 110 tirzepatide participants and 48 controls. The adjusted between-group reduction was 6.5 mmol/mol, equivalent to about 0.59 percentage points. More treated participants finished below HbA1c thresholds of 8.0%, 7.5% and 7.0%^[2]. Continuous glucose monitoring showed a 7.3-point greater increase in time in range and a 0.8 mmol/L greater reduction in mean sensor glucose versus controls.

Total daily insulin dose fell by 24.5 units within the treated group and differed by 29.9 units from controls in the smaller paired analysis. Both basal and bolus insulin declined, with a larger proportional reduction in bolus insulin^[2]. These are clinically relevant observations, but the study does not supply a dosing algorithm. Insulin adjustment in type 1 diabetes can affect both hypoglycaemia and ketoacidosis risk, so the figures describe this cohort rather than instructing readers how to change treatment.

Cardiometabolic markers are not outcomes

Blood pressure and lipids also improved relative to controls. The reported between-group differences were 12.6 mmHg for systolic blood pressure, 4.1 mmHg for diastolic pressure, 0.8 mmol/L for total cholesterol and 0.6 mmol/L for non-HDL cholesterol. Triglycerides differed by 0.7 mmol/L^[1]. The analyses excluded people with relevant medication changes where possible, which reduces one source of confounding.

Those measurements remain surrogate risk markers. The cohort did not test whether tirzepatide prevents heart attacks, strokes, kidney failure or death in people with type 1 diabetes. A one-year change in blood pressure or cholesterol can support a hypothesis about risk, but it is not a counted cardiovascular outcome. The authors call for larger randomized trials and longer follow-up for that reason.

What the safety data can support

Adverse effects were recorded for 28.2% of tirzepatide participants. Nausea or vomiting appeared in 19.7%, abdominal pain in 8.5%, diarrhea in 4.9% and gallstones in 1.4%. The records contained no pancreatitis, gastroparesis or bowel obstruction during follow-up^[2]. Mild events may have been missed because the study relied on clinical notes rather than scheduled research visits.

Rates of diabetic ketoacidosis and severe hypoglycaemia did not differ significantly between groups, but both outcomes were rare. Ketoacidosis ran at 0.04 events per patient-year with tirzepatide and zero in controls; the p value was 0.182. Severe hypoglycaemia rates were 0.01 and zero, with a p value of 0.405^[2]. A nonsignificant result in groups of 142 and 50 does not establish that risks are equal.

The current US Zepbound label lists chronic weight management and moderate to severe obstructive sleep apnea in adults with obesity as indications. It does not list glycaemic management in type 1 diabetes^[4]. The cohort therefore should not be read as label-level evidence for tirzepatide as an insulin adjunct. Our tirzepatide safety overview separates approved uses from emerging research, and the tirzepatide evidence guide covers the wider trial record.

Why the ongoing randomized trial matters

SURPASS-T1D-1 is a phase 3, randomized, double-blind, placebo-controlled study in an estimated 905 adults with type 1 diabetes and obesity or overweight. Its primary endpoint is change in HbA1c at week 40. Secondary endpoints include continuous glucose monitoring time in range, body weight, blood pressure, lipids and total daily insulin dose^[3]. ClinicalTrials.gov listed it as active but not recruiting, with primary completion estimated for November 2026 when the record was last updated in May.

Randomization should reduce the treatment-selection problem that this cohort cannot escape. Blinding and standardized follow-up should also improve adverse-event capture. Even then, a 40-week trial will not answer long-term cardiovascular outcomes. It should provide a firmer estimate of glycaemia, weight and common safety events in the population studied.

What we still do not know

This cohort cannot establish causality, define safe insulin-adjustment practice or rule out uncommon harms. It came from one UK center, 87% of participants were White British and missing data reduced several outcome analyses. We also do not know whether benefits extend to people with type 1 diabetes who do not have obesity, whether they persist over several years, or whether changes in surrogate markers translate into fewer cardiovascular or kidney events.

Medical disclaimer: This article is for educational and journalistic purposes only and does not constitute medical advice. Peptides discussed may be classified as prescription medicines or research chemicals depending on your jurisdiction. Always consult a qualified healthcare professional before using any peptide product. PeptideMethods.com does not sell, distribute, or facilitate the sale of any peptide product.