Tirzepatide in type 1 diabetes: new data

A July 2026 retrospective cohort study published in Diabetes, Obesity and Metabolism reports that tirzepatide, added to existing insulin therapy in adults with type 1 diabetes and overweight or obesity, was associated with a 10% mean body-weight reduction and a substantial fall in daily insulin requirements over roughly 28 weeks^[1]. The drug is not approved for type 1 diabetes in any jurisdiction; this paper is among the first to report real-world outcomes in that population.

What the study found

Purcell and colleagues at the University of Sydney compared 23 adults with type 1 diabetes who received tirzepatide as an off-label adjunct to insulin against 23 matched controls who did not^[1]. The tirzepatide group lost a mean 10.01% of body weight (standard deviation 4.74%), while the control group gained a mean 0.69% (SD 3.77%), an adjusted difference that reached p less than 0.0001. Total daily insulin dose fell by a mean 21.82 units per day in the tirzepatide group compared with an increase of 5.62 units per day in controls (p = 0.002). The most common dose was 5 mg weekly, used by 52.2% of participants. Both glucose management indicator and glucose variability improved with tirzepatide. Blood pressure, lipid, liver, and kidney markers showed no significant between-group differences over the observation window.

How this fits with the wider evidence

Two systematic reviews published earlier in 2026 reach similar conclusions, though with consistent warnings about certainty. A review by Acucella and Caponio covering eight studies (one randomised trial and seven observational studies) found that the single randomised trial administered tirzepatide to type 1 diabetes patients and observed a mean body-weight reduction of 10.3 kg and a placebo-adjusted 35.1% fall in total daily insulin dose^[2]. The authors rated the overall evidence quality as low to very low and noted that durable glycaemic benefit and long-term safety had not been established. A meta-analysis by Soliman and colleagues pooling six observational studies and 248 participants reported mean reductions at six months of 9.9 kg in body weight, 0.61 percentage points in HbA1c, and 23.73 IU per day in insulin requirements^[3]. The pattern across all three bodies of work is consistent: meaningful weight loss and reduced insulin burden, modest glycaemic improvement, and a safety profile dominated by gastrointestinal adverse events.

What this is not

Tirzepatide is not approved for type 1 diabetes by the FDA, the EMA, the MHRA, or any regulator tracked on this site. The full tirzepatide regulatory status by country reflects approvals for type 2 diabetes (Mounjaro) and chronic weight management (Zepbound, US only). Any use in type 1 diabetes is off-label and outside the studied and authorised indication. The Purcell cohort has 23 treated participants observed for under eight months: that is too small and too short to establish durable glycaemic benefit, to characterise long-term safety, or to rule out serious late adverse events. Tirzepatide is not insulin-replacement therapy. In type 1 diabetes, basal-bolus insulin coverage is still the standard of care; tirzepatide in this setting is adjunctive, not substitutive.

Why this matters for the tirzepatide story

The tirzepatide peptide page flags type 1 diabetes as a population the drug has not been studied in, which has been accurate until recently. That is starting to shift. A randomised controlled trial with the explicit aim of evaluating tirzepatide in type 1 diabetes and obesity was registered in BMJ Open in May 2026 (Al Ozairi et al.), and the July 2026 retrospective cohort from Purcell and colleagues adds real-world data from clinical practice ahead of those results. None of this constitutes a basis for prescribing tirzepatide in type 1 diabetes outside a clinical study setting, but it does mean the population can no longer be described as simply unstudied.

What we do not yet know

Several questions remain unanswered. Whether the insulin-dose reduction is durable past six months, and whether it introduces hypoglycaemia risk that is not captured in short observational windows, is unknown. The glycaemic improvement is modest and the certainty is low across all current evidence. Autoimmune considerations specific to type 1 diabetes, including any interaction with the residual beta-cell function some adults with type 1 diabetes retain, have not been systematically assessed. The randomised trial registered by Al Ozairi and colleagues, when it reads out, will be the first adequately controlled piece of evidence in this setting.