Tirzepatide in Adults With Type 1 Diabetes: A Phase 2 Randomized Placebo-Controlled Clinical Trial

Abstract

OBJECTIVE

Overweight and obesity are prevalent in type 1 diabetes and contribute to cardiovascular risk. Tirzepatide, a gastric inhibitory polypeptide and glucagon-like peptide 1 receptor coagonist, has not been studied in type 1 diabetes.

RESEARCH DESIGN AND METHODS

We conducted a 12-week, phase 2, double-blind, placebo-controlled trial in adults with type 1 diabetes and BMI >30 kg/m². Participants were randomized to once-weekly subcutaneous tirzepatide (2.5 mg for 4 weeks, 5.0 mg for 8 weeks) or placebo. The primary end point was change in body weight at 12 weeks.

RESULTS

Twenty-two of 24 adults with type 1 diabetes completed the study. After 12 weeks, the mean change in weight was −10.3 kg (95% CI −12.8 to −7.7 kg) in the tirzepatide group and −0.7 kg (95% CI −1.4 to 2.8 kg) in the placebo group, with an estimated treatment difference of −8.7 kg (95% CI −12.0 to −5.5 kg; P < 0.0001), representing 8.8% weight loss. In the tirzepatide group, 100% and 45% of participants experienced weight loss of ≥5% and ≥10% respectively, compared with 9% and 0% in the placebo group. Tirzepatide improved HbA_1c (mean difference −0.4% [95% CI −0.7 to 0.0%] vs. placebo; P = 0.05) and reduced total daily insulin dose (−24.2 units/day tirzepatide and −0.3 units/day placebo; difference from baseline vs. placebo −35.1% [95% CI −46.5 to −21.3%; P = 0.0002]). There were no significant adverse events in either group.

CONCLUSIONS

Among adults with type 1 diabetes and obesity, tirzepatide was superior to placebo for weight loss over 12 weeks.

Graphical Abstract

Introduction

Globally, >8 million people live with type 1 diabetes, a chronic condition characterized by pancreatic β-cell destruction, leading to a life-long requirement for exogenous insulin (1). People living with type 1 diabetes have excessive cardiovascular mortality risk, which is partially attributed to traditional risk factors such as dyslipidemia, hypertension, hyperglycemia, and overweight/obesity (2). Strict glucose targets achieved with intensive insulin therapy has been shown to reduce the risk of cardiovascular events and microvascular complications in type 1 diabetes (3). Yet, insulin therapy does not fully address the cardiometabolic issues faced by people with type 1 diabetes.

Studies of population-level registry data have reported that up to 79% of adults with type 1 diabetes do not achieve glycemic targets (4–6). Furthermore, almost two-thirds of adults with type 1 diabetes have concurrent overweight or obesity (6,7). Insulin-induced weight gain has been shown to offset the cardiovascular benefits achieved by intensive insulin therapy (8). Insulin delivered subcutaneously may also exacerbate dyslipidemia, insulin resistance, hypoglycemia, and glycemic variability (9–12). Hence, there is a strong need to identify therapies to reduce weight, improve glycemia, lower insulin dose, and address other cardiometabolic risk factors beyond what insulin-only therapy can achieve in type 1 diabetes (13).

Tirzepatide, a dual agonist of the gastric inhibitory polypeptide (GIP) and glucagon-like peptide 1 (GLP-1) receptors, improves glycemia and weight in people with type 2 diabetes and obesity without diabetes (14,15). Tirzepatide is associated with improvements in insulin sensitivity and body composition in type 2 diabetes, but its metabolic effects have not been established in type 1 diabetes (16–18).

Recent real-world retrospective studies capturing off-label use of tirzepatide in type 1 diabetes have documented dramatic impacts on weight, glycemia, and insulin dose requirements (19,20). The need for formal placebo-controlled randomized trials to control for behavioral effects in studies with a weight end point prompted the design of Tirzepatide in Type 1 Diabetes: Cardiometabolic Effects (TIRTLE1), the first randomized placebo-controlled trial to test the hypothesis that tirzepatide improves weight, glycemia, and other metabolic and behavioral measures in adults with type 1 diabetes.

Research Design and Methods

Trial Design and Oversight

TIRTLE1 was a single-site, phase 2, double-blind, randomized placebo-controlled trial conducted between May and December 2024 and sponsored by the Garvan Institute of Medical Research. The trial was approved by the St. Vincent’s Hospital Human Research and Ethics Committee (Sydney) and was prospectively registered (Australian New Zealand Clinical Trials Registry no. ACTRN12624000111572). The study was advertised to the public and to local hospital diabetes clinics. Participants provided written informed consent. The full trial protocol and statistical analysis plan are provided in the Supplementary Material.

Participants

Eligible participants were adults aged 18–60 years with at least a 2-year history of type 1 diabetes and a BMI ≥30 kg/m². We excluded individuals who had used a GLP-1 receptor agonist (GLP-1 RA), sodium–glucose cotransporter inhibitor, or glucocorticoids in the 6 weeks prior to recruitment; with a history of diabetic ketoacidosis, severe hypoglycemia, coronary event, or stroke in the preceding 3 months; with a history of gastroparesis, pancreatitis, cholecystitis, previous weight loss surgery, estimated glomerular filtration rate <45 mL/min/1.73 m², or evidence of significant liver disease (known cirrhosis or liver function tests greater than three times the upper limit of normal); who were currently pregnant or breastfeeding; and with a history of proliferative diabetic retinopathy or macular edema.

Intervention

Participants were randomly assigned 1:1 to once-weekly tirzepatide 5.0 mg or identical-appearing, volume-matched, placebo saline subcutaneous injection. Tirzepatide or placebo were initiated at 2.5 mg for 4 weeks and then administered at 5.0 mg for the remaining 8-week period research facility nursing staff. A submaximal dose was permitted if adverse effects limited tolerance of the target dose. Randomization was performed by an independent clinician using a computerized minimization procedure (MinimPy version 0.3), stratified according to sex (21). The participant, outcome assessors, and data analysts were blinded to the allocation.

Prior to commencing study medication, diabetes self-management principles were discussed, including advice to monitor for change in insulin dose requirements, hypoglycemia, and ketone risk. Access to glucagon was confirmed. Participants received a weekly phone call following the first 2 doses after drug initiation, and uptitration and postrandomization insulin dose adjustment was supported by blinded study investigators using clinical discretion based on glucose response. Outside of scheduled phone calls, participants were encouraged to contact the study team for additional support if required. Participants were asked to maintain their usual diet and physical activity practices and did not receive formal lifestyle counseling. In the event of complete appetite suppression, participants were encouraged to maintain consumption of a serving of carbohydrates at mealtimes.

End Points and Assessments

The primary end point was change in body weight after 12 weeks. Prespecified secondary end points included change in body composition (fat mass and fat free mass [FFM]) assessed by air displacement plethysmography (BodPod; COSMED, Rome, Italy) and total body water measured by bioelectrical impedance (Tanita, Tokyo, Japan). Arterial stiffness was assessed by carotid femoral pulse wave velocity (cfPWV) (SphygmoCor; ATCOR Medical, Sydney, Australia) and the augmentation index (AIx) using applanation tonometry of the radial artery (SphygmoCor). Fasting plasma and serum samples were collected for metabolite assessment on pre- and posttreatment study visits, including glycated hemoglobin (HbA_1c), fasting lipid profile, glucagon, and GLP-1 and GIP (exploratory). Resting energy expenditure and respiratory quotient were assessed during the fasted state by indirect calorimetry (Q-NRG; COSMED).

We recorded continuous glucose monitoring (CGM) data, insulin dosing, and dietary intake for a 7-day period prior to baseline and end-of-study visits via insulin pump download or by insulin dose diary in participants managed with insulin injections. Insulin dose was captured at 2 and 6 weeks postrandomization for exploratory analyses in insulin pump users. Diet records (Easy Diet Diary) were analyzed using FoodWorks 10 Professional (Xyris Pty Ltd, Brisbane, Australia).

Validated questionnaires were used at baseline and 12 weeks to assess gastrointestinal symptoms (Patient Assessment of Upper Gastrointestinal Disorders Symptom Severity Index), eating behavior (Three-Factor Eating Questionnaire–R18), and physical activity (International Physical Activity Questionnaire Short Form) (22–24). Treatment satisfaction was assessed using a questionnaire modified from published methods (25) (Supplementary Material).

Statistical Analysis

Primary and secondary end points were assessed using an ANCOVA model adjusted for baseline value, with treatment as a fixed effect and the baseline measure as a continuous covariate. Regression residuals were used to assess model assumptions. Data were analyzed by intention-to-treat principles. For the primary end point, multiple imputation chained equation was applied according to missing-at-random assumptions. Secondary end points were analyzed without imputation. We used Spearman correlation to assess the relationship between changes in total daily insulin dose and body weight. Two-sided tests and a 0.05 significance level were applied to all analyses. The widths of the CIs were not adjusted for multiplicity and should not be interpreted as confirming definitive treatment effects. We calculated that 20 participants required randomization (10 per group) to demonstrate a 5-kg between-group body weight difference from baseline to 12 weeks (α = 0.05, power 80%), assuming an SD of 3.7 kg. We increased the number of participants to 24 in anticipation of an up to 20% dropout rate. Statistical analyses were performed using RStudio 2023.06.1 + 524 Mountain Hydrangea release for macOS and power calculation using G*Power version 3.1.9.4.

Data and Resource Availability

The data generated in this study have not been deposited in a public repository due to absence of consent from study participants and restrictions placed by our study’s ethics approval. Deidentified processed data are available from the corresponding author under restricted access for collaborative analyses per a data sharing agreement.

Results

Participants

Of 32 participants screened, 24 were randomized and received at least one dose of tirzepatide or placebo (Supplementary Fig. 1). Baseline demographic and clinical characteristics are shown in Table 1 and were generally similar across treatment groups. Participants in both groups had long-duration diabetes, with a shorter duration in the tirzepatide group and greater high-sensitivity C-peptide detectability relative to the placebo group. Baseline total daily insulin dose requirements were higher in the tirzepatide group, but insulin dosage was similar when expressed as units per kilogram. Overall, the 24 participants (42% female) had a mean age of 41 ± 10 years, duration of diabetes of 23 ± 12 years, body weight of 101 ± 14 kg, BMI of 33.7 ± 3 kg/m², HbA_1c of 7.3 ± 1.2%, and total daily dose of insulin of 69 ± 33 units (0.68 units/kg). More than half of the participants (54%) used insulin pumps. All pump users had automated insulin delivery systems.

Table 1.

Baseline demographic and clinical characteristics by treatment group

Characteristic	Tirzepatide (n = 12)	Placebo (n = 12)
Age (years)	40.5 ± 10.9	41.7 ± 10.2
Sex
Male	7 (58)	7 (58)
Female	5 (42)	5 (42)
Weight (kg)	104.8 ± 14.6	97.7 ± 12.9
BMI (kg/m2)	34.2 ± 3.4	33.2 ± 3.0
BMI category
30–34.9 kg/m2	7 (58.3)	8 (66.7)
35–39.9 kg/m2	4 (33.3)	4 (33.3)
≥40 kg/m2	1 (8.3)	0 (0.0)
Waist-to-hip ratio	1.0 ± 0.1	0.9 ± 0.1
Blood pressure (mmHg)
Systolic	131 ± 15	133 ± 11
Diastolic	75 ± 8	74 ± 9
Lipid profile
Total cholesterol (mmol/L)	4.8 ± 0.9	4.5 ± 0.8
LDL cholesterol (mmol/L)	3.1 ± 0.7	2.7 ± 0.7
HDL cholesterol (mmol/L)	1.3 ± 0.2	1.4 ± 0.3
Triglycerides (mmol/L)	0.9 ± 0.4	1.0 ± 0.7
Medication use
Antihypertensive	1 (8.3)	2 (16.7)
Statin	4 (33.3)	8 (66.7)
Metformin	2 (16.7)	2 (16.7)
Active smoking history	1 (8.3)	1 (8.3)
Caucasian ethnicity	11 (91.7)	11 (91.7)
HbA1c (%)	7.3 ± 1.3	7.2 ± 1.2
Insulin delivery method*
Multiple daily injections	6 (50.0)	5 (41.7)
Insulin pump therapy	6 (50.0)	7 (58.3)
Medtronic 780G	3 (25)	3 (25)
Tandem T-slim	2 (17)	3 (25)
Ypsomed	1 (8)	1 (8)
CGM time in glucose range (%)
<3.0 mmol/L (<55 mg/dL)	0.2 ± 0.4	0.8 ± 2.6
<3.9 mmol/L (<70 mg/dL)	1.7 ± 1.8	2.1 ± 4.1
3.9–10.0 mmol/L (70–180 mg/dL)	68.4 ± 14.4	61.5 ± 21.6
>10 mmol/L (>180 mg/dL)	30.0 ± 14.0	36.4 ± 22.9
>13.9 mmol/L (>250 mg/dL)	7.6 ± 5.4	16.7 ± 18.6
Mean glucose (mmol/L)	8.7 ± 1.1	9.5 ± 1.7
Mean overnight glucose (2201 h to 0600 h) (mmol/L)	8.5 ± 1.5	9.0 ± 2.9
Mean daytime glucose (0601 h to 2200 h) (mmol/L)	8.8 ± 1.1	9.6 ± 2.4
CV (%)	34 ± 7	35 ± 5
CGM sensor
Medtronic G4	3 (25)	2 (17)
Dexcom G6	7 (58.3)	6 (50)
Freestyle Libre	2 (17)	4 (33.3)
Insulin (units/day)	74.1 ± 26.8	65.2 ± 38.5
Insulin (units/kg/day)	0.7 ± 0.2	0.7 ± 0.4
Diabetes history
Diabetes duration (years)	17.1 ± 10.9	29.4 ± 11.1
History of microalbuminuria	2 (16.7)	0 (0)
History of retinopathy	1 (8.3)	2 (16.7)
History of peripheral neuropathy	1 (0)	0 (0)
Fasting high-sensitivity C-peptide detectable†	6 (54.5)	3 (25.0)

Data are mean ± SD or n (%). To convert the values for HbA_1c to mmol/mol, multiply the value by 10.93 then subtract 23.5. To convert cholesterol values to mg/dL, divide by 0.02586. CV, coefficient of variation.

*No participants used ultra–rapid-acting insulin.

†High-sensitivity C-peptide data available for n = 11 in the tirzepatide group and n = 12 in the placebo group.

Two individuals dropped out of the study, one from the tirzepatide group due to increased anxiety coinciding with difficulties attending study visits and one from the placebo group due to loss of interest in the study after a single dose of study medication. A single participant in the tirzepatide group reduced the dose from 5.0 mg to 2.5 mg following an episode of gastroenteritis coinciding with the timing of dose escalation. All remaining participants reached the target dose. Accordingly, data from 11 participants in the tirzepatide and 11 from the placebo group provided end-of-treatment measures.

Efficacy End Points

Body Weight and BMI

After 12 weeks, the mean change in body weight was −10.3 kg (95% CI −12.8 to −7.7 kg) in the tirzepatide group and −0.7 kg (95% CI −1.4 to 2.8 kg) in the placebo group, with an estimated treatment difference of −8.7 kg (95% CI −12.0 to −5.5 kg; P < 0.0001). This corresponded to a between-group difference of −8.8% body weight (95% CI −11.6 to −5.9%; P < 0.001) (Table 2 and Fig. 1A) and −3.0 kg/m² BMI points (95% CI −4.0 to −2.0 kg/m²). Adjusting for age, sex, insulin pump use, and C-peptide status did not alter the primary outcome (data not shown). In the tirzepatide group, 100% and 45% of participants experienced clinically significant weight loss of >5% and >10%, respectively, compared with 9% and 0%, respectively, in the placebo group (Fig. 1B). The single participant in the placebo group with >5% weight loss increased physical activity and restricted calorie intake.

Table 2.

Primary and secondary end points at baseline and end of treatment, by treatment group

	Tirzepatide group (n = 12 for primary outcome)			Placebo group (n = 12 for primary outcome)
Variable	Baseline	12 weeks	Difference	Baseline	12 weeks	Difference	Difference between groups	P
Primary outcome
Body weight (kg)	106.3	96.0	−10.3 (−12.8 to −7.7)	98.1	97.5	−0.7 (−1.4 to 2.8)	−8.7 (−12.0 to −5.5)	<0.0001
Anthropometric and body composition measures
BMI (kg/m2)	34.6	31.2	−3.3 (−4.0 to −2.6)	33.6	33.3	−0.3 (−1.0 to 0.5)	−3.0 (−4.0 to −2.0)	<0.0001
Total fat mass (kg)	43.5	35.2	−8.3 (−11.0 to −5.5)	41.3	40.3	−1.0 (−3.0 to 1.0)	−7.2 (−10.5 to −3.9)	0.0002
FFM (kg)	62.6	60.6	−2.0 (−4.1 to 0.2)	56.7	57.0	0.4 (−0.3 to 1.1)	−1.8 (−3.8 to 0.1)	0.06
Total fat mass (%)	41.4	37.0	−4.5 (−6.8 to −2.1)	42.4	41.5	−0.9 (−2.4 to 0.5)	−3.5 (−6.1 to −0.9)	0.01
Fat free mass (%)	58.6	63.0	4.5 (2.1 to 6.8)	57.6	58.5	0.9 (−0.5 to 2.4)	3.5 (0.9 to 6.1)	0.01
Total body water (kg)	48.6	46.8	−1.8 (−2.4 to −1.1)	44.4	45.5	1.1 (−0.8 to 3.0)	−3.1 (−1.1 to −5.0)	0.004
WHR	0.99	0.91	−0.08 (−0.13 to −0.03)	0.94	0.90	−0.04 (−0.08 to 0.01)	−0.03 (−0.09 to 0.03)	0.35
Glycemic measures
HbA1c (%)	7.01	6.50	−0.5 (−0.8 to −0.2)	7.30	7.10	−0.2 (−0.5 to 0.1)	−0.4 (−0.7 to 0.0)	0.05
TIR (3.9–10.0 mmol/L)	69.4	72.6	3.2 (−8.1 to 14.5)	60.6	63.2	2.6 (−7.0 to 12.2)	8.4 (−4.0 to 20.9)	0.17
TBR (<3.0 mmol/L)	0.24	0.72	0.47 (−0.3 to 1.2)	0.87	0.57	−0.3 (−1.1 to 0.5)	0.51 (−0.17 to 1.2)	0.14
TBR (<3.9 mmol/L)	1.81	3.42	1.6 (−1.2 to 4.4)	2.16	1.82	−0.3 (−1.4 to 0.7)	1.8 (−0.9 to 4.5)	0.17
TAR (>10 mmol/L)	28.8	24.0	−4.8 (−16.5 to 6.9)	34.6	37.5	2.9 (−7.4 to 13.2)	−10.2 (−23.4 to 3.0)	0.12
Mean glucose >24 h (mmol/L)	8.6	8.2	0.4 (−0.62 to 1.46)	9.46	9.49	−0.03 (−1.05 to 1.00)	0.81 (−0.43 to 2.05)	0.19
Mean daytime glucose (mmol/L)	8.68	8.28	0.41(−0.55 to 1.37)	9.64	9.66	−0.02 (−1.14 to −1.09)	0.87 (−0.35 to 2.10)	0.15
Mean overnight glucose (mmol/L)	8.41	7.99	0.42 (−0.95 to 1.78)	9.01	9.04	−0.03 (−1.20 to −1.15)	0.69 (−0.85 to 2.22)	0.36
Glycemic variability (CV, %)	33	31	2 (0 to 4)	35	34	0 (−4 to 4)	−2 (−6 to 2)	0.31
MAGE*	102.2	89.5	−12.8 (−24.0 to −1.5)	115.8	114.3	−1.4 (−17.1 to 14.2)	−16.5 (−31.6 to −1.4)	0.03
Insulin dose‡
Total daily insulin dose (units)†	67.42	43.24	−24.20 (−31.55 to −15.44)	55.12	54.79	−0.33 (−3.54 to 3.96)	0.65 (0.53 to 0.79)	0.0002
Total daily insulin dose (units/kg)†	0.64	0.48	−0.19 (−0.27 to −0.10)	0.57	0.64	0 (−0.04 to 0.05)	0.71 (0.60 to 0.85)	0.0009
Basal insulin dose (units)†	31.43	24.04	−7.39 (−11.00 to −4.47)	30.25	31.21	0.95 (−1.00 to 3.03)	0.75 (0.63 to 0.88)	0.002
Bolus insulin dose (units)†	33.8	15.8	−18.0 (−23.5 to −9.7)	27.5	24.6	−3.0 (−5.8 to 0.2)	0.51 (0.31 to 0.81)	0.008
Cardiovascular measures
cfPWV (m/s)	8.2	8.9	0.8 (−1.0 to 2.5)	8.5	8.5	0 (−0.6 to 0.6)	0.6 (−0.6 to 1.7)	0.33
AIx (%)	12.2	10.6	−1.6 (−10.9 to 7.8)	14.6	12.0	−2.6 (−7.7 to 2.4)	0.3 (−8.4 to 8.9)	0.95
Pulse rate (beats/min)	71.5	76.2	4.7 (−0.4 to 9.8)	71.0	68.3	−2.7 (−8.2 to 2.7)	7.7 (2.1 to 13.3)	0.01
SBP (mmHg)	131.5	128.0	−3.5 (−9.6 to 2.6)	133.86	130.82	−3.1 (−11.2 to 5.1)	−1.3 (−10.1 to 7.6)	0.77
DBP (mmHg)	74.8	74.8	0.05 (−5.6 to 5.7)	73.7	72.6	−1.0 (−5.4 to 3.4)	1.7 (−3.6 to 7.0)	0.51
Biochemistry
Lipids
Total cholesterol	4.9	4.4	−0.5 (−1.5 to 0.5)	4.46	4.42	−0.1 (−0.3 to 0.2)	−0.2 (−1.0 to 0.7)	0.66
LDL	3.1	2.7	−0.4 (−1.2 to 0.4)	2.61	2.64	−0.02 (−0.3 to 0.2)	−0.1 (−0.8 to 0.7)	0.85
HDL	1.3	1.2	−0.1 (−0.22 to 0.02)	1.41	1.48	0.1 (−0.2 to 0.4)	−0.2 (−0.5 to 0.1)	0.22
TG†	0.9	0.9	−0.02 (−0.2 to 0.2)	0.82	0.83	0.01 (−0.1 to 0.1)	1.0 (0.8 to 1.3)	0.97
AST	24.6	27.3	2.6 (−12.0 to 17.2)	22.4	22.5	0.1 (−2.6 to 2.8)	1.5 (−12.7 to 15.6)	0.83
ALT	28.7	22.4	−6.4 (−14.3 to 1.5)	21.5	19.0	−2.5 (−5.6 to 0.7)	2.7 (−11.2 to 5.8)	0.51
GGT	26.2	15.9	−10.2 (−16.8 to −3.7)	17.4	14.9	−2.5 (−4.8 to −0.1)	−3.2 (−6.9 to 0.5)	0.08
Fasting C-peptide (pmol/L)	51.4	55.5	4.1 (−34.3 to 42.5)	125.5	114.2	−11.2 (−36.3 to 13.9)	−8.24 (−63 to 46.5)	0.71
Fasting glucagon (pmol/L)†	5.2	3.6	−1.6 (−0.4 to −2.4)	5.4	4.0	−1.5 (−0.5 to −2.2)	1.1 (0.8 to 1.4)	0.73
Fasting GLP-1 (pmol/L)†	14.9	12.2	−2.7 (−5.4 to 0.6)	13.13	15.32	2.2 (−0.6 to 5.6)	1.4 (1.1 to 1.8)	0.01
Fasting GIP (pmol/L)*	21.0	22.2	−1.2 (−5.8 to −3.4)	16.3	16.0	0.3 (−2.4 to −3.0)	−3.8 (−8.3 to −0.7)	0.09
Energy expenditure
Energy expenditure (kcal/day)	1,986	1,812	−174 (−333 to −15)	1,857	1,789	−68 (−238 to 101)	−120 (−306 to 66)	0.19
Energy expenditure (kcal/kg FFM/day)	32.69	30.13	−2.56 (−6.11 to 0.99)	33.21	31.68	−1.53 (−4.78 to 1.71)	−1.39 (−4.59 to 1.82)	0.37
RQ	0.76	0.69	−0.07 (−0.10 to −0.03)	0.74	0.69	−0.06 (−0.10 to −0.02)	0.00 (−0.02 to 0.03)	0.72

Data are mean (95% CI). The between-group difference is the difference between the tirzepatide and placebo groups at 12 weeks, adjusted for baseline value, by ANCOVA. To convert the values for HbA_1c to mmol/mol, multiple the value by 10.93 then subtract 23.5. Bolded values refer to statistically significant findings. Data available for n = 12 tirzepatide and n = 12 placebo for the primary outcome with imputation, with n = 11 tirzepatide and n = 11 placebo for all other outcomes. DBP, diastolic blood pressure; GGT, γ-glutamyl transferase; MAGE, mean amplitude of glycemic excursion; RQ, respiratory quotient; SBP, systolic blood pressure; TAR, time above range; TBR, time below range; TG, triglycerides; TIR, time in range; WHR, waist-to-hip ratio.

*Fasting GIP, MAGE, and RQ measures were exploratory. Due to resource constraints, secondary outcomes not analyzed or reported included fasting free fatty acids and serum hs-CRP.

†Ratio of geometric means (log-transformed data; back-transformed to ratio scale due to skewed distribution). A value <1.0 indicates a reduction and >1.0 an increase in the treatment group compared with placebo.

‡For insulin dose, between-group differences are ratio of geometric means (95% CI).

Figure 1.

Effect of once-weekly tirzepatide compared with placebo on body weight (primary outcome) and body composition in adults with type 1 diabetes. A: The change from baseline in body weight at 12 weeks as assessed by ANCOVA, adjusted for baseline, with multiple imputation for missing data at 12 weeks. B: Participant-level percent change in body weight ordered on the x-axis from greatest decrease to increase, plotted by treatment group. All participants randomized to tirzepatide experienced a decreased body weight by ≥5% and 45% a decreased body weight by ≥10% over the 12-week treatment period compared with 9% and 0% in the placebo group. C: Change in fat mass. D: Change in FFM. Error bars represent the 95% CI.

Body Composition

There was a greater decrease in fat mass in the tirzepatide group compared with the placebo group (estimated treatment difference −7.2 kg [95% CI −10.5 to −3.9 kg]; P = 0.0002), while the between-group difference in FFM was not significant (Table 2). Figure 1C and D shows fat mass and FFM at baseline and 12 weeks.

Glycemia and Insulin Dose

At 12 weeks, changes in the mean HbA_1c level were −0.5% (95% CI −0.8 to −0.2%) with tirzepatide, and −0.2% (95% CI −0.5% to 0.1%) with placebo (estimated treatment difference −0.4% [95% CI −0.7% to 0.0%]; P = 0.05).

The change in glycemia was observed alongside profound reductions in total daily insulin dose in the tirzepatide group (between-group difference −35% of baseline total daily insulin dose [95% CI −47 to −21%]; P = 0.0002). Expressed as insulin units per kilogram of body weight, the between-group difference was −29% (95% CI −40 to −15%; P = 0.0009) (Table 2). Percentage change in total, basal, and bolus insulin are shown in Fig. 2.

Figure 2.

Effect of once-weekly tirzepatide compared with placebo on insulin dose in adults with type 1 diabetes. Insulin dose was determined via download from participants managed with insulin pump therapy or by a logbook for those managed with multiple daily insulin injections. A: Percent change from baseline in total daily insulin dose at 12 weeks. B: Change in basal insulin dose. C: Change in bolus insulin dose. D: Change in total daily insulin dose from baseline at 2, 6, and 12 weeks after randomization in 14 participants using insulin pumps. At 6 weeks, total daily insulin dose was −42% from baseline (95% CI −59 to −18%) vs. placebo (P = 0.007). Insulin dose data at 2 and 6 weeks are exploratory only. Error bars represent the 95% CI. **P < 0.01, ***P < 0.001.

In an exploratory subgroup analysis according to insulin delivery method, significant total daily insulin dose reduction was observed in both insulin pump (−31% [95% CI −48 to −8%]; P = 0.02) and multiple daily injection (−34% [95% CI −54 to −6%]; P = 0.03) users from baseline to 12 weeks. In an exploratory analysis of the insulin pump users, insulin dose reductions were observed in the tirzepatide group as early as 2 weeks, with a statistically significant between-group difference demonstrated 6 weeks after start of treatment (−42% of baseline insulin dose [95% CI −59 to −18%]; P = 0.007) (Fig. 2D). In an exploratory analysis, change in total daily dose was associated with change in body weight (r = 0.745; P < 0.001) and remained significant when total daily dose was expressed as units per kilogram (r = 0.605; P = 0.003).

There was no difference in time-in-range, time-below-range, or other CGM metrics (Table 2). In an exploratory analysis, tirzepatide was associated with reduced glucose variability (mean amplitude of glycemic excursion estimated treatment difference −16.5 [95% CI −31.6 to −1.4]; P = 0.03).

Energy Expenditure, Food Intake, and Physical Activity

Although resting energy expenditure decreased in the tirzepatide group, there was no overall difference compared with placebo or after adjustment for FFM (Table 2). There was no significant between-group difference in substrate use as assessed by the respiratory quotient (Table 2). Tirzepatide reduced energy intake compared with placebo (mean difference −429 kcal/day [95% CI −852 to −5 kcal/day]; P = 0.05) (Supplementary Table 1). There were no differences between groups in measures of cognitive restraint, uncontrolled or emotional eating, and self-reported physical activity levels (Supplementary Table 1).

Other Metabolic and Cardiac End Points

There was no significant between-group difference in fasting lipids, blood pressure, cfPWV or AIx (Table 2). At 12 weeks, pulse rate was unchanged within both groups; however, the between-group comparison showed an overall increase in the tirzepatide group (Table 2).

Safety and Tolerability

Adverse events are described in Supplementary Table 2. There were no episodes of diabetic ketoacidosis, severe hypoglycemia, or serious adverse events. Nonserious adverse events were reported in 75% and 8.3% randomized to tirzepatide and placebo, respectively. Gastrointestinal-related symptoms were the most frequently reported adverse events. While nausea was reported in 50% randomized to tirzepatide, most cases were mild and transient during the week following dose initiation or escalation, and there was overlap with an episode of gastroenteritis in a participant with infective household contacts. Yet, based on the self-reported psychometric evaluation of the Patient Assessment of Upper Gastrointestinal Disorders Symptom Severity Index questionnaire, there was no between-group treatment difference in nausea or other symptom subscales adjusted for baseline symptoms, aside from a significant increase in fullness and early satiety from baseline in the tirzepatide group versus placebo group (Supplementary Fig. 2 and Supplementary Table 3). In the tirzepatide group, a diagnosis of depression occurred in an individual with preexisting symptoms. Increased anxiety was associated with treatment discontinuation in a participant who had a preexisting history and was thought to be related to difficulties in meeting the demands of weekly in-facility dosing during the trial.

Based on patient-reported outcomes, 100% of participants in the tirzepatide group felt that the treatment addressed their diabetes treatment needs compared with 37.5% in the placebo group. Participants receiving tirzepatide also reported higher levels of moderate or complete satisfaction across several domains, including mental fatigue and emotional distress and satisfaction with insulin dosing (Supplementary Table 4).

Conclusions

In this first phase 2 randomized placebo-controlled trial in individuals with type 1 diabetes, treatment with tirzepatide titrated to 5.0 mg for 12 weeks led to significantly greater weight loss than placebo. At 12 weeks, tirzepatide induced an 8.7-kg greater reduction in weight over placebo from baseline, equating to an 8.8% reduction in body weight during the trial period. Reassuringly, there were no episodes of severe hypoglycemia or ketoacidosis during our study.

The weight-reducing effect of tirzepatide in type 1 diabetes is consistent with findings observed in other populations. The Study of Tirzepatide (LY3298176) in Participants With Obesity or Overweight (SURMOUNT-1) reported that tirzepatide 5.0 mg over 72 weeks led to at least 5% weight loss in 85% of participants with obesity (without diabetes) vs. 35% in the placebo group (15). Ten-percent weight loss was achieved by 69% and 19%, respectively. Similarly, tirzepatide 5.0 mg induced at least 5% weight loss in 56% of participants with insulin-treated type 2 diabetes over 40 weeks (mean baseline BMI 33.4 kg/m²) and 10% weight loss in 24% with tirzepatide vs. 7% and 1%, respectively, with placebo (26). Remarkably in our study, 100% and 45% of participants randomized to tirzepatide achieved at least 5% and at least 10% body weight reduction, respectively, over a relatively short treatment period, without serious adverse events necessitating treatment discontinuation. Together, these data indicate that low-dose tirzepatide effectively induced clinically meaningful weight loss in people with type 1 diabetes, yet the observed magnitude of effect surpassed outcomes reported in other populations.

Although ours is the first randomized controlled trial of a dual GIP and GLP-1 receptor agonist, prior studies have investigated the metabolic effects of selective GLP-1 RAs as an adjunct to insulin in type 1 diabetes. In two longer and larger phase 3 trials over 26 and 52 weeks, the short-acting GLP-1 RA liraglutide reduced body weight by approximately 5 kg, HbA_1c by 0.2%, and insulin dose by 10% compared with placebo, albeit with increased rates of hypoglycemia and ketosis (27,28). In a 15-week study, the longer-acting GLP-1 RA semaglutide (up to 1.0 mg) was associated with 5 kg of weight loss, a 4.8% improvement in time in range, and a 22% reduction in insulin dose compared with placebo when used alongside an investigational automated insulin delivery system for the final 4 weeks of the trial period (29). In a longer trial with similar BMI characteristics to our study, semaglutide over 26 weeks increased time in range by 8.8% and reduced HbA_1c by 0.3%, weight by 8.8 kg, and insulin dose by 22.3 units/day (30). Therefore, the effects of tirzepatide in our study were comparable to, or exceeded, those reported with liraglutide and semaglutide but within a shorter treatment time frame.

Our findings also align with real-world studies of tirzepatide in type 1 diabetes. In a cohort of 84 adults followed up for 21 months, body weight decreased by ∼23% and insulin dose by 38 units (0.2 units/kg/day), with a median tirzepatide dose of 10 mg (31). These findings suggest the potential for durable weight loss beyond the 12-week treatment period assessed in our trial, and tolerance of higher doses of tirzepatide than tested in our study.

We assessed for several possible mechanisms for tirzepatide-driven weight loss in type 1 diabetes. Overall energy intake was reduced, aligning with reported effects of tirzepatide on appetite and food consumption in type 2 diabetes (17). However, the degree of weight loss exceeded the ∼5 kg of weight loss expected for the 429-kcal reduction in calorie intake (32). Insulin dose reduction may also have contributed to the marked weight loss observed in our study by alleviating insulin-associated weight-promoting effects. We found no difference between treatment groups in self-reported physical activity levels, thereby excluding the confounding effects of exercise on weight loss. Furthermore, there was no observable difference in energy expenditure after adjusting for FFM, mirroring a study of tirzepatide 15 mg compared with placebo performed in obesity without diabetes (33). Indeed, no human study has replicated the increases in energy expenditure and blunted metabolic adaptation reported in tirzepatide-treated mice (33).

The degree and acuity of reduction in insulin dose requirements with tirzepatide were observed as early as the 2nd week of treatment in insulin pump users. Notably, the reduction in insulin dose was primarily driven by glucose response and insulin pump algorithm. Weight loss can improve insulin sensitivity. However, the rapidity in change in total daily insulin dose and the reduction in insulin dose adjusted for body weight (units/kg) suggest that tirzepatide-induced metabolic effects may occur independent of weight loss. We speculate that an acute reduction in food intake, coupled with rapid improvements in insulin sensitivity, may account for the observed early insulin dose reduction in our study. However, given the bidirectional relationship among energy intake, insulin dose requirements, and weight, our study cannot disentangle the independent effects of one from the other.

A greater number of nonserious adverse events were reported in the tirzepatide group. Gastrointestinal effects were common after drug commencement or dose escalation in keeping with reports of tirzepatide in type 2 diabetes and obesity without diabetes (14,15). Yet, these gastrointestinal effects did not lead to treatment discontinuation. Depression and anxiety in the tirzepatide group occurred in participants with a preexisting history and were thought to be exacerbated by the time demands of protocolized weekly facility visits. Although there is mixed evidence suggesting a link between GLP-1 RA use and anxiety and depression, the same association has not been reported with tirzepatide (34,35).

Our trial had several strengths. First, we included comprehensive anthropometric, glycemic, hormonal, and mechanistic measures, spanning food intake, physical activity, eating behavior, and tolerability. Second, we designed our study to be generalizable to a population with long-standing type 1 diabetes and excess body weight and did not exclude participants based on HbA_1c or use of metformin. Importantly, our satisfaction surveys strongly favored tirzepatide over placebo, highlighting patient-reported outcomes of adjunctive tirzepatide that were not fully captured by metabolic outcomes.

A limitation of our study was that we only performed fasting assessments of incretin hormones and glucagon. Although we did not demonstrate a change in fasting glucagon, glucagon excess in type 1 diabetes occurs more prominently in the prandial state, and meal studies are required to further assess this (36). Additionally, the weekly dose of tirzepatide or placebo was delivered in our research unit, which does not reflect home environments. Despite random assignment to treatment groups, the tirzepatide group had higher baseline insulin dose requirements and weight. However, expressing change as a percent change reduces the bias introduced by this imbalance. Similarly, while HbA_1c showed a statistically borderline improvement, CGM-derived metrics of glycemia, such as time in range, did not change. Our study was likely underpowered to assess glycemic superiority and may have been underpowered for other secondary outcomes. As this was a phase 2 trial, we did not define end points according to Food and Drug Administration phase 3 guidance for weight management interventions and did not counsel research participants regarding diet or lifestyle recommendations. Finally, this trial was designed to assess low-dose tirzepatide over a 3-month period, and a longer study duration was constrained by the burdens of weekly study visits. Reduction in weight would need to be assessed for a longer period and in a larger cohort to determine sustainability of effect, alongside long-term safety and glycemic and other metabolic outcomes. This trial serves as strong rationale for further evaluation of tirzepatide in type 1 diabetes in upcoming phase 3 trials (A Study of Tirzepatide [LY3297176] Compared With Placebo in Adults With Type 1 Diabetes and Obesity or Overweight [SURPASS-T1D-1] and SURPASS-T1D-2) (37,38).

In summary, we found that tirzepatide resulted in a greater reduction in weight than placebo among adults with type 1 diabetes and obesity. The results from our study demonstrate that tirzepatide has profound and clinically significant metabolic effects in type 1 diabetes.

This article contains supplementary material online at https://doi.org/10.2337/figshare.30455297.

Funding Statement

J.R.S is supported by Breakthrough T1D (formerly JDRF) Type 1 Diabetes Clinical Research Network grant 3-SRA-2023-1296-M-N and is the recipient of the Commonwealth of Australia grant for Accelerated Research under the Medical Research Future Fund. R.F. is supported by the University Postgraduate Award (University of New South Wales). The project received funding from St. Vincent’s Clinic Foundation, the Australian Diabetes Society Lindsey Baudinet Rising Star Award in Type 1 Diabetes Research, and Melissa and Jonathon Green.