Dulaglutide: what the trials show, status and safety

Summary

Dulaglutide is a laboratory-made protein that copies a natural gut hormone and switches on the same docking point. It has been an approved, prescription-only medicine since 2014, licensed for type 2 diabetes and, in America, for lowering heart risk. In a trial of 9,901 adults followed for a median of 5.4 years, serious heart events were about 12 percent less common. It is licensed nowhere for weight loss, and in the one trial that placed it beside semaglutide it came off worse on both blood sugar and weight.

Key findings at a glance

• Approved since 2014, and on prescription in every market here. America licensed it on 18 September 2014 as an entirely new active substance, and the European Union followed on 21 November 2014 ^[1], ^[2].
• About 12 percent fewer serious heart events. In 9,901 adults with type 2 diabetes, such an event struck 12.0 percent against 13.4 percent. They were followed for a median of 5.4 years (hazard ratio 0.88, range 0.79 to 0.99, p = 0.026) ^[3].
• No weight-loss licence anywhere. The approved uses are type 2 diabetes and, in America, the lowering of heart risk. Neither the American nor the European text mentions weight management ^[4], ^[5].
• Semaglutide beat it in the one head-to-head trial. Over 40 weeks in 1,201 adults, weight fell by 6.5 kg against 3.0 kg in the higher pair of strengths, a gap of 3.55 kg (p < 0.0001) ^[6].
• In children and teenagers, blood sugar improved but weight did not. In 154 young people aged 10 to 17, the blood sugar marker fell while body mass index showed no difference against the dummy treatment ^[7].
• A framed American warning that Europe does not carry. The American insert warns about thyroid tumours seen in rats. The European text carries no such box and names only allergy as a bar ^[4], ^[5].
• It is a biologic, so it cannot legally be copied by a compounding pharmacy. The FDA answers the question "Can biologics be compounded?" with a flat "No" ^[8], ^[1].

What it is

Dulaglutide is a large made-to-order protein that imitates the gut hormone GLP-1. Eli Lilly developed it under the code LY2189265 and sells it under a single brand name worldwide, Trulicity ^[4], ^[9]. The brand is named here only so that readers can match the substance to the product they have seen.

Its build explains almost everything else about it. Two identical chains are locked together by sulphur bridges. Each chain carries a copy of the gut hormone at one end, joined by a short link to the tail of a human antibody ^[4].

That antibody tail is the trick. It keeps the molecule in circulation for about five days, where the natural hormone survives minutes. The whole thing weighs roughly 63 kilodaltons and is brewed in cultured hamster ovary cells, because a protein this size cannot be built by chemistry ^[4], ^[9].

Quick facts

How it works

The approved insert sets out the mechanism, because it was assessed alongside the trials ^[4].

• The GLP-1 docking point. Dulaglutide attaches to the receptor for the gut hormone GLP-1 and switches it on, as the natural hormone does ^[4].
• Insulin, but only when needed. Inside the insulin-producing cells of the pancreas the switch raises a messenger molecule, and insulin is released only while blood sugar is actually high ^[4].
• Less of the opposing hormone. Glucagon, which pushes blood sugar up, is released in smaller amounts ^[4].
• The stomach empties more slowly, which flattens the rise in blood sugar after a meal ^[4].
• What the body does with it. It is thought to be taken apart into its building blocks by the ordinary machinery that breaks proteins down ^[4].
• Why it lasts. The antibody tail, not a fatty chain, gives it a half-life of about five days, with steady levels reached after two to four weeks ^[4].

Semaglutide and liraglutide belong to the same family but are built differently. They are short peptides with a fatty chain that binds them to a carrier protein in the blood; dulaglutide uses an antibody tail instead ^[15], ^[4].

What the trials found

The heart

Phase 3 RCT, built to show a difference The heart trial enrolled 9,901 adults aged 50 and over with type 2 diabetes, at 371 centres in 24 countries. Only 31.5 percent had established heart disease; 62.8 percent merely carried several risk factors. They were followed for a median of 5.4 years ^[3].

The combined endpoint counted heart attack, stroke and death from a heart cause. It struck 594 people of 4,949 in the treated group, 12.0 percent, against 663 of 4,952 on the dummy treatment, 13.4 percent. That is 12 percent less risk (hazard ratio 0.88, range 0.79 to 0.99, p = 0.026) ^[3].

Taken apart, the result rests largely on strokes people survived: 2.7 percent against 3.5 percent (hazard ratio 0.76, range 0.61 to 0.95). Heart attacks people survived showed no clear difference (0.96, range 0.79 to 1.16), nor did death from a heart cause (0.91, range 0.78 to 1.06) ^[3].

Death from any cause did not fall clearly either: 10.8 percent against 12.0 percent (hazard ratio 0.90, range 0.80 to 1.01, p = 0.067). Stomach and bowel complaints were far commoner in the treated group, 47.4 percent against 34.1 percent (p < 0.0001) ^[3].

This result was written into the American labelling on 21 February 2020, as a new approved use ^[16].

The kidneys

Exploratory analysis of a randomised trial The same 9,901 people were examined for kidney damage over 51,820 person-years. The combined kidney endpoint occurred in 17.1 percent against 19.6 percent (hazard ratio 0.85, range 0.77 to 0.93, p = 0.0004) ^[17].

Nearly all of that came from one component, new heavy protein loss in the urine (hazard ratio 0.77, range 0.68 to 0.87, p < 0.0001).

The rest did not move. A sustained fall in kidney filtration of at least 30 percent missed significance (0.89, range 0.78 to 1.01, p = 0.066). So did the need for long-term dialysis (0.75, range 0.39 to 1.44, p = 0.39) ^[17].

The authors themselves call this analysis exploratory, which is the reason it never became an approved use ^[17].

Blood sugar

Phase 3 RCT The registration programme compared dulaglutide against a long list of alternatives. Against sitagliptin over 104 weeks in 1,098 adults, the blood sugar marker HbA1c fell by 0.99 and 0.71 percentage points in the two treated arms. Under sitagliptin it fell by 0.32 points (p < 0.001) ^[18].

Against metformin alone in 807 adults, the fall at 26 weeks was 0.78 and 0.71 points against 0.56 ^[19]. Added to a sodium-glucose transporter blocker in 424 adults, it was 1.34 and 1.21 points against 0.54 under the dummy treatment (p < 0.0001) ^[20].

Against insulin glargine in 810 adults on two tablets, the higher arm was superior and the lower arm merely no worse: 1.08 and 0.76 points against 0.63 ^[21]. Alongside meal-time insulin in 884 adults, both arms edged ahead of glargine at 26 weeks, by 0.22 points (p = 0.005) and 0.17 points (p = 0.015) ^[22].

Phase 3 RCT Added to glimepiride in 300 adults over 24 weeks, HbA1c fell 1.3 points further than under the dummy treatment (range 1.0 to 1.6, p < 0.001). A value below 7.0 percent was reached by 55.3 percent against 18.9 percent. Weight fell against baseline, but the difference between the groups was not significant ^[23].

Higher against lower strengths

Phase 3 RCT with error control A trial in 1,842 adults on metformin tested three strengths against each other over 36 weeks. The highest arm beat the lowest on HbA1c, 1.77 against 1.54 percentage points, a gap of 0.24 points (p < 0.001) ^[24].

Weight followed the same order: 4.6 kg against 3.0 kg, a gap of 1.6 kg (p < 0.001). The middle arm beat the lowest only on the analysis that ignores people who stopped (gap 0.17 points, p = 0.003), not on the stricter one (gap 0.10 points, p = 0.096) ^[24].

Sickness rose with strength as well, from 13.4 to 16.4 percent, and vomiting from 5.6 to 9.3 percent ^[24]. This trial supported the American approval of the additional strengths on 3 September 2020 ^[25].

Against semaglutide

Phase 3b RCT, open-label One trial placed the two substances side by side in 1,201 adults on metformin, over 40 weeks at 194 centres in 16 countries. It was run by the maker of the comparator, which is worth holding in mind ^[6].

Dulaglutide lost on both counts. In the higher pair of strengths HbA1c fell 1.4 points against 1.8 under semaglutide, a gap of 0.41 points (range 0.57 to 0.25, p < 0.0001). In the lower pair the gap was 0.40 points ^[6].

Weight fell by 3.0 kg against 6.5 kg in the higher pair, a gap of 3.55 kg (range 2.78 to 4.32, p < 0.0001). In the lower pair it was 2.3 kg against 4.6 kg ^[6]. Stomach and bowel complaints were the commonest reason for stopping in both groups; 72 of 1,201 people left the trial ^[6].

Against liraglutide

Phase 3 RCT, non-inferiority design A second head-to-head trial enrolled 599 adults on metformin at 62 centres in nine countries. Here the weekly substance merely had to prove itself no worse than the daily one, and it did ^[26].

HbA1c fell by 1.42 points against 1.36, a difference of 0.06 points (range 0.19 lower to 0.07 higher, p < 0.0001 for non-inferiority). Sickness affected 20 percent against 18 percent, and 18 people in each group, 6 percent, stopped because of side effects ^[26].

Children and teenagers

Phase 3 RCT, double-blind A trial in 154 young people aged 10 to 17 ran for 26 blinded weeks. Their average age was 14.5 years, 71 percent were girls, and their average weight was 90.5 kg ^[7].

HbA1c rose by 0.6 points on the dummy treatment and fell by 0.6 and 0.9 points in the two treated arms (p < 0.001 for both). Reaching a value below 7.0 percent was managed by 51 percent of the treated group against 14 percent ^[7].

The important negative result sits in the same trial. Body mass index showed no difference at all between the groups, in young people who by entry criterion were above the 85th percentile ^[7]. This trial supported the American extension to age 10 on 17 November 2022 ^[27].

Kidney disease

Phase 3 RCT, open-label against insulin A trial in 577 adults with moderate to severe kidney disease compared the substance with insulin glargine. Both treated arms were no worse on HbA1c at 26 weeks. Average filtration at entry was 38 millilitres per minute ^[28].

At 52 weeks filtration was higher in the treated arms, 34.0 and 33.8 against 31.3 (p = 0.005 and p = 0.009). But protein loss in the urine did not differ significantly, at 22.5 and 20.1 percent against 13.0 percent ^[28].

The American insert adds a finding that cuts the other way. Fasting blood sugar rose by 6 and 14 mg per decilitre in the treated arms while falling by 23 under insulin ^[4].

What is still unknown

• Whether the rat tumours matter for people. Neither clinical nor laboratory work has settled it, and the insert says so ^[4].
• What to do about stomach contents reaching the lungs during anaesthesia. The label states outright that the evidence is too thin to draw practical conclusions from ^[4].
• Anything below age 10. Safety and effect have not been established, and no data exist ^[4].
• Use in end-stage kidney failure. The European text calls the experience very limited and does not recommend it ^[5].
• Passage into breast milk. No data exist on milk, on the nursing infant or on milk production ^[4].
• Damage to genetic material. No such studies were run, because the substance is a recombinant protein ^[4].
• A possible eye signal discussed for another substance in this class. The dulaglutide label says nothing about it, and nothing may be read into that silence ^[4].

Side effects and safety

Stomach and bowel complaints dominate. In the pooled placebo-controlled adult trials, sickness affected 12.4 percent in the lower and 21.1 percent in the higher strength arm, against 5.3 percent on the dummy treatment ^[4].

Diarrhoea affected 8.9 and 12.6 percent against 6.7 percent, vomiting 6.0 and 12.7 percent against 2.3 percent, and abdominal pain 6.5 and 9.4 percent against 4.9 percent. Complaints of this kind of any sort affected 32 and 41 percent against 21 percent ^[4].

Stopping over them was uncommon: 1.3 percent in the lower and 3.5 percent in the higher arm, against 0.2 percent. Investigators rated 7 and 11 percent of the episodes as severe ^[4].

Further points from the inserts and the regulatory record:

• The framed warning about thyroid tumours. In male and female rats, dulaglutide causes tumours of a particular thyroid cell, more often the more they get and the longer they get it. Whether it does so in people is unknown ^[4].
• Two human cases in the record. One medullary thyroid cancer occurred in a trial participant whose marker was already about eight times the upper limit before treatment. One case of cell overgrowth with a raised marker came from the heart trial ^[4].
• What the animal work showed. Rat tumours appeared from three times the human exposure upwards. In a six-month study in a transgenic mouse strain, no such increase occurred at all ^[4].
• Who must not have it in America. Anyone with that thyroid cancer in their own or their family history, anyone with a particular inherited hormone syndrome, and anyone who has reacted severely to the substance ^[4].
• Europe bars far less. The European text names one contraindication only, allergy to the substance or an ingredient, and carries no framed warning ^[5].
• Sudden inflammation of the pancreas. Cases including bleeding and tissue death, some fatal, have been seen with this class. In the registration trials confirmed cases ran at 1.4 per 1,000 patient-years against 0.88 ^[4].
• Blood sugar dropping too low. The risk rises sharply in combination with insulin or with tablets that push insulin out. With meal-time insulin over 52 weeks, low readings affected 77 and 69 percent; without such a partner the rate was under 1 percent ^[4].
• A faster pulse and a slower electrical signal. Resting heart rate rose by 2 to 4 beats a minute. A first-degree block of the heart's electrical conduction occurred in 1.7 and 2.3 percent against 0.9 percent ^[4].
• Pancreatic enzymes. Lipase and pancreatic amylase rose by 14 to 20 percent on average, against up to 3 percent on the dummy treatment ^[4].
• Kidney injury after fluid loss. Sudden kidney failure, including cases needing dialysis, has been reported since approval, mostly after vomiting or diarrhoea ^[4].
• Eyes and gallbladder. In the heart trial, retinal complications occurred in 1.9 percent against 1.5 percent, and gallstones at 0.62 against 0.56 per 100 patient-years ^[4].
• Severe allergic reactions. Anaphylaxis and swelling beneath the skin have been reported since approval ^[4].
• Antibodies against the drug. Of 3,907 adults, 1.6 percent developed them, and 0.9 percent developed neutralising ones. No effect on safety or benefit was found ^[4].
• Not for type 1 diabetes. The European text states plainly that it is no substitute for insulin ^[5].

The European frequency table sorts the same picture differently. Sickness, diarrhoea, vomiting and abdominal pain count as very common, meaning more than 1 person in 10. Gallstones and injection-site reactions count as uncommon, and sudden pancreatic inflammation as rare ^[5].

Why this page lists no doses

Dulaglutide is a prescription-only biological medicine in every market examined here. Choosing an amount belongs to the approved product information and to the doctor writing the prescription. This page therefore refers to trial arms only as the lower, middle or higher tested strength, and reports what was measured at each.

Development and approval status

• 2014 — America approves it on 18 September as an entirely new active substance, and the European Union follows on 21 November ^[1], ^[2].
• 2015 — Switzerland authorises it on 6 May and Japan on 3 July ^[11], ^[29].
• 2018 and 2019 — American supplements add data on kidney disease and on combination with a sodium-glucose transporter blocker ^[31], ^[32].
• 2020 — the heart use is added on 21 February, and further strengths on 3 September ^[16], ^[25].
• 2022 — the approved age drops to 10 years on 17 November ^[27].
• 2023 — the European agency publishes a supply shortage notice on 28 August ^[30].
• 2026 — that shortage is recorded as resolved, and the American insert tightens its warning on severe stomach and bowel reactions in March ^[30], ^[4].

Three entries need a word of explanation. The British authorisations were carried over from the European ones and sit under four national numbers today ^[33], ^[34]. The Australian date is that of the current poisons instrument, not of first registration, because the register itself was unreachable ^[35], ^[36].

The Canadian date is the marketing date of the pen product; the date of first regulatory clearance could not be retrieved ^[37], ^[38]. Two markets could not be settled at all. The Russian register returned nothing even for a control query, and China was not checked against an official source ^[39].

The approved uses differ between markets. America lists two: type 2 diabetes from age 10, and the lowering of serious heart events in adults with diabetes and either established heart disease or several risk factors ^[4]. Europe lists type 2 diabetes from age 10 and nothing else ^[5].

Switzerland is the only other market here whose register names the prevention of heart events in its indication field ^[11].

There is no obesity licence in any market, so the German question of who pays for a weight treatment does not even arise here. Dulaglutide is a diabetes medicine and falls outside the statutory exclusion for lifestyle products ^[40].

Shortages, and what does not follow from them. The European shortage notice opened on 28 August 2023 and covered all four strengths of the pen. It was marked resolved, with a last update of 17 February 2026 ^[30].

For America nothing comparable can be shown. The substance is absent from the shortage database today, and resolved entries are dropped from that database over time. An earlier American shortage can therefore be neither proven nor ruled out ^[41], ^[42].

Prescribing data show what shortages do to a whole class. In Australia, dulaglutide prescriptions rose 53 percent between April and July 2022 as semaglutide fell, then dropped 17 percent from August ^[43].

A British time-series analysis dates the start of the squeeze there to July 2023 ^[44]. A review describes a forced pause across every available substance in the class ^[45].

Stopping has consequences. In 69 people who had to come off the drug during the worldwide shortage, HbA1c rose from 7.0 to 8.1 percent within three months. Fasting blood sugar rose from 129 to 156 mg per decilitre (p < 0.001 for both) ^[46].

Why no compounded version can be legal. This is the sharpest difference from semaglutide, liraglutide and tirzepatide. Those are licensed in America as drugs, and while they sit on a shortage list, the law lets pharmacies copy them ^[8].

Dulaglutide is licensed as a biological product instead, under a different statute ^[1]. The FDA puts it plainly: "Can biologics be compounded? No. Biological products are not eligible for the exemptions for compounded drugs under sections 503A and 503B of the FD&C Act." ^[8]

So there is no legal route to a compounded dulaglutide, shortage or no shortage. The substance is likewise absent from the list of bulk substances used in compounding, and from the American warning page about unapproved GLP-1 products, which names other substances ^[47], ^[48]. The wider framework is set out under are peptides legal and FDA-approved peptides.

The grey market, and why it looks different here. A targeted literature search for counterfeit or illegally sold dulaglutide returned nothing. The same searches for semaglutide return several papers ^[49]. The reason is structural rather than reassuring: a 63-kilodalton protein grown in cell culture cannot be made in the workshops that supply short research peptides ^[4].

Where it does leave the regulated supply chain, the literature describes diverted or falsified finished pens and online-pharmacy purchase, not raw powder ^[50], ^[51].

A safety-report analysis found 3,348 reports of accidental overdose across this class. Reporting ratios were raised for every substance in it, dulaglutide included (range 2.64 to 61.12, all p < 0.008) ^[50]. A separate study of suspected counterfeits found 234 case reports, 89.3 percent of them serious, though that work examined semaglutide ^[52].

No biosimilar has been approved anywhere, and the American ingredient search returns exactly one application ^[1]. Development is under way: one group reports a production cell line for a dulaglutide-like protein reaching 1.05 grams per litre ^[53].

Every entry in this section describes the position on 10 September 2026.

Anti-doping

Dulaglutide is not banned in sport. It appears nowhere on the 2026 prohibited list, in or out of competition, and no medical exemption is needed for the substance as such ^[54].

It is not even watched. The 2026 monitoring programme names "Markers of Semaglutide and Tirzepatide" and no other substance from this class ^[55]. The German anti-doping annex does not list it either ^[56].

Nothing about safety or medicines law follows from that. The prescription requirement and the framed American warning apply regardless. The wider picture is at peptides banned in sport.

Compared with related peptides

Two of those comparisons were run properly, side by side, and they point in opposite directions. Against liraglutide, dulaglutide was designed only to prove itself no worse, and it did ^[26]. Against semaglutide it was clearly beaten, on blood sugar and on weight alike ^[6].

What sets dulaglutide apart is not the size of its effect but its build. It is the only substance in this table made as an antibody fusion protein, which is why it is regulated as a biological product ^[4], ^[1].

That single fact does more legal work than any trial result. It is the reason no compounded version can be lawful in America and the reason a copy would need its own biosimilar approval rather than a generic one ^[8], ^[1].

Common misconceptions

• "Dulaglutide, semaglutide and tirzepatide are interchangeable weight-loss injections." Dulaglutide has no weight licence in any market. In the head-to-head trial its weight effect was less than half that of semaglutide, 3.0 kg against 6.5 kg ^[4], ^[6].
• "It is a peptide like the others." It is a fusion protein of about 63 kilodaltons, roughly seventeen times the weight of liraglutide at 3,751 daltons. That is why it is regulated as a biologic and cannot be built by chemistry ^[4], ^[15].
• "Weekly means stronger than daily." The interval follows the half-life of about five days, not the strength. Against daily liraglutide the weekly substance was merely no worse, with a gap of 0.06 points ^[4], ^[26].
• "What applies to compounded semaglutide applies here too." It does not. Semaglutide is licensed as a drug and dulaglutide as a biologic, and the FDA states that biologics cannot be compounded at all ^[8], ^[1].
• "The heart use is approved everywhere." The American label has listed it since 21 February 2020, and the Swiss register names it in its indication field. The European text confines itself to type 2 diabetes ^[16], ^[11], ^[5].
• "It is unproblematic in sport because no list mentions it." It sits on neither the prohibited list nor the monitoring programme, but it remains prescription-only everywhere and carries a framed American warning ^[54], ^[55], ^[4].

Frequently asked questions

Is dulaglutide a prescription medicine?

Yes, in every market on this page. It is licensed in the United States, the European Union, Germany, the United Kingdom, Australia, Canada, Switzerland and Japan, and everywhere it is dispensed on prescription only. It is also a biological product, not a research chemical.

Is dulaglutide approved for weight loss?

No, nowhere. The approved uses are type 2 diabetes from age 10 and, in America, the lowering of serious heart events in adults with diabetes. In the trial in children and teenagers, body mass index showed no difference at all against the dummy treatment.

What did the dulaglutide heart trial show?

In 9,901 adults with type 2 diabetes followed for a median of 5.4 years, a serious heart event struck 12.0 percent against 13.4 percent on the dummy treatment. That is about 12 percent less risk. Death from any cause did not fall clearly, at 10.8 against 12.0 percent.

How does dulaglutide compare with semaglutide?

Badly, in the one trial that placed them side by side. Over 40 weeks in 1,201 adults, weight fell by 3.0 kg on dulaglutide against 6.5 kg on semaglutide, and the blood sugar marker fell 0.41 points less. That trial was run by the maker of the comparator.

What are the most common side effects of dulaglutide?

Stomach and bowel complaints. In the pooled placebo-controlled trials, sickness affected 12.4 and 21.1 percent in the two strength arms against 5.3 percent, vomiting 6.0 and 12.7 percent against 2.3 percent. Complaints of any kind affected 32 and 41 percent against 21 percent.

What is the boxed warning on dulaglutide about?

The American insert carries a framed warning about tumours of a particular thyroid cell. In male and female rats they appear more often the more the animals get and the longer they get it. Whether the same happens in people is unknown. The European text carries no such warning.

Why does this page list no dulaglutide doses?

Because dulaglutide is prescription-only wherever it is licensed, and choosing an amount belongs to the approved product information and to the doctor writing the prescription. This page refers to trial arms only as the lower, middle or higher tested strength.

Can dulaglutide be compounded like semaglutide?

No. Dulaglutide is licensed in America as a biological product, and the FDA states that biologics are not eligible for the compounding exemptions at all. That holds whether or not the substance is in short supply, which makes it different from semaglutide and liraglutide.

Is dulaglutide banned in sport?

No. It appears nowhere on the 2026 prohibited list of the World Anti-Doping Agency, in or out of competition, and no medical exemption is required. It is not in the monitoring programme either, which names markers of semaglutide and tirzepatide only. Nothing about safety follows from that.

Are there generic versions of dulaglutide?

No. No generic and no biosimilar has been approved anywhere, and the American ingredient search returns exactly one application. Because it is a biological product, a copy would need its own biosimilar approval rather than the generic route. Several groups are working on candidates.

Cite this page

The facts on this page were checked on 10 September 2026, and the register searches behind the status table were run on that date. Approved medicines gain new uses and new warnings over time, so the version and the date matter as much as the text.

myPeptides Research & Editing. (2026). Dulaglutide: what the trials show, status and safety. Version 1.0, 10 September 2026. myPeptides Peptide Register. Retrieved from https://mypep.app/peptides/dulaglutide

How pages in this register are compiled and graded is described under methodology; the full register is at peptides.

Last verified: 10 September 2026. Next review: on a labelling change in any covered market, on any change to a shortage listing, or on any change in the anti-doping classification.