Cagrilintide's side effect profile resembles the GLP-1 class, with two differences: injection-site reactions are more common than with semaglutide, and GI event rates rise noticeably when it is combined with semaglutide.
Here is the breakdown by type, with all figures drawn from published trials.
[Gastrointestinal events]
This is the dominant category.
Phase II monotherapy: 41–63% of cagrilintide-treated participants reported GI adverse events versus 32% on placebo. Nausea, constipation and diarrhoea dominated. Nausea alone ranged from 20% to 47% depending on dose, versus 18% on placebo (Lau et al., 2021, PMID 34798060).
Combination: in REDEFINE 1, GI adverse events occurred in 79.6% of the CagriSema arm versus 39.9% on placebo; in REDEFINE 2 the figure was 72.5%. Most events were transient and mild to moderate (Garvey et al., 2025, PMID 40544433; Davies et al., 2025, PMID 40544432).
In REIMAGINE 2, rates were 86.9% for CagriSema and 81.2% for semaglutide alone — the combination does add tolerability cost (Buse et al., 2026, PMID 42251859).
[Discontinuation rates]
In the Phase II monotherapy trial, 10% discontinued permanently due to adverse events, similar to placebo. Most GI reactions are manageable, but not everyone tolerates them.
[Injection-site reactions]
This is where cagrilintide stands out relative to GLP-1 drugs. Exact rates vary by trial but are generally higher than with semaglutide alone. Most are local redness or mild pain that resolves on its own.
[Fatigue]
This is the amylin-class signal to watch.
Fatigue does not feature prominently in published cagrilintide summaries. But Eli Lilly's selective amylin agonist eloralintide produced fatigue in 43% of participants at 9 mg versus 12% on placebo (Billings et al., 2025, PMID 41207310).
Whether this is a class effect or a molecule-specific one is unresolved. For cagrilintide, more data is needed before drawing a conclusion.
[Hypoglycaemia]
In the Phase II type 2 diabetes trial, including the cagrilintide monotherapy arm, no clinically meaningful hypoglycaemia was reported (Frias et al., 2023, PMID 37364590).
Amylin analogues lower glucose in a meal-dependent way, so hypoglycaemia risk is low. This distinguishes them from insulin and sulfonylureas.
[The biggest unknowns]
First, maximum exposure. The longest documented exposure is 84 weeks, and that trial is not yet formally published. There are no discontinuation studies and no maintenance studies.
Second, combination safety. There is no human tolerability data for cagrilintide plus retatrutide. The rat study was not designed to assess safety.
Stacking two appetite suppressants that both cause nausea will, at minimum, worsen nausea. That is pharmacological common sense, not a trial finding.
Third, long-term data. The thyroid, pancreatic and gallbladder signals that matter for long-term GLP-1 use have not been tracked long enough for cagrilintide.
[Conclusion]
Cagrilintide's safety dataset is more complete than most investigational peptides, but clear gaps remain: nothing beyond 84 weeks, no combination data, no post-discontinuation data. What is known: GI events are common but mostly transient, injection-site reactions exceed semaglutide, hypoglycaemia risk is low, and whether fatigue is a class effect remains undetermined.