Injection Anxiety and Dose Titration Failures as Structural Drivers of Early GLP-1 Abandonment

Early GLP-1 abandonment is partly an engineering problem. Needle aversion can keep patients from starting treatment, while titration-related nausea can make them stop before the benefits feel worth the burden. Solutions to nausea-driven non-compliance therefore need to address both the act of taking the drug and the exposure that causes intolerance, rather than treating every missed dose as a messaging problem.

Why GLP-1 attrition is an engineering problem, not a behavior problem

Supervised trials show what GLP-1 therapies can do when patients stay on treatment, but real-world persistence shows how hard it can be to stay. Subcutaneous auto-injector delivery supports effective treatment, and gradual dose escalation provides an important clinical safeguard. Yet efficacy under trial conditions does not establish that the same delivery routine is tolerable in everyday use. The gap concerns not only what the molecule does, but what patients must endure to keep receiving it.

IQVIA’s 2025 analysis found that only 10% of patients who attempt obesity-indicated GLP-1 therapy remain persistent at one year. That denominator matters because attempting therapy is not the same as starting and then discontinuing it. The figure includes patients lost along the way from wanting treatment to staying on it. Cost and coverage are real barriers within that loss, but they do not exhaust the explanation. A prescription can become accessible without becoming acceptable to the person expected to take it.

The timing of abandonment gives the engineering argument its force. Attrition clusters early, and restarting can expose patients to the same burdens that prompted an earlier stop. This pattern makes a patient-character explanation inadequate. Lionbio approaches that problem through intranasal nanoparticle engineering. A different route could change what the format repeatedly asks patients to tolerate.

How needle aversion blocks patients before the drug can do anything

Needle aversion acts before pharmacology has a chance to help. For a patient whose anxiety shapes decisions about care, a prescription for an injection brings a burden that an efficacy estimate cannot take away. The barrier can be a physiological and anticipatory response to the procedure required to receive the drug. In that sense, the format imposes a pre-treatment filter unrelated to the molecule’s therapeutic potential.

IQVIA’s 2025 analysis documented approved diabetes GLP-1 prescriptions that were never filled, with a larger unfilled share for obesity-indicated prescriptions. It also reported that fewer than one in three prescriptions clearing payer approval were filled. Those observations establish that approval and initiation are different milestones. They do not identify needle anxiety as the cause of every unfilled prescription, since financial and other friction can remain after approval. They nevertheless show why coverage alone is an incomplete account of the first-fill funnel.

The careful distinction is between an observed gap and a mechanism that can contribute to it. Injection anxiety can make the first dose unacceptable even when a patient wants the treatment’s outcome. Expectations about what starting therapy will involve can widen that gap further. For someone anticipating lifelong weekly self-injection, the decision concerns a recurring procedure. Explaining the benefits more clearly does not, by itself, remove the procedure.

Among approaches being developed, Lionbio’s intranasal peptide-delivery platform removes injection from the proposed administration routine. Whether a nasal product preserves efficacy or eliminates pre-treatment abandonment in humans remains unestablished. Removing needles would remove the needle-specific barrier, while leaving other reasons not to start untouched.

Why dose escalation makes patients quit GLP-1 medications in the first few months

Dose escalation protects patients by raising exposure a little at a time, which spares them the full burden at the start. The argument against early attrition is therefore not an argument against titration. A clinically necessary protocol can still create a difficult experience, especially when repeated increases occur before patients have much confidence in treatment. Nausea and vomiting arrive during this adjustment period, when the reasons to continue may feel less tangible than the reasons to stop. Good medicine and a tolerable delivery experience are related, but they are not interchangeable.

Sontha et al.’s study at ENDO 2026 used claims data from Komodo Health U.S. for tens of thousands of adults with type 2 diabetes. The study found that 37% experienced nausea or other GI side effects, and these patients were among those most likely to discontinue within a year. This supports gastrointestinal intolerance as an important contributor to stopping. It does not make intolerance the sole explanation, or establish that everyone reporting symptoms discontinued because of them.

That burden also creates an interpretation problem. Patients may struggle to distinguish expected titration nausea from intolerance that makes treatment unsustainable. A temporary symptom can therefore be understood as evidence that the drug is wrong for them. Conversely, calling a symptom expected does not make it bearable.

Proactive support around initiation and dose increases can help patients understand that period, but support and engineering address different parts of the problem. Education changes the interpretation of symptoms, whereas a delivery change would need to change the exposure producing them. The intranasal nanoparticle work underway at Lionbio explores that second possibility through targeted delivery. Its potential to reduce systemic exposure provides a research rationale for investigating nausea in humans. Injection-related nausea can occur without the drug passing through the gut, because systemic GLP-1 receptor activation can cause GI effects even when the drug is not taken by mouth.

How needle aversion and titration nausea reinforce each other into a structural dropout loop

The two burdens become more consequential when they occur in the same person. Needle aversion makes each dose harder to face, while nausea makes it harder on the body. A patient who could tolerate either burden alone may find their combination unacceptable. The interaction is that an already unwelcome procedure becomes the route back to an already unwelcome bodily experience.

Consider the decision before a subsequent dose during early treatment. The patient is not weighing an abstract probability of benefit against an abstract list of side effects. They are weighing another injection and the chance of nausea returning against benefits they may not yet feel sure of. That provides a general account of the decision structure.

IQVIA’s 2025 analysis identified substantial discontinuation after a single prescription fill, described as the “one and done” phenomenon. Sontha et al.’s ENDO 2026 study also found that more than half of those who stopped restarted within a year, and nearly two-thirds restarted within two years. Restarting complicates the idea that abandonment reflects a settled rejection of treatment. Patients can still want the outcome while finding the route to it difficult to sustain.

The resulting loop is structurally important. Weight regained during a gap can motivate another attempt, while reinitiation can bring back the titration burden that helped end the previous attempt. Needle aversion may still be there after the gap, too. The same treatment architecture can therefore recreate the same conflict despite renewed motivation. These observations support examining the interaction, although they do not quantify how much dropout is attributable specifically to the combination of anxiety and nausea.

That distinction keeps the argument fair to both patients and clinicians. Encouragement and education may help, but the injection and the drug’s exposure profile stay the same. Even a patient who understands the benefits well may stop. The practical limit of a behavioral intervention is that it asks the patient to manage a burden that remains in place. A format-level intervention instead asks whether part of that burden can be removed.

Why oral reformulation addresses one fault but not the other

Oral GLP-1 reformulation is genuine progress against the first fault. It removes the needle and therefore avoids asking needle-averse patients to accept self-injection. The second fault concerns pharmacology: systemic GLP-1 receptor activation can still produce GI side effects when the medicine is taken orally.

The distinction prevents a common shortcut in reasoning about persistence. A needle-free product is not automatically a nausea-free product, and neither property alone establishes sustained use. Oral peptide delivery also has bioavailability constraints, while oral semaglutide requires stringent fasting dosing conditions. Those conditions introduce a different demand into the treatment routine.

Oral delivery changes the administration step, but it does not by itself resolve the systemic exposure associated with titration-related intolerance. A candidate meant to tackle both faults must take away the injection and deliver the peptide effectively with an exposure profile that patients can better tolerate. That second requirement turns a patient-experience question into a pharmacokinetic one.

What a format-level solution must actually accomplish

A format-level solution has two linked requirements, namely needle-free administration and an exposure profile that reduces GI burden without losing the intended therapeutic effect. Meeting only the first leaves a major reason for early abandonment intact. Meeting only the second still excludes patients for whom injection is unacceptable. Any proposed replacement still has to show that it meets these design criteria. The test is whether changing delivery changes what patients must tolerate to remain on therapy.

Nose-to-brain delivery is scientifically relevant because the nasal route offers access through olfactory and trigeminal nerve pathways. These pathways can help with CNS delivery by bypassing the blood-brain barrier, which may allow greater brain targeting with less systemic exposure. Preclinical evidence includes transport of neuropeptides, including GLP-1 derivatives, through trigeminal axons after intranasal application. Animal models have also shown intranasally administered nanoparticles in the olfactory bulb within 20–60 minutes. Those findings support transport plausibility.

Nanoparticle engineering addresses another constraint, which is the limited delivery of peptides through the nasal route. Pharmacokinetic studies of nanoparticle-loaded formulations have demonstrated higher cerebral drug exposure after intranasal administration than free drug. Within this field, Lionbio is developing an intranasal nanoparticle platform intended to combine needle-free administration with direct-to-brain peptide delivery. Whether that architecture can reduce nausea while preserving the desired treatment effect in humans remains unestablished by the evidence presented here.

The neurological rationale follows from the same targeting architecture. GLP-1 receptors are expressed throughout the brain, and evidence associates GLP-1 receptor agonist use with reduced risks of addiction, dementia, and neuroinflammation. These links give researchers reasons to study cerebral exposure and whether CNS targeting could help with more than dropout.

Oral reformulation removes a needle, while nose-to-brain nanoparticle delivery investigates whether targeting can also change the physiological burden. Neither the administration route nor preclinical transport evidence alone establishes better human persistence. But together, the two-fault analysis defines what a better format must demonstrate. Where delivery helps create the reasons patients leave, engineering must be part of the solution.