Why GLP-1 treatment response varies

People taking GLP-1 receptor agonists or dual GLP-1/GIP medicines do not all get the same result. Some people lose more weight. Others do not. A new narrative review finds that no single baseline characteristic can reliably predict who will respond. The most useful signal available now is the early response after treatment has started, assessed with the clinician who prescribed it^[1].

That answer is less tidy than a genetic test or a personality profile, but it is closer to the evidence. Response can mean different things in type 2 diabetes, obesity without diabetes, or both conditions together. The review separates glycaemic response from weight change and warns against treating early research on genes, metabolism, the microbiome, or hormones as a ready-made prediction tool^[1]. Readers looking for background on the medicines can start with our GLP-1 receptor agonist explainer.

What counts as a response

A treatment response is not one number for every person. In type 2 diabetes, clinicians may focus on changes in glycaemic control. In obesity care, weight change is often the outcome of interest. For people with both conditions, the relevant outcome can include both metabolic and weight outcomes. That difference matters because a person may see a meaningful change in one measure without seeing the same pattern in another. Our semaglutide guide explains the medicine's role in this class^[1].

The review describes weight non-response in controlled obesity trials as commonly defined by less than 5% total body weight loss. It estimates that this applies to about 10% of trial participants, while noting that routine-practice cohorts may report a higher frequency. Those figures describe groups, not a rule that can determine what will happen to one person^[1].

Why response varies

The review treats variation as a clinical problem with several possible contributors rather than one hidden switch. It examines biological, genetic, metabolic, hormonal, behavioural, and psychosocial factors. That is not evidence that each factor can be used to forecast an individual result. It is a map of questions researchers are still trying to answer^[1].

This distinction is especially relevant when comparing semaglutide and tirzepatide. They are related incretin-based medicines, but a response to one person or one medicine cannot be projected onto another. Our pages on semaglutide and tirzepatide cover the evidence and safety context for each molecule separately.

Early response is the practical signal

Across the three clinical contexts in the review, early on-treatment response is the most readily actionable predictor currently available. In plain language, the observed change after treatment begins tells clinicians more than an unvalidated baseline marker. The paper does not turn that finding into a self-management protocol. It supports structured follow-up and decisions made in clinical context. Readers can also review our tirzepatide explainer^[1].

That approach also avoids a common mistake: treating slower change as proof that a medicine will never help, or treating an early change as a guarantee of a lasting result. The review identifies substantial variation in response and does not offer a reliable cutoff that readers can use on their own. If you are considering a change to a prescribed medicine, discuss the response you are seeing with the clinician responsible for your care. Our tirzepatide safety guide discusses safety questions separately.

What genetics can and cannot tell us

The review discusses a recent genome-wide association study that linked variation in the GLP1R drug-target gene with self-reported weight-loss efficacy and gastrointestinal tolerability. It also reports agent-specific associations between GIPR variants and nausea and vomiting with tirzepatide. The review calls these findings hypothesis-generating because they came from a self-reported single cohort and have not yet been replicated^[1].

That is why the paper does not support individual genetic testing as a way to select a GLP-1 or GLP-1/GIP medicine. The same caution applies to proposed metabolic, microbiome, and neuroendocrine markers. Each is a research lead, not a validated clinical prediction service. For a wider view of the class, see our guide to how semaglutide works and our overview of tirzepatide.

What this review does and does not show

This is a narrative review, not a new treatment trial. Its value is in bringing together human research published through 25 May 2026 and separating signals that may matter from tools that are ready for clinical use. It does not establish that a person should switch medicines, change a prescribed dose, or order a genetic test. Our semaglutide safety page covers separate safety considerations^[1].

The most defensible takeaway is modest. There is no single baseline profile that reliably predicts response. Early response during clinician-led care is more informative than the proposed biomarkers reviewed so far. That may feel unsatisfying, but it is more useful than a prediction claim that the evidence cannot support.

What we do not yet know

Researchers still need prospective replication of genetic findings and better tests of metabolic, microbiome, and neuroendocrine markers. They also need to establish whether any marker improves real clinical decisions, rather than merely showing an association in a cohort. Until then, response prediction remains a research goal rather than a precise tool for choosing treatment. See our tirzepatide evidence overview^[1].

Medical disclaimer

This article is for educational and journalistic purposes only and does not constitute medical advice. Peptides discussed may be classified as prescription medicines or research chemicals depending on your jurisdiction. Always consult a qualified healthcare professional before using any peptide product. PeptideMethods.com does not sell, distribute, or facilitate the sale of any peptide product.