Genetic predictors of GLP-1 response: Promising, but not there yet

Nishu Uppal, MD, MBA

Nishu Uppal, MD, MBA

Principal Physician, Population Health

Dr. Uppal is a practicing internal medicine physician and healthcare policy researcher trained at Brigham and Women's Hospital. He has published in leading journals including NEJM, JAMA, and Health Affairs, and holds an MD from Harvard Medical School and an MBA from Harvard Business School.

Akl Fahed, MD, MPH

Akl Fahed, MD, MPH

Co-Founder and Senior Advisor

Dr. Fahed is a leading physician-scientist in the prevention of cardiometabolic disease and practicing interventional cardiologist at Mass General Brigham. He is the author of more than 120 peer-reviewed publications, and directs the Fahed Lab at Massachusetts General Hospital and the Broad Institute of MIT and Harvard – an NIH-funded research group that uses genomics, data science, and artificial intelligence to advance the prevention of cardiometabolic disease.

By Nishu Uppal, MD, MBA, and Akl Fahed, MD, MPH

This article is part of The Signal, a monthly column from Goodpath’s clinical team on the emerging research, clinical data, and broader learnings and trends shaping the future of obesity and cardiometabolic treatment.

One of the clinical realities of GLP-1 medications is that treatment responses vary widely from person to person. Similar to other potent drugs like antidepressants or blood pressure medications, studies show a vast variability in how people respond to GLP-1 medications. For example, a recent systematic review and meta-analysis showed that women and men taking GLP-1 medications experience similar pounds of weight loss, but women lose a greater percentage of body weight than men (10.9% vs. 6.8% on average).

Among patients taking semaglutide or tirzepatide, some lose less than 5% of their body weight. Others lose more than 20%. Some experience debilitating nausea starting with their first dose. Others feel almost nothing. 

This isn’t random variability, it’s a signal — and we’re starting to understand more about what could be behind it. 

A study published in Nature in April, conducted by 23andMe Research Institute using self-reported data from nearly 28,000 individuals who had used GLP-1 medications represents one of the largest (by sample size) attempts yet to identify potential genetic drivers of GLP-1 responses. 

It is interesting new research and the findings are worth unpacking, both in terms of what they tell us about the biology of GLP-1 response and what they suggest about where prescribing could be headed. 

The findings and why they matter for GLP-1 care 

In the Nature study, researchers performed a large-scale genome-wide association study – an analysis designed specifically to identify genetic variants associated with specific outcomes across a large population - on nearly 28,000 people who had taken a GLP-1 medication. 

Here are the key findings to note, and why they matter. 

Weight loss efficacy. Researchers identified a variant in the GLP1R gene – the gene encoding the protein that GLP-1 medications directly target – that is associated with greater weight loss when taking a GLP-1 medication. Carriers of one copy of this variant lost an average of 1.7 pounds more over a median of 8 months of treatment compared to non-carriers. Importantly, the association of this genetic variant with weight loss differed based on whether an individual was taking tirzepatide or semaglutide. 

Why this matters: The effect of this variant alone is modest. An extra 1.7 to 3 pounds over eight months is not a dramatic difference. But this research proves emerging genetic evidence that variation in the gene encoding the molecular targets of GLP-1 medications may influence how well these medications work, which could open the door to more refined models and further genetic research that will help us understand who is most likely to experience meaningful weight loss on GLP-1 drugs. Furthermore, the study reported that weight loss among individuals with the genetic variant may be different depending on whether they took tirzepatide (BMI decline of 0.95%) vs. semaglutide (BMI decline of 0.51%). Although this difference is similarly modest to the overall associated weight loss findings, it may serve to facilitate future research that uncovers additional genetic predictors that collectively become more clinically meaningful in identifying GLP-1 medication responsiveness.

Nausea and vomiting side effects. The study also identified genetic variants that are associated with nausea and vomiting while taking GLP-1 medications and looked at differences between those taking semaglutide, which only targets the GLP-1 receptor, and tirzepatide, which also targets GIP. Individuals taking tirzepatide with specific genetic variations within the GLP1R and GIPR genes had up to 15-fold higher odds of experiencing vomiting compared to those without the variants conferring this risk. 

Why this matters: This is significant because nausea and GI side effects are a primary reason that patients discontinue GLP-1s before they can realize meaningful outcomes. The ability to identify individuals who are at elevated risk of experiencing nausea and vomiting from GLP-1 medications before they are prescribed carries the potential to change the outlook for adherence and outcomes. 

It’s important to note that the authors developed predictive models to see whether incorporating these newly identified genetic factors alongside standard demographic factors (i.e. age, sex, education) and clinical factors (i.e. drug type, dosage, time on treatment) carried the potential to be clinically actionable at the patient level, and unfortunately found that incorporating genetic factors offered relatively little additional predictive value. In other words, these findings on their own are not going to change the GLP-1 prescribing picture for an individual patient with these genetic variants. 

Therefore, the study findings, while interesting in methodology and potentially informative for future research on drug response biology and pharmacodynamics, likely don’t add much new clinical meaningfulness yet. 

Why we find this research promising 

The 23andMe study is a large and rigorous step toward understanding genetic variability in GLP-1 response that the field has long observed but struggled to explain. With nearly 28,000 participants, the scale gives the findings real credibility, and the drug-specific signals in particular will likely drive targeted follow-up research. 

What we’re watching for: Studies that utilize longitudinal data to identify interactions between drug and dosage changes with clinical outcomes and genetic variation, replicate and extend these findings in more diverse populations (the limitations of consumer genetics datasets on ancestry diversity are real and acknowledged by the authors) and with other GLP-1 medications aside from semaglutide and tirzepatide, and expanded models that integrate biomarkers alongside genetic data. Major manufacturer-sponsored trials of GLP-1 medications and other anti-obesity pharmacologic treatments may also concurrently collect clinical and genetic data from participants that could serve as an enriched data source for future work on genetic predictors of GLP-1 responses.

While this Nature study does not enable the clinical capability to precisely predict the likelihood of weight loss and side effects for a specific person receiving a GLP-1 medication, the identification of novel genetic variants of biological plausibility suggest that genetic factors still hold promise for improving prediction of treatment response. 

Future research will help develop the evidence base needed to fully test the utility of genetic predictors as well as broader prediction models that aspirationally seek to get patients on the right medication based on their individual profile at the start of their weight loss journey, rather than the traditional trial-and-error approach, and with more realistic expectations around side effects and results.

References