Genetic Variants May Influence How Well Patients Respond to Semaglutide and Metformin for Type 2 Diabetes | Research

The takeaway

In a small real-world study, oral semaglutide produced significantly more weight loss than metformin (-6.5 vs -1.6 kg), and genetic variants in the GLP-1 receptor and drug transporter genes showed preliminary signals of influencing treatment response.

-6.5 kg vs -1.6 kg
Oral semaglutide 14 mg produced four times more weight loss than metformin 2000 mg over three months in type 2 diabetes patients

What the researchers found

Over three months, oral semaglutide 14 mg (n=10) significantly outperformed metformin XR 2000 mg (n=17) for weight loss: -6.5 ± 3.6 kg vs. -1.6 ± 2.5 kg (95% CI: -7.6 to -2.2; P=0.001). BMI reduction was also superior: -2.0 ± 1.2 vs. -0.3 ± 0.9 kg/m² (P=0.001).

Genetic findings (exploratory):

- OCT1 rs34130495 was associated with HDL cholesterol change in metformin-treated participants (β=+0.340 mmol/L per minor allele; P=0.0026; q=0.063 at 10% FDR)

- GLP1R rs6923761 showed nominal trends for weight and BMI change in semaglutide users (P≈0.06-0.07) but did not survive FDR correction

The authors describe these as hypothesis-generating data supporting the feasibility of genotype-guided diabetes treatment studies.

Why it matters

As GLP-1 drugs like semaglutide become first-line treatments for millions of patients, understanding why some people respond better than others is increasingly important. Pharmacogenomics — using genetic testing to predict drug response — could help clinicians choose the right medication for each patient from the start. While this study is too small to draw firm conclusions, it demonstrates that genotype-guided studies are feasible even in local clinical settings and identifies specific gene variants worth testing in larger populations.

How the study worked

Twenty-seven Bulgarian adults with type 2 diabetes (BMI ≥25 kg/m², mean HbA1c 8.3 ± 0.9%) received either metformin XR 2000 mg (n=17) or oral semaglutide 14 mg (n=10) for three months. Three common genetic polymorphisms were genotyped by Sanger sequencing: SLC22A1 rs628031, SLC47A1 rs2252281, and GLP1R rs6923761. Primary endpoints were 3-month changes in weight and HbA1c. ANOVA and ordinary least squares regression assessed genotype and treatment effects, with covariate-adjusted linear models.

What this study cannot tell us

The sample size is very small (27 total; only 10 on semaglutide), severely limiting statistical power for genetic associations. The genetic findings did not survive multiple comparison correction (FDR). This was not a randomized trial — treatment assignment was clinical, introducing potential selection bias. Only three genetic variants were tested; comprehensive pharmacogenomic profiling was not performed. The 3-month follow-up is short. The single-ethnicity (Bulgarian) cohort limits generalizability. HbA1c results are not reported in the abstract.

How to read the evidence

This is a small, non-randomized clinical cohort study with 27 participants. While it provides real-world data on semaglutide vs. metformin, the genetic findings are exploratory and did not reach statistical significance after correction for multiple comparisons. The evidence is preliminary and hypothesis-generating.

When this study was published

Published in 2026, this is very recent research addressing the timely question of pharmacogenomics in the era of widespread GLP-1 agonist prescribing.

The bigger picture

Pharmacogenomics is transforming many areas of medicine (cancer treatment, psychiatry, cardiology), and diabetes is next. Large biobank studies have identified variants in GLP1R, SLC22A1, and other genes that may influence drug response, but real-world validation has been limited. This small study adds to the growing push to personalize diabetes treatment — selecting semaglutide vs. metformin vs. other options based on a patient's genetic profile rather than trial-and-error prescribing.

Questions still open

Would GLP1R rs6923761 show a significant effect on semaglutide response in a larger, adequately powered study?
Could genetic testing before starting diabetes treatment meaningfully improve outcomes by matching patients to the best medication?
Do these genetic variants affect response to other GLP-1 agonists (liraglutide, tirzepatide) in the same way as semaglutide?

Common questions

Could a genetic test tell me if semaglutide will work for me?

Not yet. This study found hints that a variant in the GLP-1 receptor gene might influence how well semaglutide works, but the finding was too weak to be conclusive — partly because only 10 people were on semaglutide. Larger studies are needed before genetic testing could reliably predict individual drug response. For now, the clinical approach remains starting the medication and monitoring its effects.

Why did semaglutide produce so much more weight loss than metformin?

Semaglutide and metformin work through completely different mechanisms. Semaglutide mimics the gut hormone GLP-1, which powerfully suppresses appetite through brain signaling and slows stomach emptying. Metformin primarily reduces glucose production in the liver and improves insulin sensitivity, with only modest effects on appetite and weight. The four-fold difference in weight loss (-6.5 vs -1.6 kg) in this study is consistent with larger clinical trials comparing these drug classes.

Read the original research

GLP1R and OCT1 variants modulate semaglutide and metformin response in type 2 diabetes.

Pharmacogenetics and genomics, 36(1), 9-17

View on PubMed ↗Publisher’s paper ↗

Citation

Tourtourikov, Ivan; Kalinkova, Maria; Ivanov, Peter; Mileva-Popova, Rene; Tafradjiiska-Hadjiolova, Radka; Handjieva-Darlenska, Teodora; Kadiyska, Tanya. (2026). GLP1R and OCT1 variants modulate semaglutide and metformin response in type 2 diabetes.. Pharmacogenetics and genomics, 36(1), 9-17. https://doi.org/10.1097/FPC.0000000000000577