Can GLP-1 Treatment Be Personalized? What New Penn and Mayo Studies Actually Show

Can GLP-1 treatment be personalized today?

Not from these studies alone. A Penn-led network meta-analysis synthesized 19 randomized trials with 13,117 participants across seven cardiometabolic outcomes (Penn study on PubMed, 2026). A Mayo study phenotyped 483 adults, but assessed tirzepatide response retrospectively in only 61 participants (Mayo study on PubMed, 2026).

The studies examine different sides of personalization. Penn compared treatment outcome profiles at the trial level. Mayo investigated a possible patient phenotype involving gastric emptying, hunger and GLP-1 biology. Neither validated a consumer test, a prescribing algorithm or a rule that identifies which GLP-1 is best for one person.

The practical answer to which GLP-1 is best for me still requires a licensed clinician. Approved indications, contraindications, side effects, other medicines, coverage and patient preferences remain central. Research can improve that conversation. It cannot replace an individual evaluation or turn one promising subgroup into a clinical test.

What is the Two-Axis GLP-1 Match framework?

The Two-Axis GLP-1 Match is an editorial evidence map, not a clinical selector. Axis One describes patient phenotype, using Mayo's three statistical clusters among 483 adults (Mayo, 2026). Axis Two describes treatment outcome profiles, using Penn's seven-domain Cardiometabolic Efficacy Index (Penn, 2026).

<!-- [UNIQUE INSIGHT] --> The framework's value is the separation. A drug can have an attractive average profile without being best for a specific phenotype. A phenotype can be biologically interesting without supporting a validated treatment decision. Keeping both axes visible prevents a study-level ranking from masquerading as precision prescribing.

Why add an access overlay? Research doses and commercial products do not always share the same number or formulation. The paper's orforglipron trials used capsule doses, for example, while the approved Foundayo tablets use bioequivalent strengths. Access, coverage and format matter, but they sit outside both biological axes.

What did the Penn study measure?

Penn's team analyzed 19 randomized controlled trials covering 13,117 adults with overweight or obesity (Penn full text, 2026). The Cardiometabolic Efficacy Index, or CEI, combined rankings across seven endpoints: body-weight percentage, waist circumference, HbA1c, systolic blood pressure, triglycerides, HDL cholesterol and LDL cholesterol.

The CEI runs from zero to one and summarizes relative ranking patterns. Semaglutide 7.2 mg had the highest reported CEI at 0.86, followed by the orforglipron 36 mg trial capsule at 0.68 and semaglutide 2.4 mg at 0.66. The paper reports the orforglipron trial dose as bioequivalent to a Foundayo 17.2 mg tablet.

The authors warn that small CEI differences do not establish clinically meaningful superiority. The index weights all seven endpoints equally, excludes safety and tolerability, and does not measure cardiovascular or kidney events. Network inconsistency appeared for several outcomes, while trial populations, durations and lifestyle programs varied.

The funding also belongs beside the result. The study states that Eli Lilly supported it in part, while the funder had no role in design, data collection, analysis or interpretation (Penn full text, 2026). The authors declared no conflicts of interest.

Did the Penn analysis include tirzepatide?

No. Penn restricted the network to GLP-1 receptor mono-agonists and excluded dual incretin agonists such as tirzepatide (Penn full text, 2026). That means its CEI rankings cannot answer whether Zepbound ranks above, below or alongside the analyzed semaglutide, liraglutide and orforglipron regimens.

This exclusion preserved the study's chosen mechanistic comparison, but narrows its usefulness for a real 2026 decision. Tirzepatide activates GIP and GLP-1 receptors and is central to the Wegovy versus Zepbound comparison. A mono-agonist network should not be stretched into a market-wide ranking.

The status overlay prevents another common error: calling every study dose experimental. FDA approved Wegovy HD 7.2 mg on March 19, 2026 (FDA, 2026). FDA approved Foundayo on April 1, with tablet strengths up to 17.2 mg (FDA, 2026).

What phenotype did the Mayo study find?

Mayo researchers identified three statistical clusters among 483 adults with obesity (Mayo full text, 2026). One cluster included 130 people, or 26.9%, with faster solid-meal gastric emptying, greater post-meal hunger and unexpectedly low post-meal GLP-1 levels relative to the gastric-emptying pattern.

The researchers called that cluster discordant gastric emptying and GLP-1, shortened to dc-GE/GLP-1. It also showed lower peptide YY and cholecystokinin. A separate 31-person biopsy cohort showed lower intestinal expression of genes related to GLP-1 and peptide YY, adding biological support without turning the cluster into a routine diagnosis.

Microbiome analysis did not show a significant difference in fecal microbial composition between the discordant and concordant clusters. That negative finding is useful. It narrows the observed signal and discourages a leap from this paper to microbiome supplements, stool tests or claims that gut bacteria can select tirzepatide treatment.

Can a test predict who responds best to tirzepatide?

No validated clinical test emerges from this study. Mayo retrospectively reviewed tirzepatide outcomes in 61 participants, a small subset of the 483-person phenotype cohort (Mayo, 2026). At six months, the discordant cluster showed 21.5% weight loss versus 11.7% in the concordant clusters.

The difference is hypothesis-generating, not a prescribing rule. Treatment was not randomized by phenotype, the response analysis was retrospective, and only 61 participants contributed. A larger prospective trial would need to predefine the test, assign treatment appropriately, reproduce the result and show that using the test improves clinical decisions.

<!-- [UNIQUE INSIGHT] --> Axis One currently has biological depth but weak decision validation. The phenotype was characterized across hundreds of people, while the treatment-response link rests on a much smaller retrospective subgroup. That mismatch is exactly why the result can motivate research without supporting a consumer matching product.

Conflicts require equal visibility. PubMed reports that two authors and Mayo Clinic co-founded and hold intellectual property licensed to Phenomix Sciences. It also lists consulting, advisory and research relationships involving several drug companies (Mayo PubMed record, 2026). These disclosures do not invalidate the findings, but they matter when evaluating commercialization claims.

Which findings can patients act on now?

Patients can act on the questions, not the experimental matching claims. Penn evaluated seven outcome domains across 19 trials (Penn, 2026), so a clinician discussion can name the outcomes that matter. Mayo's 61-person response subgroup cannot justify ordering a test or choosing tirzepatide from a phenotype claim.

The semaglutide hub and tirzepatide hub explain approved molecules and access pathways. The Foundayo guide covers approved orforglipron, while the Foundayo versus Wegovy pill comparison focuses on two oral options. These resources organize public evidence, not personal treatment selection.

A weight-loss predictor can illustrate population trial averages, but it cannot forecast an individual's result. The result should be treated as an educational range, then discussed with a clinician. Side effects, dose tolerance, discontinuation and adherence can move real outcomes far from an average curve.

How do access, budget and preferences fit without choosing a medicine?

Access is a separate decision layer because Penn's 19 trials and seven CEI outcomes did not score insurance, provider quality or total price (Penn full text, 2026). Once a clinician defines medically appropriate options, readers can compare legitimate access, budget and service preferences without pretending those factors predict biology.

The provider comparison organizes available services, while the cost calculator models recurring expenses. A low advertised medication price may exclude membership, laboratory or follow-up costs. Insurance coverage may also change the practical choice among medically reasonable options, even when research evidence looks similar.

The provider matching quiz asks about provider access, budget and preferences. It does not choose a medicine, diagnose a phenotype or predict response. That boundary is useful: provider matching can reduce shopping friction after clinical eligibility, while medication selection remains a shared decision with a licensed prescriber.

• •Use the evidence axis to understand what studies measured.
• •Use a clinician to evaluate indications, risks and suitable treatment options.
• •Use provider tools only to compare access, support, budget and format preferences.
• •Reject any service claiming the Mayo phenotype already identifies your best GLP-1.

Frequently Asked Questions

The two studies answer research questions at different levels. Penn pooled 19 trials and 13,117 participants to compare seven outcome domains (Penn, 2026). Mayo found a 130-person phenotype, but its tirzepatide response analysis included only 61 people (Mayo, 2026). Neither selects individual treatment.

Which GLP-1 is best for me?

Neither 2026 study can answer that for one person. Penn ranked seven study-level outcomes across 19 trials, while Mayo's treatment-response comparison involved 61 retrospective participants (Penn, 2026; Mayo, 2026). A licensed clinician must assess individual benefits, risks and alternatives.

What is the Cardiometabolic Efficacy Index?

The CEI is Penn's zero-to-one composite of rankings across seven measures, including weight, waist, HbA1c, blood pressure and lipids (Penn full text, 2026). It summarizes relative trial-level patterns. It is not a calibrated benefit score, safety score or individualized prediction.

Can low GLP-1 levels show that tirzepatide will work better?

Not in current practice. Mayo identified a 130-person cluster with discordantly low post-meal GLP-1 relative to gastric emptying, then analyzed tirzepatide response in 61 people retrospectively (Mayo, 2026). That is not a validated test, cutoff or prospective treatment-selection rule.

Does the provider quiz choose a GLP-1 medicine?

No. The quiz matches provider access, budget and service preferences. It does not prescribe a medicine, diagnose Mayo's 26.9% phenotype or predict treatment response (Mayo, 2026). Medication decisions belong with a licensed clinician after an individual medical assessment.

Sources