GLP-1 Diabetes Medications Linked to Higher Recorded Rates of Eyelid Surgery

In a large multicenter retrospective investigation drawing on real-world electronic health records, adult women treated with glucagon-like peptide-1 receptor agonists for diabetes and hypertension experienced more than double the recorded hazard of undergoing upper eyelid blepharoplasty compared to matched control patients who did not receive the medications. The primary analytical finding demonstrated that recorded upper eyelid surgical excision occurred in 96 patients prescribed a GLP-1 receptor agonist, representing an absolute proportion of 0.41 percent, compared to 36 individuals in the unexposed control cohort, representing 0.15 percent. This relative distribution translated to a statistically significant adjusted hazard ratio of 2.22, with a 95 percent confidence interval spanning from 1.51 to 3.26 and a significance level of p less than 0.001. Survival analysis via Kaplan-Meier curves confirmed a shorter cumulative time to procedural intervention among individuals initiated on incretin mimetic therapy, providing empirical electronic record documentation of subsequent periorbital surgical utilization patterns.

Despite the pronounced relative hazard observed across the five-year tracking window, the investigators emphasized that the absolute clinical incidence of upper eyelid blepharoplasty remained rare across the entire study population. The overall absolute risk difference between cohorts was 0.26 percentage points, illustrating that while relative surgical utilization rates were noticeably elevated among drug recipients, less than one-half of one percent of patients in either matched arm underwent recorded operative repair. The research team cautioned that these retrospective associations derived from billing and diagnostic registries should not be interpreted as definitive proof that glucagon-like peptide-1 receptor agonists directly accelerate or trigger periorbital aging, skin redundancy, or tissue degradation. Instead, the authors framed the measured difference as a documented variation in documented surgical procedure utilization within formal healthcare health organizations, urging careful contextualization against baseline metabolic health and healthcare encounter frequencies.

The study cohort was derived from a massive retrospective screening pool of 821,792 adult female patients who were at least 18 years of age and had verified concurrent clinical diagnoses of both diabetes mellitus and essential hypertension recorded between January 1, 2015, and January 1, 2020. From this broader population, 28,970 women were identified as having received an active prescription or administration of a GLP-1 receptor agonist within Anatomical Therapeutic Chemical classification class A10BJ. The pharmaceutical agents accessible during this designated enrollment timeframe included exenatide, liraglutide, dulaglutide, lixisenatide, and semaglutide. Investigators explicitly restricted eligibility to female patients to minimize known anatomical, cosmetic referral, and aesthetic care-seeking differences between sexes, while the dual diagnostic inclusion criteria of diabetes and hypertension were applied to establish a biologically comparable cardiometabolic baseline, accounting for the historic clinical prescribing environment prior to the widespread commercial expansion of incretin therapies solely for cosmetic weight loss.

To rigorously control for baseline demographic discrepancies, systemic comorbidities, and concurrent pharmaceutical therapies, the investigators performed one-to-one greedy nearest-neighbor propensity score matching without replacement, yielding two highly balanced cohorts of 23,349 patients each. Matching criteria encompassed index age, racial background, baseline body mass index, and a broad spectrum of medical diagnoses and pharmacotherapies, including chronic kidney disease, hyperlipidemia, chronic obstructive pulmonary disease, metformin, exogenous insulin, and sodium-glucose cotransporter 2 inhibitors. Post-matching assessments confirmed robust covariate balance, evidenced by standardized mean differences remaining below the standardized 0.10 threshold across clinical and laboratory metrics. At the index baseline, the matched cohorts exhibited comparable mean body mass indexes of 30.7 plus or minus 5.2 in the GLP-1 receptor agonist cohort and 31.2 plus or minus 6.5 among controls. Over the five-year follow-up, mean body mass index decreased to 27.97 plus or minus 4.18 in the exposure cohort relative to 29.90 plus or minus 5.89 in the unexposed cohort.

Surgical intervention was tracked in the TriNetX analytic database through standardized mapping of Current Procedural Terminology code 15823, which specifies upper eyelid blepharoplasty. The authors noted that this administrative coding designation encompasses functional reconstructive eyelid surgery indicated for mechanically obstructing dermatochalasis or visual field impairment, as well as procedures pursued for cosmetic rejuvenation. Within the available electronic records data extracted for the study, granular clinical documentation detailing the precise surgical indication, eyelid laterality, operating surgeon specialty, hospital or ambulatory facility type, anesthesia modality, and concurrent facial rejuvenation procedures was unavailable. Consequently, the researchers could not differentiate between functional procedures reimbursed by third-party health insurance to relieve superior visual obstruction and elective, patient-financed aesthetic refinements undertaken to correct age-related contour changes.

To evaluate whether the elevated surgical frequency was uniquely tied to receptor agonist pharmacology or represented a generalized consequence of body weight reduction, the researchers executed an exploratory secondary analysis evaluating weight loss achieved without GLP-1 receptor agonists. This secondary evaluation matched 24,690 patients who achieved weight loss through other means against 24,690 control subjects maintaining stable body mass index values over comparable timeframes. Baseline body mass index was closely aligned between these comparative groups, measuring 34.1 plus or minus 4.9 in the non-pharmacologic weight-loss cohort and 33.8 plus or minus 3.3 among stable controls, with follow-up body mass indexes recorded at 31.68 plus or minus 5.6 and 33.62 plus or minus 2.22, respectively. Within this non-incretin comparison, the incidence of upper eyelid blepharoplasty showed no statistically significant divergence, yielding a hazard ratio of 1.31 with a 95 percent confidence interval of 0.89 to 1.93 and a p-value of 0.56, though the authors classified this secondary assessment as strictly hypothesis-generating due to unmeasured variations in weight reduction velocity and durability.

External three-dimensional imaging studies and anatomical reviews cited in the paper have suggested that rapid systemic adipose depletion may accentuate midfacial volume deflation, alter superficial fat compartments, or reveal latent dermatochalasis, descent of adjacent brow tissues, and upper eyelid skin laxity. However, the study authors expressly stated that none of these cellular changes, tissue-level mechanics, facial compartment volumes, or skin elasticity parameters were directly measured or evaluated within their electronic health record dataset. These proposed physical transformations remain hypothetical contributing factors to care-seeking behavior rather than verified biological mechanisms established by the clinical record study.

The study authors highlighted multiple substantial limitations and confounding factors inherent to large retrospective electronic health record studies. Most prominently, socioeconomic status, private insurance coverage parameters, and individual disposable income levels were unrecorded in the dataset, creating potential confounding because expensive incretin medications and elective periorbital procedures are both historically concentrated among patients with higher financial means and comprehensive healthcare access. Because TriNetX exclusively monitors coding within participating healthcare networks, operations conducted outside participating institutions or performed as private cash transactions were absent from the registry, underscoring that prospective clinical studies with standardized anatomical examinations, visual field evaluations, and objective three-dimensional imaging will be required to clarify these findings.

Subject of Research: Association between GLP-1 receptor agonist therapy and upper eyelid blepharoplasty rates in clinical records

Article Title: Use of GLP-1 receptor agonists and upper eyelid blepharoplasty in women with diabetes and hypertension: a retrospective TriNetX cohort study

Article References: Almeida, V. F. A., Ha, J., Duraes, F. R., Dantas, M., Hashem, A. M., Djohan, R. S., & Duraes, E. F. R. (2026). Use of GLP-1 receptor agonists and upper eyelid blepharoplasty in women with diabetes and hypertension: a retrospective TriNetX cohort study. BMC Plastic and Reconstructive Surgery, 2(1), Article 31. https://doi.org/10.1186/s44452-026-00044-6

Image Credits: AI Generated

DOI: 10.1186/s44452-026-00044-6

Keywords: GLP-1 receptor agonists, blepharoplasty, type 2 diabetes, hypertension, plastic surgery, clinical epidemiology, dermatochalasis, TriNetX, GLP-1, receptor, agonists, upper