“Ozempic Face” Is Real. The Question Is Whether It’s the Ozempic.

A patient told me last month that she’d lost forty pounds and felt better than she had in a decade — and that she couldn’t look at photographs of herself. She was proud of what her body had done. She could not reconcile it with a face that seemed to have aged five years without warning.

I want to be blunt about our position before I go further, because it’s an uncomfortable one. I prescribe tirzepatide. Our practice also sells dermal filler, Sculptra and Morpheus8 — the treatments for exactly the problem I’m about to describe. We profit at both ends of this. You should read everything below knowing that, and you should be suspicious of anyone in my position who writes about this without saying so.

Where does the term come from?

Not from a study. The phrase was coined by Dr. Paul Jarrod Frank, a cosmetic dermatologist in New York, in a New York Times piece in early 2023, describing what he was seeing in his own practice. It spread quickly from there, and within months it had become the shorthand everyone used for a face that had thinned during weight loss.

It is not a diagnosis. It doesn’t appear in any classification of disease, and no medical body has defined it. That doesn’t make it fictional — a clinician noticing a pattern in their own practice is often where research starts. But there’s a difference between a pattern a doctor notices and a phenomenon someone measured, and for the first two years this was firmly the former.

What did anyone actually measure?

In 2025, a team at Vanderbilt published the first objective attempt. Sharma and colleagues, in Otolaryngology–Head and Neck Surgery, searched their hospital records for patients who had CT or MR imaging of the head both before and after starting a GLP-1, between 2017 and 2024, and measured midfacial volume on the scans.

Twenty patients met the criteria. Median age 54. Average weight loss 11 kg over an average of 321 days.

Median total midface volume fell 9.0%. Broken down by layer, the superficial fat compartments — the ones that give cheeks their fullness — lost a median 11.0%. The deep compartments lost a median 7.0%. Linear regression put it at roughly 7% of midfacial volume per 10 kg lost.

Those are real numbers from real imaging, and they confirm that patients aren’t imagining this.

So is it the drug, or is it the weight loss?

Here’s what the study can’t tell you, and it’s the part that gets left out of every article I’ve read on this.

There was no control group. Nobody compared these twenty patients against twenty people who lost the same weight through diet, exercise or bariatric surgery. Without that comparison, the study cannot separate “this is what semaglutide does to your face” from “this is what losing 11 kg does to your face.”

And the study’s own internal numbers point toward the second answer. Superficial fat loss correlated strongly with how much weight a patient lost. Deep compartment loss showed essentially no correlation with weight at all — and its spread was enormous, wide enough that some patients in the sample actually gained deep volume. If the drug were doing something specific to facial tissue beyond driving weight loss, you’d expect a cleaner signal than that.

There is a second hypothesis in circulation — that GLP-1 receptors exist in skin tissue and the medications may affect the skin directly. A 2025 review in Endocrine discusses it. I’d file it as interesting but still unproven. It is a proposal for future research, not a finding, and I’d want to see it tested before I built a treatment plan around it.

Fact vs. Fad: The honest version is that rapid weight loss thins the face, GLP-1 medications produce rapid weight loss, and nobody has yet demonstrated a facial effect independent of the weight loss itself. “Ozempic face” is a memorable name for something that probably would have happened at that rate of loss regardless of how you achieved it. Bariatric surgeons have described the same thing for decades without a catchy label.

Practically, this changes very little about what to do. It changes quite a lot about who to blame.

Can you prevent it?

Partly, and I’d rather be careful here than sell you certainty.

Slower loss gives skin and soft tissue more time to accommodate. Adequate protein and resistance training during treatment protect lean mass, though neither has been shown to prevent the facial fat pads from emptying. Those are sound recommendations that I make to every patient on a GLP-1, and they stand on firm ground for preserving muscle and body composition.

What I can’t tell you is that any of it has been shown to prevent facial volume loss specifically. Nobody has run that trial. I recommend these things because the physiology is sound and the downside is zero, not because I have evidence they protect your cheeks.

The one thing that genuinely reverses it is regaining the weight, which is not a recommendation.

What actually restores volume — and what doesn’t?

This is where being precise matters, because the three treatments people lump together do different things.

Hyaluronic acid filler replaces volume immediately. It’s structural — you put material where fat used to be. Results are visible the same day, which is both its advantage and its risk, because a deflated midface is easy to overfill in a way that looks obvious.

We place filler under Clarius ultrasound guidance. In a face that has lost 9% of its midfacial volume, the anatomy is not where it was, and the vessels are closer to the surface than the textbook depictions suggest. Seeing the plane you’re injecting into rather than inferring it is worth more in a deflated face than in a full one. Very few practices in this city utilize ultrasound guidance.

Sculptra is the one with the most interesting regulatory history for this problem. Poly-L-lactic acid received FDA approval in the United States in 2004 specifically for facial lipoatrophy — the facial fat loss seen in patients on older HIV antiretroviral therapy. That is nearly the same clinical problem: fat pads that emptied faster than the face could adapt. Sculptra doesn’t fill; it promotes your own collagen growth over months. Results build gradually across a series of treatments.

Morpheus8 does not restore volume. I want to be clear about that, because it gets sold into this conversation constantly. Radiofrequency microneedling addresses skin laxity and quality — loose skin, texture, mild tightening. If your complaint is hollowness, Morpheus8 is the wrong tool. If your complaint is that skin no longer fits the face beneath it, it’s reasonable. A 2026 scoping review in JMIR Dermatology cited work finding that a single session of fractional bipolar RF microneedling achieved roughly 37% of the skin laxity improvement seen with a surgical facelift. That’s a genuine effect, and it comes without the downtime of surgery.

When should you actually do any of this?

Later than most people want to, and this is the part that costs us bookings.

If your weight is still dropping, filler placed today is being placed into a face that will keep changing. Volume that looks correct at 190 pounds can look overdone at 170. I’d rather see you at a stable weight for two or three months before we replace anything structural — with the exception of Sculptra, which builds slowly enough that starting during the tail end of loss can work reasonably well.

I’d also say this, as the person prescribing the medication: a face that looks thinner is not a reason to stop a treatment that’s working on your metabolic health. Those are not equivalent stakes. If the facial change is distressing you, that’s worth addressing directly rather than by abandoning the weight loss.

Come in and let us look at what’s actually deflated. Sometimes the answer is a syringe of filler in two specific places, not a full-face plan. Occasionally the opposite is the problem — submental fullness that persisted through the weight loss, which is a question about Kybella for a double chin rather than about replacing volume. Consultations are complimentary, and the provider you book is the provider who treats you.

References

Sharma RK, Vittetoe KL, Barna AJ, Takkouche S, Varelas AN, Yang SF, Stephan SJ, Patel PN. Radiographic midfacial volume changes in patients on GLP-1 agonists. Otolaryngology–Head and Neck Surgery. 2025;173(2):360–366. doi:10.1002/ohn.1209

Paschou IA, Sali E, Paschou SA, et al. GLP-1 receptor agonists and possible skin aging. Endocrine. 2025;89:680–685. doi:10.1007/s12020-025-04293-w

Wollina U, Wetzker R, Abdel-Naser MB, Kruglikov IL. Role of adipose tissue in facial aging. Clinical Interventions in Aging. 2017;12:2069–2076. doi:10.2147/CIA.S151599

Panlilio M, Bolen R, Martini O, Bonk A, Tedesco J. Navigating the intersection of radiofrequency microneedling and surgical facelifts: scoping review. JMIR Dermatology. 2026;9:e78385.