Faces looking hollow after weight loss picked up a name of its own, "Ozempic face", as Wegovy spread. In August 2025 an American otolaryngology journal published a study that measured the change on imaging. Midfacial volume fell by about 7% for every 10 kg of weight lost.
That the face hollows with weight loss has been said for a long time, but there has been little data putting a number on it. The researchers searched records at Vanderbilt University Medical Center for patients prescribed a GLP-1 drug between 2017 and 2024. Twenty of them had CT or MR imaging covering the face both before and after the prescription.
The imaging was not taken for this study. Scans done for other reasons were collected afterwards. The researchers measured the volume of the midfacial region on each scan and set the difference against the change in body weight.
The midface is the region from below the eyes to above the corners of the mouth, taking in the front of the cheekbone and the upper cheek. Fat there sits in a superficial layer near the skin and a deep layer against the bone, and the two do different work and shrink at different rates.
For every 10 kg of weight lost, midfacial volume fell by about 7%. Across all 20 patients the median reduction in the midface was 9%, and 11% for the superficial fat layer alone.
| Region measured | Finding |
|---|---|
| Midface overall | Median 9% reduction |
| Superficial fat pads | Median 11% reduction |
| Deep compartment | Not correlated with weight lost (rho 0.115, p=.629) |
Rho in the last row describes how closely two values move together. Near 1 they move together; near 0 they move independently. The deep layer came out at 0.115, close to zero.
The p value is the probability of a result like this arising by chance. Below 0.05 is the usual threshold for calling something unlikely to be chance, and the deep layer was 0.629. That does not mean the deep layer did not shrink. It means its change did not track the amount of weight lost.
The layer that visibly thinned with weight loss was the superficial fat. It sits directly under the skin, so when it goes the surface contour follows it down. Hollow cheeks and a more pronounced nasolabial fold come from there.
The deep layer rests on bone and acts as the structural support of the face. In this study its volume moved independently of how much weight was lost.
This is data from 20 patients. It does not establish that the deep layer is spared, only that its change was not explained by weight lost, and that is as far as this data goes.
The largest limitation is that no comparison was made with people who lost weight without the drug. So this is less a measurement of what a GLP-1 drug does to the face than of what happens to the face when weight comes off.
The records span 2017 to 2024, so the drug used, the dose and the duration also differ from patient to patient.
The usable part of this study is that the change tracks weight. Lose 10 kg and about 7% of midfacial volume goes with it. While weight is still coming off, the face keeps changing.
It is better to look at the face once weight has held steady for a couple of months. Volume placed before weight loss is finished has to be adjusted again after more comes off.
It also matters that the layer that shrank was the superficial one. A surface that has dropped because the superficial layer thinned is handled differently from a face that was short on deep support to begin with. Which one it is cannot be told by eye and needs checking in person.
What this study shows is that facial volume moves with body weight. Around 7% of the midface went with every 10 kg.
So filling during active weight loss means going back for adjustment a few months later. Looking at the face once weight has been steady for a few months saves that round trip.
At the same time it is early to read the finding as an effect of the drug. Without a comparison against people who lost the same weight without it, the drug and the weight loss are not separated.
This study cannot separate the two, because there was no comparison with people who lost weight without the drug. What it confirms is that midfacial volume fell in proportion to the weight lost.
In this study midfacial volume fell by about 7% per 10 kg. With 20 patients it does not capture how much that varies from person to person.
The layer that visibly thinned here was the superficial fat near the skin, down a median of 11%. The deep layer against the bone did not track the amount of weight lost.
This study did not look at timing. But volume moved with weight, so results keep shifting while weight loss is ongoing. Timing is something to discuss at the appointment.
