What Causes a Sunken Face After Braces and Is It Permanent?

The phenomenon sometimes called “braces face” is real and measurable, not just a trick of the mirror. Research using 3D facial scanning shows that during orthodontic treatment, the overall projection of facial soft tissues in adult women decreases by roughly 2 mm on average, whether or not teeth are extracted as part of the plan.¹ The hollowed cheeks, more prominent cheekbones, and thinner-looking temples that patients notice are driven by a combination of soft-tissue retraction, muscle changes, and subtle bone remodeling, and the degree to which any individual experiences them depends on face shape, treatment plan, and age.

What “Sunken” Actually Looks Like on a Scan

When researchers map the face in three dimensions before, during, and after braces, the areas that lose volume follow a consistent pattern. The infraorbital region, cheekbone and zygomatic areas, the tissue below the cheekbones, the temples, the buccal (cheek) region, and the area around the jaw angle all show negative changes, meaning the tissue sits closer to the bone than it did before treatment. On average these areas flatten by about 1 mm.² One millimeter sounds trivial on paper, but spread across the mid-face and temples it is enough to shift the visual balance of the face, making cheekbones look sharper and cheeks look hollower.

The lips tend to show the most dramatic change in absolute terms. In patients who had four premolars extracted, the upper lip retracted nearly 1 mm and the lower lip about 1.4 mm relative to a control group after accounting for normal aging effects.³ That retraction pulls the perioral tissues backward and can make the area between the nose and the upper lip look flatter, compounding the sunken impression even in regions where the actual tissue loss is smaller.

Why It Happens Even Without Extractions

Many patients assume “braces face” is caused solely by pulling teeth, since extraction allows the remaining teeth to be moved further backward. Extraction cases do tend to show more lip retraction: a 3D morphometric study found the upper and lower lips retruded significantly in extraction groups, with the temple and cheek areas depressing by about 1 mm in patients who had maximum anchorage control (the most aggressive retraction).⁴ But the temple and cheek changes in that study did not reach statistical significance, and the BMC Oral Health scanning study found the roughly 2 mm average loss of facial projection happened regardless of whether teeth had been extracted.⁵

That finding suggests something else is going on beyond tooth retraction. Part of the answer lies in how the muscles respond to treatment, and part in how wearing braces changes the way you chew.

The Muscle Factor

Your masseter muscle, the thick slab of tissue that runs from the cheekbone down to the jaw angle, is one of the strongest muscles in the body relative to its size. It also happens to sit right over the area people point to when they say their face looks sunken. During orthodontic treatment, chewing efficiency drops. Brackets and wires cause discomfort after adjustments, patients often shift to softer foods, and the bite itself is temporarily destabilized as teeth move. The result is that the masseter gets less of a workout than it is used to.

A one-year follow-up study in pediatric orthodontic patients measured this directly with ultrasound. The group wearing braces lost an average of 0.6 mm of masseter thickness over the year, while a matched control group gained about 0.5 mm over the same period, as you would expect from normal growth.⁶ A difference of roughly 1 mm in muscle bulk on each side of the face is visible, particularly in people whose faces were already lean. And because the masseter overlaps the cheek and jaw angle, its thinning contributes to the hollowed look in exactly the areas patients complain about.

A systematic review comparing clear aligners to traditional brackets found that fixed appliances had a more noticeable effect on masseter muscle coordination, though values generally stayed within normal ranges for both types of treatment.⁷ After three months of retainer use following bracket removal, muscle activity returned to normal in both groups. That detail matters because it hints that at least the muscular component of the sunken look may be partly reversible once treatment ends and normal chewing resumes.

What Happens Under the Bone

Tooth retraction does not just move the teeth; it remodels the bone around them. A study using cone-beam CT scans measured how the alveolar bone (the ridge that holds your teeth) changes when the front teeth are pulled back. In the upper jaw, the tips of the upper incisors moved lingually by an average of about 6 mm, and the bone landmark known as “point A,” which sits at the base of the upper jaw, followed to a degree.⁸ The lower jaw showed a similar pattern: the lower incisor tips moved back about 4.7 mm, and the corresponding bone landmark shifted with them.

This bone remodeling means the skeletal scaffold supporting the lower face actually moves backward during retraction. Soft tissue follows bone, so even if fat pads and muscle stayed exactly the same, the face would look slightly less projected. In practice, all three are changing at once: bone remodels, muscle thins, and the overlying soft tissue compresses. The combination is what produces the appearance patients find alarming.

Who Gets Hit the Hardest

Not everyone ends up looking gaunt after braces. The scanning study identified specific facial proportions that predict a higher risk of unflattering soft-tissue changes. People with a shorter vertical face height (measured from the bridge of the nose to the chin), a wider mid-face and lower face, and a high ratio of face height to face width were more likely to experience negative aesthetic changes in the cheek and zygomatic areas.⁹ In plainer terms, if you start with a relatively short, wide face and not much subcutaneous fat, you are more likely to notice hollowing.

Age plays a role too, though not in the way you might expect. Younger adults (under 30) who had incisor retraction actually experienced improvements in the appearance of their nasolabial folds, the creases that run from the nose to the corners of the mouth. A study found that young women in the retraction group were about 2.4 times more likely to show a visually perceptible improvement of the nasolabial fold compared to controls.¹⁰ Older adults, who have less tissue elasticity and less subcutaneous fat to begin with, may not get that benefit and are more susceptible to the hollowed look because their tissue does not “spring back” as readily.

Body composition matters independently of age. Someone who loses weight during treatment, whether from the dietary changes braces impose or for unrelated reasons, will see more facial hollowing simply because there is less fat cushioning the contours. The braces themselves did not cause the weight loss, but the timing makes it easy to blame the orthodontics.

Extraction Plans and the Lip Connection

Extraction cases deserve special attention because they involve the most dramatic tooth movement. When four premolars are removed and the front teeth are pulled back to close the gaps, the soft tissue over the lips follows the teeth. The structured-light scanning study found that the upper lip landmark retracted about 1.4 mm and the lower lip landmark about 1.9 mm in the extraction group, after subtracting normal aging changes seen in controls.¹¹ The correlation between how far the incisors moved back and how much the soft tissue changed was moderate but consistent, with correlation coefficients between 0.45 and 0.55.

Here is the piece that reassures some patients, though: that same study found no significant differences between the extraction and control groups in the temporal, cheek, and buccal regions.¹² The sunken temples and hollow cheeks, in other words, were not significantly worse in the extraction group than in people who did not have teeth removed. Extractions clearly flatten the lip profile, but the broader mid-face hollowing that bothers patients appears to be driven more by the universal aspects of wearing braces (reduced chewing, muscle thinning, duration of treatment) than by whether teeth were taken out.

Can Palatal Expansion Help?

One orthodontic intervention that moves the face in the opposite direction, outward rather than inward, is maxillary skeletal expansion. This uses a device anchored to mini-implants in the palate to widen a narrow upper jaw. Because the expansion physically pushes the two halves of the upper jaw apart, it also widens the overlying soft tissue. A retrospective study found that the cheek points moved outward by about 1.1 to 1.4 mm and forward by about 0.4 to 0.6 mm after expansion.¹³ Another study documented increases in upper and lower face heights and a wider nose following skeletal expansion.¹⁴

A meta-analysis pooling data from 22 studies confirmed that mini-implant-assisted expansion significantly widened the nasal cavity and nasal floor, and also increased the alar width (the width of the nostrils) by about 1.5 mm.¹⁵ The forward and lateral movement of the cheeks from expansion can counteract some of the flattening that braces produce. This is not a cosmetic add-on, though: skeletal expansion is done when the jaw is genuinely too narrow and there is a crossbite or airway issue. The facial filling effect is a side benefit, not an indication for the procedure.

Is the Sunken Look Permanent?

This is the question most patients actually want answered, and the honest response is “partly.” The muscle thinning component appears to be at least partially reversible. Once braces come off and normal chewing resumes, the masseter and temporalis muscles can rebuild. The muscle coordination study found that values returned to normal ranges within three months of switching to retainers.¹⁶ Regaining muscle bulk takes longer than regaining coordination, but anecdotal reports from orthodontists and patients suggest that the cheeks fill out somewhat over the 6 to 12 months after debonding, particularly if the person returns to a normal, varied diet.

The bone remodeling component is largely permanent. Once point A and point B have shifted backward following incisor retraction, the skeleton does not spontaneously return to its original position.¹⁷ Similarly, the soft-tissue retraction over the lips is stable: if the teeth stay retracted, the lips stay retracted. For patients whose sunken appearance is driven mostly by lip and lower-face changes from extraction and retraction, the new profile is generally there to stay.

The fat-loss component is independent of orthodontics entirely. If you lost weight during treatment, regaining it will restore facial fullness. Some patients who underwent treatment through their late teens or early twenties also experienced normal age-related facial fat redistribution at the same time. Separating what braces did from what growing up did is genuinely difficult, and this ambiguity fuels a lot of frustration in online forums.

Predicting Your Face Before Treatment Starts

One of the more promising developments for patients worried about facial changes is the emergence of deep-learning models that can predict what your face will look like after orthodontics. A model trained on cone-beam CT datasets achieved a mean prediction error of 1.2 mm across facial landmarks, with about 81% accuracy, and more than half of experienced orthodontists could not tell the predicted images apart from real post-treatment photos.¹⁸ A newer diffusion-model approach improved this further, reaching a successful prediction rate of about 91% within a clinically acceptable 2 mm threshold, and orthodontists in a visual test identified the real face only about 52% of the time, essentially a coin flip.¹⁹

These tools are not widely available in general orthodontic offices yet, but they signal a future where a patient considering braces could see a realistic 3D preview of the facial changes before committing to a treatment plan. For someone whose primary concern is aesthetics rather than bite function, that preview could be the deciding factor in choosing between extraction and non-extraction approaches, or between braces and an alternative treatment.

Practical Steps If You Are Concerned

If you are already in braces and noticing a sunken look, a few things are worth knowing. First, the change is likely at its most dramatic while the braces are still on, because that is when chewing is most disrupted and muscle thinning is at its peak. Many patients report that their face fills out somewhat in the months after debonding. Second, maintaining your weight during treatment matters more than most people realize. The dietary shift toward soft foods can cause unintentional weight loss, and the face is one of the first places it shows.

If you have not started treatment yet and are worried about facial changes, ask your orthodontist specifically about how much incisor retraction the plan involves. The correlation between incisor retraction and lip retraction is well established.²⁰ A plan that involves minimal retraction, or one that uses expansion rather than extraction to create space, will generally produce less flattening of the lower face. That said, the clinical decision about whether to extract is driven by the severity of the crowding or protrusion, and declining extractions when they are genuinely needed can compromise the functional outcome of treatment.

Some patients explore dermal fillers after treatment to restore volume in the temples, cheeks, or nasolabial folds. This is a cosmetic decision rather than a medical one, and it treats the appearance rather than the cause. The results are temporary, lasting anywhere from several months to a couple of years depending on the product and location. For patients whose sunken appearance is driven primarily by bone remodeling and permanent soft-tissue repositioning rather than reversible muscle thinning, fillers may be the most realistic option for restoring a pre-treatment look.

The Role of the Molar Width Change

An underappreciated factor in the hollowing of the lower cheeks specifically is what happens to the width between the upper first molars during treatment. The scanning study found a moderate correlation between changes in molar width and negative changes in the lower cheek region, with a correlation coefficient of 0.41.²¹ When orthodontic mechanics narrow the upper arch (which can happen during retraction as the arch is consolidated after extraction), the cheek tissue that drapes over the molars loses some of its outward support. The effect is subtle, but it adds to the overall impression of a thinner face.

An increase in posterior facial height was also flagged as a risk factor for negative changes in the zygomatic arch area.²² This means that in patients whose back teeth move downward during treatment, elongating the lower portion of the face, the tissue stretches over a longer vertical distance and appears thinner across the cheekbone area. Orthodontists sometimes refer to this as a “long face” tendency, and it is one of the mechanical side effects that careful bite management tries to minimize.