Mandibular Advancement Device for Sleep Apnea

A mandibular advancement device is a mouth guard-like appliance worn during sleep that holds the lower jaw slightly forward, physically opening the airway behind the tongue and soft palate to reduce or eliminate obstructive sleep apnea events. Research consistently shows these devices cut breathing disruptions during sleep by roughly half or more in most users, making them a genuine treatment option rather than a consolation prize for people who cannot tolerate CPAP. But who they work best for, what trade-offs they carry for your teeth over time, and how they actually stack up against the gold-standard pressurized mask are all more nuanced than the marketing brochures suggest.

How the Device Opens Your Airway

The basic idea is mechanical: by pushing the lower jaw forward a few millimeters, the tongue and surrounding soft tissue follow, widening the space at the back of the throat where collapse typically occurs during sleep. Imaging studies confirm this. One large-scale study using computational modeling found that mandibular repositioning changed the pharyngeal passage, and that the change in the narrowest cross-sectional area was the best predictor of how much airway resistance dropped.¹ In a prospective clinical study, wearing a MAD significantly increased total upper airway volume, with the most dramatic enlargement occurring in the area behind the soft palate.²

What about the jaw joint itself? That is a reasonable worry, given that the device is essentially holding your jaw in an unnatural position for hours every night. Biomechanical modeling has examined this question directly. One study found that stress on the temporomandibular joint structures did not increase substantially at rest regardless of how far the jaw was advanced, though closing the jaw against the device did require extra muscle force as advancement increased.³ A separate simulation study using finite element modeling concluded that the stress on the jaw joint and the ligaments holding the teeth in place remained within safe ranges even for long-term use.⁴

How Well Do They Actually Work

The headline number most sleep specialists track is the apnea-hypopnea index, which counts how many times per hour your breathing partially or fully stops during sleep. Across studies, MADs reliably bring that number down. In one randomized crossover trial, the device achieved at least a 50% reduction in breathing events in about 46% of patients, and brought the index below the normal threshold of five in about 32% of cases.⁵

Those numbers look modest, but they represent averages across all severity levels. When researchers break results down by severity, the picture gets more interesting. A study examining outcomes across mild, moderate, and severe categories found statistically significant improvements in all sleep parameters for the moderate and severe groups. In patients with severe apnea, the median index dropped from 49 to 9 events per hour, and blood oxygen levels improved significantly.⁶ A separate retrospective study focusing specifically on very severe cases (starting indexes above 57 events per hour) found a mean reduction of about 73%, with over 95% of patients achieving at least a 50% reduction.⁷ These severe-case results are small studies and should be interpreted cautiously, but they challenge the old assumption that MADs only belong in the mild-to-moderate bucket.

Beyond raw numbers, the question patients usually care most about is whether they actually feel better. Most studies show improvements in daytime sleepiness scores, though the size of that improvement varies. One trial in mild apnea patients found no difference between MAD and CPAP for daytime sleepiness, mood, or sustained attention.⁸ That finding surfaces repeatedly across the literature: objective breathing metrics favor CPAP, but subjective feeling-of-improvement often looks similar for both treatments.

MAD Versus CPAP

CPAP is technically the more powerful tool. An updated meta-analysis comparing the two found that CPAP lowered the breathing-event index by about six more events per hour than MADs did, and produced slightly better oxygen saturation numbers.⁹ On paper, that makes CPAP the clear winner. In practice, it is more complicated.

The reason is adherence. A systematic review found that people used CPAP about 1.1 hours less per night than they used their oral appliance, and that gap was enough to wash out the efficacy advantage across quality of life, cognitive function, and other real-world outcomes.¹⁰ A prospective non-randomized study made this point starkly: nightly adherence was significantly higher with the oral device (about 6.7 hours per night versus 5.4 for CPAP), and the overall real-world effectiveness metric, which combines efficacy with how much the device is actually used, was virtually identical between the two treatments.¹¹ A systematic review in the Journal of Sleep Research summarized the dynamic well: MADs offer better adherence, especially for mild to moderate cases, and despite being less effective at raw index reduction, they produce significant improvements in sleep quality and cognitive function.¹² When a treatment sits unused on the nightstand, its superior efficacy does not matter.

Long-term dropout rates between the two treatments appear to be similar. A meta-analysis found no significant difference in dropout rates between MAD and CPAP users over extended follow-up periods.¹³ That suggests the adherence advantage of the MAD manifests more in hours used per night than in whether people abandon treatment altogether.

Effects on Blood Pressure

Sleep apnea is strongly linked to high blood pressure, particularly the nighttime spikes that happen when breathing repeatedly stops. Both CPAP and MADs lower blood pressure, and the surprising finding from a large network meta-analysis published in JAMA is that the two treatments are essentially equivalent on this front. CPAP was associated with about a 2.5 mmHg reduction in systolic blood pressure and MADs with about a 2.1 mmHg reduction, with no significant difference between them.¹⁴

A more recent head-to-head trial found that the MAD group experienced a 2.5 mmHg drop in 24-hour mean arterial blood pressure at six months, while the CPAP group showed no significant change. The MAD group’s advantage was most pronounced in nighttime blood pressure readings.¹⁵ An interesting sex-specific finding emerged from a randomized trial: women using the device saw nighttime systolic blood pressure drop by about 11 mmHg compared to a sham group, while men showed no significant difference.¹⁶ The reasons for this sex difference are not well understood, but it suggests that women with sleep apnea and hypertension may benefit disproportionately from oral appliance therapy.

When it comes to broader cardiovascular and metabolic markers, the evidence is thinner. A randomized trial following patients with moderate apnea for 12 months found no major changes in cardiovascular risk factors with either MAD or CPAP treatment.¹⁷ The blood pressure benefits appear more reliable than broader cardiovascular remodeling, at least over the study durations available so far.

What Happens to Your Teeth Over Time

This is the part that gets underplayed in sleep clinic brochures. Long-term MAD use gradually shifts your bite. The changes are slow and individually small, but they accumulate. A systematic review and meta-analysis found that on average, the vertical overlap of the front teeth (overbite) decreased by about 0.6 mm and the horizontal overlap (overjet) decreased by about 0.8 mm, with both changes progressing steadily the longer treatment continued.¹⁸ Upper front teeth tilted slightly backward and lower front teeth tilted slightly forward, and the lower jaw itself shifted a tiny amount posteriorly. None of these individual numbers are alarming, but they trend in one direction.

A decade-long follow-up study provides the most sobering data. Over roughly ten years, overbite decreased by about 2.3 mm and overjet by about 1.9 mm, with corresponding increases in lower-jaw dental arch width and lower front tooth crowding.¹⁹ Some changes slowed over time, but overjet and lower arch crowding continued decreasing at a constant rate, meaning they did not plateau even after many years. Another study measuring objective dental changes versus what patients noticed found a median overjet decrease of 1.6 mm and increased irregularity of the lower front teeth, though patients did not always perceive these changes subjectively.²⁰

For most patients, these bite changes are cosmetically minor and clinically acceptable given the seriousness of untreated sleep apnea. Some patients actually welcome the shift, finding that the slight forward repositioning of lower teeth improves their profile. But anyone starting MAD therapy should understand that their bite will likely change over the years. Periodic dental monitoring is part of the deal.

Jaw Pain and Temporomandibular Concerns

Early jaw discomfort is common. Many new MAD users experience soreness in the jaw joint or chewing muscles during the first weeks. A systematic review found that some studies reported temporary increases in jaw-related pain or symptoms early in treatment that later subsided.²¹ The same review noted that no patients in any included study discontinued MAD therapy because of temporomandibular problems.

The longer-term picture is actually reassuring. A meta-analysis examining both short and long-term effects found that while early-treatment odds of jaw or muscle pain did not clearly change, long-term follow-up showed significantly reduced odds of temporomandibular pain, suggesting the jaw adapts and symptoms improve over time rather than worsen.²² A five-year follow-up study found no significant change in the prevalence of temporomandibular disorders among continuing users.²³

Who Responds Best

Not everyone gets the same benefit from a MAD, and predicting who will respond well is an active area of research. One consistent predictor is whether someone’s apnea is position-dependent, meaning it is mainly a problem when sleeping on the back. A prospective study found that position-dependent apnea and a low non-supine apnea index were the strongest independent predictors of treatment success, with specificity and sensitivity both around 70%.²⁴ X-ray measurements of the skull and jaw (cephalometry) were not helpful in the same study.

A more invasive but increasingly used approach involves drug-induced sleep endoscopy, where a doctor looks at the airway while the patient is sedated and simulates the jaw advancement by hand. A prospective study found that a positive response to a jaw-thrust maneuver during this procedure was significantly associated with good MAD outcomes, while persistent sideways collapse of the throat walls during the maneuver predicted treatment failure.²⁵ This makes anatomical sense: if the collapse is happening at the level where jaw advancement opens things up, the device works; if it is happening from the sides, where jaw position has less influence, it does not.

Interestingly, one imaging study found that the enlargement of the narrowest airway cross-section with a MAD in place was not significantly different between patients who responded to treatment and those who did not.²⁶ That counterintuitive result suggests that a simple static image of how much the airway opens may not be the whole story, and dynamic factors during sleep matter more.

Getting the Advancement Right

A MAD is not a one-size-fits-one device even after it is custom-made. The amount of forward jaw advancement needs to be dialed in over time, and the protocols for doing this vary. A common starting point is 60% of the patient’s maximum comfortable protrusion. One stepwise protocol starts there and advances in stages to 75% or 90% if symptoms have not improved, or backs off to 60% or 45% if side effects become bothersome.²⁷

A different approach starts with no advancement at all and increases by 1 mm at a time, checking the apnea index at each step with a home sleep test, seeking the smallest effective protrusion to minimize side effects.²⁸ Another classic protocol advances by 1 mm per week, combining symptom resolution with overnight oximetry readings to decide when to stop.²⁹ The general principle across all these approaches is the same: advance gradually, verify with objective testing, and find the minimum effective position rather than assuming more advancement is always better.

Custom Versus Off-the-Shelf Devices

You can buy thermoplastic boil-and-bite oral appliances over the counter for a fraction of the cost of a custom device. They work on the same principle but with far less precision in fit. A systematic review and meta-analysis found that custom-made devices outperformed ready-made ones by a meaningful margin on the breathing-event index and on sleep-related quality of life scores, and patients also reported substantially better adherence and expressed a clear preference for the custom devices.³⁰

That said, one randomized trial comparing a specific non-custom device to a custom one found no significant difference in apnea index reduction or self-reported outcomes between the two.³¹ The discrepancy probably reflects the fact that “non-custom” covers a wide range, from flimsy boil-and-bite devices to semi-custom prefabricated designs that sit closer in quality to fully custom ones. The literature broadly favors custom, but certain higher-end prefabricated devices may narrow the gap.

Among custom devices, you will encounter two broad designs: one-piece (monoblock) and two-piece (twin-block or bibloc) devices. Monoblock appliances lock the jaw in a fixed position, while two-piece designs allow some lateral and vertical jaw movement. A retrospective comparison found that response rates and improvements in apnea severity were similar between the two types, with monobloc responders at 61% and bibloc at 56%.³² One comparative study found the monoblock produced a larger increase in airway volume, and a narrative review suggested monoblock devices may have better short-term outcomes and higher patient tolerance.³³ ³⁴ The trade-off is that two-piece designs allow titration (gradual adjustment of advancement) without making a new device, which makes the fitting process more flexible.

Combining Treatments When One Is Not Enough

For patients whose apnea is partly position-dependent, combining a MAD with positional therapy (a method that prevents back-sleeping) can be more effective than either alone. A study of this combination found that the overall apnea index dropped to about 5.7 events per hour with the combined approach, significantly lower than either treatment by itself. The treatment response rate jumped to 95% with the combination, compared to 45% for the MAD alone and 55% for positional therapy alone.³⁵ This is a relatively practical add-on: positional therapy devices range from specialized sleep vests to simple tennis-ball techniques, and combining them with a MAD does not add much complexity.

Cost and Real-World Value

MADs cost less upfront than CPAP in most healthcare systems, but the total economic picture is more complicated. A randomized controlled trial comparing the two in moderate apnea found that total societal costs after 12 months were actually higher for MAD, driven by dental visits and device-related expenses. However, when the analysis looked at quality-adjusted outcomes instead of raw apnea reduction, the MAD performed slightly better per unit of quality-of-life gained.³⁶ A UK health technology assessment found that at a willingness-to-pay threshold of £20,000 per quality-adjusted life year, both treatments were cost-effective, with CPAP having a very slight edge at 52% probability versus 47% for MAD.³⁷ In practical terms, the cost-effectiveness difference between the two is small enough that patient preference and adherence reasonably tip the decision.

Children and MADs

Pediatric use is a smaller and less mature area of evidence. In children, sleep apnea often traces to enlarged tonsils and adenoids, and surgery is the first-line treatment. But for children with residual apnea after surgery or those with contributing jaw-structure issues, mandibular advancement devices have shown promise. An umbrella review found that all included studies reported improvements in the apnea index after MAD use in children.³⁸ A second umbrella review confirmed the finding but noted that the evidence quality remains low, with significant differences across studies in device designs, treatment duration, and patient populations.³⁹

An interesting wrinkle in the pediatric context is that because children’s jaws are still growing, the advancement may actually promote favorable skeletal development. A systematic review of devices combining mandibular advancement with palatal expansion found decreases in the apnea index and improvements in oxygen saturation along with the expected orthodontic changes.⁴⁰ In adults, the bite changes from MAD use are side effects to be managed. In a growing child, those same forces might be therapeutic.

Digital Fabrication and Where the Technology Is Heading

Traditional MAD fabrication involves taking physical impressions of the teeth, sending them to a dental lab, and waiting. The process typically takes eight to ten days. A fully digital workflow using intraoral scanning, computer-aided design, and 3D printing has been developed that shortens this to two to three days.⁴¹ Digital workflows also eliminate the gag-inducing putty impressions and allow precise iterative adjustments to the design before printing. Clinical reports have confirmed the feasibility of producing functional custom MADs entirely through digital scanning and additive manufacturing.⁴²

This shift matters beyond convenience. Faster turnaround means patients start treatment sooner, and the precision of digital manufacturing may improve fit consistency from one device to the next. As 3D printing materials improve and dental practices invest in scanning equipment, the cost and accessibility barriers for custom devices should continue to drop, potentially closing the gap with off-the-shelf alternatives while maintaining the clinical advantages of a custom fit.