Subscribe 

Good Morning.

We've all had that patient.

You deliver a well-fitted oral appliance. You titrate carefully. They come back at their follow-up and say they feel about the same. The sleep study confirms it: minimal improvement. You refer back to sleep medicine, they end up on CPAP, and you're left wondering what went wrong.  Nothing went wrong with the appliance. Nothing went wrong with the appliance. The problem started at the consultation, before the impression was ever taken, and over the next three issues, we're going to fix that.

For decades, oral appliance therapy has been positioned as the treatment for OSA patients who can't tolerate CPAP. What the research is now telling us is more nuanced: oral appliances work brilliantly for some patients and marginally for others, and the difference is almost entirely predictable from your initial examination. The science of OSA phenotyping is giving us the tools to identify those patients before we start.    

5-minute read👇

 OSA Is Not One Disease

In 2013, sleep researcher Danny Eckert and his team at Brigham and Women's Hospital published one of the most cited papers in sleep medicine. Their conclusion was straightforward but underappreciated: obstructive sleep apnea is not a single condition. 

It's four different physiological problems that produce the same symptom, interrupted breathing during sleep.                                                                                                              

They called the framework the PALM model, identifying four distinct phenotypes:       

1. Anatomical — The airway physically collapses due to structural factors: a narrow palate, enlarged tonsils, excess soft tissue, and a retruded mandible. This is the phenotype we've been treating for years with oral appliances, and it responds exceptionally well.        

2. Arousal Threshold — These patients wake up too easily. Any mild dip in oxygen triggers a full arousal, which fragments sleep and perpetuates the cycle of obstruction. An oral appliance moves the jaw forward — but it can't change how sensitively the brain responds to airway resistance.                    

3. Loop Gain — The respiratory control system overreacts. Think of a thermostat that keeps overshooting: oxygen drops, breathing drives up dramatically, oxygen spikes, then the system overcorrects and breathing falters again. Mechanical jaw advancement doesn't stabilize that feedback loop.

4. Muscle Responsiveness — The genioglossus and pharyngeal dilator muscles simply don't activate reliably during sleep. Without that muscular support, the airway collapses even when the anatomy is adequate.                                                                  

Here's the critical number: 69% of OSA patients have at least one non-anatomical phenotype, arousal threshold, loop gain, or muscle responsiveness, in addition to or instead of a structural problem. That's the population where oral appliances show the most variable results.                                                                                                                   

What the Success Numbers Actually Look Like. When patient selection is carefully performed, oral appliance therapy achieves 70–85% success rates, defined as a reduction of at least 50% in AHI. That's an excellent outcome.                                            

But when patients are selected without phenotype awareness? That number drops significantly. A 2022 systematic review comparing OAT responders to non-responders confirmed that approximately 1 in 3 patients show negligible improvement with oral appliance therapy when screening criteria aren't applied thoughtfully.                               

Moderate OSA (AHI 15–30) — somewhat counterintuitively — actually shows the highest oral appliance success rate — 85% — outperforming both mild and severe categories. Positional OSA patients, where breathing worsens significantly when lying on the back, achieve 77% response rates, compared to 68% for non-positional cases. Female patients have 2.4 times higher odds of treatment success.

These aren't random findings. They map directly onto phenotype. Positional OSA is a hallmark of a predominant anatomical phenotype with positional sensitivity. It responds to mechanical advancement because the problem is structural.        

The Toolkit Is Coming

Identifying phenotypes is only half the job. The other half is doing something about them, and the literature is clearer than most clinicians realize on what that "something" looks like. Combination therapies, augmentation strategies, and phenotype-aware consent processes are well-supported in the evidence base but remain underused in dental sleep practice.

Over the next two issues, we'll build out the practical toolkit:

Issue 2 — A one-page screening worksheet to phenotype your next patient from their existing PSG report, plus the augmentation playbook for each PALM trait (what's in dental scope, what requires sleep-medicine co-management).

Issue 3 — A phenotype-aware informed consent template, a co-management referral letter template, and the escalation decision tree for when titration alone isn't getting you there.

Phenotyping doesn't make oral appliance therapy harder. It makes it more honest — and more effective.

 Key Takeaways

• OSA is not one disease. The PALM model identifies four distinct physiological causes — anatomical (P), arousal threshold (A), loop gain (L), and muscle responsiveness (M) — that produce the same clinical picture.

• 69% of OSA patients have at least one non-anatomical trait. Oral appliances directly address only the anatomical contribution.

• Patient selection drives outcomes. Carefully selected patients achieve 70–85% success; unselected patients show roughly 1 in 3 negligible responders.

• Moderate OSA (AHI 15–30), positional OSA, and female sex are positive predictors. Recognize them in your pre-treatment review.

• Phenotype-aware practice is not selection bias against treatment — it is treatment design with augmentation built in.

CLINICAL CORNER 

Your 5-Minute In-Office Phenotype Screening

You don't need a sleep laboratory to identify likely OAT responders. These five assessments take under five minutes and significantly sharpen your case selection:        

 1. Mallampati Score:

Ask the patient to open wide and say "ahhh" without phonating. Score I–II (uvula and tonsil pillars visible) indicates a favorable anatomical profile. Score III–IV suggests a crowded, higher-risk airway. Obese patients with Mallampati Class IV carry a 92% negative predictive value for OAT failure.   

2. Neck Circumference 

Measure at the level of the thyroid cartilage. Men under 40 cm and women under 37 cm are favorable. Larger neck circumference correlates with non-anatomical obstruction and reduced OAT response.                                                                                 

3. Positional History

Ask directly: "Is your snoring or apnea worse when you sleep on your back?" Patients who answer yes, or whose sleep study shows a supine AHI significantly higher than lateral AHI, are among your best OAT candidates. These are predominantly anatomical patients.                                                                                                             

4. Sleep Fragmentation Pattern

Ask: "Do you wake up frequently during the night, even when you haven't had an obvious apnea event?" Multiple unexplained awakenings are a red flag for low arousal threshold, the phenotype least responsive to oral appliance therapy alone.  

5. Review the Sleep Study AHI Breakdown. If a diagnostic study is available, look for REM vs. NREM and supine vs. lateral AHI breakdown. REM-dependent OSA correlates with loop-gain and muscle-responsiveness phenotypes, which means over OAT success. NREM-predominant, supine-predominant OSA is your green light.                                                                                                                           

 Quick Reference: Favorable vs. Unfavorable Signs 

One more tool worth knowing: For uncertain cases, Drug-Induced Sleep Endoscopy (DISE) performed by an ENT or sleep medicine specialist can identify collapse patterns that predict OAT outcomes with high precision. Tongue base collapse on DISE gives 3.7x better odds of OAT success. Complete concentric palatal collapse increases the odds of failure by 5.3x. When you're on the fence, a DISE referral is worth it.             

Before You Blame the Appliance, Check the Nose

Here's a clinical pearl that doesn't get said enough: for any case where OAT underperforms, your first audit step should be a nasal evaluation.

Mouth breathing during sleep largely defeats the airway-opening effect of mandibular advancement, regardless of which PALM phenotype is driving the OSA. It's the most common preventable cause of disappointing OAT outcomes — and the one most often missed, because we don't ask the patient about it and they often can't tell us.

Nasal patency cuts across every phenotype:

• Anatomical — nasal obstruction adds to the upper airway collapse the appliance is trying to fix

• Arousal threshold — nasal congestion is a known arousal trigger

• Loop gain — turbulent nasal airflow contributes to ventilatory instability

• Muscle responsiveness — chronic mouth breathing weakens orofacial muscle tone over time

Before you reach for any other diagnostic, ask the patient or their bed partner: "Is your mouth open when you sleep? Do you wake with a dry mouth?" If the answer is yes, no amount of titration will get you where you want to be until the nose is open.

We'll lay out the full tiered nasal protocol — Tier 1 conservative, Tier 2 pharmacologic, Tier 3 ENT referral, plus a cautious approach to mouth taping — in Issue 2.

A note on this issue's sponsor: More Than Teeth is sponsored in part by Xlear, a Utah-based company whose xylitol-based saline nasal sprays we use as part of our Tier 1 conservative nasal protocol at Advanced Dental Care. Xylitol's mucosal benefits make it a clinically meaningful upgrade over plain saline for patients dealing with chronic congestion and mouth breathing. We'll cover product-level recommendations in Issue 2. 

COACH CATHY'S TAKE

 By Cathy Bennett, MS, NBCHWC

When Mike comes home talking about the patient whose appliance "should have worked," I almost always ask the same question: What was that patient eating?

We focus so much on the device that we forget the body it's going into. A patient eating a high-inflammatory diet — refined sugars, ultra-processed foods, dairy if they're sensitive, alcohol within three hours of bed — shows up with congested nasal passages, swollen mucosa, and a gut microbiome actively working against good sleep. We can advance the mandible all we want, but if the airway tissues are inflamed and the nose is closed, we're swimming upstream.

Before your next OAT case, ask one question: "Walk me through what you ate and drank yesterday." You'll be surprised how often the appliance isn't the variable that needs to change.

Want to dig deeper before Issue 2? I've put together a short reading list of the foundational PALM and OAT phenotyping papers — Eckert 2013, Edwards 2014, the 2015 AASM/AADSM guideline, and a few key DISE outcome studies. Free to all subscribers.

Next week: the screening worksheet and the augmentation playbook. If you've ever delivered a perfect appliance to a patient who didn't respond, Issue 2 is the one to share with your sleep-medicine referral partners.

REFERENCES

1. Eckert et al. (2013). Defining phenotypic causes of obstructive sleep apnea: Identification of novel therapeutic targets.
   https://doi.org/10.1164/rccm.201303-0448OC

2. Camañes-Gonzalvo et al. (2022). Comparison of the phenotypic characteristics between responders and non-responders to obstructive sleep apnea treatment using mandibular advancement devices: Systematic review and meta-analysis.
   https://doi.org/10.1016/j.smrv.2022.101644

3. Manetta et al. (2024). Relationship between obstructive sleep apnea pathophysiological phenotypes and mandibular advancement device treatment response.
   https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11294123/

4. Nanda et al. (2025). Predictors of treatment response in obstructive sleep apnea using mandibular advancement devices.
   https://pubmed.ncbi.nlm.nih.gov/39306592/

5. (2024). Polysomnographic phenotypes and mandibular advancement device treatment response.
   https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11294123/

6. Sutherland et al. (2015). Oral appliance treatment response and polysomnographic phenotypes of obstructive sleep apnea.
   https://doi.org/10.5664/jcsm.4934

7. McNicholas (2023/2026 concept paper). Translation of obstructive sleep apnea pathophysiology and endotyping toward clinical practice.
   https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10483231/

Click here for related podcast topics.

Until next week,
Dr. Michael & Cathy Bennett
More Than Teeth | A Mission for Generational Health