Do snoring exercises work? The 8-minute trial routine
Snoring exercises work, and they are the one cheap fix with a randomized trial that measured snores rather than asked about them. In São Paulo, 39 snorers were assigned to three months of daily tongue, palate and cheek exercises or to nasal strips plus breathing exercises. The exercise group's snore index fell from 99.5 to 48.2 snores per hour; the control group did not change (Ieto et al. 2015). The catch is doing them: 8 minutes, three times a day, for weeks before anything moves.
- In a randomized trial of 39 snorers, 3 months of oropharyngeal exercises cut the median snore index from 99.5 to 48.2 snores per hour and the total snore power index from 60.4 to 31.0; the control group did not change (Ieto et al. 2015).
- In 31 patients with moderate sleep apnea, 3 months of about 30 minutes of daily exercises lowered the apnea-hypopnea index from 22.4 to 13.7 events per hour and neck circumference from 39.6 to 38.5 cm (Guimarães et al. 2009).
- Across 9 adult studies with 120 patients, myofunctional therapy roughly halved the apnea-hypopnea index, 24.5 to 12.3 per hour, and cut time spent snoring from 14.1% to 3.9% of the night (Camacho et al. 2015).
- Four months of didgeridoo practice, about 25 minutes on 6 days a week, lowered the apnea-hypopnea index by 6.2 events per hour relative to a waiting list (Puhan et al. 2006).
- In a pilot of 20 snorers, 20 minutes of daily singing exercises for 3 months reduced recorded snoring, mostly in people who practised accurately and were not overweight (Ojay and Ernst 2000).
This article is part of a series on snoring fixes you can check with a recorder instead of a review. The numbers come from published sleep-lab trials; we do not yet have Snorly data on exercise nights, and will publish it when enough users tag them.
Do throat exercises really stop snoring?
They reduce it, by about half in the only randomized trial that recorded snores objectively. Ieto's group recruited people whose main complaint was snoring, with no or mild to moderate sleep apnea, and counted every snore over 36 dB during a sleep-lab night before and after three months. The exercise group halved its snore index; the control group, on nasal strips and breathing exercises, stayed where it started.
| Snores over 36 dB per hour (median) | Baseline | After 3 months | Change |
|---|---|---|---|
| Snore index | 99.5 | 48.2 | -52% |
The earlier trial from the same lab explains why. Guimarães assigned 31 people with moderate sleep apnea to the exercises or sham therapy: the apnea-hypopnea index fell from 22.4 to 13.7 events per hour and the neck shrank by about a centimetre, with no change in weight. The exercises toned the muscles that hold the airway open in sleep. Whether that is your bottleneck depends on your cause; see why do I snore.
What is the exact routine from the trial?
Ieto's team cut the 30-minute Guimarães protocol to half the exercises so people would actually do it: about 8 minutes, three times a day. The set below is the one the review of these trials lists for it (de Felício et al. 2018). Do it in front of a mirror for the first week.
| # | Exercise | How | Dose per session |
|---|---|---|---|
| 1 | Saline rinse | 10 mL of saline in each nostril before the set | Once |
| 2 | Palate lift with sound | Say a short "A" repeatedly, lifting the soft palate and uvula each time | 20 reps |
| 3 | Tongue slide | Tip of the tongue on the front of the palate, slide it backward along the roof of the mouth | 20 reps |
| 4 | Tongue suck | Press the whole tongue up against the palate as if sucking it to the roof of the mouth, then release | 20 reps |
| 5 | Tongue down | Tip of the tongue against the lower front teeth, push the back of the tongue down to the floor of the mouth | 20 reps |
| 6 | Cheek press | Index finger inside the cheek, tighten the cheek muscle against it | 10 reps each side |
| 7 | Chew and swallow | At every meal, chew on alternate sides and swallow with the tongue on the palate, teeth together, lips relaxed | Every meal |
After three to five weeks, once you can feel the palate move, the trial added a silent version of exercise 2: lift the palate without any sound and hold it for 5 seconds. Everything else stays the same for the full three months.
Does singing or playing the didgeridoo help snoring?
Both train the same throat muscles, with thinner evidence. Ojay and Ernst taught 20 snorers singing exercises for 20 minutes a day and recorded seven nights before and after with a voice-activated tape recorder; snoring time fell on average, mostly in people who practised accurately and were not overweight. Choir singers also score lower on a snoring scale than non-singers after adjusting for age and BMI (Pai et al. 2008). Neither is a randomized trial with objective snore counts.
The didgeridoo has one. Puhan's group in Zurich randomized 25 people with moderate sleep apnea to four months of lessons and daily practice or a waiting list. The players practised 5.9 days a week for 25 minutes on average; their apnea-hypopnea index fell 6.2 events per hour more than controls, and partners rated their sleep 2.8 points less disturbed on a 10-point scale (Puhan et al. 2006). Circular breathing is a sustained workout for the palate and tongue base; the sound has nothing to do with it.
| Method | Trial | Daily dose | Measured result |
|---|---|---|---|
| Oropharyngeal exercises | Ieto 2015, 39 snorers, randomized | 8 min, 3 times a day, 3 months | Snore index 99.5 to 48.2 per hour |
| Oropharyngeal exercises | Guimarães 2009, 31 patients, randomized | about 30 min, 3 months | AHI 22.4 to 13.7 per hour |
| Didgeridoo | Puhan 2006, 25 patients, randomized | 25 min, 6 days a week, 4 months | AHI 6.2 per hour lower than controls |
| Singing exercises | Ojay and Ernst 2000, 20 snorers, no control group | 20 min a day, 3 months | Recorded snoring reduced on average |
How long before snoring exercises work?
Weeks, not nights. Every trial ran three to four months, and the silent palate lift came in only after three to five weeks because that is how long it takes to control the muscle. Expect nothing in week one, a possible change by week four, and the trial-sized change by week twelve.
Adherence is the whole game. The routine was cut from 30 minutes to 8 because the longer version did not fit into people's days, and in the singing pilot only those who practised accurately and daily improved. The pooled adult studies halved the apnea-hypopnea index, and every one had a therapist checking in weekly (Camacho et al. 2015). At home, your recorder is the therapist.
| Percent of total sleep time | Before therapy | After therapy | Change |
|---|---|---|---|
| Snoring time | 14.1 | 3.9 | -72% |
An 8-week plan with recorded checkpoints
Eight weeks is shorter than the trials, so treat week 8 as a decision point, not the finish. Record with the phone on the nightstand, 0.5 to 1 m from your head, and change nothing else: same pillow, same bedtime within 30 minutes, no alcohol on recording nights.
| Week | Exercises | Record | Checkpoint |
|---|---|---|---|
| 0 | None yet | 7 nights | Baseline: median snore minutes, clip count, loudest clip's label |
| 1 to 3 | Full set, 3 sessions a day, mirror for week 1 | 3 nights a week | Log sessions done out of 21; expect no change in the recordings |
| 4 | Add the silent 5-second palate hold | 7 nights | Compare the week-4 median to baseline |
| 5 to 7 | Full set plus the hold | 3 nights a week | Sessions done out of 21; watch the loudest clip's label |
| 8 | Full set | 7 nights | Compare the week-8 median to baseline and decide |
In Snorly, tag exercise nights with a remedy so the app lines them up against the baseline week; the loudness label on each clip shows whether snores got quieter as well as fewer, which is what Ieto's "total snore power" captured. Any recorder with a nightly summary works.
Rule for calling it: if the week-8 median snore minutes are below your quietest baseline night and you did at least 15 of 21 sessions most weeks, keep going to week 12. If you did the sessions and the median did not move, muscle tone was not your bottleneck; test position, mouth tape if you breathe through your mouth, or weight next. Fewer than 10 sessions a week means you have not tested the exercises yet.
A phone recording cannot tell you whether your oxygen dropped. Silences of 10 seconds or more followed by a gasp are worth a doctor's visit first, not instead. Every fix ranked by evidence is in how to stop snoring.
- Put the phone on the nightstand, plugged in, and record the whole night. Baseline night 1 of 7, no exercises yet.
- Skip alcohol this evening so the baseline is honest.
- Do the 8-minute set once in front of a mirror, so you know what a palate lift feels like before week 1 starts.
- Put three daily reminders in your phone: after breakfast, after lunch, before bed.
Hear your own snoring tonightSnorly records only when it hears snoring and plays it back in the morning. Free, no account, no night limit.
Download on the App StoreRelated questions
Do mouth exercises for snoring work?
Yes, in a randomized trial with objective snore counts: three months of daily tongue, palate and cheek exercises halved the snore index, 99.5 to 48.2 snores per hour, while nasal strips plus breathing exercises did nothing. About 8 minutes three times a day, for weeks.
Which tongue exercises stop snoring?
The three from the trial: slide the tongue tip backward along the palate, suck the whole tongue up against the palate, and push the back of the tongue down with the tip against the lower front teeth, 20 repetitions each, three times a day.
Does singing help with snoring?
Probably, with weaker evidence. A pilot of 20 snorers who sang 20 minutes a day for 3 months recorded less snoring, mainly in those who practised properly and were not overweight. No randomized trial has counted snores objectively.
How long do snoring exercises take to work?
Three months in every trial; the silent palate hold was added only after 3 to 5 weeks because it takes that long to control the muscle. Record a 7-night baseline, check at weeks 4 and 8, and judge by the weekly median.
Can throat exercises help sleep apnea?
They lowered the apnea-hypopnea index from 22.4 to 13.7 per hour in moderate apnea and roughly halved it across nine pooled studies, so they are used alongside CPAP or a mouthpiece, not instead. Apps cannot diagnose apnea; a snore-silence-gasp pattern on a recording is a reason for a sleep study.
Sources
- Ieto V, Kayamori F, Montes MI, et al. (2015). Effects of Oropharyngeal Exercises on Snoring: A Randomized Trial. Chest 148(3):683-691. https://doi.org/10.1378/chest.14-2953
- Guimarães KC, Drager LF, Genta PR, Marcondes BF, Lorenzi-Filho G (2009). Effects of Oropharyngeal Exercises on Patients with Moderate Obstructive Sleep Apnea Syndrome. American Journal of Respiratory and Critical Care Medicine 179(10):962-966. https://doi.org/10.1164/rccm.200806-981OC
- Camacho M, Certal V, Abdullatif J, et al. (2015). Myofunctional Therapy to Treat Obstructive Sleep Apnea: A Systematic Review and Meta-analysis. Sleep 38(5):669-675. https://doi.org/10.5665/sleep.4652
- de Felício CM, da Silva Dias FV, Trawitzki LVV (2018). Obstructive sleep apnea: focus on myofunctional therapy. Nature and Science of Sleep 10:271-286. https://pmc.ncbi.nlm.nih.gov/articles/PMC6132228/
- Puhan MA, Suarez A, Lo Cascio C, Zahn A, Heitz M, Braendli O (2006). Didgeridoo playing as alternative treatment for obstructive sleep apnoea syndrome: randomised controlled trial. BMJ 332(7536):266-270. https://doi.org/10.1136/bmj.38705.470590.55
- Ojay A, Ernst E (2000). Can singing exercises reduce snoring? A pilot study. Complementary Therapies in Medicine 8(3):151-156. https://doi.org/10.1054/ctim.2000.0376
- Pai I, Lo S, Wolf D, Kajieker A (2008). The effect of singing on snoring and daytime somnolence. Sleep and Breathing 12(3):265-268. https://doi.org/10.1007/s11325-007-0159-1
Snorly is a snore recorder and tracker, not a medical device. It does not diagnose sleep apnea or any other condition. If you notice pauses in breathing, gasping or daytime sleepiness, talk to a doctor.