At-Home Treatments for Sagging Jowls: What Works and What Doesn't

The short version: most at-home treatments for sagging jowls do something real. Almost none of it happens where a jowl is made. The most generous study ever run on the most popular at-home method followed women for 20 weeks of daily practice, used blinded raters and a validated scale, measured the jawline specifically, and found nothing there. Here is the whole category, sorted by what was actually measured.

At-home approach	What the research measured	Effect on a jowl
Face yoga / facial exercise	Cheek fullness improved in one uncontrolled study. Jawline tested separately at p = 0.31	None measured
LED and red light masks	Good sham-controlled evidence for crow's feet. Retail masks deliver 8 to 500 times less energy than the trials	Never tested
Gua sha	Fourfold rise in local blood flow, observed over 25 minutes	Temporary de-puffing
Facial rollers	Cheek blood flow for about 10 minutes. No cosmetic outcome was ever measured	Temporary de-puffing
Facial massage	A controlled test of massage after moisturiser was negative	None for laxity
Microcurrent devices	400 microamps sits below the threshold that can contract a muscle. No trials for laxity	None demonstrated
Home radiofrequency	All published studies measured the eye or mouth area. No study has measured what temperature these devices reach in skin	Never tested
Jaw exercisers	No controlled evidence. Added jaw muscle bulk widens the lower face	Works against definition
Facial taping, chin straps	No cosmetic evidence of any kind	Lasts as long as the tape
Sleep position	Tested directly against facial descent in 100 women. Null	None
Oral collagen	Meta-analysis of 26 trials shows improved instrument-measured elasticity	Never tested on a jowl

The One Study Everyone Cites, Read Properly

Nearly every article recommending face yoga traces back to the same source: a 2018 paper in JAMA Dermatology. It is a good-faith study and it deserves to be read rather than summarised.

Twenty-seven women enrolled and 16 completed. They did 32 exercises, 30 minutes a day for eight weeks, then every other day out to 20 weeks. Two dermatologists who did not know which photo was which rated the results on a validated facial ageing scale. Blinded estimated age fell from 50.8 to 48.1 years, a reduction of 2.7 years, and that result was significant.

Then look at where the change happened.

Two areas moved, and they were both the cheek. The pattern makes sense: muscle building up, adding volume higher in the mid-face. A genuine effect. A pleasant one. Nothing along the jaw line changed. The study measured exactly the thing you came here about and returned a null result.

The study was also uncontrolled, meaning there was no comparison group. Eleven of the 27 women dropped out and were not counted. Around twenty comparisons were run without correcting for multiple testing. And one of the co-authors is the founder of the face yoga Program being tested, which was disclosed.

There is a controlled study, and it is the one almost nobody cites. In 2013, researchers ran nine people through seven weeks of daily facial exercise against nine controls. A blinded lay panel rated five areas, including the jawline and the area under the chin. Exactly one area beat the control group: above the upper lip. The authors wrote that it cannot be concluded that facial exercises are effective. A systematic review registered in PROSPERO and published in 2025, covering 12 studies and 321 women, reached the same verdict: the evidence remains insufficient.

In eight years, nobody has replicated the 2018 result. That is worth saying plainly, because the internet has treated it as settled since the week it came out.

None of this makes face yoga a bad thing to do. No study has reported a single adverse event, it costs nothing, and facial exercise has decent evidence for other purposes such as recovery after nose surgery and jaw joint dysfunction. It is just not a treatment for the thing it is most often sold for.

Gua Sha, Rollers and Massage: Real, and Temporary

These get grouped together because they share a mechanism and a ceiling.

The gua sha mechanism study is solid. Eleven healthy subjects, a fourfold increase in microcirculation, sustained across a 25-minute observation window. It was performed on the back rather than the face, but the physiology carries. Gua sha moves blood.

The facial roller study is the only one of its kind. Blood flow rose in the treated cheek and not the untreated one, for roughly ten minutes. Note what was not measured: any cosmetic outcome at all. Not wrinkles, not firmness, not contour. One of its authors is affiliated with a beauty device manufacturer. And the lymphatic drainage claim attached to rollers everywhere is an inference. That study did not measure lymphatic flow.

There is one controlled trial comparing gua sha against a roller over eight weeks, published in 2025 in women aged 20 to 50. Both groups showed small reductions in surface measurements, on the order of two to three millimetres. But both arms were active. With no untreated group, those numbers cannot be separated from the passage of time, expectation, and the normal drift of the measuring method. A two millimetre change in a surface path is entirely consistent with slightly less fluid in the face, which is what you would expect from massage.

Fluid and blood flow effects work over minutes to hours. Collagen remodelling works over weeks to months. Gua sha, rollers and facial massage have evidence for the first and none for the second. If your face looks better after ten minutes with a stone, that is real. It is also mostly water.

One caution about a study you will see cited for jade rollers. A 2017 paper is regularly used as proof that facial massage boosts collagen protein expression. It tested a powered oscillating device running at around 75 hertz, all of its authors worked for a cosmetics company, and it is not evidence about a stone you push across your cheek by hand.

The Devices: Where the Dose Falls Short

LED and red light masks

This is the at-home category with the most legitimate science behind it, and also the widest gap between the science and the product.

The best home mask evidence is genuinely good. A multicentre, randomised, double-blind, sham-controlled trial published in 2025 followed 60 people over 16 weeks using a combined red and infrared mask. Independent raters found significant improvement against sham at 8, 12 and 16 weeks, and no conflicts of interest were declared. The endpoint was crow's feet. Not jowls, not laxity.

The problem is dose. In 2026, researchers put four retail masks on a laboratory-grade spectroradiometer. The masks delivered between 0.017 and 1.03 joules per square centimetre per session. The most-cited red light study delivered 8.5 to 9.6, and the best-designed clinical trial delivered 126. That is a shortfall of roughly eight to five hundred times. The same measurements found up to a fortyfold difference in output between different parts of the same mask, and one mask advertised a wavelength that is physically implausible.

One more thing to carry in. The study most often quoted as the proof of red light therapy used gas discharge lamps in full-body tanning cabinets, was fully funded by the manufacturer, and drew its control group from the sponsor company's own employees without randomisation. Its control group visibly worsened during the study, which inflates every comparison. It measured no laxity endpoint at all.

Red light therapy may well turn out to be useful. At home mask doses, for a jowl, there is nothing to point to.

Microcurrent devices

Microcurrent is defined by being below the threshold of sensation, under one milliamp. That definition is also the problem. If a current is too small to feel, it is too small to make a muscle contract. The market-leading device's own regulatory filing lists an output of 400 microamps, and the company's 2021 submission states that no clinical testing was performed. Clearance in that pathway means a device is similar enough to one already on the market, not that it does anything.

There are no randomised controlled trials of microcurrent for jowls, jawline or facial laxity. For scale, a review of the entire home device field found 18 studies across every technology combined, only three with more than 50 participants, and a maximum follow-up of 12 weeks.

Why does it feel like it works? In one properly double-blinded microcurrent trial, the fake device produced 55 percent of the effect of the real one. That is not a criticism of anyone who has felt a difference. It is how faces, mirrors and expectation work, and it is exactly why blinded studies exist.

One safety note. A 2026 report reviewed adverse event filings for at-home microcurrent devices and found cardiac rhythm disturbances among the complaints logged, along with one retinal detachment. That reporting system has no denominator, so no rate can be calculated from it and none should be. But if you have an implanted cardiac device, ask your doctor before using one.

Home radiofrequency

Collagen contracts when it is heated to around 65 to 70 degrees, and at that temperature the cells that make new collagen die. The window where fibroblasts are stimulated rather than destroyed is roughly 43 to 45 degrees held for 10 to 15 minutes, and 47 degrees at the skin surface causes burns. Clinic radiofrequency reaches useful dermal temperatures only because the surface is actively cooled while the deeper tissue heats.

No published study has ever measured the temperature an at-home radiofrequency device actually reaches in skin. Not one of the home device papers reports it. Any specific number you read is invented.

The more useful number is the ceiling in clinic. A blinded, split-face, randomised trial of radiofrequency measured laxity improvement of 4.6 percent with one configuration and 7.3 percent with another, and neither reached statistical significance. An unblinded satisfaction survey of 5,700 patients in the same category reported 94 percent. Both of those numbers are real. The distance between them is the single most useful thing to carry into any consultation, including ours.

Every published home radiofrequency study measured the area around the eyes or mouth. None measured a jowl. The definitive systematic review of radiofrequency for facial rejuvenation excluded home devices from its evidence base entirely.

Why the Ceiling Exists, and Why It Isn't About Muscle Tone

Anatomists do not agree on what creates a jowl. There are three serious explanations in the literature and they genuinely conflict. A 49-cadaver study led from Melbourne found that the connective fibres running through the jowl are not thinner or weaker with age, they are longer: the short, springy tissue that lets a young face move freely without sagging simply stretches out. A second account, from the same senior author 15 years earlier, describes a space over the chewing muscle expanding as its boundaries weaken. A third argues the cheek fat compartments descend. Each has cadaver evidence. Anyone who tells you it is settled has not read the other two.

Now notice what none of the three says.

Not one of them identifies weak facial muscles as the cause. Not connective tissue lengthening, not the expanding space, not fat descent. Yet muscle tone is the premise on which face yoga, microcurrent, jaw exercisers and home facial stimulation are all sold. The entire at-home category rests on a mechanism that does not appear in any of the competing anatomical accounts of how a jowl forms. Whichever explanation eventually wins, that argument still stands.

The depths make the same point. Skin over the jaw measures about 2.1 millimetres and the fat layer beneath it about 5.2 millimetres. The scaffold that a facelift repositions sits under both. The jaw bone itself loses around 4 to 5 millimetres of height with age, and facial bone density starts falling between the young and middle age groups, earlier than the spine does. The skin along the jaw line is among the thinnest on the whole face, measuring about 1.2 millimetres at the angle of the jaw, which is a useful thing to know: the sag is not caused by heavy skin. Skin also loses its snap-back long before it loses its substance, with cheek elastic recovery roughly halving between early adulthood and the seventies.

Every structure that could be making your jowl sits below the millimetre and a half of dermis that is the entire working range of a cream or a low-powered home device.

Jowls at 25: Usually Not Ageing at All

If you are in your twenties or early thirties and you can see softness along your jaw, the most likely explanation is the shape of the bone underneath, and that is worth hearing clearly because the internet will tell you that you are ageing early.

The facial ageing literature is direct about this. In patients with a congenitally lighter or set-back jaw structure, the skeleton itself is described as the primary cause of what looks like premature ageing. A study of 40 jaw surgery patients with a mean age of 22 quantified it: roughly every millimetre of change in chin position shifted the angle under the jaw by about a degree. That figure comes from surgery rather than from faces ageing naturally, but it shows how directly the bone sets the line. Two people the same age, same weight and same skincare routine can have visibly different jawlines for reasons decided before either of them was born. Softness under the jaw in people under 40 is recognised in the surgical literature as its own clinical entity, attributed to a congenital lack of definition rather than to ageing.

Genetics accounts for a large share of how old a face looks. In a study of Danish twins, heritability of perceived age came out at 61 percent, though that cohort was aged 59 to 81, so the number describes older faces and should not be read across to a 25-year-old. No study has ever produced a heritability figure for jowls specifically, so be sceptical of any article that gives you a percentage.

Two things are worth knowing about weight. In 186 pairs of identical twins, a four-point higher body mass index made the twin look older before 40 and younger after 40. Extra facial volume works against you young and for you later. And after major weight loss, facial change does not track the scale: in patients scanned before and after bariatric surgery, total weight lost showed no significant correlation with how much volume the neck lost.

The one modifiable factor with jowl-specific evidence is smoking. Seventy-nine pairs of identical twins, one smoker each, judged blind on a validated scale. The smoking twin scored significantly worse for jowls specifically, along with the eyelids and the area around the mouth, while the forehead and crow's feet showed no difference. A five-year difference in smoking history was enough to show up. Genetics held constant by design. That is about as clean as evidence gets in this field.

The At-Home Treatments for Sagging Jowls That Are Worth Your Time

An honest article should not end at what fails. Several things at home have real evidence. They are just not the things being marketed at your jawline.

Two claims worth setting aside. Sleep position was tested directly against facial descent in 100 women, and side preference showed no significant correlation with wrinkles or facial descent. There is no trial of silk pillowcases; the study usually cited is about copper and has an industry co-author. As for tech neck, a review of 41 articles concluded there is no scientific evidence linking it to neck pain at all, let alone to a jawline.

Oral collagen deserves a fair hearing. A meta-analysis of 26 randomised trials covering 1,721 people found a real improvement in measured skin elasticity. But no trial in the literature has ever used a jowl, jawline or lower-face endpoint, and the instruments used measure how quickly the skin surface springs back, which is not the same as the multi-layer descent that produces a jowl. It moves a device that measures the top layer of skin. It has never been shown to move a jowl.

Where Clinical Treatment Starts, and What It Costs

The line is simpler than the marketing makes it. If you can see or pinch a soft pouch along your jaw line, that tissue sits below the reach of anything topical or low-powered. If what you are seeing is puffiness in the morning that settles by lunchtime, keep the gua sha and save your money.

What we do at Kaizen Therapy is WonderFace, which combines bipolar radiofrequency with synchronised neuromuscular emissions in the same session. The current offer is a 6-Week Facial Rejuvenation Program at $1,699, twelve sessions delivered twice weekly, including skin analysis, progress photos and one bonus maintenance session, and backed by our satisfaction guarantee. Payment plans start at $142 per week on the 12-week option. Ongoing maintenance starts at $199 per month with no lock-in. The full breakdown is on our pricing page.

We should hold ourselves to the same standard as everything above. In-clinic energy treatments have published evidence, and the blinded effect sizes are more modest than the industry advertises. The gap between a 94 percent satisfaction survey and a 4.6 percent blinded measurement is not unique to home devices. In-clinic treatment reaches layers a home device cannot. Early to moderate jowling is where it makes the most sense. Severe jowling with redundant skin is still a surgical conversation.

If you want the full comparison of the clinical categories, including HIFU, radiofrequency microneedling and threads, that is the companion guide to this one. If your concern is specifically under the chin rather than along the jaw, see our breakdown of double chin removal costs in Melbourne.

If you want to see what change actually looks like rather than reading about it, the before and after comparisons in our results gallery are the most complete reference we hold publicly. And the free 15-minute assessment is where we look at your face and tell you which layer is actually driving what you see. Sometimes the answer is that nothing needs treating yet.

Frequently Asked Questions