Gua Sha for Nasolabial Folds: What a 2025 Trial Found, and 4 Moves to Try

Most gua sha tutorials measure nothing.

One 2025 trial did measure something: Ahn et al., Journal of Cosmetic Dermatology, June 2025, which tracked facial surface distances around the nasolabial folds and jawline in a gua sha arm and a facial roller arm.

The headline number from that 2025 trial: gua sha shortened four facial surface distances, three across the mid face and one along the jawline, by 2.23 to 2.40 mm. The facial roller arm's reductions were larger on paper, 2.75 to 3.26 mm, but the gap between the two arms was not statistically significant. Each arm's change from its own starting point was. Neither was a wrinkle eraser. But both were real, measurable, and cheaper than a single tox appointment.

What did the 2025 RCT actually measure?

The trial randomized 34 women aged 20 to 50 into two arms of 17: a facial roller group and a gua sha group. Both used the same dual function tool, a roller on one side and a gua sha edge on the other, and one woman in the roller group dropped out, so 33 finished. Every participant did 10 minutes of the assigned technique, 5x a week, for eight weeks. The researchers measured three things at week zero and week eight: facial surface distances using a three dimensional facial scanner, muscle tone using a Myoton handheld device, and skin elasticity using a Cutometer.

The contour result is worth pausing on, and worth reading precisely. Ahn and colleagues report that the gua sha group's four surface distances shrank by 2.23 to 2.40 mm, while the roller group's shrank by 2.75 to 3.26 mm. These are not fold depths. Three run over the skin from points under the nose and on the upper lip to just below the earlobes, which the authors used to gauge the nasolabial folds, and one runs along the jawline. Each is about 125 to 131 mm long, so the gua sha change is under 2 percent of the distance. The authors note that the reductions exceed a 2.0 mm threshold earlier work linked to visible changes in facial contour.

What the trial also found, and what the gua sha influencer videos almost never mention: gua sha did not improve skin elasticity. The roller arm did. Gua sha did improve muscle tone, sharply. Oscillation frequency dropped 2.02 Hz and dynamic stiffness dropped 56.46 N/m, measured over the masseter, the jaw muscle. Translation: that muscle relaxed. The authors read this as gua sha working mainly through muscle properties, and the roller mainly through skin elasticity.

It is not collagen.

It is a muscle letting go.

This matters for who should be doing it. If your lower face feels tight and clenched, gua sha is the arm that lowered muscle tone. If your concern is skin that has lost its bounce, the roller arm was the one that improved elasticity, while on contour change the two arms were not significantly different. I'll come back to that comparison below.

Who is this guide for?

I wrote this for the person who already owns a gua sha stone and wants to know which moves are doing the work. Not the person looking for an overnight fix. If you have deep nasolabial folds from years of sun damage, weight loss, or genetic bone structure, no facial massage is going to undo that. You need filler or surgical lift. A gua sha stone can soften the visible depth of a fold by relaxing the muscles tugging on it and improving local circulation. That's the realistic ceiling.

It is also for the person whose fold deepens after a stressful week, after sleeping with their face mashed into a pillow, after a flight. That fold is mostly muscle tension and fluid. It's the most responsive to gua sha. The Ahn trial enrolled women aged 20 to 50 without recent facial procedures, not people selected for that profile.

How were the 4 moves chosen?

The 2025 trial used one fixed 10 minute sequence with a little facial oil and moderate pressure, firm enough to move the skin without bruising: strokes on the neck and shoulders, work near the ears and jaw, upward strokes from the chin toward the ears, strokes from the chin to the ears and down the neck, strokes from the cheeks to the temples, and outward strokes along the forehead. The trial tested that whole sequence, not the four moves below. Moves 1 and 4 loosely echo the direction of its strokes; Moves 2 and 3 are not part of its protocol. Anything that targets the wrong fascial vector for the fold, and anything that needs a special tool the average reader doesn't already own at home, is left out.

That leaves four moves, listed in the order you do them in a session, not by how good they look on TikTok.

Move 1: The undercheek sweep

Tool: the curved long edge. Glide oil first. Place the curve against the underside of the cheekbone, just above the fold, with the stone almost flat to the skin. Sweep upward and outward toward the ear in one slow motion, about three seconds per pass. Lift, reset, repeat. Five passes per side.

This is the move closest in direction to the Ahn protocol's upward strokes toward the ears, though the trial never tested it on its own. The cheekbone is the anchor for the deep zygomaticus muscles that pull on the corner of your mouth. Relaxing the soft tissue underneath them is what creates the visible lift.

Move 2: The marionette zigzag

Tool: the serrated edge if your stone has one, otherwise the same long curve. From the corner of your mouth, gently zigzag along the marionette line down to the jaw. Light pressure. About four passes.

It does two things. First, it works the fascia just under the skin, which can adhere to itself in patterns that pull the corners of the mouth downward. Second, it pushes lymph fluid toward the jawline drainage points. It works as a follow up to the undercheek sweep, not a replacement for it.

Move 3: The cheek lift hold

Tool: the heart shaped notch if you have one, otherwise the rounded edge. Place the notch around the corner of the mouth where the fold begins. Press inward and upward and hold. Count to ten. Lift. Repeat three times each side. This is the move nobody on TikTok actually does, because it doesn't look like anything on camera.

The mechanism is sustained pressure on the trigger area of the orbicularis oris and the lower zygomaticus minor. Holding longer than a stroke teaches the muscle to release. A 2016 study using three dimensional computed tomography on facial massage in 12 women reported marked changes in the nasolabial folds and cheeks after massage, but it did not compare sustained pressure with stroking, so it does not test this hold specifically.

Move 4: The lymph drain finisher

Tool: the flat edge. After the first three moves, sweep gently from the side of the nose down the side of the cheek, along the jaw, and down the neck to the collarbone hollow. Three passes per side. This isn't glamorous. It is also not optional. The first three moves dislodge fluid; this is the move that gets the fluid out of the face.

If you skip the drain step, you can wake up the next morning with a fold that looks slightly deeper, not shallower, because the fluid you mobilized settled back into place overnight. Treat the lymph drain finisher as the rule you don't break.

Gua sha vs facial roller: what the trial said about each

If you only own one tool and you're choosing between a gua sha stone and a jade roller, here's the actual head to head from the trial, with one caveat: both arms used the two sides of the same dual function device, not a stone or a jade roller.

The honest read: if you have a tense, slightly puffy lower face, gua sha is the right tool. If your concern is skin that has lost elasticity, the roller improved elasticity scores and gua sha did not, while the two arms' contour reductions were not significantly different. Gua sha is often assumed to be better for sculpting because of how it is marketed. The trial does not back that up.

The honest caveat

I want to be careful here. The Ahn trial is a single small RCT. It randomized 34 participants and 33 finished, all women recruited in and around Seoul, with no men and no participants over 50. A bigger trial could move these numbers. A 2022 review on gua sha, jade roller, and facial massage in dermatology concluded that the evidence base is small, the effect sizes are modest, and most published studies have methodological limits. That's the honest state of the science.

Here's what that means for you. If you read this and expect a fold to vanish, you'll be disappointed. If you read this and expect changes of about 2 mm in measurements like the trial's, over two months of 10 minute sessions five days a week, that is roughly what its gua sha arm averaged. The gua sha stone is not skincare in the sense of a serum that changes biology. It's a tool that relaxes muscle, pushes fluid, and improves local circulation.

The result is real. It is also bounded.

If your fold is structural and deepening fast, this isn't the right intervention. A dermatologist consultation about filler, a retinoid for skin remodeling, and ideally a posture and sleep audit will compound. Our writeup on how chronic stress accelerates collagen breakdown and the breakdown of what your collagen cream can and cannot do cover the layers the stone alone cannot reach. The peptides explainer covers the topicals that can.

What to know before you start

If you're buying a stone, a few choices matter. Buy one with a heart notch and a serrated edge, not just the simple curve. The notch makes Move 3 possible; the serration makes Move 2 sharper. The version with the notch and serration covers all four moves.

Learn Move 4 first. Lymph drainage is the foundation. Without it, you push fluid around and watch it pool. With it, every other move compounds.

Skip the heated electric version. Heated gua sha tools don't appear in the Ahn trial protocol or in any peer reviewed work I could find. Heat feels nice. There's no evidence it changes the fold result. If you have a budget, spend it on a real cold pressed jade or rose quartz stone and an oil that glides well.

And don't skip the muscle tone reading. Press the underside of your cheekbone with your finger before and after a session. If it feels softer and your jaw feels less clenched, the session worked. If it doesn't, you went too fast or used too much pressure. The 2025 trial measured this with a Myoton; your finger is a fine approximation for the home version.

One more honest reframe before I stop. The trial's contour change is real but small, a couple of millimeters that a stranger might not spot in a before and after photo without being told where to look. The muscle relaxation is the part of the gua sha pitch that the trial actually backs. Photograph a stone or its box and Skinventry shows the materials its maker lists, who makes it, and where it is made when that is known, each with its source.

Where these numbers come from: the Ahn et al. 2025 randomized trial (34 randomized, 33 finished, 8 weeks, 10 minutes 5x a week) and a 2016 PubMed paper using three dimensional computed tomography on facial massage outcomes.