Why nasolabial folds deepen, and what actually improves them

In 1989 a study cut vertically through the nasolabial folds of four fresh cadavers and examined them histologically. This is what it established.

• Beneath the nasolabial fold there is dense fibrous tissue
• Muscle fibres of the levator labii superioris group pass through the orbicularis oris and attach directly into the dermis
• The nasolabial fold is absent on the face of a newborn, disappears on a paralysed face, and persists after death

Putting those three together gives a conclusion. The nasolabial fold is not a crease left by folded skin but a structural boundary formed along the points where muscle pulls on the dermis. It appears when muscle moves, and disappears when muscle stops.

Which muscle governs which part has also been established. The levator labii superioris alaeque nasi (LLSAN) defines the medial nasolabial fold and the levator labii superioris (LLS) the middle third. A 2017 study dissecting 12 hemifaces additionally reported a distinct muscle between the orbicularis oculi and the LLSAN, present in every specimen and attaching to the cheek fat.

So the accurate phrasing is not “a nasolabial fold has appeared” but “a boundary that was always there has come to look deeper”. This difference determines everything in the treatment discussion that follows.

A 2023 study of 50 cadaveric heads explains this part in the most detail. Alongside sectioned specimens · histology · micro-CT, it also carried out mechanical tensile testing.

This is what it established.

• The fat compartments of the cheek take the nasolabial fold as their medial boundary. That is, the fat stops at that line
• The fibres joining skin to deep tissue (retinacula cutis) are arranged in a honeycomb pattern
• With age, the position of the nasolabial fold itself stays fixed by its muscular attachment while the fat above it atrophies and descends and the fibres lengthen, so that it comes to drape in a teardrop shape over a fixed boundary

The volume of the nasolabial fold has actually been measured. In a study using three-dimensional stereophotogrammetry on 87 normal volunteers (aged 13–84), nasolabial fold volume ranged from 0.0026 mL to 0.2306 mL, and the correlation with age was r = 0.727 (p < 0.0001). Men had larger volumes at all ages, and there was no left-right asymmetry.

In the same study, simply going from lying down to standing changed nasolabial fold volume significantly (p = 0.0012). In the mothers’ generation it increased by 32% on standing. There is a reason why looking in the mirror and taking a selfie lying down do not look the same.

The last row matters. We could not find a study apportioning contributions numerically, such as “fat atrophy 40%, bone resorption 30%”. If you see an article giving such numbers, it is worth checking its source.

And there is one point the field has not settled. Is it that the midface ‘descends’, or that it ‘deflates’? A 2007 study that scaled, aligned and superimposed old and current photographs of the same individuals found almost no inferior migration of upper midface landmarks and questioned the gravitational descent account. Cadaveric CT and the 50-head study above, on the other hand, support descent. Both exist.

On “filling the cheek lifts the fold”

This belief has actually been tested. 77 patients received 1–3 cc of hyaluronic acid above the malar eminence, and changes in the positions of their pores were tracked to measure three-dimensionally in what direction and how far the skin moved. The result, analysing the 37 whose expressions were consistent, was this.

Malar injection alone produced no planar movement of the tissue towards the nasolabial fold, and no traction on it. In the authors’ own terms, even 3 cc into the cheek left the skin between the injection site and the nasolabial fold unmoved. Improvement appeared when injection was made directly into the fold.

There is, however, a paper with the opposite result. A 2017 report described an additional positive effect on the nasolabial fold when the malar area was augmented with a particular hyaluronic acid volumiser, but we could not verify that paper’s design or sample size. Citing only one side would not be accurate, so we record both.

On thread lifting

The conclusion of the 2018 systematic review is worth quoting — that over ten years little or no substantial evidence supporting efficacy and safety had been added, that in all the literature except two papers the durability of the lifting effect was very limited at best, and that those two positive papers were funded by thread manufacturers.

The tensile testing in the 50-cadaver study above explains why. Even when sutures were placed deep in the cheek fat, tension was transmitted through the skin rather than through the fat layer. In other words, what ultimately takes the pulling force is the skin, which stretches easily.

It would be wrong to discuss nasolabial fold filler and leave this out. Article 56(2) of the Medical Service Act also prohibits advertising that omits important information such as adverse effects.

First the anatomy. In a 2024 study that contrast-imaged 52 Asian cadavers and reconstructed them three-dimensionally, the facial artery continued as the angular artery in 83.7%. In a 2025 ultrasound study of 45 Korean patients, 38% showed individual variation in the course of the facial artery.

So the explanation that “experience lets you avoid it” is not accurate. In more than one person in three, the vessel runs somewhere other than where the textbook says.

Do not pass over the figure of 81.3% HA in the second row. The statement that “hyaluronic acid is safe because it can be dissolved” concerns treatability, not frequency of occurrence. In the same review, complete blindness was 52%.

There are reports on the skin necrosis side as well. In a 2025 retrospective analysis of 12 patients who developed cutaneous ischaemia after nasolabial fold filler, 75% had an ischaemic pattern spanning the nasolabial fold and the whole nose.

Recommendations on injection plane have also been published. The Asian cadaver study above recommended the supraperiosteal plane for the upper third of the nasolabial fold and the dermal plane along the fold for the lower two-thirds. In the study of 45 Korean patients, the 20 injected under ultrasound guidance had zero vascular complications over 12 weeks.

The incidence is still unknown. We could not find statistics with a denominator such as “one case per so many syringes”. Everything available is based on case reports and case series. There is no basis for calling it either rare or common.

One addition on posture: it is true that gravity changes nasolabial fold volume — in the study of 87 subjects above, volume changed significantly on going from lying to standing, increasing 32% in the mothers’ generation. But this is the immediate effect of “it shows less when you lie down”, not evidence for “sleeping on your side creates wrinkles”.

A 2025 ultrasound study of 45 Korean patients classified the nasolabial fold into three types. It is not far from how we divide things in consultation, so we set it out here.

• Volume-deficient type — the tissue above has deflated, making the boundary relatively prominent
• Tissue-laxity type — the tissue above has come down and draped over it
• Muscular-traction type — strong muscular pull on expression is the main factor

This classification is a proposal, however, and not a study comparing treatment effects for each type in a randomised trial. In particular, using toxin for the muscular-traction type is offered as an option, but as written above we could not find a toxin trial with the nasolabial fold as its primary endpoint.

What we actually do is this — we look at how much it changes between sitting and lying, how the line moves on smiling, and whether the line becomes shallower when the tissue above is lifted by hand. The third check matters most. If the line stays as it is even when lifted by hand, that line is likely not one created by draped tissue but the part fixed by muscular attachment.

We also often make the judgement to do less. The nasolabial fold is not a line that can be eliminated entirely, and overfilling in an attempt to eliminate it makes the face deviate from its own expression when smiling. Setting the goal as ‘lightening the shadow’ rather than ‘removing it’ gives better results.

Apart from blindness, the frequency of everyday reactions has also been measured. These are the figures from a multicentre randomised double-blind split-face trial carried out in Korea (93 analysed, 48 weeks of follow-up).

Duration also differs by product class. In the network meta-analysis of 13 randomised trials, the main assessment point for hyaluronic acid was 6 months, whereas PLLA was superior from 3 to 13 months with the effect persisting to 25 months. In the same analysis, hyaluronic acid had a significantly lower risk of nodule formation than bovine collagen (relative risk 0.593).

So generalising “filler lasts one to two years” to hyaluronic acid is inaccurate. The 13 to 25 months belongs to the PLLA data.

Reduced to a sentence: the nasolabial fold is not something to erase but a matter of sorting out the tissue draped over it, and the answer changes depending on which tissue is the problem. That is why we cannot give a single “for nasolabial folds, do this”.

Written and reviewed by Lee Chi-Hak, MD, medical director of Miso Clinic in Daegu, South Korea. Every study cited above is given together with its design, its size and the limitations the authors themselves recorded, and where we could not find data, we have said that we could not find any.