Hyaluronic Acid Injections: Lips, Under-Eye Bags, Duration and Risks, According to Studies

★ OUR PRODUCT

Hyaluronic Acid Serum

A facial serum with sodium hyaluronate, aloe vera leaf juice and glycerin, designed for mature, dry and wrinkled skin. A few drops morning and evening on clean, thoroughly cleansed skin, alone or before your usual cream.

View the Hyaluronic Acid Serum

Hyaluronic acid is a naturally occurring molecule in the skin: a 70 kg adult contains approximately 15 g, half of which is in the skin, according to a 2024 review on filler product manufacturing. In its free form, its half-life is approximately one day: when injected as is, it disappears within several days.

Hence the cross-linking : hyaluronic acid, produced by bacterial fermentation, is chemically bonded into a network that resists degradation. The linking agent is almost always BDDE (1,4-butanediol diglycidyl ether), whose residues must remain below strict thresholds. A 2024 review notes that sensitization to BDDE has been described, particularly in occupational settings, and judges its long-term effects insufficiently evaluated; the 2016 consensus considered biocompatibility data partly reassuring for certain products.

The elastic modulus, denoted G', measures gel firmness: a high G' is suitable for volume, such as cheekbones or chin, a low G' for fine lines. Monophasic gels, more cross-linked, are smooth and cohesive; biphasic gels, less cross-linked, are granular and less cohesive. On nasolabial folds, a meta-analysis of 22 trials (1,848 participants) found more marked and more durable improvement with monophasic gels, from 2 to 52 weeks, with less pain.

Most trials compare a given gel to no treatment, with an evaluator blinded to treatment status. Each focuses on a specific product, whose results do not automatically transfer to another gel.

Nasolabial Folds. A meta-analysis of 51 randomized trials (4,097 patients), judged to be of low methodological quality by the authors, covers all filler products (hyaluronic acid, collagen, polylactic acid, fat, etc.). In the 19 trials (1,155 patients) using the 5-point severity scale, the score changed, across all products combined, from 3.23 before treatment to 1.79 at one month, 2.02 at 6 months, and 2.46 at 12 months; for hyaluronic acid alone, the gain at 6 months was 1.15 points. Improvement peaks around 3 to 5 months, generally using less than 2 mL per fold.

Lips. In 180 adults with thin lips, 93% of patients treated with a small-particle gel (1.5 mL maximum per lip) had gained at least one volume grade at 8 weeks, compared to 29% of untreated controls; 74% still considered themselves improved at 24 weeks. In another trial comparing two gels for the lips, responders decreased from 88% at 8 weeks to 60% at 48 weeks, using an average of 1.82 mL.

Cheekbones. In the pivotal trial of a volumizing gel at 20 mg/mL, 235 adults aged 35 to 65 were treated and 47 remained as controls: at 6 months, 85.6% of treated patients had gained at least one point on the volume scale, and nearly half maintained the correction at 24 months.

Chin. In 140 adults with a receding chin, 83.3% of patients injected with a firm gel at 25 mg/mL were responders at 3 months, compared to 10.8% of controls, with the effect maintained up to 12 months.

Nose. A systematic review of 37 studies (7,339 patients) on nasal injections reports 13.34% of mild complications (bruising, hematomas, edema, asymmetry) and 3 serious complications per 1,000 patients (blindness, necrosis, stroke). The nose is classified, along with the glabella, the base of the nasal wings, and the temples, among zones that the 2016 consensus classifies as higher vascular risk areas.

The result is visible immediately, but swelling initially masks the actual shape. In trials, a touch-up was offered approximately 4 weeks after the first injection when the correction was not optimal.

Duration then depends on the area: for the lips, 74% of patients still considered themselves improved at 24 weeks in one trial, 60% remained responders at 48 weeks in another; under the eyes, 63.5% were at 12 months; for the cheekbones, duration even varies from one sub-area to another.

In this same trial, 92.8% of patients considered their cheeks improved at 6 months, and still 79.0% at 2 years. The second session required less product: in 167 patients retreated 12 to 24 months later, the new session required an average of 3.13 mL, compared to 6.8 mL for the initial treatment including touch-up, with 91.1% responders 12 months later and fewer local reactions.

The disappearance of the visible effect does not mean the gel has disappeared. In three patients consulting for swelling around the eyes appearing years after an injection, imaging and tissue analysis found hyaluronic acid up to 9 years later. These isolated cases suggest always disclosing past injections.

Type A botulinum toxin, commonly called botox, temporarily relaxes muscles whose contraction deepens expression lines: glabellar lines, forehead lines, crow's feet. Hyaluronic acid compensates for volume loss or fills a visible fold at rest, such as the nasolabial fold, under-eye hollow, or thin lips.

Duration also differs. A pooled analysis of 4 trials (621 patients) treated with 20 units of toxin for glabellar lines found 84.2% responders at 30 days and a median duration of effect of 120 days on wrinkles in contraction, 131 days at rest.

The 2016 international consensus reserves botulinum toxin as first-line treatment for cases where excessive muscle contraction is the primary cause of the defect, and notes increasingly frequent combinations with hyaluronic acid.

In France, hyaluronic acid injection is a medical procedure. The consultation collects your medical history (past injections and their complications, allergies, diseases, medications) and your written consent, as recommended by international consensus.

The price. No published study provides a reference rate. The price depends mainly on the gel volume, thus on the number of syringes: in trials, approximately 1.8 mL for both lips, less than 2 mL per nasolabial fold, 6.8 mL on average for a complete initial treatment of the mid-face, approximately half that afterwards. Add to this the number of areas, the type of gel, and a possible touch-up.

Precautions. The 2016 consensus recommends skin without makeup, cleansed with antiseptic; preventive antiviral in case of herpes history if the area is exposed, such as the lips, and postponement of the session in case of herpes outbreak; a case-by-case decision for medications and supplements that thin the blood. An anticoagulant, a simple relative contraindication, should never be stopped without the advice of the prescribing physician; slow injection and prolonged compression limit bruising. The 2024 expert consensus on late reactions also cites recent dental work and active autoimmune disease as risk factors: discuss this to choose the date.

In the 39 trials (3,281 patients) of the meta-analysis on nasolabial folds that reported complications, across all filler products, the most frequent were small palpable irregularities (43%), sensitivity (41%), swelling (34%), bruising (29%), pain (28%), and redness (26%), most often mild and temporary.

Lip edema. In the 180-adult trial, 58% of treated patients experienced swelling and 44% bruising, mild in 88% of cases. In the 48-week trial, symptoms noted by patients lasted generally 7 days at most, and ice could be applied against swelling; 5% of patients experienced late events, 21 days or more after injection, all mild.

Under the eyes and on the cheekbones. In the under-eye hollows trial, 12.7% of patients reported an adverse effect, most often mild (4.9% with needle, 20.9% with cannula). Treated patients reported recovery from the session in 3 to 5 hours. On the cheekbones, reactions, mostly mild to moderate, lasted at most 2 weeks. Serious effects remain exceptional: in a series of 7,659 patients treated over 11 years by three dermatologists, across all filler products (74% of volume in hyaluronic acid), 4 were recorded (3 late nodules, 1 vascular involvement with necrosis), all resolved.

The 2016 international consensus retains three directives: no makeup after injection, no intense physical effort for 24 hours, to limit bruising and swelling, and immediate compression in case of bleeding, supplemented if necessary with cold. Small irregularities in the first few days may respond to massage, only if the physician requests it. Avoiding intense heat, alcohol and sun in the first few days, or sleeping with head elevated the first night, is practical advice, not evaluated in trials.

Nodules, redness, swelling or induration: their frequency varies by gel type. In a series of 4,500 patients treated with a volumizing gel, delayed effects occurred in 0.98% of patients, at a median of 4 months after injection, more often between October and January. Approximately one-third were preceded by an identifiable immune event, and all resolved.

The 2024 expert consensus proposes three explanations: gel structure, particularly its low molecular weight fragments; an infection or reactivated biofilm; an immune imbalance. Treatment follows the cause: hyaluronidase for a non-inflammatory reaction without infection, monitoring or corticosteroids for an inflammatory reaction, and for infection, drainage, sampling and antibiotics, with hyaluronidase not being a first-line treatment.

The Tyndall effect, described in studies on under-eye hollows, is an undesirable bluish discoloration of the lower eyelid. The 2016 consensus advises against very firm gels with high G' in the tear trough, and treats this effect with hyaluronidase injected around the gel, followed by gentle massage.

Migration refers to gel extending beyond the intended area, such as a lip that overflows above the lip line. An ultrasound study of 126 people showed that the subcutaneous layer of the lips measures less than 1 mm: in most treated patients, the gel was located in the muscle, and vertical injections were associated with deeper deposition and signs of migration. These displacements are also corrected by hyaluronidase.

If the gel is injected into an artery or compresses it, blood no longer circulates downstream and the skin suffers, up to necrosis. If the gel reaches the eye artery, it can cause vision loss, sometimes a stroke. The international consensus reminds that negative aspiration before injection is not sufficient to exclude this risk.

These accidents are rare: only one vascular injury with necrosis in the series of 7,659 patients mentioned above, 3 serious complications per 1,000 patients in nasal injections. But in worldwide reviews of blindness cases, hyaluronic acid has become the product most often implicated in published cases: 23.5% of the 98 cases in the 2015 review, behind autologous fat (47.9%), then 81.3% of the 48 new cases published between 2015 and 2018, among which 20.8% of patients had recovered complete vision.

According to the 2016 consensus: immediate cessation of injection, then hyaluronidase without delay over the entire threatened area, at least 200 to 300 units, repeated daily for at least 2 days until circulation returns (up to 1,500 units), with reassessment every 24 hours. For eye involvement, a 2026 systematic review identifies hyaluronidase as the only intervention consistently associated with better outcomes, although in the 2024 review of 365 cases, no treatment was significantly associated with vision recovery. The 2026 review also notes that neurological complications occur in up to 24% of cases: transfer to specialized care must be immediate.

Excess volume, asymmetry, lump, Tyndall effect, migration: the advantage of hyaluronic acid, highlighted by the 2016 consensus, is its reversibility. The hyaluronidase, an enzyme that breaks down hyaluronic acid, dissolves the gel, which other filling products cannot do. For minor irregularities, the consensus recommends 10 to 20 units per injection point, to be repeated if necessary.

Not all gels respond the same way. In a laboratory and animal study, biphasic gels dissolved faster and at lower enzyme concentrations than monophasic ones. Injection into the bulk of the gel was more effective than surface application, and enzyme activity decreased within 30 minutes in tissues: multiple sessions may be necessary.

Hyaluronidase can cause allergic reactions, and the consensus recommends limiting its use outside emergency situations. In a survey of 264 practitioners, 92% had never observed an acute reaction, and just over 1% had already observed anaphylaxis. The 2016 consensus also mentions needle aspiration or a small incision to evacuate the gel.

No identified study shows that a serum, cream, or supplement prolongs the duration of a filling product. The skincare products below have been studied for skin quality, not for the gel.

Sun protection. This is the skincare evaluated by the longest and largest trial on this list. In a randomized trial lasting 4.5 years among 903 adults under 55 years old, the group that applied a broad-spectrum sunscreen daily showed no detectable worsening of skin aging, and this aging was 24% less than in the group that used it as needed.

Retinoids. In 36 people with an average age of 87 years, a 0.4% retinol lotion applied to one arm up to 3 times per week for 24 weeks improved fine lines (−1.64 points versus −0.08 on a scale of 0 to 9). This result concerns retinol applied topically, not oral retinol (see oral retinol or cream); retinoids should be avoided during pregnancy (see retinol and pregnancy).

Hyaluronic acid topically. A Raman spectroscopy study on harvested human skin shows that low molecular weight forms (20 to 300 kDa) penetrate the stratum corneum, unlike high molecular weight forms (1,000 to 1,400 kDa). A serum is therefore part of a hydration routine: our Hyaluronic Acid Serum, with sodium hyaluronate, is applied morning and evening to clean skin, before cream, without replacing or prolonging an injection. Our comparison serum or capsules details both options.

Collagen taken orally. Another meta-analysis of 19 trials (1,125 participants aged 20 to 70 years, 95% women) found favorable results for hydration, elasticity, and wrinkles after approximately 90 days. By convention, an effect size of 0.2 is small, 0.5 moderate, and 0.8 large: the measured effects are on the order of a moderate effect, in trials that are often small and short.

Several of these trials used 5 g per day. Our Novagen provides 5 g of Naticol® x3Peptide marine collagen per dose, including 1,250 mg of tripeptides, and 120 mg of vitamin C, in a recommended course of at least 8 weeks; this form has not been tested as such in these trials. Without hyaluronic acid, it is reserved for adults and not recommended during pregnancy, breastfeeding, and in case of fish allergy; if taking anticoagulants, consult your physician or pharmacist. Our guide collagen, skin and wrinkles details the trials.

★ OUR FORMULA

Novagen, premium marine collagen

5,000 mg of Naticol x3Peptide marine collagen per dose, including 1,250 mg of collagen tripeptides, with 120 mg of vitamin C. A powder derived from fish skins certified Friend of the Sea, produced in France, to be dissolved in a large glass of water, coffee, tea or smoothie, at any time of day.

View Novagen

For a liquid format, our Drinkable marine collagen provides 5,000 mg of Naticol marine collagen, 60 mg of vitamin C, 20 mg of coenzyme Q10 and 10 mg of hyaluronic acid per 25 ml shot; it is not recommended for children, during pregnancy and breastfeeding, and in case of allergy to salicylate derivatives.

Frequently Asked Questions

Sources

To learn more