# Cellulite Treatments and What the Evidence Actually Supports

_Twelve cellulite approaches graded on mechanism, study design, durability and verdict — including the modality this studio operates, which has no independent randomised literature at all. The one randomised test of the closest comparable modality was negative. If your dimpling is deep and fixed, this article tells you to see a dermatologist rather than book twelve sessions._

Source: https://lanasculptstudio.com/journal/cellulite-treatments-evidence
Publisher: Lana Sculpt Studio — Body Contouring & Wellness, https://lanasculptstudio.com
Section: Cellulite
Updated: 2026-08-18

> This page is free to quote and cite. Please attribute to Lana Sculpt Studio and link to https://lanasculptstudio.com/journal/cellulite-treatments-evidence.
>
> Not medical advice. The publisher is a wellness studio, not a medical
> provider, and nothing here diagnoses or treats any condition.

## Key takeaways

- The largest systematic review to date (Lim et al., Aesthetic Plastic Surgery, 2025) pooled 24 randomised controlled trials and 2,084 patients, with a mean follow-up of 3.33 ± 13.4 weeks — a standard deviation four times the mean, which tells you most randomised trials stopped measuring inside a month while a few ran far longer. Where year-plus data do exist, they come almost entirely from single-arm manufacturer studies of the invasive procedures, not from randomised comparisons.
- Two reviews disagree because they applied different thresholds to an overlapping literature ten years apart. A 2015 review of 67 articles concluded that no clear evidence of good efficacy existed for any cellulite treatment. The 2025 review graded the overall evidence as moderate and named shockwave therapy, radiofrequency and injectables as the most promising.
- Two device families currently hold FDA clearance under the cellulite-specific product code OUP (21 CFR 878.4790, "powered surgical instrument for improvement in the appearance of cellulite"): Cellfina (Ulthera; K150505, K153677, K161885, K192185), which descends from the Cabochon Aesthetics de novo DEN110004 that created the classification in 2013, and Avéli (K212399 under Nc8, Inc.; K221336 and K232153 under Revelle Aesthetics). Both work by cutting connective bands under the skin. A second cellulite-specific code exists — OYW, "Laser, Cellulite Appearance", 21 CFR 878.4810, held by Cellulaze.
- The only randomised controlled trial of endermologie, the best-studied vacuum-roller device, was negative. Collis and colleagues (Plastic and Reconstructive Surgery, 1999) treated one leg and left the other as a within-patient control in 52 completers over 12 weeks, found no statistically significant difference between legs (p > 0.4), and concluded that the treatment was not effective for the appearance of cellulite.
- QWO, the only injectable ever FDA-approved for cellulite — indicated for moderate to severe cellulite in the buttocks of adult women — was withdrawn from production by Endo in December 2022. Bruising after treatment was near-universal rather than occasional, and Endo's own APHRODITE mitigation trial produced only a modest reduction in bruise area and severity.
- Compressive Microvibration® is the manufacturer's registered trademark rather than a generic scientific term. Europe PMC returns zero indexed records for that phrase, zero for "endosphere therapy", and nothing on cellulite from either of the two investigators the manufacturer's journal page names, Raoul Saggini and Pier Antonio Bacci. The modality this studio operates has no independently indexed randomised trial literature, and we are not going to pretend otherwise.
- The American Academy of Dermatology says of endermologie that "the cellulite tends to return within 1 month of stopping treatment." Cleveland Clinic goes further on the home-use end, saying of massage devices and foam rollers: "There isn't any evidence that suggests these tools improve the long-term appearance of cellulite."

---

Cellulite is not a disease, not a sign of poor health, and not a failure of discipline. It is a structural feature of how female subcutaneous tissue is built, present in an estimated 80 to 90 percent of women past puberty. In men it is uncommon: Cleveland Clinic puts men under 10 percent, and the 2023 surgical review by Gabriel and colleagues puts it nearer 2 percent, typically in the context of androgen deficiency or hormone therapy. That prevalence is the first honest thing to say about it, because it reframes the question. You are not fixing a defect. You are asking whether any available intervention meaningfully changes the surface appearance of normal anatomy, and for how long.

Before anything else: this is a beauty studio, not a medical provider. Nothing here is medical advice, we do not diagnose or treat any condition, and none of our treatments are medical procedures.

This article grades twelve approaches in a table — the proposed mechanism, what studies actually exist and how good they are, how long any effect lasts, and a plain verdict. The sections after the table take the ones people actually ask about. That includes the modality we operate at this studio, which is graded on the same terms as everything else.

## What is actually happening under the skin when cellulite appears?

The leading explanation is mechanical tethering, though Cleveland Clinic is careful to say the exact cause is not known. Fibrous septa — strands of connective tissue anchoring skin to deeper layers — run roughly vertically in women and pull the surface down while fat lobules push outward between them, producing dimpling. In men the septa cross at about 45 degrees and distribute load more evenly.

That anatomy explains most of the sex difference. Women have fewer, larger fat lobules; men have more, smaller ones. The 2023 review by Gabriel and colleagues in *Aesthetic Surgery Journal Open Forum* describes the mechanism as an imbalance of forces: outward pressure from lobules against inward tethering by short, unstable septa, contained by a dermis that thins with age. Oestrogen appears to matter, with progression associated with pregnancy, oral contraceptives and hormone therapy. Higher body fat worsens the appearance without being its cause — thin women get cellulite too.

Two things follow. First, anything that does not change either the septa or the dermis is working on appearance only. Second, cellulite has nothing to do with toxins, and nothing you can do at home clears anything from the tissue.

Cellulite is also unrelated to *cellulitis*, despite the similar name. Cellulitis is a spreading bacterial infection of the skin — hot, red, tender, often with fever, usually spreading over hours. It is a medical emergency, not a cosmetic concern, and no body treatment should be performed on or near it.

## Why are cellulite studies so hard to run well?

Cellulite research has four structural weaknesses: the endpoint is a photograph, the samples are small, the follow-up is short — a mean of 3.33 weeks across 24 randomised trials — and the funding usually comes from the company selling the device. These are not minor caveats. They explain why two systematic reviews of the same field reached materially different conclusions ten years apart.

The primary endpoint in most trials is a severity score read off standardised photographs. The Hexsel Cellulite Severity Scale rates five features — number of depressions, their depth, morphology, laxity and Nürnberger-Müller grade — each 0 to 3, giving a 0 to 15 score, where 1–5 is mild, 6–10 moderate and 11–15 severe. Its validation is genuinely good (intraclass correlation coefficients 0.881–0.922, Cronbach's alpha 0.851–0.989), but the review by Gabriel and colleagues notes it does not capture the patient perspective and has been validated only for buttocks and thighs.

Photography compounds this. Dimpling shifts with lighting angle, posture, muscle contraction, hydration, time of day and how recently the skin was compressed by clothing. A blinded panel is the correct control, and it is not always used.

Then there is follow-up. The 2025 systematic review by Lim and colleagues pooled 24 randomised trials and 2,084 patients and reported a mean follow-up of 3.33 ± 13.4 weeks. A standard deviation four times the mean means the distribution is dominated by a few long outliers: most randomised trials stopped measuring inside a month, and the mean on its own describes the field poorly. Where year-plus data do exist, they come almost entirely from single-arm manufacturer studies of the invasive procedures rather than from randomised comparisons. Sample sizes are often under 30. Industry sponsorship is the norm rather than the exception, and negative device trials rarely reach publication.

## Which cellulite treatments actually have evidence behind them?

Three approaches have evidence worth the name: mechanical subcision (Cellfina, Avéli), which cuts the septa and holds up for one to three years; laser-assisted subcision (Cellulaze), durable to about a year; and acoustic wave therapy, the best-supported non-invasive option at two to six months. Everything else — creams, massage, radiofrequency, cryolipolysis — changes appearance briefly or not at all. The table below grades all twelve.

"Evidence" refers to the best independently published study design we could locate, not to marketing claims. Circumference changes are reported as published; a change in limb circumference over a few weeks reflects tissue fluid and is not evidence of fat loss or of a change in body composition.

| Approach | Proposed mechanism | Best evidence located | Durability | Verdict |
|---|---|---|---|---|
| Topical caffeine | Phosphodiesterase inhibition, local microcirculation | Small, short, placebo-controlled trials, almost all of multi-ingredient formulations in which caffeine cannot be isolated. No systematic review we located reports a pooled effect for caffeine alone, and the one within-patient randomised trial of a xanthine cream (aminophylline, Collis 1999) found no difference from placebo | Requires daily use; reverses on stopping | Cosmetic. Skin looks temporarily firmer; septa unchanged |
| Topical retinol (~0.3%) | Increased dermal thickness, neocollagenesis | No randomised trial of retinol for cellulite specifically that we could locate; the dermal-thickening evidence is borrowed from photoaging research. AAD notes 6+ months before any visible change | Ongoing daily application | Weakly plausible, slow, modest. Avoided in pregnancy |
| Dry brushing | Claimed lymphatic stimulation | None. Cleveland Clinic: no proof it reduces the appearance of cellulite | Hours | Exfoliation only. The smoothing is transient flush |
| Vacuum-roller massage (endermologie) | Mechanical tissue mobilisation, matrix remodelling | The single within-patient randomised controlled trial (Collis 1999, 52 completers, 12 weeks) found no statistically significant difference between treated and untreated legs (p > 0.4) and concluded the treatment was not effective. Positive reports are observational and mostly uncontrolled | AAD: cellulite tends to return within 1 month of stopping | Widely used; the one randomised controlled test of it was negative. Any benefit is short-term appearance and comfort, not structural |
| Compressive Microvibration® (Endospheres) | Roller-matrix microvibration, tissue compression, circulatory effect | None indexed. Europe PMC returns zero records for the trademarked phrase, zero for "endosphere therapy", and nothing on cellulite from either investigator named by the manufacturer, Raoul Saggini or Pier Antonio Bacci; published material sits in non-indexed cosmetology literature linked to the technology's originators | Not established in independent literature | Mechanistically adjacent to vacuum-roller massage, whose one randomised trial was negative. Independently unproven |
| Acoustic wave / shockwave therapy | Acoustic pulses, microcirculation, neocollagenesis | One double-blinded sham-controlled randomised trial (Knobloch 2013): CSS 10.9 → 8.3 versus sham 10.0 → 10.1 at 12 weeks (P = 0.001), both arms doing daily gluteal strength training. The 2025 review reports a pooled reduction score of 2.07 ± 0.39 across four studies, scale unspecified | Six sessions at one- to two-week intervals in the randomised trials; Cleveland Clinic puts expected duration of benefit at two to six months, and nothing in the controlled literature runs past 12 weeks | The best-supported non-invasive option, on the strength of that one sham-controlled trial and its identification as a standout in the 2025 review |
| Rapid acoustic pulse (Resonic) | Acoustic disruption of septa | Multicentre single-arm trial, 42 women, single session; all 42 had severe cellulite at baseline, the population in which a one-point CSS change is easiest to achieve. 70.4% had >1-point CSS reduction, mean 1.09, at >52 weeks | Held past 52 weeks in that trial | Promising, single manufacturer-run uncontrolled dataset |
| Radiofrequency | Dermal heating, collagen remodelling | 2025 review: thigh circumference −2.09 cm, subcutaneous thickness −2.23 cm (both p<0.001), pooled across heterogeneous measurement methods | AAD: short lived, repeat sessions needed | Short-term change at best. The pooled thickness figure is not physiologically plausible and should not be quoted as an expected result |
| Laser-assisted subcision (Cellulaze, 1440 nm) | Fibre inserted under skin cuts septa and heats dermis | 2015 review named it one of only two modalities with signal; ~90% of sites improved at 1 year. Cleared under cellulite-specific code OYW | 1 year or longer | Invasive, but among the better-evidenced |
| Mechanical subcision (Cellfina, Avéli) | Blade releases individual septa under the dimple | Cellfina: 93% ≥1-point CSS improvement, 94% satisfied at 1 year; 3-year extension in 45 subjects, open-label, participants as their own controls. Avéli CONFFIRM: single-arm, open-label, blinded scoring; CSS improved 1.50 at 3 months, 1.54 at 6, 1.48 at 12 (P <.0001) | 1–3 years reported | Strongest evidence in the field, and that evidence is uncontrolled. Also a physician-performed surgical procedure |
| Injectable collagenase (QWO) | Enzymatic digestion of septal collagen | Phase 2–3 RCTs; 2025 review reports clinician-scored improvement 17.0%, patient-scored 25.7% | Reported to 2 years | Worked, then withdrawn. Bruising was near-universal, and Endo's own mitigation trial produced only a modest reduction |
| Cryolipolysis | Controlled cooling of adipocytes | AAD: it "can be very effective for getting rid of small pockets of unwanted fat" but "currently cannot get rid of cellulite" | Not applicable | Not a cellulite treatment |

## Do creams with caffeine or retinol change cellulite?

Cellulite creams change how skin looks, slightly, for only as long as you keep using them. Caffeine transiently dehydrates surface tissue; retinol at around 0.3 percent gradually thickens the dermis, and the American Academy of Dermatology says you need six months or longer of daily use before you would know. Neither touches a fibrous septum, and no topical cellulite product is FDA-approved.

The honest framing is that a good retinol is a skin-quality product that may make the overlying surface marginally less translucent and less lax over six months or more. We could not locate a randomised trial of retinol for cellulite specifically; the dermal-thickening evidence comes from photoaging research and is being borrowed. The AAD is explicit that results require daily application, and notes the possibility of irritation or allergic reaction. Topical retinoids are also avoided in pregnancy and when trying to conceive — check with your doctor before using one.

Caffeine's effect is largely surface dehydration of the tissue. Real enough to photograph, gone by the next morning. The placebo-controlled trials that exist are small, short, and mostly test multi-ingredient formulations in which caffeine cannot be separated from everything else in the tube. The one within-patient randomised trial of a xanthine cream — aminophylline, in Collis 1999 — found no difference from the untreated leg.

## Does massage — dry brushing, hand massage, or a roller device — do anything lasting?

No. The only randomised controlled trial of endermologie, the best-studied roller device, found no statistically significant difference between treated and untreated legs after 12 weeks (Collis 1999, p > 0.4) and concluded it was not effective. Cleveland Clinic says there is no proof dry brushing reduces cellulite and no evidence massage tools improve it long term. What massage does change is short-term comfort.

The Collis trial is worth describing properly, because it is the strongest single piece of evidence anywhere in the massage literature and it points the wrong way for the industry. Sixty-nine women were randomised; 52 completed 12 weeks. Each woman acted as her own control — one leg treated, one leg not — which removes almost every confounder that plagues this field. Only 10 of 35 endermologie-treated legs were rated as improved by the patients themselves. The authors wrote that they did not believe either of the two treatments tested improved the appearance of cellulite.

The regulatory position matches. LPG's devices are cleared by FDA under product code ISA, a Class I "therapeutic massager" category under 21 CFR 890.5660 — a general physical-medicine classification, not a cellulite-specific indication. That detail matters more than most marketing: a Class I massager clearance is not evidence of efficacy for dimpling.

Massage is generally understood to move interstitial fluid, and many people report their legs feel lighter afterwards. That is a comfort effect, and it is not the same as changing the dimpling. If your legs feel heavy and look worse by evening, that is the part a session may address, and it is the part we are willing to describe.

Manual lymphatic drainage is a clinical therapy for diagnosed lymphoedema, delivered by trained therapists on medical referral. We are describing it here to explain what it is and is not; it is not what we do. It moves fluid. It does not remove toxins, and no reputable source claims otherwise.

## What about compressive microvibration — the technology this studio uses?

Compressive Microvibration® — the manufacturer's registered trademark, not a generic scientific term — delivers rhythmic compression and vibration through a rotating matrix of silicone spheres. It has no independent randomised trial literature: Europe PMC returns zero records for the phrase, zero for "endosphere therapy", and nothing on cellulite from either investigator named by the manufacturer, Raoul Saggini or Pier Antonio Bacci. Mechanistically it is a close relative of vacuum-roller massage, whose one randomised trial was negative.

The trademark point is not pedantry. A zero-hit biomedical search for a proprietary marketing term is close to expected, and on its own it would prove much less than it appears to. That is why we also searched "endosphere therapy", "microvibrocompression" and the two named investigators, Raoul Saggini and Pier Antonio Bacci — all of which returned nothing on cellulite either. The conclusion holds; the single phrase search alone would not have carried it.

We think you should know that before booking. The published material on this technology appears in cosmetology journals that are not indexed in MEDLINE or Europe PMC, and the visible article page for the main one lists no named authors, no study design, no sample size, no outcome measures, no results, no funding and no conflict-of-interest statement. Non-indexed cosmetology journals are a normal publication route in European aesthetics. They are not a substitute for independent randomised evidence.

The manufacturer's own site describes the technology as "non-surgical and without contraindications for guaranteed results from the very first session", and says results are "clinically proved by Universities and Research Institutes" without naming one. We do not repeat any of that. Every mechanical body treatment has situations in which it should not be performed, and a device marketed as having none has simply not published its screening criteria — ours are set out in the table further down. And no body treatment can be guaranteed. A studio that repeats a supplier's guarantee has adopted it as its own claim.

We make no FDA claim about this device. We have not been able to confirm a clearance record, and until the distributor supplies documentation we will not use FDA language in any of our marketing.

What clients most often tell us after a session is that their legs feel lighter and less tight, and that skin looks smoother for a while. That is client-reported experience, recorded in the room, and it is not clinical evidence of a treatment effect. We label it that way deliberately, because neither an advertising regulator nor an honest reader should accept in-house observation as substantiation.

## Which cellulite treatments have the strongest evidence?

Mechanical subcision has the strongest evidence: a small blade is passed under a dimple to release the band tethering it, and improvement holds for one to three years. Cellfina and Avéli are cleared under FDA product code OUP, written specifically for cellulite subcision. The important caveat is that their pivotal studies were single-arm and open-label — no cellulite procedure has been tested against a sham at this scale.

It has the longest published durability data of anything in this article, and OUP is not the only cellulite-specific classification: laser-assisted subcision (Cellulaze) is cleared under OYW, "Laser, Cellulite Appearance", 21 CFR 878.4810.

Under OUP (21 CFR 878.4790, "used for controlled release of subcutaneous tissue for improvement in the appearance of cellulite") sit Cellfina (Ulthera; K150505, K153677, K161885, K192185), which descends from the Cabochon Aesthetics de novo DEN110004 that created the classification in 2013, and Avéli (K212399 under Nc8, Inc.; K221336 and K232153 under Revelle Aesthetics).

In the Avéli CONFFIRM multicentre pivotal study — single-arm and open-label, with no control group, though scored by three independent blinded physicians — Cellulite Severity Scale scores improved by 1.50 points at three months, 1.54 at six and 1.48 at twelve (P <.0001). That design is the ceiling of this field, and it is worth saying plainly: the best-evidenced cellulite procedures have never been tested against a sham.

Cellfina data reported in the 2023 review show 93 percent achieving at least a one-point improvement and 94 percent satisfied at one year. The three-year extension (Kaminer et al., *Dermatologic Surgery*, 2017) followed 45 subjects in an open-label design with participants serving as their own controls, and the AAD states that in a small study many patients had less cellulite for up to three years.

Reported adverse effects go beyond bruising: haematoma, pain during the procedure, post-inflammatory hyperpigmentation, and — reported in the literature after vacuum-assisted subcision — anetoderma-like depressed lesions. These are physician-managed procedures for a reason, and they are not what a non-medical studio offers. We mention them because a fair article has to tell you what the ceiling looks like.

Acoustic wave therapy is the best-supported non-invasive option, on the strength of one double-blinded sham-controlled trial and its identification as a standout in the 2025 systematic review. In that trial (Knobloch 2013), six sessions of focused shockwave — 2,000 impulses at 0.35 mJ/mm², every one to two weeks — moved Cellulite Severity Scale scores from 10.9 to 8.3 at 12 weeks, while the sham arm went from 10.0 to 10.1 (P = 0.001). Both arms also did daily gluteal strength training, so the shockwave effect sits on top of exercise, not instead of it. The 2015 review found signal for only two modalities, and this was one of them.

Radiofrequency shows measurable short-term change and poor durability. And the one injectable that worked, QWO, is gone: Endo ceased production in December 2022, citing the extent and variability of bruising after initial treatment and the potential for prolonged skin discolouration. That is the company's commercial framing. In substance, bruising was near-universal rather than occasional, and Endo's own APHRODITE mitigation trial produced only a modest reduction in bruise area and severity. The withdrawal was commercial in form and safety-tolerability in substance.

## Who should not book a body session, and what counts as a red flag?

One thing overrides everything else on this page. If one leg is newly swollen, warm, red or painful — especially if only one — do not book a body treatment. That combination can indicate a deep vein thrombosis, which massage or compression can make dangerous. Go to urgent care or your GP the same day.

Two more patterns are worth knowing before you scrutinise your own skin any further.

Cellulite develops gradually, symmetrically, and over years. Dimpling that appears suddenly, affects one side only, is warm or tender, or comes with skin thickening over a breast is not cellulite and needs a doctor, not a treatment room.

Two medical conditions are regularly mistaken for cellulite: lipoedema, a symmetrical painful fat distribution disorder that spares the feet, and lymphoedema, persistent swelling from lymphatic damage. Neither responds to cosmetic body treatment, both have real medical pathways, and if your legs are painful, bruise easily, or the swelling does not resolve overnight, get assessed before you book anything. Any swelling that is new, one-sided, painful, or still there in the morning should be looked at by a doctor before you book a session.

The table below is our screening list for treatments involving compression and vibration. It reflects standard studio practice rather than a published clinical guideline, and it is not exhaustive. If you have any medical condition, ask your own physician first.

| Situation | Why it matters | What we do |
|---|---|---|
| Pregnancy, or trying to conceive | No safety data for mechanical compression treatment in pregnancy | We do not treat |
| Active infection, cellulitis, or inflamed skin | Mechanical work can spread infection and worsen inflammation | We do not treat; cellulitis is same-day medical care |
| History of blood clots, or taking anticoagulants | Compression over a vessel carries clot and bleeding risk | We do not treat without written clearance from your doctor |
| Lymphoedema, or any limb after lymph-node dissection or radiotherapy | Lymphatic anatomy is altered; this is a medical condition needing clinical management | We do not treat; we refer you back to your medical team |
| Significant varicose veins or venous insufficiency | Compression over incompetent veins can cause pain and vessel damage | We avoid the affected area; clearance needed for anything more |
| Recent surgery or an unhealed scar in the area | Tissue is still remodelling and can be disrupted | We wait until your surgeon says the area is healed |
| Bleeding disorders or low platelets | Compression and vibration bruise readily | We do not treat without written clearance |
| Undiagnosed lumps or new skin changes | A lump that has not been examined must not be worked over | We do not treat until it has been assessed by a doctor |
| Active cancer, or current cancer treatment | Care belongs with the oncology team, not a beauty studio | We do not treat |
| Implanted electronic devices (pacemaker, ICD, neurostimulator) | Vibration and proximity risks are not characterised for this equipment | We do not treat |
| Open or broken skin, active eczema or psoriasis in the area | Mechanical work damages a compromised barrier | We avoid the area until the skin is intact |
| Hernia in the treatment field | Pressure over a hernia can worsen it | We do not treat the area |
| Uncontrolled cardiovascular disease or uncontrolled hypertension | Circulatory load during treatment is not appropriate | We do not treat without written clearance |
| Severe osteoporosis | Mechanical loading over fragile bone is not appropriate | We do not treat without written clearance |
| Directly over the spine, kidneys or thyroid | These areas are not appropriate targets for compression and vibration | We do not treat over them, in anyone |

## Where is the evidence thin, and what does none of this do?

No treatment on this page permanently eliminates cellulite, and no non-invasive one changes body composition. Only the invasive procedures that cut septa have durability published past a year, and even they release the dimples treated rather than the tendency to form them. Every non-invasive modality here requires ongoing maintenance, and the gap between what devices are cleared to claim and what they are marketed as doing is wide across this whole industry.

Specific thin spots worth naming.

The 2025 review's pooled radiofrequency figure of a 2.23 cm reduction in subcutaneous tissue thickness is not physiologically plausible as a mean effect of external heating over a few weeks, and almost certainly reflects pooling of heterogeneous measurement methods. We report it because it is what the review published. We do not present it as an expected result, and nobody selling radiofrequency should either.

The shockwave pooled "cellulite reduction score of 2.07 ± 0.39" is reported without the underlying scale being specified in the source abstract, which makes it uninterpretable on its own. The single sham-controlled trial's numbers, quoted above, are the checkable ones.

The best evidence in the field is uncontrolled. Cellfina, Avéli and Resonic all rest on single-arm, open-label, manufacturer-run studies. The randomised comparisons that do exist are small and short.

Mesotherapy and oral supplements are not on the table above because there is nothing to grade. AAD's position on mesotherapy is "few studies show this works", and on supplements, "there is no evidence that any supplement can reduce cellulite".

The modality we operate has no randomised trial at all. Not a weak one, not a small one — none that is independently indexed.

Weight change cuts both ways. Smalls and colleagues, in *Plastic and Reconstructive Surgery* in 2006, found that most women improved with weight loss but some worsened, with worsening associated with a lower starting BMI, a smaller weight reduction and increased tissue compliance.

### What does each claim in this article rest on?

| Claim | Source | Study design | Independent of the manufacturer? |
|---|---|---|---|
| Subcision holds 1–3 years | Kaminer 2017; CONFFIRM 2023 | Single-arm, open-label, participants as own controls | No |
| Shockwave beats sham at 12 weeks | Knobloch 2013 | Double-blinded sham-controlled RCT; sample size not stated in the record retrieved | Investigator-led; the same investigator also authored the field's meta-analysis |
| Endermologie is not effective | Collis 1999 | Within-patient randomised controlled trial, 52 completers | No manufacturer sponsorship stated — and the result was negative |
| Prevalence and pathophysiology | Cleveland Clinic; Gabriel 2023 | Patient resource; review article | Not a device study; author disclosures not checked |
| Radiofrequency circumference figures | Lim 2025 | Systematic review pooling heterogeneous methods | Yes, but the pooling is the problem |
| QWO withdrawal and reasons | Endo press release, Dec 2022 | Corporate announcement | No |
| Compressive Microvibration® efficacy | None located | None | Not applicable |

## What does a realistic goal look like?

A realistic goal is skin that looks and feels better in the short term, maintained by repetition, alongside the two things with the longest track record: stable body weight and muscle underneath. The AAD notes that added muscle makes overlying skin look smoother and firmer. That is unglamorous and it is also the most durable lever available.

We cannot promise you a result, and we are not going to. What clients most often report is that legs feel lighter and skin looks smoother for a period after a session, with the effect fading once sessions stop. Some people notice little. There is no published trial of this modality that would let us put a number on it. Not fewer fat cells. Not released septa. Not a permanent change.

If your dimpling is deep, discrete and unchanged by posture — the kind that stays put when you lie down — the honest advice is that a studio treatment will not release it, and a dermatologist or plastic surgeon offering subcision is the conversation with the best published outcomes. We would rather tell you that than sell you twelve sessions.


## Where the evidence is thin

- The mean follow-up across the 24 randomised trials in the 2025 Lim systematic review was 3.33 ± 13.4 weeks. The standard deviation is four times the mean, so the figure describes a skewed distribution rather than a typical trial, and it is quoted with its SD in the key takeaways and in the body's discussion of follow-up.
- The 2015 Luebberding review and the 2025 Lim review reach materially different conclusions about an overlapping literature ten years apart; both are cited and the disagreement is stated in the body rather than resolved.
- The strongest evidence in this field is uncontrolled. The Resonic, Cellfina and Avéli pivotal studies are single-arm, open-label and manufacturer-run, and this is stated in the body rather than left to the reader.
- Primary endpoints in this literature are photograph-based severity scores that vary with lighting, posture, hydration and time of day.
- The pooled radiofrequency figure of a 2.23 cm reduction in subcutaneous tissue thickness is reported as published and named in the body as not physiologically plausible for external heating over a few weeks.
- The shockwave pooled 'cellulite reduction score of 2.07 ± 0.39' is reported without the underlying scale being specified in the source abstract, and is therefore kept in the table and caveats rather than used as the headline shockwave result.
- The shockwave meta-analysis (Knobloch & Kraemer, 2015) is authored by the same investigator who ran the sham-controlled trial it pools. It is not an independent appraisal.
- The contraindication table reflects standard studio screening practice, not a published clinical guideline, and is not exhaustive.
- Sample size for the Smalls 2006 weight-loss study was not available in the record retrieved, so no participant count is stated.
- The Hexsel validation statistics and severity bands, the anetoderma adverse-event report, and the two null-result Europe PMC author searches come from the fact-checking record rather than from a search run during final editing, and each source URL should be resolved once more before publication.


## Questions people ask

### What is the single most effective cellulite treatment?

By published evidence, mechanical subcision — cutting the connective bands under each dimple. Cellfina and Avéli are cleared under the cellulite-specific FDA product code OUP, and Avéli's pivotal trial showed a 1.48-point Cellulite Severity Scale improvement at twelve months. That trial was single-arm and open-label with no control group, which is the ceiling of this field. It is a physician-performed surgical procedure with bruising, not a spa treatment.

### Do cellulite creams work at all?

Marginally and temporarily. Caffeine transiently dehydrates surface tissue; retinol around 0.3 percent may thicken the dermis over six months or more, though that evidence is borrowed from photoaging research rather than cellulite trials. No topical cellulite product is FDA-approved. Creams change skin quality, not the fibrous septa that create dimpling, and results reverse when you stop. Topical retinoids are avoided in pregnancy.

### Why did the cellulite injection QWO disappear?

Endo ceased production and sale of QWO in December 2022, citing the extent and variability of bruising after initial treatment plus the potential for prolonged skin discolouration. In substance, bruising was near-universal rather than occasional, and Endo's own APHRODITE mitigation trial produced only a modest reduction in bruise area and severity. It had been approved for moderate to severe cellulite in the buttocks of adult women, and remained approved when withdrawn.

### How long do results from massage-based cellulite treatments last?

The only randomised controlled trial of endermologie found no difference between treated and untreated legs at 12 weeks and concluded it was not effective (Collis 1999). The American Academy of Dermatology says of endermologie specifically that cellulite tends to return within one month of stopping treatment, and Cleveland Clinic says there is no evidence that home massage tools improve cellulite long term. Any benefit is short-term comfort and appearance.

### Does losing weight get rid of cellulite?

Usually it helps, but not always. Smalls and colleagues, in Plastic and Reconstructive Surgery in 2006, found most women improved with weight loss while some worsened. Worsening was associated with a lower starting BMI, a smaller weight reduction and increased tissue compliance — essentially, losing support without losing thigh fat. Building muscle underneath is the more consistently useful lever.

### Is compressive microvibration proven to reduce cellulite?

Not by independent published evidence. Compressive Microvibration® is the manufacturer's registered trademark rather than a scientific term, so we searched more widely: Europe PMC returns zero records for that phrase, zero for "endosphere therapy", and nothing on cellulite from either named investigator. The available material sits in non-indexed cosmetology journals linked to the technology's developers. Mechanistically it sits close to vacuum-roller massage, whose one randomised trial was negative.

### Can cryolipolysis or CoolSculpting treat cellulite dimpling?

No. The American Academy of Dermatology states that while cryolipolysis "can be very effective for getting rid of small pockets of unwanted fat", it "currently cannot get rid of cellulite". Cooling targets fat cells; cellulite is produced by fibrous septa tethering skin downward. The mechanism does not address the structure causing dimpling, and it should not be sold as a cellulite treatment.