How much evidence is behind each looksmaxxing procedure? We counted the papers

You are deciding whether something you saw on TikTok is a real procedure, a research-backed intervention, or a rumour with confident packaging. We counted the papers for 22 appearance interventions and put every query in the table below.

Key numbers

• Bone smashing: 4 PubMed records, 0 RCT-tagged records, and 0 registered studies. Source
• Teeth whitening: 601 RCT-tagged PubMed records, the highest count in this index. Source
• Rhinoplasty: 14,658 PubMed records and 369 RCT-tagged records. Source
• Buccal fat removal: 727 PubMed records, 15 RCT-tagged records, and 7 registered studies. Source
• Mewing / orthotropics: 58 combined-query PubMed records, 2 RCT-tagged records, and 0 registered studies. Source

Table of contents

• Which appearance procedures have the most research behind them?
• What does “four papers” actually look like?
• Why does mewing’s number overstate the evidence?
• The best-studied thing on this list is teeth whitening
• What an evidence count cannot tell you
• How to rerun this yourself
• The bottom line
• Sources

Which appearance procedures have the most research behind them?

Teeth whitening has the most RCT-tagged PubMed records, while rhinoplasty has the most total PubMed records. The table measures published volume across three databases, not quality or outcomes.

Two ClinicalTrials.gov rows need a warning label. Counts above roughly 100 for chemical peel and jaw exercise are loose matches: the term appears as one intervention among many. They are not precise counts of trials devoted only to that intervention.

What does “four papers” actually look like?

The exact PubMed search for “bone smashing” returned four records, and none were randomised trials or registered studies. Bone smashing’s literature consists of warnings; the fourth record is historical rather than a warning about the trend.

The four records were:

• 2024 warning: PMID 38757990, “Risks of 'bone smashing' trend on social media.”
• 2024 warning: PMID 38458546, “The bone smashing and the excesses of social networking trends.”
• 2024 warning: PMID 38278441, “Urgent concern regarding 'bone smashing', a dangerous trend on TikTok.”
• 1968 paper: PMID 5650960, “Bone smashing by late Miocene Hominidae.”

For a look at the phrase, see our bone smashing explainer. The point here is narrower: the forum's most aggressive technique has no medical literature.

Caveat: a narrow keyword can miss papers that use different terminology. That limitation is exactly why the query is published, so you can test the definition instead of trusting our headline.

Why does mewing’s number overstate the evidence?

The combined mewing / orthotropics query returned 58 PubMed records, but that is not the same as 58 studies of the online technique. The raw count includes broader tongue-posture literature and unrelated records that match the string.

The bare-term checks show the problem. “Mewing” returned 22 records, “orthotropics” returned 11, and “tongue posture” AND (facial OR maxillofacial) returned 58. The table uses the combined query, so its 58 is a reproducible search result, not a clean intervention-specific total.

Among the records returned by the bare term “mewing” are PMID 31005620 (2019), “Mewing: Social Media's Alternative to Orthognathic Surgery?”, PMID 41739658 (2026), “The mewing trend,” and PMID 42580741 (2026), “From Mewing to Peptides: Looksmaxxing as an Emerging Pediatric Concern.” Another returned record is PMID 42674114 (2026), “Dangerous trends on social media related to facial esthetics: Analysis of looksmaxxing and its ...”; the title is truncated.

The orthodontic literature it borrows from, tongue posture and facial growth, has 58 records in the broader search. Orthognathic surgery, the thing mewing is pitched as an alternative to, has 8,192 records and 231 randomised trials. That comparison does not settle the question, but it shows why a search count needs context; our mewing safety guide covers the separate question of risk.

The best-studied thing on this list is teeth whitening

Teeth whitening has 601 RCT-tagged PubMed records, more than any other intervention in this index.

The next RCT-tagged counts are rhinoplasty at 369 and botulinum toxin, cosmetic at 286. Teeth whitening also has 3,679 total PubMed records and 127 registered studies. Those figures overturn the forum instinct that the most extreme intervention must be the most seriously studied.

Attention is not evidence volume. Search interest rewards novelty, visible transformation, and a strong claim. Research volume is shaped by many other things, including how long a field has existed and how widely its terminology is used. Our guide to whitening teeth can help you separate the procedure name from the evidence question.

Buccal fat removal makes the opposite point. It has 727 PubMed records, only 15 RCT-tagged records, and 7 registered studies. It is a warning against treating a heavily discussed procedure as though discussion itself were a research base; our buccal fat removal overview keeps those questions separate.

What an evidence count cannot tell you

An evidence count cannot tell you whether a procedure works, whether it is safe, or whether it is right for anyone. It only tells you how many records matched a defined search at a particular time.

Read the columns as three different signals:

• PubMed records measure retrieved publication volume. They do not measure study quality, agreement, or practical relevance.
• RCT-tagged records measure records carrying the randomized controlled trial[pt] tag after the same query was applied. They do not tell you what those trials found.
• ClinicalTrials.gov studies measure registered studies matching the intervention field. Registration is not a result, and broad terms can create loose matches.

A large number can mean a mature literature or a crowded, low-quality one. A small number can mean a new idea or a dead end. This page is a map of attention, not a verdict.

Use it to ask questions: What exactly was searched? Does the term catch adjacent topics? Are the registered matches specific? What would I need before making a personal decision? This material is for orientation, not feeding appearance anxiety.

If your real question is how your current photos land in ordinary perception, you can use /test, which is free and shows how a photo reads rather than scoring you. That is a different question from whether a procedure has a large literature, and it should stay a different question.

How to rerun this yourself

You can reproduce the index by copying each query exactly, checking the same fields, and recording the date.

Method: PubMed counts via the NCBI E-utilities esearch endpoint (https://www.ncbi.nlm.nih.gov/home/develop/api/), run 7 September 2026. "RCT" column = the same query ANDed with randomized controlled trial[pt]. Registered-study counts via the ClinicalTrials.gov API v2 (https://clinicaltrials.gov/), field query.intr (intervention), same date. Counts move as records are added; treat them as a snapshot.

1. Copy the query: Use the exact PubMed string in the table, including quotation marks, field tags, parentheses, and Boolean operators.
2. Record PubMed: Save the total result and then rerun the same query with randomized controlled trial[pt] added.
3. Check registration: Search ClinicalTrials.gov API v2 using the intervention field query.intr.
4. Save the date: A later rerun may produce a different count without either search being wrong.

Do not silently clean the data. The messy mewing result and the loose ClinicalTrials.gov matches for chemical peel and jaw exercise are part of what a reader needs to know.

The bottom line

The evidence landscape is uneven. Teeth whitening has the largest RCT-tagged count here, rhinoplasty has the largest PubMed total, buccal fat removal has a much smaller trial and registration footprint than its online attention suggests, and bone smashing has four exact-term records, including one about Miocene apes.

None of those findings tells you what to do with your face. They tell you how much searchable attention exists under a stated definition, and where the online confidence is running far ahead of the published record. Before spending anything, use /test to see how a photo actually reads, then keep that perception question separate from any medical decision. If a term in the table is new to you, the cosmetic procedures glossary defines it without selling it.

Sources

• NCBI E-utilities API: https://www.ncbi.nlm.nih.gov/home/develop/api/
• ClinicalTrials.gov API v2: https://clinicaltrials.gov/
• PMID 38757990, “Risks of 'bone smashing' trend on social media.” https://pubmed.ncbi.nlm.nih.gov/38757990/
• PMID 38458546, “The bone smashing and the excesses of social networking trends.” https://pubmed.ncbi.nlm.nih.gov/38458546/
• PMID 38278441, “Urgent concern regarding 'bone smashing', a dangerous trend on TikTok.” https://pubmed.ncbi.nlm.nih.gov/38278441/
• PMID 5650960, “Bone smashing by late Miocene Hominidae.” https://pubmed.ncbi.nlm.nih.gov/5650960/
• PMID 31005620, “Mewing: Social Media's Alternative to Orthognathic Surgery?” https://pubmed.ncbi.nlm.nih.gov/31005620/
• PMID 41739658, “The mewing trend.” https://pubmed.ncbi.nlm.nih.gov/41739658/
• PMID 42580741, “From Mewing to Peptides: Looksmaxxing as an Emerging Pediatric Concern.” https://pubmed.ncbi.nlm.nih.gov/42580741/
• PMID 42674114, “Dangerous trends on social media related to facial esthetics: Analysis of looksmaxxing and its ...” (title truncated in our capture). https://pubmed.ncbi.nlm.nih.gov/42674114/