3,954 vs 0papers tagged as trials, a depth count and not effect or quality: resistance training against BPC-157
Evidence · The ladder
Which self-improvement topics have real trials behind them, and which run on hype
Resistance training has 3,954 randomized-trial papers, creatine 732, sunscreen 321, and BPC-157 none, while BPC-157 draws 308K Wikipedia pageviews a year. Counted 30 October 2026 from public registries; these count papers tagged as trials, not effect or quality.
On 30 October 2026, PubMed held 3,954 randomized-trial-typed papers on resistance training, 732 on creatine supplementation and 321 on sunscreen, against none on BPC-157, a compound English Wikipedia readers opened about 308K times in the year to 30 September 2026. These figures count papers tagged as trials, not effect or quality.
Counts from published PubMed queries by publication type, and Wikimedia pageview data, read 30 October 2026.
The short version
3,954resistance trainingtrial-typed papers on PubMed
732creatinetrial-typed papers on PubMed
321sunscreentrial-typed papers on PubMed
0BPC-157trial-typed papers, next to 308K yearly wiki views
The ladder: claims against the recordwhat gets said, and what the record shows, read 30 October 2026
They sayBPC-157 is as well studied as creatine.
Thin evidence
Creatine has 732 papers typed as randomized trials on PubMed. BPC-157 has 0. See the receipt
They sayLots of trials means it works.
Depth, not effect
A trial count is depth of record. It does not measure the size of an effect, the quality of a study or anyone's safety. See the receipt
They sayEveryone is talking about it, so it's proven.
Pure copium
BPC-157 draws about 308K Wikipedia pageviews a year with no trial-typed papers behind it. Attention is not evidence. See the receipt
What is the evidence ladder?
A way to put very different self-improvement topics on one scale: how many randomized controlled trials the public record holds for each. The scale is deliberately blunt.
Read the reasoning
It does not say any topic works; it says how much careful human testing exists to argue about. The top rungs are well tested and unglamorous, and the bottom rung is where the most discussed arguments are.
The counts come from PubMed publication-type tags (Randomized Controlled Trial), using the query linked on each topic below. The exact queries are printed in full in the method section.
PubMed papers whose publication type is "Randomized Controlled Trial", from a published query per topic, read 30 October 2026. The lime bar is the one with none. The count is of papers tagged as trials, not of effect or quality: a topic can have many trials and a small benefit, or the reverse.
Reviews tell the same story. Resistance training has 1,852 systematic reviews or meta-analyses, creatine 173, sunscreen 98, and BPC-157 1.
Papers typed as a systematic review or meta-analysis for the same queries. Reviews pool trials, so a topic needs trials before it can have them: BPC-157 shows 1.
Does attention follow the evidence?
Not closely. Creatine draws the most Wikipedia attention and also has a deep trial record, which is how it should work.
BPC-157 draws about 308K pageviews a year, more than minoxidil or sunscreen, with no trials typed as such behind it, and melanotan II draws 112K against a handful.
Yearly English Wikipedia pageviews (user agent only), 1 October 2025 to 30 September 2026. A curiosity proxy, not search volume. Compare the lime bars with their rows in the first figure.
The pageview data comes from the Wikimedia pageviews tool, which is secondary, derived data: a rough measure of curiosity, not of use, and not search volume. That gap between attention and evidence is the reason this site exists.
How to read one rungwhat a count tells you, and what it leaves out
This page measures only the first box. A deep record is a reason to look closer, not a reason to act; the other three boxes are where a clinician earns their fee.
How to use a trial count without fooling yourself
1. A big number means a deep record, not a big benefit.
2. A zero means "no trials of this type found", which is different from "tested and failed".
3. Check what the trials measured. Hair counts, skin scores and strength tests are different outcomes.
4. Prefer a systematic review to a single trial.
5. Prescription items on this ladder (tretinoin) are listed for education; a clinician decides, not a page.
Where does BPC-157 sit, and what is its status?
At the bottom, with 226 papers and no trials typed as such, though ClinicalTrials.gov lists 4 interventional studies matching the name, which we have not assessed. Its regulatory status is moving and has its own page: see BPC-157: status, trials and harms, and for how to read FDA's category labels, how FDA's peptide categories work. Sunscreen has its own regulator page at FDA.
How should you use the ladder?
If you are reading about the best-recorded topicsDepth of record: the top rungsTraining, sleep and sunscreen have the deepest trial records here. They are also the least glamorous topics on the page. Whether any suits you is a question for a clinician.
If you are reading about a bottom-rung topicDepth of record: a status page existsA zero here is not proof of failure, but it means the claims rest on animal studies and case reports. The regulator's date is on the status page. Whether any suits you is a question for a clinician.
If you are reading about skin or hair careDepth of record: the middle rungsTretinoin and minoxidil have real trial records, and one of them is a prescription item. A count cannot tell you what fits your skin. Whether any suits you is a question for a clinician.
How the ladder was counted
Method
Each count is the number of PubMed records matching a published title and abstract query plus a publication-type filter ("Randomized Controlled Trial", or "Systematic Review" or "Meta-Analysis"), read 30 October 2026. Queries are broad and will include some off-topic papers.
Publication-type tags can be missing or loose, so treat the counts as depth of record, not exact tallies, and as papers tagged as trials, not effect or quality. Pageviews are English Wikipedia article views summed from 1 October 2025 to 30 September 2026. Nothing here measures effect size, study quality or safety, and nothing is medical advice.
The exact queries
Resistance training: resistance training[tiab]
Creatine supplementation: creatine[tiab] AND supplementation[tiab]
Sleep extension or hygiene: (sleep extension[tiab] OR sleep hygiene[tiab])
Sunscreen use: (sunscreen[tiab] OR sunscreens[tiab])
Topical tretinoin: tretinoin[tiab] AND (skin[tiab] OR acne[tiab] OR photoaging[tiab])
Minoxidil for hair loss: minoxidil[tiab] AND (alopecia[tiab] OR hair[tiab])
Microneedling: microneedling[tiab]
Topical niacinamide: niacinamide[tiab] AND (topical[tiab] OR skin[tiab])
BPC-157: "BPC-157"[tiab] OR "BPC 157"[tiab] OR "Body Protection Compound-157"[tiab]
FAQ
What is the evidence ladder?
A comparison of how many randomized controlled trials PubMed holds for each self-improvement topic, counted 30 October 2026. It measures depth of the trial record, not whether anything works.
Which self-improvement topic has the most randomized trials?
Resistance training, with 3,954 trial-typed papers on PubMed on 30 October 2026, then creatine supplementation with 732.
How many randomized trials are there on BPC-157?
None are typed as randomized or clinical trials on PubMed as of 30 October 2026, out of 226 papers matching the name. ClinicalTrials.gov lists four interventional studies matching it, which we have not assessed.
Does a high trial count mean something works?
No. It means the record is deep. Effect size, study quality and your own situation are separate questions, and a clinician is the right person for them.
Why use Wikipedia pageviews?
They are open data that anyone can re-run, which makes them a rough proxy for curiosity. They are secondary, derived data and a rough measure of curiosity, not of use. They are not search volume and we do not treat them as such.
Bottom line
The best-tested things are the least glamorous: training, sleep, sunscreen, creatine. The most discussed thing here, BPC-157, has no trials typed as such behind it. Use the count to decide where to look first, then ask a clinician about you.