One rule, two questions, and the documents behind every answer
How a grade is worked out, why the worse of two answers wins, what the rule deliberately cannot measure, and the full list of documents each answer is read from.
Follicle Index grades hair-loss treatments. It is written by people with no medical training at all, from documents that are public, and no clinician has read it. That sentence is the masthead: there is no doctor on this site, no dermatologist checking the copy, no advisory board, and nothing here is ever labelled reviewed, approved or checked by anyone medical, because the label would be false.
What can be done without a medical qualification is read a regulator's notice carefully and report it without decoration, and read a meta-analysis and report how many trials it pooled, how many people were in them, and who funded them. What cannot be done is tell you what is happening to your scalp. That is a job for a GP or a board-certified dermatologist, and no amount of reading replaces the appointment.
The rule, stated once
Every treatment is scored on two questions. Each question has four answers, ranked. The grade is the worse of the two, and there is no averaging, no weighting and no tie-breaker.
Question one: standing
What has a regulator actually said about this product, for hair loss?
- Approved by the FDA for androgenetic alopecia. The highest answer, and only two entries have it.
- Approved for a different condition and prescribed off-label, or cleared as a device under the 510(k) route. Clearance means a maker showed substantial equivalence to a device already on sale; the FDA's own consumer explainer distinguishes that from the approval it gives new drugs and the highest-risk devices.
- Neither approved nor cleared for hair loss: compounded medicines, cosmetics used for an effect they are not sold for, clinic procedures.
- Outside drug approval altogether. Dietary supplements and cosmetics, which the FDA says plainly it does not approve for safety and effectiveness.
Below all four sits one further answer: the FDA has published a safety warning about the category and no approved product exists. That is tier X, and it does not average with anything.
Question two: evidence
What randomised controlled trials exist, have they been pooled, and who paid?
- Pooled in a published meta-analysis of randomised controlled trials that counted hairs, by authors independent of any single seller.
- Randomised trials exist and have been pooled, but they are small, inconsistent with each other, or largely funded by the people selling the product.
- One or two small controlled trials.
- No controlled trial in pattern hair loss, or seller-funded work that nobody independent has repeated.
Why the worse answer wins
Because the two weaknesses are different, and neither fixes the other. A drug with excellent trials and no approval for this use leaves you carrying a risk a regulator has not examined. An approved product with thin evidence leaves you carrying a different one. Averaging them would hide both. Taking the worse answer states the weaker half out loud, which is the only part that ever costs anybody anything.
The rule also means a grade can move without any new trial being run — an approval, a clearance or a safety notice changes an answer on its own.
What the rule refuses to do
It does not measure how well something works for a person. It cannot: a tier A treatment does nothing for some people, and roughly half the reasons hair falls out are not the condition this scale is about.
It does not grade price, and it does not grade brands. A grade belongs to an active ingredient or a category, never to a bottle, and no individual product is ranked against another.
It does not grade surgery or cosmetic camouflage. Both sit off the scale, labelled as such, because neither treats the condition and grading them against drug trials would answer a question nobody asked.
| Document | What it settles here | Read |
|---|---|---|
| FDA prescribing information for the approved oral treatment (DailyMed) | The approval, who it is and is not for, the reported harms, and what stopping does | 24 Sep 2026 |
| Drug-facts label for over-the-counter topical minoxidil (DailyMed) | What the approved topical is sold to do, and where it says it was not intended to be used | 24 Sep 2026 |
| FDA alert on compounded topical finasteride, 22 April 2025 | That no topical formulation is approved, and what was reported to the agency | 24 Sep 2026 |
| EMA news item on finasteride and dutasteride, May 2025 | The confirmed side effect, the patient card, and the precaution added to dutasteride | 24 Sep 2026 |
| MHRA Drug Safety Update on finasteride and dutasteride, 11 May 2026 | The UK position and when the patient card was introduced | 24 Sep 2026 |
| FDA consumer explainer on approved, cleared and unapproved products | The difference between approval and 510(k) clearance, and what is outside both | 24 Sep 2026 |
| FDA public safety notification on exosome products | That no exosome product is approved, and that harms have been reported | 24 Sep 2026 |
| Network meta-analysis of monotherapies for male androgenetic alopecia, J Cosmet Dermatol, 6 Oct 2025 | Which treatments have pooled randomised evidence, and how they ranked | 24 Sep 2026 |
| Meta-analysis of platelet-rich plasma in alopecia, Dermatol Ther, 13 Sep 2025 | Trial count, participant count and the authors' own caution | 24 Sep 2026 |
| Meta-analysis of combined microneedling therapy, J Cosmet Dermatol, May 2024 | Trial count, participant count, and that most trials tested it as an add-on | 24 Sep 2026 |
| MedlinePlus Genetics: androgenetic alopecia | Prevalence, the role of DHT, and what is and is not known about inheritance | 24 Sep 2026 |
| Classifications of Patterned Hair Loss, J Cutan Aesthet Surg, 2016 | The Hamilton-Norwood types and the A variants, as originally described | 24 Sep 2026 |
| American Academy of Dermatology: hair loss diagnosis, and shedding versus loss | What an appointment involves, and the line between a shed and a loss | 24 Sep 2026 |
Corrections
If a grade is wrong, the fastest way to change it is to name the document that makes it wrong. Every grade on this site is arithmetic on a published source, so a better source beats an argument every time, and a correction that lands gets made and dated rather than quietly edited. Send it here.
The most likely way this site goes out of date is an approval. A candidate treatment with published phase III results and a filing its maker has announced is not a tier at all until a regulator decides; on the day one does, a row changes.