The
ledger
Notes from the bench. Mostly about dose, method, and the specific numbers this category has decided not to print on the outside of the box.
Latest
Proven in whom
Across seventeen registration programmes now in the register, three publish a mean or average age above thirty — and two of them are for diseases of midlife. Almost everything a forty-year-old is prescribed for their skin was measured in teenagers.
Archive
- L-092026-08-02
The vehicle is a dose
One molecule, four approved vehicles, and irritation rates that differ by more than the concentrations do. A percentage without a base is not a specification.
FORMULATIONRETINOIDSREGISTER7 MIN - L-082026-08-02
Ask what the vehicle arm did
In sixteen controlled trials now in this register, between 3% and 58% of people improved on the base cream or the placebo tablet alone. Not one compounded product sold direct to consumers has ever published that number, because not one has ever measured it.
EVIDENCECONTROLSREGISTER7 MIN - L-072026-07-28
Where the warning lives
Across thirty-three reviews the safety copy is almost always written, and almost never on the page taking the money. Placement is now the finding, not absence.
DISCLOSUREPLACEMENTREGISTER6 MIN - L-062026-07-27
A course is not a subscription
Hydroquinone, topical steroids and oral antibiotics all require a stop. A subscription's default behaviour is that it continues. One company in the register has built the stop into the product.
DURATIONMODELREGISTER7 MIN - L-052026-07-26
Compounded is not approved
Most of the prescription skincare sold online is compounded, which means it has never been through FDA review. Four companies in this register give four different answers to that fact, and one of them gives the wrong one.
REGULATIONDISCLOSUREREGISTER6 MIN - L-042026-07-02
The number nobody publishes
Every sunscreen on earth is tested at 2 mg/cm². Almost nobody applies 2 mg/cm². The gap is the whole product.
SPFDOSEMETHOD6 MIN - L-032026-05-19
Against the actives arms race
Ten percent niacinamide is not twice as good as five. The dose–response curve flattens and the complaint rate does not.
DOSE-RESPONSEFORMULATION5 MIN - L-022026-03-08
Retinisation is not a result
The flaking at week three tells you a retinoid is present. It tells you nothing about whether it is working.
RETINOIDTOLERANCE4 MIN - L-012026-01-14
What thirty actually changes
Not a cliff. A set of slopes that were already running, becoming visible at the same time.
AGEINGMODEL7 MIN