NEW SCIENCE
Can a cream change your hair colour or curl?
In short
A small study changed the colour and curl of new hair with low-dose scalp creams. All the drugs are prescription. This is not medical advice, just handy to know. Talk to a doctor if you want to explore this option further.
Do this
-
1
This is early science, not a product you can buy.
-
2
Never put a prescription drug on your scalp or make your own cream.
-
3
Know the dial: cAMP, a cell signal, turns pigment making up in follicle cells.
-
4
What the science points to: caffeine is unlikely to darken hair.
-
5
For colour or curl now, use a stylist's dye or curl treatment.
-
6
Talk to a doctor if your hair changes after you start a medicine.
The numbers behind it
-
157%
Melanin in the best volunteer's new hair on the dipyridamole patch versus the placebo patch, blond hair turning medium to dark brown (3 volunteers).
-
Up to +120%
Straightening of very curly hair with midodrine in the best volunteer. The average across volunteers who responded was 76% (4 volunteers).
-
Over 80%
Share of the topiramate group whose hair changed direction, for example the start of a curl in straight hair (6 volunteers).
-
Growth slowed
Raising cAMP, the signal behind the pigment boost, slowed human hair follicle growth in lab dishes. The pilot saw no growth change.
For your hair
Blond or light brown hair
Dipyridamole darkened new hair in two blond to dark blond volunteers. The one brown-haired volunteer showed no change.
Wavy and very curly hair
Midodrine straightened new hair in 3 of 4 volunteers. A second patch on each head got a placebo.
Straight hair
Topiramate started a curl in some straight-haired volunteers. The direction of the twist varied from person to person.
Gray or white hair
Outside the study. Pigment is made as hair grows, so at most only new hair could change.
Myths to ignore
-
Myth A cream that changes your hair colour or curl already exists.
Truth Not yet. This was one pilot of 33 adults, with 3 to 6 people per drug, 15 applications over 5 weeks on 1 cm2 patches. It is a first result that still needs repeating.
Matama 2024 ↗ -
Myth Roflumilast, a skin cream that repigments vitiligo, will darken hair.
Truth It repigmented skin in four children, and it raises melanin in cell dishes, mostly with a second signal. Its data cover skin, not hair colour.
Warren 2025 ↗ -
Myth Caffeine, adenosine or Coleus extract in a cosmetic darkens hair the same way.
Truth What the science points to: unlikely. Their hair data cover growth and thickness, not colour. Forskolin, the Coleus compound, raised pigment in follicle dishes but also slows growth.
Michelet 2009 ↗
Go deeper
- Gray hair and premature grayingGraying is mostly genetic, but single-hair mapping shows some gray hairs can regain colour, and the famous stress-causes-gray mechanism was shown in mice.
- Curl patternHow straight, wavy, curly or coily your hair is gets decided in the follicle, mostly by genes, and it shapes shine, tangling, breakage and how a routine should be built.
- Drug-induced hair changesMedicines can thin hair, regrow it curly, darken or whiten it, or lengthen lashes; this table sorts the best-known links by how strong the evidence is.
All studies in this guide (7)
Matama T et al. Changing human hair fibre colour and shape from the follicle. J Adv Res. 2024
Tested: Low-dose scalp emulsions of dipyridamole, rivastigmine or paroxetine (colour) and midodrine, ethacrynic acid, topiramate or entacapone (curl), three times a week; the other patch got a placebo emulsion
Result: A Portuguese split-scalp pilot found colour changes in 8 of 14 volunteers and curl changes in 16 of 19. Dipyridamole darkened new blond hair (best volunteer 157% of placebo melanin, medium to dark brown), midodrine straightened curly hair (average 76%, best +120%) and topiramate made straight hair start to curl (over 80% of its group). Rivastigmine and paroxetine lowered melanin slightly. Hair growth did not change and no side effects were reported. Groups were 3 to 6 people per drug.
Read the study ↗Harmon CS, Nevins TD. Evidence that activation of protein kinase A inhibits human hair follicle growth and hair fibre production in organ culture. Br J Dermatol. 1997
Tested: agents that raise cAMP: db-cAMP, the phosphodiesterase inhibitors IBMX and Ro 20-1724, and forskolin
Result: Agents that raise cAMP slowed human hair follicle growth in culture (db-cAMP: half-maximal at 100 micromol/L, 85% inhibition at 1 mmol/L). IBMX, Ro 20-1724 and forskolin also slowed growth.
Read the study ↗Michelet JF et al. Human hair follicle pigmentary unit as a direct target for modulators of melanogenesis, as studied by [14C]-2-thiouracil incorporation. Exp Dermatol. 2009
Tested: forskolin 1 micromol/L, kojic acid 200 micromol/L, alpha-MSH 0.2 micromol/L
Result: In cultured pigmented human hair follicles, forskolin (1 micromol/L) raised melanin production and tyrosinase, kojic acid lowered it, and alpha-MSH (0.2 micromol/L) did not raise it significantly.
Read the study ↗Chen Z et al. Roflumilast enhances the melanogenesis and attenuates oxidative stress-triggered damage in melanocytes. J Dermatol Sci. 2023
Tested: roflumilast (a PDE4 inhibitor), alone and with forskolin
Result: Roflumilast alone raised melanin only slightly in cultured pigment cells, while roflumilast plus forskolin raised cAMP and melanin much more. PDE4 levels were higher in active vitiligo skin.
Read the study ↗Warren K, Sanchez S. Recalcitrant pediatric facial vitiligo successfully treated with roflumilast cream 0.3% once daily. J Clin Aesthet Dermatol. 2025
Tested: roflumilast cream 0.3% once daily
Result: All four children with stubborn facial vitiligo showed skin repigmentation on roflumilast cream 0.3%. The report is about skin; it describes no change in hair colour.
Read the study ↗Fischer TW, Hipler UC, Elsner P. Effect of caffeine and testosterone on the proliferation of human hair follicles in vitro. Int J Dermatol. 2007
Tested: caffeine 0.001-0.005% with and without testosterone
Result: Testosterone suppressed hair-shaft growth; caffeine reversed this and also stimulated growth on its own.
Read the study ↗Watanabe Y et al. Topical adenosine increases thick hair ratio in Japanese men with androgenetic alopecia. Int J Cosmet Sci. 2015
Tested: adenosine lotion vs niacinamide lotion
Result: Adenosine was superior (P<0.05) for global improvement and share of thick hairs.
Read the study ↗