Cica decoded: madecassoside vs asiaticoside, what trials show

The first time I really read a centella asiatica ingredient list, the marketing made no sense. Two products from the same brand both said “powered by cica” on the front. One listed centella asiatica extract at the eleventh slot. The other listed madecassoside at the third. They were sold side by side, at the same price, with the same green leaf graphic. Only one of them actually had a chance of doing anything.

This is the cica problem. The plant has been one of the most heavily marketed botanicals in skincare since roughly 2018, and almost nothing on the front of the bottle tells you which of its four active compounds is actually present, in what concentration, or whether there’s any evidence behind the form being used.

The four active triterpenoids

Centella asiatica is not a single ingredient. The aerial parts of the plant contain four major triterpenoid actives, each with a different molecular shape and a different evidence base:

Asiaticoside. A glycoside (a sugar molecule is attached to the active core). Historically the most-studied compound in centella, with wound-healing data dating back to French clinical use in the 1950s. According to a 2021 review in the journal Antioxidants, asiaticoside is the form most active in collagen synthesis and angiogenesis (the regrowth of small blood vessels during healing).

Madecassoside. The other major glycoside, and the form Chemist Confessions called “the real hero” of modern cica skincare in a 2023 deep-dive. A 2025 clinical study on 100 adults with sensitive skin, cited by Personal Care Insights, found that a madecassoside-led cream measurably reduced visible redness, improved pH regulation, and rebalanced sebum over four weeks.

Asiatic acid. The non-glycoside form (the sugar is removed). Penetrates the skin more readily than asiaticoside because it is smaller, but is also more irritating in higher concentrations.

Madecassic acid. The non-glycoside form of madecassoside. The compound most often credited in lab studies with anti-inflammatory action on the NF-kB pathway, which is the same pathway that ibuprofen interrupts.

The takeaway is unspectacular: the four are related, they overlap in function, but they are not interchangeable. A product listing “centella asiatica extract” might contain all four in unknown ratios. A product listing “madecassoside 0.1 per cent” tells you exactly what you are getting.

What the recent clinical data actually shows

A 2023 split-face trial on patients with mild-to-moderate atopic dermatitis compared a 0.1 per cent madecassoside serum against a 2 per cent whole-plant centella extract. Both groups improved on the primary endpoint (clinician-rated redness at four weeks). The interesting wrinkle was the eight-week follow-up: the centella group had a 32 per cent lower relapse rate. The authors’ read was that the synergy of the four compounds together provides longer-lasting barrier remodelling, while isolated madecassoside acts faster but doesn’t hold as well.

The 2025 trial I mentioned above used a finished cream rather than a serum, and the active was madecassoside (no breakdown of the rest of the formula was published). Over four weeks, participants showed reduced transepidermal water loss, improved skin smoothness on Visia imaging, and the pH and sebum rebalancing that gets quoted in marketing copy. The trial was sponsored by a brand, which is worth noting, but the methodology was clean and the effect sizes were modest enough to be plausible.

A 2022 PMC review summarised the broader mechanism: these compounds act on NF-kB, TGF-beta/Smad, MAPK, Wnt/beta-catenin, and STAT signalling. That is a long list of pathways, which usually means a small effect spread across many systems rather than a strong effect on one. Which lines up with what cica feels like in practice. It is not dramatic. It is quietly correct.

Reading a cica ingredient list

A few specifics that separate the formulas worth using from the ones coasting on the leaf graphic:

Position matters more than wording. “Centella asiatica leaf extract” listed within the first five ingredients (after water, glycerin, and the basic emollients) means the active is functional. Listed below ingredient nine or ten, it is window dressing.

Look for a percentage. Brands that have done the trial work usually advertise the percentage on the bottle, often 0.1 per cent madecassoside or 5 per cent centella extract. Dr Jart+‘s Cicapair line lists 0.05 per cent madecassoside, which is on the low end but consistent with European cosmetic norms. La Roche-Posay’s Cicaplast Baume B5 lists 5 per cent panthenol with madecassoside as a secondary active, and the panthenol is doing most of the visible work.

Combinations beat singles. A formula with both madecassoside and asiaticoside, plus panthenol or beta-glucan, will outperform a single-active product at the same total concentration. The 2023 split-face data is the strongest evidence for this.

Beware “centella complex”. This is a marketing phrase, not a regulated term. It can mean a clean blend of all four triterpenoids or it can mean a single low-grade extract. Ask the brand for a breakdown; if they cannot provide one, treat the claim as unverified.

A useful sense check for the products on my own shelf: the COSRX Centella Blemish Cream is unambiguously cica-led, listing the extract second after water. Skin1004’s Madagascar Centella Ampoule is a 100 per cent extract product (the gimmick is real, the texture is sticky). Beauty of Joseon’s Calming Serum pairs centella with green tea and aloe and is more of an everyday hydrator than a treatment. Three formulas, three different reasons to reach for the bottle.

What cica is good for, and what it is not

Cica is the right ingredient for: visible redness, post-acne calming, supporting the barrier through a retinol introduction, soothing skin after a chemical exfoliant or laser, and general winter recovery. Pair it with a no-makeup-makeup base and you have a face that looks reset rather than treated. The same logic applies to the glass skin and dolphin skin finishes, where the priority is a clear, low-inflammation surface rather than added pigment.

Cica is the wrong ingredient for: pigmentation (use tranexamic acid or alpha arbutin), wrinkles (use a retinoid), texture (use a hydroxy acid). It also will not, despite the marketing, function as a sunscreen alternative or replace your moisturiser. It is an adjunct, not a system.

The bigger structural point: the next time a formula promises cica on the front, flip it over. The molecule that matters is the one listed in the top five, with a number next to it. Everything else is the leaf graphic.