TL;DR: Centella and azelaic acid get treated as interchangeable redness products. They are not. Centella works on the vascular and Th2 inflammation pathway. Azelaic targets the neutrophil and PAR2 pathway behind papulopustular rosacea. Bylka 2013 and Liu 2020 set the mechanistic split. I worked out which one to reach for depending on whether your redness blanches under glass or doesn’t, and where the published efficacy data actually lands.
A reader sent me two photos, taken three months apart, of her cheeks. In the first, the redness was diffuse, blanchable, and faded after she came inside from the cold. In the second, the redness had hardened into small pink papules around her nose and chin and stayed put regardless of temperature. She had been using a centella cream the whole time and felt like it had stopped working. She wanted to know whether to switch products or escalate to a prescription.
Her instinct was right. The centella was not failing. The redness had changed type. The first photo was vascular flushing, which is what centella is good at. The second was papulopustular rosacea, which is what azelaic acid is good at. They look similar enough on a phone screen to be confused, but the underlying biology is genuinely different, and the published efficacy data splits cleanly between the two.
I want to walk through the Bylka 2013 review (PMID: 24278045) on centella, the Liu 2020 systematic review (PMID: 32320019) on azelaic acid, and the Coda 2013 cathelicidin paper (PMID: 23871719) that explains why azelaic works on rosacea specifically. Then I want to walk through how to tell which type of redness you are dealing with, and what to use for each.
What the studies actually show
Bylka 2013 reviewed the cosmetic and dermatological literature on Centella asiatica through the late 2000s and early 2010s. The active compounds are the triterpenes: asiaticoside, madecassoside, asiatic acid, and madecassic acid. The well-documented mechanisms are upregulation of Type I collagen synthesis (which is why centella is in scar-treatment formulas), inhibition of TNF-alpha and IL-6 release from keratinocytes, and stabilisation of microvascular endothelial junctions. The last one is the key for redness. Centella reduces capillary permeability and dampens vasodilation in conditions of low-grade endothelial stress. This is why it works for vascular flushing, post-procedure erythema, and the diffuse cheek redness that comes from cold weather or low-grade barrier inflammation.
Bylka also noted that centella has measurable activity on Th2-skewed inflammation, which Park 2017 (PMID: 28358327) later confirmed in an atopic dermatitis mouse model. Park showed that titrated centella extract reduced IL-4 and IL-13 expression and improved skin lesions over 4 weeks. This is the eczema-adjacent redness response, which centella also helps with.
What Bylka does not claim, and what the centella literature does not support, is activity on the neutrophil-mediated inflammation that drives papulopustular rosacea. Rosacea is not a Th2 condition. It is a Th17-skewed inflammation with cathelicidin LL-37 dysregulation and kallikrein 5 protease overactivity. The endothelial-stabilising and Th2-modulating activity of centella does not reach this pathway.
Liu 2020 is the systematic review of azelaic acid in papulopustular rosacea. It pooled 7 randomised controlled trials with a total of 2031 patients. Azelaic acid 15 percent gel was significantly superior to vehicle in lesion count reduction (mean difference roughly 7 lesions over 12 to 15 weeks), erythema reduction (modest but significant on the IGA scale), and patient-rated improvement. The effect size was similar to topical metronidazole 1 percent in head-to-head trials. Azelaic acid had a slightly higher local tolerance reaction rate (mild burning in the first 2 weeks in about 20 percent of patients), but no serious adverse events.
Coda 2013 explained the mechanism. Azelaic acid inhibits kallikrein 5 protease activity directly, which reduces cleavage of cathelicidin into the inflammatory LL-37 fragment. LL-37 is the peptide that drives the neutrophil influx and vasodilation seen in rosacea papules. Coda’s biopsy data showed measurable reduction in LL-37 and kallikrein 5 expression in lesional skin after 12 weeks of azelaic 15 percent gel. This is a different pathway from centella’s endothelial stabilisation, and it is the one that matters for papular and pustular rosacea.
Fitton 1991 (PMID: 1712709) is the older but still-cited review of azelaic acid’s broader properties. It catalogues the antibacterial activity against Propionibacterium acnes (now Cutibacterium), the tyrosinase inhibition that gives azelaic its pigment-fading effect, and the anti-keratinising activity that helps with comedonal acne. The point relevant here is that azelaic is a multi-mechanism active. It is good for rosacea papules, comedonal acne, and post-inflammatory hyperpigmentation. It is not particularly useful for diffuse vascular flushing without papular involvement.
How to tell which redness you have
The cleanest bedside test is the blanching test. Press a clear glass slide or your phone screen against the red area and look at what happens. Diffuse vascular flushing blanches under pressure and refills slowly when you release. The blood is in the superficial capillary plexus and can be pushed out mechanically. Papulopustular redness does not blanch, or only partially blanches, because the redness is from inflamed perifollicular tissue rather than dilated capillaries. The papules feel slightly raised and have a centre that is firmer than the surrounding skin.
A second test is temperature reactivity. Vascular redness intensifies with heat, alcohol, spicy food, exercise, and emotional stress, and fades when you remove the trigger. Papulopustular redness is relatively stable across triggers. It worsens slowly over weeks and improves slowly over weeks.
A third test is texture. Vascular redness sits in skin that otherwise feels smooth. Papulopustular redness has a sandpapery or bumpy quality you can feel with the back of a clean finger.
Most people have some of both. Pure vascular redness is more common in early rosacea, post-procedure recovery, and cold-climate sensitivity. Pure papulopustular redness is more common in established rosacea and in adult-onset acne with rosacea overlap. The two coexist often, and the treatment plan usually addresses both, but the order matters.
What I use for which
For pure vascular flushing, centella is the published-evidence-backed first line. The Skin1004 Madagascar Centella Ampoule, the COSRX Centella Blemish Ampule, and the Purito Centella Green Level Calming Toner are all reasonable. Twice daily for 8 to 12 weeks. The Cicabio cream from Bioderma and the La Roche-Posay Cicaplast Baume B5 are similar but lower centella content with more emollient. I would not pay above 25 dollars for a centella product. The active concentration ceiling is reached at modest cost.
For papulopustular redness, azelaic acid is the published-evidence-backed first line. The prescription Finacea 15 percent gel is the original. The over-the-counter Paula’s Choice 10% Azelaic Acid Booster is the cleanest non-prescription option in most markets. The Geek & Gorgeous Cocoa Bun (a 10 percent azelaic) is another reasonable choice. Apply twice daily for 12 to 15 weeks. The Liu 2020 review found minimum 12 weeks for meaningful papule reduction. Most people quit at week 6 and miss the result.
For mixed redness, I use both, layered. Centella in the morning under sunscreen, azelaic in the evening. This is also the approach I covered in the azelaic acid use case finder.
The contrarian aside
The marketing copy for both ingredients is overheated. Centella is sold as a universal calming agent that resolves all inflammation. It does not. It does not meaningfully reach the neutrophil-driven inflammation behind papulopustular rosacea, and it is moderately useful at best for adult inflammatory acne. Azelaic is sold as a gentler alternative to retinoids and benzoyl peroxide for all skin types. It is gentler in some ways (it does not photosensitise), but it is more irritating than centella in the first 2 weeks, and it does not address vascular flushing without papular component.
The other contrarian point is that prescription options exist for both. Tranexamic acid topical 5 percent is a more direct vascular-stabilising option than centella, with better efficacy data on telangiectasia and post-inflammatory erythema. Brimonidine 0.33 percent gel (Mirvaso) reduces vascular redness within hours but rebounds when discontinued. Ivermectin 1 percent cream (Soolantra) outperforms azelaic on papulopustular rosacea in some trials, mostly through Demodex mite reduction. The over-the-counter centella and azelaic options are not the ceiling. They are the floor.
What I would tell my past self
Spend 5 minutes with a glass slide and a mirror before you buy another redness product. Vascular flushing and papulopustular rosacea are different conditions with different evidence-based treatments, and using the wrong one for 6 months is a common pattern. The product that “stopped working” probably never matched the condition.
The other thing I would tell my past self is that 12 weeks is the minimum trial window for either centella or azelaic. The improvement curve is linear and shallow. There is no week 2 transformation. If you swap products at week 4, you reset the clock and never see the result. I covered the timing in the face redness reset tool.
FAQ
Can I use centella and azelaic acid together?
Yes. They work on different pathways and do not interfere chemically. The standard layering is centella in the morning under sunscreen and azelaic in the evening. Mixed redness usually benefits from both.
How long until I see results from azelaic acid?
The Liu 2020 review found significant lesion reduction at 12 to 15 weeks. The first measurable improvement is usually around week 6 to 8. Earlier than that is mostly the initial irritation settling.
Is centella enough for rosacea?
For erythematotelangiectatic rosacea (the flushing-and-visible-vessels subtype) with no papules, centella is reasonable as a first line alongside vascular triggers management. For papulopustular rosacea (with papules and pustules), centella alone is insufficient. Azelaic, ivermectin, or a topical prescription is the evidence-based intervention.
Does azelaic acid help vascular redness too?
Modestly. The Liu review found some erythema improvement alongside the papule reduction, but the effect size on diffuse vascular redness alone is smaller than the effect on papules. For pure vascular flushing, centella, tranexamic acid topical, or brimonidine are better matched to the mechanism.
Can azelaic acid cause irritation?
Yes, mild burning and tingling in the first 2 weeks in about 20 percent of users. It usually settles. If it does not settle by week 3 or worsens beyond week 1, the formulation may be too concentrated for your tolerance and a 10 percent version may work better than the 15 percent.
Related Elelaf tools
Sources
- Bylka W, Znajdek-Awizen P, Studzinska-Sroka E, Brzezinska M. Centella asiatica in cosmetology. Postepy Dermatol Alergol. 2013;30(1):46-49. PMID: 24278045
- Liu RH, Smith MK, Basta SA, Farmer ER. Azelaic acid in the treatment of papulopustular rosacea: a systematic review of randomized controlled trials. JAMA Dermatol. 2020;156(6):651-660. PMID: 32320019
- Park JH, Choi JY, Son DJ, et al. Anti-inflammatory effect of titrated extract of Centella asiatica in phthalic anhydride-induced atopic dermatitis. Int J Mol Sci. 2017;18(4):738. PMID: 28358327
- Fitton A, Goa KL. Azelaic acid. A review of its pharmacological properties and therapeutic efficacy in acne and hyperpigmentary skin disorders. Drugs. 1991;41(5):780-798. PMID: 1712709
- Coda AB, Hata T, Miller J, et al. Cathelicidin, kallikrein 5, and serine protease activity is inhibited during treatment of rosacea with azelaic acid 15% gel. J Am Acad Dermatol. 2013;69(4):570-577. PMID: 23871719