Natural Products Show Promise Against Atopic Dermatitis, But Clinical Proof Remains Thin

Atopic dermatitis, the itchy, relapsing inflammatory skin condition that affects between 1 and 20 percent of the world’s population, remains one of dermatology’s most stubborn therapeutic challenges. A wide-ranging review published in Chinese Medicine surveys decades of laboratory and animal research on natural products—from flavonoids and saponins to essential oils and polysaccharides—and maps the mechanisms by which these compounds appear to counter the disease.

The review frames the problem by first cataloguing the limitations of existing therapy. Mild disease is typically managed with topical corticosteroids or calcineurin inhibitors, moderate cases with continued topical treatment, and severe disease with systemic immunosuppressants, phototherapy or biologics. Each option carries drawbacks. Long-term topical steroid use is linked to skin atrophy and hypothalamic-pituitary-axis changes; antihistamines lack anti-inflammatory action; phosphodiesterase-4 inhibitors cause burning and stinging at the application site. Newer biologics such as dupilumab, tralokinumab, lebrikizumab and nemolizumab, along with oral JAK inhibitors like upadacitinib and abrocitinib, have transformed severe disease management, but they bring injection-site reactions, conjunctivitis, infections and—for JAK inhibitors—a boxed warning plus documented risks including thrombosis and gastrointestinal perforation. The authors argue this safety gap justifies the search for adjunctive agents with gentler profiles.

The mechanistic core of the review organizes natural products around the pathogenesis of atopic dermatitis itself. The disease follows what immunologists describe as Th2-dominated immune dysregulation: CD4+ T cells polarize toward a Th2 phenotype in the acute phase, flooding the skin with interleukin-4, interleukin-13, interleukin-31 and other cytokines that degrade barrier proteins such as filaggrin, loricrin and involucrin, activate IgE-producing B cells, and sensitize sensory neurons to itch. Chronic disease adds Th1 and Th17 activation. Because nuclear factor kappa B, MAPK and JAK-STAT signaling sit upstream of much of this cascade, compounds that quiet these pathways can, in principle, intervene at several points simultaneously.

The single-compound evidence is the most mechanistically detailed. Magnolol from Magnolia officinalis, evodiamine from Evodia rutaecarpa, glycyrrhizic acid from licorice, luteolin, baicalin from Scutellaria baicalensis, mangiferin and notoginsenoside R1 have each been shown in mouse models of DNCB- or MC903-induced dermatitis to shift the Th1/Th2/Th17 balance, suppress pro-inflammatory cytokines and, in several cases, restore filaggrin and loricrin expression in lesional skin. Baicalin, for example, reduced dermatitis scores, ear thickness and serum IgE in DNCB-treated BALB/c mice while modulating NF-κB and JAK/STAT signaling and even reshaping gut microbial composition. Glycyrrhizic acid restored barrier proteins in MC903-induced C57BL/6 mice. Prim-O-glucosylcimifugin from Saposhnikovia divaricata selectively inhibited Th2 cell differentiation without affecting Th1, Th17 or regulatory T cells, an unusual specificity the authors highlight as therapeutically attractive.

Beyond single molecules, the review covers bioactive fractions and crude extracts. Essential oils of Artemisia annua, rosemary, lavender and Mentha arvensis suppressed inflammatory cytokines in AD mouse models and in LPS-stimulated macrophage and keratinocyte assays. Total flavonoids of Hippophae rhamnoides, or sea buckthorn, reduced inflammatory markers and upregulated filaggrin, and the authors note that in direct comparison the flavonoid emulsion preserved keratinocyte morphology better than dexamethasone, which caused marked cellular changes. Extracts of Terminalia chebula, Isatis tinctoria, Grifola frondosa, red ginseng, Pyrus ussuriensis leaves and Curcuma longa leaves and pseudostems all showed anti-AD activity in vivo, typically reducing dermatitis scores, inflammatory infiltration and cytokine mRNA in spleen and lymph node tissue.

Barrier repair emerges as a second major mechanism. The stratum corneum’s permeability function depends on ceramides, cholesterol and free fatty acids, plus filaggrin-derived natural moisturizing factors; barrier defects raise transepidermal water loss and lower the threshold for inflammation. Quercetin and isorhamnetin in Eruca sativa extract activate PPAR-α ligand binding, promoting barrier-related protein restoration in cell experiments. Valencene from Cyperus rotundus, diosmetin from Lobelia chinensis, barley saponarin and Dictamni Cortex extract each upregulated barrier proteins or protease inhibitors in preclinical systems. Notably, Tribulus terrestris and Spirodelae Herba extracts were identified as the first botanical agents modulating TRPV3, an ion channel implicated in epidermal homeostasis and cornification—a mechanism the review presents as a genuinely novel approach to barrier dysfunction.

Itch, the most burdensome symptom for patients, is addressed through a third mechanistic category. Osthole suppressed IL-4 and IL-13 secretion in skin models, upregulated TRPM8 and showed stronger TRPV1 inhibition than clobetasol propionate in vitro, while in vivo work linked its effects to PPARα/γ-mediated suppression of IL-31 receptor expression. Curcumin supplementation outperformed betamethasone cream and pimecrolimus in suppressing substance P in a human study of sulphur mustard-induced chronic pruritus—one of the few human data points in the review. Rosmarinic acid bound the MRGPRX2 receptor with high affinity, cimifugin competitively blocked the chloroquine receptor MrgprA3 to relieve non-histaminergic itch, and Artemisia argyi volatile oil inhibited TRPA1 and reduced CGRP release. The authors caution, however, that TRPA1 antagonism may hinder barrier repair, since the channel also contributes to barrier restoration, requiring a careful balance in future drug design.

Ancillary mechanisms round out the picture. Rosemary extracts containing carnosic acid suppressed Staphylococcus aureus quorum sensing and virulence, addressing the bacterial over-colonization characteristic of AD skin; α-mangostin disrupted bacterial membranes; berberine inhibited S. aureus growth and toxin production. On oxidative stress, curcumin, α-mangostin and sinomenine scavenged reactive oxygen species and boosted glutathione and superoxide dismutase. Perhaps most intriguingly, the gut-skin axis features prominently: ginsenoside F2 reshaped gut microbiota in AD mice, enriching short-chain fatty acid producers and increasing propionic acid, which suppressed inflammation through the GPR43/NF-κB pathway; baicalin restored beneficial gut bacterial abundance; and Senecio scandens polysaccharides ameliorated intestinal dysbiosis correlated with reduced AD severity. The review also discusses metabolic transformation—deglycosylation, methylation and hydroxylation—as a determinant of efficacy, noting that metabolites such as orobol and piceatannol can be more potent than their parent compounds.

The translational caveats are unusually candid for a review of this scope. The authors stress that natural origin does not guarantee safety: dose-dependent cytotoxicity, hepatotoxicity from certain components, skin irritation from topical preparations, potential interactions with corticosteroids, JAK inhibitors and antihistamines via cytochrome P450 modulation, and contamination with heavy metals or pesticides all demand systematic evaluation. More fundamentally, nearly all the efficacy data come from induced mouse models and two-dimensional cell cultures that imperfectly recapitulate human AD. Hapten-based models initially drive Th1/Th17 contact hypersensitivity rather than the Th2-dominated chronic mixed inflammation of human disease; inbred mouse strains carry immune biases—C57BL/6 is Th1-leaning, BALB/c Th2-leaning—that cannot reproduce human immune heterogeneity; and most preclinical studies use prophylactic rather than therapeutic dosing, which the authors say overestimates anti-inflammatory effects. Human-validated biomarkers such as TARC/CCL17, IL-13, IL-31 and total serum IgE are rarely measured across these models.

Despite these limitations, the review identifies baicalin and glycyrrhizic acid as the most clinically actionable candidates on current evidence, citing favorable safety profiles, barrier-protein restoration, anti-inflammatory activity and itch suppression across multiple models, with fermented soybean isoflavones suggested as possible adjunctive oral therapies. The authors call for cross-validation across multiple animal models, systematic measurement of human biomarkers, better isolation techniques for low-abundance compounds, rigorous toxicology and integration of network pharmacology and metabolomics to map structure-activity relationships. Only through such rigorous evaluation, they conclude, can traditional Chinese medicine-derived compounds become genuine lead compounds or adjunctive options for atopic dermatitis—rather than remaining a large but largely preclinical evidence base awaiting clinical translation.

Subject of Research: A review of the mechanisms and translational potential of natural products for treating atopic dermatitis

Article Title: The potential effects, mechanisms and translational perspectives of natural products in the treatment of atopic dermatitis

Article References: Shao, C., Zhang, Y., Bai, Y., Wu, Y., Cong, L., Zhao, M., & Ren, T. (2026). The potential effects, mechanisms and translational perspectives of natural products in the treatment of atopic dermatitis. Chinese Medicine, 21(1), Article 219. https://doi.org/10.1186/s13020-026-01497-8

Image Credits: AI Generated

DOI: 10.1186/s13020-026-01497-8

Keywords: atopic dermatitis, natural products, traditional Chinese medicine, baicalin, skin barrier, Th2 inflammation, gut-skin axis, pruritus, JAK-STAT signaling, NF-kB, preclinical models, drug safety