Atopic dermatitis (AD) is a chronic, relapsing inflammatory skin disorder with substantial global burden, affecting approximately 11% of children and adolescents and 6–10% of adults depending on the population studied.^1,2 Its clinical and biological heterogeneity, including distinct phenotypes and immunologic endotypes characterized by differences in inflammatory signaling, barrier dysfunction, and microbiome composition, likely influences response to therapy.³

Increasing interest has turned toward the gut–skin axis as a contributor to AD pathobiology and a potential therapeutic target. Individuals with AD exhibit a high prevalence of intestinal dysbiosis, characterized by reduced microbial diversity, depletion of short-chain fatty acid (SCFA)–producing organisms, and altered intestinal permeability.^4,5 These alterations may amplify systemic inflammation and promote Th2-skewed immune responses central to AD pathogenesis.^6,7 Growing evidence also highlights dietary patterns as a modifiable factor influencing the gut–skin axis. High-fiber diets are associated with favorable gut microbiome profiles, including increased SCFA production, which may indirectly improve AD severity.⁸ Polyphenol-rich, whole-food diets may beneficially influence gut microbial composition and metabolite production, with growing emphasis on fiber variety rather than quantity alone. Restrictive elimination diets should be reserved for clearly defined allergic indications, whereas balanced, nutrient-dense dietary patterns represent a low-risk, microbiome-informed adjunct to standard AD management.

Biological plausibility for microbiome-targeted interventions derives from multiple converging pathways. Dysbiosis in AD frequently includes loss of key SCFA-producing species such as Faecalibacterium prausnitzii, with downstream reductions in butyrate and propionate.⁹ SCFAs promote regulatory T cell (Treg) differentiation, restore Th2/Treg balance, enhance epithelial barrier function, and modulate systemic inflammatory signaling, together influencing cutaneous immune tone.¹⁰ Across animal and human studies, select probiotic strains have been shown to attenuate Th2 cytokine activity, increase Treg populations, and improve skin barrier metrics, offering support for clinical application.^11,12 Probiotic effects are highly strain- and dose-specific, with commonly studied strains in AD including those within the Lactobacillus rhamnosus GG, Bifidobacterium longum, and Lactobacillus paracasei species.^13,14 Parabiotics (also referred to as non-viable or inactivated probiotics), which include heat-killed organisms or lysate-based preparations, may offer practical advantages such as improved product stability, more consistent dosing, and potentially lower infectious risk in immunocompromised populations.¹⁵ Although sometimes grouped under the broader category of “postbiotics,” this review primarily focuses on parabiotic interventions rather than metabolite-only postbiotic products, and available clinical data in AD remain limited. Prebiotics, including inulin, fructo-oligosaccharides (FOS), galacto-oligosaccharides (GOS), and dietary polyphenols, function as substrates for beneficial gut bacteria, increasing SCFA production and microbial diversity.¹⁶ Synbiotics, combining probiotics with complementary prebiotics, may confer synergistic benefits through improved colonization and enhanced immunologic effects.¹⁷

Given this foundation, interest in diet- and microbiome-directed therapies for AD has accelerated, particularly in the context of adjunctive management of established disease. This manuscript synthesizes current evidence on oral probiotics, prebiotics, synbiotics, and parabiotics, as well as dietary interventions, with an emphasis on clinical efficacy, mechanistic underpinnings, safety, and practical applications in the treatment of both pediatric and adult AD.

We performed a comprehensive literature search of PubMed/MEDLINE, Embase, Cochrane Library, and ClinicalTrials.gov from inception to October 2025, supplemented by manual review of references from relevant publications. Eligible studies included randomized controlled trials (RCTs), randomized crossovers, and prospective cohort studies assessing oral probiotics, prebiotics, synbiotics, parabiotics, or dietary fiber interventions for established AD in pediatric and adult populations. Studies focusing on primary prevention or early-life interventions were excluded. Outcomes of interest included validated severity measures such as SCORing Atopic Dermatitis (SCORAD), Eczema Area and Severity Index (EASI), and Patient-Oriented Eczema Measure (POEM), quality-of-life indices, topical corticosteroid usage, and adverse events.

In pediatric populations, multiple RCTs (n ≈ 20–25, sample sizes 30–200) have evaluated oral probiotics for established AD.^1–3 Meta-analyses demonstrate modest reductions in SCORAD, with standardized mean differences (SMD) ranging from 0.3–0.5.⁴ Effects appear more pronounced in children than adults, with several pediatric trials, particularly those examining Lactobacillus rhamnosus GG (LGG) and B. lactis, showing clearer clinical and immunologic responses. In a double-blinded RCT of infants with AD, LGG supplementation (3.4×10⁹ colony-forming units (CFU)/day for 3 months) led to significant reductions in IgA- and IgM-secreting cells (baseline-adjusted ratios 0.59 and 0.53) and increased circulating CD19⁺CD27⁺ memory B cells, reflecting strain-specific effects on mucosal immunity.⁵ Intervention duration in one systematic review of 37 RCTs ranged from 4 to 16 weeks, with wide variability in strain selection, dosing, and concurrent topical therapy, limiting comparability across trials; this pattern is consistent with meta-analysis of the same 37 RCTs (n ≈ 6,100), which found an overall relative risk (RR) of 0.83 for probiotic benefit but substantial heterogeneity (I² = 65%) (Table 1).⁶

346588 Table 1.

A 12-week double-blind RCT in 70 children with AD evaluated a multi-species probiotic formulation, Bifidobacterium lactis, Bifidobacterium longum, and Lactobacillus casei (total dose 1x10⁹ CFU/day).⁷ At week 12, the probiotic group demonstrated a mean SCORAD reduction of −5.43 compared with placebo, with 90.5% achieving Investigator Global Assessment (IGA) improvement versus 56.7% of controls, and a statistically significant reduction in topical corticosteroid use emerging between weeks 6 and 12.⁷ These findings suggest that multi-strain probiotic formulations may provide clinical benefit in AD, potentially reflecting complementary or additive biological effects; however, head-to-head comparisons with single-strain approaches are limited, and superiority of multi-strain formulations has not been established.⁸ Adult trials are smaller and more variable, typically enrolling 20–100 participants, with SCORAD reductions ranging from 4–10 points; a meta-analysis of nine RCTs reported a pooled SCORAD change of −5.93 (95% CI −8.43 to −3.43), with the largest effects observed for L. salivarius (−9.79), L. acidophilus (−5.77), and L. plantarum (−3.76), and no meaningful benefit for L. fermentum.⁹ The collective evidence indicates small-to-moderate, strain-dependent benefits, with the strongest signal observed in pediatric cohorts.

Oral probiotics are generally well tolerated in immunocompetent individuals, with the most commonly reported side effects being mild gastrointestinal effects.¹⁰ Serious adverse events are rare; however, isolated cases of bacteremia, fungemia, and sepsis linked to probiotic organisms have been reported in severely immunocompromised, critically ill, or oncology patients, indicating the need for heightened caution in these groups.¹¹ Current evidence supports the use of probiotics as adjunctive therapy in mild-to-moderate AD, while acknowledging study heterogeneity and limited long-term safety data.

Prebiotics, including oligosaccharides and polyphenols, promote growth of beneficial gut bacteria and enhance SCFA production.¹² Adult data remain limited but generally align with pediatric trends; in a systematic review of adult trials, six of nine studies reported improvements in AD symptoms, quality of life, or clinical severity with probiotic or prebiotic supplementation, though findings were strain-dependent and inconsistent across studies.¹³ Synbiotics, which combine pre- and probiotics, may outperform either intervention alone, particularly when the probiotic is matched to a complementary prebiotic substrate.¹³ Evidence indicates that treatment duration and dosing materially influence outcomes; in a meta-analysis of synbiotic trials, only studies administering mixed-strain formulations for at least 8 weeks demonstrated significant SCORAD improvements (pooled −6.56; 95% CI −11.43 to −1.68), whereas shorter or lower-intensity interventions showed no measurable benefit.¹⁴

Polyphenols, which are plant-derived bioactive compounds such as quercetin, resveratrol, curcumin, and epigallocatechin gallate (EGCG) are frequently overlooked but can function as prebiotics with anti-inflammatory and microbiome-modulating effects. Delivered through diet or supplements, they may enhance short-chain fatty acid production and augment the clinical efficacy of probiotic interventions.¹⁵ Tolerability is generally acceptable, with only sporadic mild gastrointestinal effects reported; however, the evidence base remains too limited and methodologically heterogeneous to support definitive conclusions regarding safety.

Postbiotics, which are non-viable microbial cells, cell fragments, and metabolites capable of exerting host benefits, are an emerging category within microbiome-directed therapeutics.¹⁶ Although considerably less studied than probiotics, postbiotics have attracted substantial consumer interest. Current evidence remains limited but suggests a modest, short-term symptomatic benefit in select pediatric and adult populations.

Across randomized and open-label studies, Lactobacillus acidophilus L-92 is the most consistently studied postbiotic for AD.¹⁷ Trials in children and adults demonstrate reductions in global severity indices, such as SCORAD, improvements in symptom–medication composite scores, decreases in peripheral eosinophil counts, and increases in regulatory T-cell–associated markers, with heat-killed Lactobacillus acidophilus L-92 in adults yielding significant SCORAD improvement (p = 0.002), decreased eosinophil ratios (p = 0.03), and elevated serum transforming growth factor-beta (TGF-β) (p = 0.03), supporting its potential role in suppressing Th2-dominant inflammation.^18,19 These effects, however, have been observed almost exclusively in Asian populations, and symptom improvement has not persisted after discontinuation, reflecting the potential transient nature of postbiotic activity. Lactobacillus sakei proBio65, evaluated in both live and heat-killed forms, has been shown to reduce SCORAD scores (live: p = 0.0015; heat-killed: p = 0.0017) and improve barrier metrics, including skin sebum content (p < 0.0001 for both), in a 12-week randomized, double-blind, placebo-controlled pediatric trial (ages 3–18), though evidence remains limited to this single study.²⁰

Findings reinforce several patterns: (1) reported benefits are more likely in older children and adults; (2) higher-dose preparations (≥10 billion CFU equivalents) outperform lower-dose formulations; (3) effects are strain-specific; and (4) postbiotics lack the capacity to induce durable alterations in the gut microbiome once administration ceases. Additionally, most viable probiotics do not produce sustained microbial engraftment in AD populations, with observed effects typically transient and dependent on continued supplementation. Age appears to be an important modifier, as durable colonization has been demonstrated in infants and very young children in non-AD cohorts, whereas evidence for sustained engraftment in older children and adults with AD remains limited. Any longer-term influence in AD appears strain-specific and modest, rather than indicative of stable community restructuring.²¹ Safety profiles have been favorable, with only mild gastrointestinal symptoms reported. From a mechanistic perspective, postbiotics can modulate inflammatory cytokines, enhance barrier protein expression, and alter innate immune responses, supported by emerging in vitro and animal data demonstrating downregulation of interleukin-6 (IL-6), interleukin-8 (IL-8), thymic stromal lymphopoietin (TSLP), and restoration of filaggrin-related pathways.²² However, these pathways remain exploratory and strain-dependent.

Given growing public interest, the evolving commercial landscape, and a small but maturing evidence base, postbiotics warrant inclusion within clinical counseling frameworks. At present, their utility appears adjunctive, with strain-specific, non-durable improvements and limited generalizability across age groups and geographic cohorts. Further controlled studies, particularly outside East Asian populations and using standardized, high-potency formulations are needed before postbiotics can be recommended broadly in AD management.

AD remains highly prevalent worldwide, affecting roughly 11% of children and 6% of adults.²³ Interest in microbiome-directed therapies has grown quickly, but the data remain inconsistent. Some randomized trials and meta-analyses show benefit from oral probiotics or prebiotics, yet results vary widely by strain, dose, and treatment duration. This heterogeneity makes it difficult to translate findings into routine practice.

Relevant findings also come from outside dermatology. Work in oncology, particularly in patients receiving immune checkpoint inhibitors, has raised concerns about how probiotics and diet may shape systemic immune responses. Observational studies in melanoma cohorts link higher dietary fiber intake with better immunotherapy outcomes, while over-the-counter probiotic use has been associated with reduced response rates.^24,25 Animal models suggest potential mechanisms involving dendritic cell activation, regulatory T-cell dynamics, and broader immune tone, but these ideas remain preliminary.²⁶ Limitations include self-reported probiotic use, lack of strain-level detail, variable product quality, antibiotic exposure, and confounding dietary factors which make these findings hard to interpret with confidence.

These findings carry several practical implications for clinical management. Currently, there is insufficient guidance regarding which probiotic strains, doses, or treatment durations are most effective for treating AD, and evidence is lacking on which patient subgroups—pediatric versus adult, mild versus severe disease, or immunocompromised individuals—are most likely to benefit. Long-term safety also remains uncertain, particularly in immunocompromised patients or those with active or prior malignancy. The lack of consistent regulation of over-the-counter probiotic products further complicates clinical decision-making due to variability in strain identity, potency, and viability.²⁷

Clinicians should actively inquire about over-the-counter probiotic use in all AD patients, documenting product, strain, dose, and duration whenever possible. For mild-to-moderate AD, a shared decision approach may be appropriate, including a defined trial of a well-characterized oral probiotic for 8–12 weeks with objective outcome monitoring. In patients who demonstrate clinical benefit, next steps should be individualized. Options include continued probiotic use, particularly if benefit is sustained and well tolerated, or discontinuation with an emphasis on dietary fiber and prebiotic intake to support microbial diversity. Given the limited long-term data, the decision to continue supplementation should be guided by patient preference, tolerability, cost, and ongoing response, using shared decision-making. Probiotics should generally be avoided in patients receiving systemic immunomodulators, immune checkpoint inhibitors, or with active cancer without oncology consultation. Decisions should be individualized, favoring strains with demonstrated efficacy and safety in controlled regimens, and concurrent documentation of diet, antibiotic exposure, and probiotic intake is critical for both clinical interpretation and future research. Dietary strategies emphasizing fiber-rich, whole foods are recommended, while unnecessary elimination diets should be discouraged. Combining prebiotics with probiotics may enhance therapeutic effects, and polyphenol-rich foods or supplements can serve as additional microbiome-directed adjuncts.

Future investigations should prioritize larger, multicenter RCTs stratified by age, AD phenotype, and comorbidities, testing prespecified, high-quality strains with standardized product controls. Dedicated study arms should specifically include patients receiving systemic immunomodulators (including biologics and Janus kinase (JAK) inhibitors), a population with significant immune perturbation and gut dysbiosis in whom microbiome-directed interventions may be most biologically and clinically relevant. Future trials should incorporate baseline gut microbiome assessment and stratify participants by dysbiosis status, as probiotics and prebiotics may preferentially benefit patients whose AD is accompanied by measurable microbial disruption. Failure to account for baseline microbiome heterogeneity likely contributes to inconsistent effect sizes and masks true treatment signal in existing studies. Head-to-head trials comparing prebiotics, probiotics, synbiotics, and parabiotics interventions are also needed, alongside long-term safety registries focusing on immunocompromised and oncology populations. Standardized clinical endpoints, including SCORAD, EASI, POEM, and steroid-sparing outcomes should be used with integrated biomarker and microbiome substudies, favoring shotgun metagenomics over 16S for strain-level resolution and functional insights to link gut and skin microbial changes, SCFA levels, and immune biomarkers to meaningful clinical responses.²⁸ These approaches will be essential to identify truly effective, safe, and mechanistically informed microbiome-directed interventions for AD treatment.