Probiotics for Childhood Eczema: What the Evidence Shows

Oral probiotics may help some children with mild-to-moderate atopic dermatitis, but the useful conclusion is narrower than "probiotics work for eczema." Three independent meta-analyses point in the same direction for pediatric SCORAD scores, while adult results remain too inconsistent for a reliable recommendation. The details that appear to matter are age, the probiotic formulation, and whether treatment lasts longer than three months.

SCORAD is a clinician-rated measure of atopic dermatitis severity. A lower score means less extensive or intense disease and fewer symptoms. It is a relevant patient-facing outcome, though a statistically significant change does not automatically tell us how much relief a given child will feel.

The pediatric signal appears across different reviews

The broadest analysis, Kim et al.'s systematic review and meta-analysis, pooled 25 randomized controlled trials with 1,382 participants. Across the full set of trials, probiotics were associated with a statistically significant reduction in SCORAD: standardized mean difference (SMD) −4.0, 95% confidence interval (CI) −7.3 to −0.7. The confidence interval stays below zero, so the pooled result met conventional statistical significance.

The important findings were in the subgroups. Participants with mild-to-moderate atopic dermatitis had a significant reduction in SCORAD (SMD −1.4, 95% CI −2.2 to −0.7). Courses lasting more than three months also reached significance (SMD −5.1, 95% CI −9.7 to −0.4), whereas shorter courses did not. Lactobacillus-containing combination products had a significant subgroup estimate of SMD −4.4 (95% CI −8.0 to −0.8).

A separate 2025 pediatric-only meta-analysis by Gaikwad et al. included 25 randomized trials and analyzed the 14 that reported SCORAD. Using Hedges' g, a correction designed to reduce small-study bias, it found an overall pediatric effect of about 0.65. The review described that as a moderate effect. Lactobacilli appeared in 23 of its 25 included studies, including Lactobacillus acidophilus, L. paracasei, L. rhamnosus, L. fermentum, L. pentosus, and L. johnsonii.

A third review narrows the question further. Fijan et al.'s 2023 systematic review and meta-analysis examined single-strain Lactobacillus interventions in children with atopic dermatitis. It found that individual strains could significantly reduce SCORAD, although effects were smaller and more variable than those reported for combination regimens in the broader review.

These reviews overlap in some trials, so they are not three fully independent replications of three unrelated datasets. Still, they used different eligibility criteria and analytic approaches and reached the same general pediatric conclusion: an oral Lactobacillus-based intervention can improve SCORAD in some children.

Strain matters, but no strain has won the race

"Lactobacillus" is a genus, not a treatment. Products differ by species, strain, dose, delivery vehicle, and whether they contain one organism or several. Pooling them can help detect a broad signal, but it cannot identify the best product for an individual child.

The trial literature includes L. rhamnosus, including LGG, L. casei, L. fermentum, L. acidophilus, and L. paracasei. In a randomized trial of children with both atopic dermatitis and cow's milk protein allergy, Cukrowska et al. reported SCORAD improvement with L. rhamnosus and L. casei. That result applies to a dual-diagnosis population; it does not establish that either strain will work equally well in children with atopic dermatitis alone.

Combination products may have an advantage in some analyses, but that is not a blanket rule. A randomized, double-blind, placebo-controlled trial in children and adolescents tested a multi-strain product and measured symptom severity and corticosteroid use; Feito-Rodríguez et al. reported complex results. Different mixtures, trial populations, and outcomes prevent a simple verdict that more strains are better.

No head-to-head trial series or atopic-dermatitis-specific network meta-analysis currently ranks LGG against L. paracasei, L. fermentum, or other candidates on SCORAD. A network meta-analysis in pediatric food allergy shows that comparative methods are possible, but food allergy is a different condition. Indirect comparisons across eczema trials cannot settle a strain ranking.

Adults are the unresolved part of the story

The adult evidence does not support the same confidence. Umborowati et al.'s adult meta-analysis included six randomized trials with 241 participants. The pooled SCORAD mean difference was −7.90, a number close to an estimated minimally clinically important difference of about 8.7 points. But the heterogeneity was I²=96%, meaning the trial results differed dramatically. That pooled average does not predict an adult patient's response with useful confidence.

The adult trials used different interventions, including L. plantarum IS-10506, L. salivarius LS01 alone, LS01 with BR03, and L. acidophilus L-92, in Indonesia, Japan, and Italy. Their courses lasted 8 to 16 weeks. None of the pooled secondary outcomes reached statistical significance, including itch, Dermatology Life Quality Index score, IL-4, IFN-γ, TNF-α, or serum IgE.

Duration could explain part of the discrepancy, but it cannot yet explain all of it. Kim et al. found benefit in the subgroup treated for more than three months, while adult studies lasted 8 to 16 weeks. Some adult courses therefore approached or exceeded 12 weeks, yet the adult trials still varied too much in strain and setting to isolate duration as the cause. Age-related differences in microbiome plasticity and immune development are plausible explanations, not proven explanations.

A plausible mechanism is not a prescription

The gut-skin axis has reasonable biological support. Probiotics and resident microbes may influence short-chain fatty acids, intestinal barrier function, NF-κB signaling, and regulatory T-cell pathways involving IL-10 and TGF-β. These mechanisms could affect the Th2-skewed inflammation associated with atopic dermatitis.

Human observational evidence supports relevance, not causation. In adults, Blicharz et al. found that atopic dermatitis severity correlated with gut-derived metabolites and biomarkers related to intestinal permeability. A companion adult study linked food sensitization with disease severity, gut-derived metabolites, and leaky-gut biomarkers. Those associations do not show that changing one biomarker with a probiotic will improve eczema.

There is another useful complication: a meta-analysis of oral Lactobacillus-derived postbiotics found lower SCORAD than placebo across three studies. Because postbiotics contain dead microorganisms or their components, live colonization may not be required for every effect. That finding broadens the mechanistic possibilities; it does not prove which pathway caused the clinical changes.

What to do with this evidence

For a child with mild-to-moderate atopic dermatitis, a Lactobacillus-containing probiotic can be considered as an adjunct to standard skin care and prescribed treatment, not a replacement for either.

The evidence supports a trial lasting more than three months more than it supports an eight-week experiment. Single-strain products have evidence in children, and combination products had the strongest subgroup signal in one large meta-analysis. Choose a product that identifies the organism at strain level when possible; a label that says only "probiotic blend" makes it difficult to connect the purchase to the research.

Do not interpret the research as proof that every Lactobacillus product will help, or that severe disease will respond. The broad meta-analysis found significant benefit in mild-to-moderate disease, not severe disease. Pediatric trials were also small, ranging from 33 to 203 participants, and studies differed in dose, vehicle, diagnostic criteria, co-interventions, and duration. The pediatric review was not prospectively registered, which leaves room for post-hoc analytic choices. No included trial reported outcomes beyond the intervention period, so relapse rates, durability, and long-term safety remain uncharacterized.

For adults, current evidence cannot support a confident strain-specific recommendation. A clinician can help weigh a supervised trial against the person's eczema severity, allergy history, medications, and risk factors. Probiotics are an adjunct with a credible pediatric signal, an uncertain adult signal, and no evidence-based strain leaderboard yet.