Error bars indicate 95% confidence interval.
eTable 1. Major and minor criteria in the study participants within placebo and probiotic groups
eTable 2. KIDMED test to assess the Mediterranean Diet
eTable 3. Estimation of the effect of the intervention in laboratory parameters
eTable 4. Estimation of the effect of the intervention in SCORAD components
eFigure 1. Estimated differences between the two trial arms
eFigure 2. Percentage of patients with IGA score under 2, measured at each checkup in weeks 4, 8, and 12
eFigure 3. Distribution of number of days using steroids to treat flares
Customize your JAMA Network experience by selecting one or more topics from the list below.
Navarro-López V, Ramírez-Boscá A, Ramón-Vidal D, et al. Effect of Oral Administration of a Mixture of Probiotic Strains on SCORAD Index and Use of Topical Steroids in Young Patients With Moderate Atopic Dermatitis: A Randomized Clinical Trial. JAMA Dermatol. 2018;154(1):37–43. doi:10.1001/jamadermatol.2017.3647
Can treatment with an oral probiotic reduce the SCORAD index and the use of topical steroids in children with moderate atopic dermatitis?
This randomized clinical trial of 50 children treated with a mixture of probiotics or placebo for 12 weeks found that SCORAD and topical steroid use decreased significantly in the probiotic group compared with the placebo group.
This probiotic is an effective and safe coadjuvant treatment to reduce the SCORAD index and topical steroid use in children with moderate atopic dermatitis.
Oral intake of new probiotic formulations may improve the course of atopic dermatitis (AD) in a young population.
To determine whether a mixture of oral probiotics is safe and effective in the treatment of AD symptoms and to evaluate its influence on the use of topical steroids in a young population.
Design, Setting, and Participants
A 12-week randomized, double-blind, placebo-controlled intervention trial, from March to June 2016, at the outpatient hospital Centro Dermatológico Estético de Alicante, Alicante, Spain. Observers were blinded to patient groupings. Participants were children aged 4 to 17 years with moderate atopic dermatitis. The groups were stratified and block randomized according to sex, age, and age of onset. Patients were ineligible if they had used systemic immunosuppressive drugs in the previous 3 months or antibiotics in the previous 2 weeks or had a concomitant diagnosis of intestinal bowel disease or signs of bacterial infection.
Twelve weeks with a daily capsule containing freeze-dried powder with 109 total colony-forming units of the probiotic strains Bifidobacterium lactis CECT 8145, B longum CECT 7347, and Lactobacillus casei CECT 9104 and maltodextrin as a carrier, or placebo (maltodextrin-only capsules).
Main Outcomes and Measures
SCORAD index score and days of topical steroid use were analyzed.
Fifty children (26 [50%] female; mean [SD] age, 9.2 [3.7] years) participated. After 12 weeks of follow-up, the mean reduction in the SCORAD index in the probiotic group was 19.2 points greater than in the control group (mean difference, −19.2; 95% CI, −15.0 to −23.4). In relative terms, we observed a change of −83% (95% CI, −95% to −70%) in the probiotic group and −24% (95% CI, −36% to −11%) in the placebo group (P < .001). We found a significant reduction in the use of topical steroids to treat flares in the probiotic arm (161 of 2084 patient-days [7.7%]) compared with the control arm (220 of 2032 patient-days [10.8%]; odds ratio, 0.63; 95% CI, 0.51 to 0.78).
Conclusions and Relevance
The mixture of probiotics was effective in reducing SCORAD index and reducing the use of topical steroids in patients with moderate AD.
clinicaltrials.gov Identifier: NCT02585986
Atopic dermatitis (AD) is a chronic recurrent inflammatory skin disease characterized by intense pruritus, inflammation, and skin barrier disruption. The prevalence of AD is approximately 3% to 10% in adults and up to 20% in children worldwide.1,2 The first symptoms usually develop during childhood, and approximately 50% of cases are diagnosed in the first year of life. Atopic dermatitis significantly reduces the quality of life of patients and their families. Moreover, patients with AD have an increased risk of other atopic disorders, including asthma, allergic rhinitis, and chronic sinusitis.3,4
As occurs in other atopic disorders, a predominance of T helper 2 cells rather than T helper 1 causes an imbalance that might also aggravate the pathogenesis of AD, increasing IgE and activating interleukins.5,6 Another point to consider is skin barrier integrity. Filaggrin interacts with intermediate filaments, particularly keratins, causing their aggregation into macrofibrils. Defects in filaggrin cause dysfunctions in the skin barrier, resulting in decreased protection from microbes and allergens.7
Clinically, cutaneous manifestations related to AD include erythema, edema and/or papules, exudate, excoriation, and lichenification, as well as the resulting symptoms (pruritus and loss of sleep).3,8 Topical corticosteroids have been the keystone of pharmacological treatments for mild to moderate AD. Alternatively, cases of moderate-to-severe AD are treated with long-term applications of topical corticosteroids,9 with topical calcineurin inhibitors providing second-line effective agents. However, long-term data for these medications are lacking for pediatric patients.10 In cases of severe refractory disease, patients might benefit from a short course of systemic therapy with immunosuppressants such as corticosteroids, cyclosporine, and azathioprine, but these drugs have a potentially more severe adverse effect profile and risks of rebound after treatment discontinuation.11,12
Quiz Ref IDDuring recent years, many authors have suggested an association between a disruption in intestinal barrier function and the origin of AD, mediated by immunological activation leading to a type 2 dominant inflammation.13 In this respect, gut microbiota may play an important immunomodulatory role in the development of normal immune tolerance.14,15 Recently an analysis of the gut microbiota of patients with AD has shown an intraspecies compositional change in Faecalibacterium prausnitzii that reduces the number of high butyrate and propionate producers.16 Butyrate and propionate are microbial-produced short-chain fatty acids with an anti-inflammatory role.17 Moreover, butyrate has been shown to be a key player in maintaining gut barrier integrity.18 Therefore, reduced levels in the microbiota of both butyrate and propionate producers may result in a pro-inflammatory state in the gut and a loss of barrier integrity. All these data indicate the potential role of probiotics as microbiota recovery players, and consequently as potential nutritional supplements in AD treatment.19-21
The primary goal of the present study was to determine the efficacy of a mixture of probiotics in improving the SCORAD (Scoring Atopic Dermatitis) index22 and in reducing the percentage of days with topical steroid treatment during flares in patients with moderate AD.
The study design was a double-blind, 2-arm placebo-controlled trial with stratified randomization by baseline variables (ratio, 1:1). The study received approval from the Ethics Committee for Clinical Research of the Hospital General Universitario de Alicante, and the Spanish Medicines Agency. There was no change to the trial protocol (Supplement 1) after it commenced.
Inclusion criteria of the protocol were as follows: children between 4 and 17 years old, with a diagnosis of AD that meets the Hanifin and Rajka criteria (eTable 1 in Supplement 2) and moderate SCORAD index (from 20 to 40)22 who had been prescribed topical steroids for the treatment of AD. Participants were also required to be currently consuming a high-quality Mediterranean diet with a Mediterranean Diet Quality Index (KIDMED)23 score more than 7. Written informed consent was obtained from parents or a legal representative (and the child if >12 years).
Among the exclusion criteria, patients were ineligible for the study if they had used systemic corticosteroids, methotrexate, cyclosporine, or anti–tumor necrosis factor drugs in the previous 3 months, antibiotics in the previous 2 weeks, or had a concomitant diagnosis of intolerance to gluten and/or lactose or signs of bacterial infection (the Trial Protocol in Supplement 1 provides a complete list of exclusion criteria).
Children aged 4 to 17 years were recruited between March and June 2016 from a single outpatient dermatological clinic where 3 dermatologists (A.R.-B. and 2 others) with expertise in pediatric dermatology evaluated their AD and SCORAD index. After informed consent was obtained from their legal representatives, participants completed the dietary screener KIDMED short questionnaire for quality of the Mediterranean diet (eTable 2 in Supplement 2).23
Each child was classified into 1 of the 8 possible strata combining 3 binary variables at baseline: sex, age (4-12 vs >12 years), age of onset (0-4 vs >4 years). Patients were then assigned to 1 of the 2 trial arms (probiotics or control) following a computerized randomization list that been previously prepared for each stratum by the principal investigator of the study (V.N.-L.).
All patients received treatment during the 12-week study period with topical methylprednisolone aceponate, moisturizer, and 1 oral antihistamine, according to the guidelines for the management of AD.9 Participants in the probiotic group received daily a pill containing 109 colony-forming units (CFUs) of a mixture of the 3 probiotic strains in a 1:1:1 ratio, freeze-dried powder with maltodextrin as a carrier, and participants in the placebo group received a pill containing only maltodextrin. The probiotic and placebo pills were matched for size, shape, and volume of contents (gelatin capsules of 9.85 × 16.4 mm) and were dispensed by the pharmacy department staff.
Quiz Ref IDThe product was designed not as a single strain but as a bacteria mixture with lactobacilli and bifidobacteria because previous AD clinical trials with probiotics have obtained better results with mixed products.24 The mixture was composed of Bifidobacterium lactis CECT 8145, B longum CECT 7347, and Lactobacillus casei CECT 9104. The blend was selected on the basis of previous data: oxidative stress has been associated with AD,25 and strain B lactis CECT 8145 has previously shown antioxidant properties in the Caenorhabditis elegans model.26 Strain B longum CECT 7347 has an anti-inflammatory profile, previously demonstrated in both preclinical and clinical studies,27,28 and it modulates the gut microbiota.28 Finally, strain L casei CECT 9104 has in vitro activity against certain gut pathogens (E.C.-C., S.G.-M., D.R.-V., unpublished data, June 2009).
Two primary outcomes were compared between the trial arms to assess the treatment effect: (1) change in SCORAD index between baseline and 12 weeks of follow-up and (2) the proportion of days of topical steroid use during flares within the 12 weeks of follow-up. The SCORAD index was measured at the time of inclusion and every 4 weeks until the end of the 12-week follow-up period. The number of days of topical steroid use during flares was recorded biweekly until week 12. A flare was defined as a worsening of the disease leading to use of topical corticosteroids for at least 3 consecutive days (so steroid use on only 1 or 2 isolated days was not considered a flare).
Secondary outcomes were laboratory values. Peripheral blood samples were collected at baseline and after 12 weeks of treatment and analyzed for routine biochemical laboratory values and for interleukin 4 (IL-4), IL-5, IL-10, and IL-13 levels.
According to the trial protocol, 25 patients per group would be needed to detect a difference of clinical improvement as evaluated by the SCORAD index, assuming 5% loss to follow-up, type I error of .05, and 80% power to detect a difference of 4.13 units on the SCORAD index with a standard deviation of 5 units between the 2 groups on a 2-sided t test.
Data were analyzed as intention to treat, and the last available value was not carried forward for missing observations. Quantitative variables were summarized as means and standard deviations or median and interquartile range, and categorical variables were summarized with proportions.
We used linear mixed-effects models with a group-by-week interaction term to compare the mean SCORAD score between groups at weeks 4, 8, and 12. Participant-level random intercepts accounted for the correlation due to repeated measurements. The model was adjusted for variables used to stratify randomization (age at recruitment, sex, and age at onset). Residuals were examined visually and tested for normality with the Kolmogorov-Smirnov test. To analyze the proportion of days of topical steroid use, we used logistic regression to model the odds per individual with treatment group as the main explanatory variable and adjusted for the variables used for stratified randomization.
Laboratory determinations from blood samples were log-transformed to achieve more normally distributed variables and avoid extreme values. They were then analyzed with linear mixed-effects models (as for the SCORAD although with only 2 time observations, baseline and end of study).
We followed CONSORT recommendations for reporting of randomized clinical trials (Figure 1). All statistical analyses were performed using IBM SPSS statistics, version 22 (SPSS Inc), and R, version 3.2.3.
From March to June 2016, 50 children with a diagnosis of AD were enrolled in the study. All met the criteria for a high-quality Mediterranean diet (KIDMED score >7). There were no significant differences in the baseline characteristics between the 2 groups (Table 1). Three of the patients in the probiotic group failed to provide data in some of the visits. Two of them declined contact in the second and third visit due to a work-related problem of the parents. Another patient had a concurrent disease unrelated to the treatment and could not provide data on the third visit. All reasons for missing data were unrelated to the treatment or the outcome, so we do not expect effect bias. Because missing data were only 2.5% of follow-up measurements, we decided not to use it and we used only observed data. Patients completed the KIDMED test at inclusion and repeated the test at the end of the intervention. The data show no relevant differences in the KIDMED score between the trial arms at baseline (Table 1) or end of the study (mean [SD] score, 8.7 [0.63] vs 8.6 [0.66] for the placebo and treatment groups, respectively).
Twenty-two of 23 (96%) patients in the probiotic group and 11 of 24 (46%) in the placebo group improved in the SCORAD index. The mean reduction and 95% confidence interval for the SCORAD index and its components (eczema extension, eczema intensity, and subjective symptoms) for different weeks are shown in Figure 2, whereas the estimated changes in each trial arm and the differences between the 2 arms (intervention effect) are presented in Table 2 for the total SCORAD index, eTable 4 in Supplement 2 for the subcomponents, and eFigure 1 in Supplement 2 for all. These eTables also give the changes as percentages over baseline values of the variable. At baseline, there were no large differences between arms in any of the variables. As time passed after starting treatment, the differences between arms increased, with better improvements in the probiotic group for total SCORAD, eczema intensity, and eczema spread (with CI excluding the null) but not for subjective symptoms. Quiz Ref IDAfter 12 weeks of follow-up, the mean change in the SCORAD index (main outcome) was, in relative terms, −83% (95% CI, −95% to −70%) in the probiotic group and −24% (95% CI, −36% to −11%) in the placebo group. This is a difference in effectiveness of −19.2 (95% CI, −23.4 to −15.0) SCORAD points, or −59% (95% CI, −72% to −46%; P < .001), in favor of the probiotic. Patients in both probiotic and placebo groups showed improvement in atopic dermatitis disease activity as measured with Investigator Global Assessment (IGA) over the 12-week study period. Comparison between groups showed significant differences in the number of patients with response to treatment: 21 patients (91%) in the probiotic group achieved IGA scores of 0 or 1 (95% CI, 72%-99%) and 5 patients (21%; 95% CI, 7%-42%) achieved these scores in the placebo group. The percentage of patients with IGA score less than 2, measured at each checkup in weeks 4, 8, and 12, is included in eFigure 2 in Supplement 2.
At the end of follow-up, there were 2032 patient-days of observation in the placebo group and 2084 patient-days in the probiotic group. Steroids were used to treat flares on 220 (10.8%) and 161 (7.7%) patient-days, respectively. The distribution of number of days per individual did not seem to follow a Poisson distribution (eFigure 3 in Supplement 2). Considering that follow-up was approximately 84 days, some individuals used steroids for a large proportion of them. The logistic regression adjusted for baseline stratification variables estimated an odds ratio of 0.63 (95% CI, 0.51-0.78; P < .001) for the effect of the treatment on the reduction of use of corticosteroids. Quiz Ref IDWe also did a sensitivity analysis comparing the total number of days of steroid use (including nonflare use). The placebo arm had 336 patient-days of use (16.5%) whereas the probiotic arm had 291 patient-days of use (14.0%). The adjusted logistic regression model produced an odds ratio of 0.77 (95% CI, 0.65-0.91; P < .003), a statistically significant finding of less corticosteroid use in the probiotic group.
Baseline blood marker levels of patients in the probiotic and placebo groups are given in Table 1. During the 12-week intervention period, no significant differences were observed between the 2 groups in the changes in blood levels of IL-4, IL-5, IL-10, IL-13, eosinophils, IgE, and lactate dehydrogenase (eTable 3 in Supplement 2).
No relevant adverse events were associated with drug or placebo intake.
Several important studies have explored the efficacy of certain probiotics in the prevention and treatment of AD. Overall, the current evidence suggests that probiotics could be an option to improve moderate and severe AD recovery rates in children and adults; however, to date, there is no strong experimental evidence supporting their effectiveness and safety in clinical practice.29 Importantly, evidence and clinical trials demonstrating strain-specific effects are lacking.30-33
The clinical trial reported herein explores the role of a mixture of probiotics administered to patients with moderate AD. Several variables, such as antibiotic use, diet, and other concomitant allergenic diseases were controlled for, and, to avoid bias, all cases included in the study were matched for these variables. Our results suggest that administration of this mixture of probiotics, as adjuvant treatment, may be effective in reducing the SCORAD index and, subsequently, decreasing the use of steroids during AD flares. The response rate was significant when the reduction in baseline SCORAD index was compared between groups (Table 2). The clinical response documented in the probiotic group was greater than that obtained with other probiotics tested in a previous placebo-controlled clinical trial.32-43 This study showed better response in SCORAD (83% relative reduction) than previously communicated results by Farid et al44 (68% of response), Iemoli et al45 (63%), or Yeşilova et al (64%).46 Several factors may influence the response to probiotic treatment in AD, and they were taken into account when the protocol of this clinical study was designed: treatment longer than 8 weeks might condition the positive effect of probiotic use,47 patients older than 1 year have a greater response to probiotics,35,36,38,47 patients with moderate to severe AD have a better response,20 and a mixture of probiotics has better beneficial effects than a single probiotic, especially when lactobacilli and bifidobacteria are included in the mixture.20,44,45 The final blend used in the study was selected on the basis of published results and internal unpublished data, as described in the Methods.24-28
Two of 3 subcomponents of the SCORAD index (eczema spread and intensity) showed a clear improvement in favor of the probiotic group compared with the placebo group (Figure 2, B and C). In contrast, there was no statistically significant difference in subjective symptoms (eTable 4 in Supplement 2 and Figure 2D). This finding was not due to a lack of effect in the probiotic group, which showed a clinically relevant proportional reduction of 77%, but rather because there was also a reduction in the placebo group of 53%. A possible explanation is that the placebo group was able to reduce symptoms such as pruritus by using more corticosteroids.
Quiz Ref IDAfter 12 weeks, we saw a slightly greater reduction of IL-4, IL-5, and IL-13 in the probiotic arm compared with placebo (eTable 3 in Supplement 2, last column). This may suggest decreased activity of the T helper 2 cells in the probiotic arm. We did not find statistical significance in any of these differences, but lack of statistical significance does not mean evidence of no difference, especially when the trial was not powered to test differences in cytokines. An investigation of the biological mechanism will require a specifically designed study.
The limitations of the study should be considered and clarified through further research. These limitations include the applicability of our results to patients consuming a different diet in different geographic areas, and whether the results can be extended to other population groups such as newborns (<1 year) or adults older than 17 years. In addition, the short follow-up of 12 weeks, the fact that topical corticosteroid dose was not recorded, and the inclusion of patients from a single center should also be considered limitations of this clinical trial. Finally, questions should be answered about adequate dosage, the duration of probiotic administration, and at what age the use of probiotics would be most efficacious. Future trials will be necessary and should consider all these questions to assess the probiotic mixture used in this clinical trial and other specific probiotic strains.
The results of our study indicate a strong positive effect in reducing the SCORAD index and use of topical corticosteroids in the group treated with the probiotic mixture. This evidence supports the efficacy of administering this probiotic mixture to patients with moderate AD and suggests that it could be used more extensively in clinical practice.
Corresponding Author: Vicente Navarro-López, MD, Department of Clinical Medicine, Universidad Católica San Antonio de Murcia (UCAM), Campus de los Jerónimos, s/n 30107 Guadalupe (Murcia), Spain (firstname.lastname@example.org).
Accepted for Publication: July 27, 2017.
Published Online: November 8, 2017. doi:10.1001/jamadermatol.2017.3647
Author Contributions: Dr Navarro-López had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Navarro-López, Ramírez-Boscá, Ramón-Vidal, Genovés-Martínez, Chenoll-Cuadros, Horga de la Parte, Codoñer-Cortés.
Acquisition, analysis, or interpretation of data: Navarro-López, Ramírez-Boscá, Ruzafa-Costas, Chenoll-Cuadros, Carrión-Gutiérrez, Horga de la Parte, Prieto-Merino.
Drafting of the manuscript: Navarro-López, Ramón-Vidal, Codoñer-Cortés.
Critical revision of the manuscript for important intellectual content: Ramírez-Boscá, Ruzafa-Costas, Genovés-Martínez, Chenoll-Cuadros, Carrión-Gutiérrez, Horga de la Parte, Prieto-Merino.
Statistical analysis: Navarro-López, Prieto-Merino, Codoñer-Cortés.
Administrative, technical, or material support: Ramón-Vidal, Ruzafa-Costas, Genovés-Martínez, Chenoll-Cuadros, Horga de la Parte.
Study supervision: Ramírez-Boscá, Ruzafa-Costas, Carrión-Gutiérrez, Horga de la Parte.
Conflict of Interest Disclosures: Dr Ramón-Vidal, Mr Genovés-Martínez, and Dr Chenoll-Cuadros are employed by Biopolis. No other disclosures are reported.
Funding/Support: This study was supported by Biopolis and Korott.
Role of the Funder/Sponsor: The sponsors had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: José M. Pérez-Orquín, MD, Korott SL, and Vicente García Román, MD, Universidad Católica San Antonio de Murcia, provided critical review of the final version of the manuscript. Leticia Sánchez, BSN, and Encarna Espejo, BSN, Centro Dermatológico Estético de Alicante, provided support during the clinical and monitoring phase. No compensation was received for such contributions. We also thank all the parents and participants who agreed to take part in this trial.