The Efficacy and Safety of Probiotics for Allergic Rhinitis: A Systematic Review and Meta-Analysis

Background Probiotics have proven beneficial in a number of immune-mediated and allergic diseases. Several human studies have evaluated the efficacy and safety of probiotics in allergic rhinitis; however, evidence for their use has yet to be firmly established. Objective We undertook a systematic review and meta-analysis aiming to address the effect and safety of probiotics on allergic rhinitis. Methods We systematically searched databases [MEDLINE (PubMed), Embase, and the Cochrane Central Register of Controlled Trials] from inception until June 1, 2021. Qualified literature was selected according to inclusion and exclusion criteria, the data were extracted, and a systematic review and meta-analysis was conducted. Results Twenty-eight studies were included. The results showed that probiotics significantly relieved allergic rhinitis symptoms (standardized mean difference [SMD], −0.29, 95% confidence interval (CI) [−0.44, −0.13]; p = 0.0003, I 2 = 89%), decreased Rhinoconjunctivitis Quality of Life Questionnaire (RQLQ) scores compared with the control group (SMD, −0.64, 95% CI [−0.79, −0.49], p < 0.00001, I 2 = 97%), and increased T helper cell 1(Th1)/Th2 ratio (mean difference [MD], −2.47, 95% CI [−3.27, −1.68], p < 0.00001, I 2 = 72%). There was no significant change in overall or specific IgE levels between probiotic-treated and placebo-treated subjects (SMD, 0.09, 95% CI [−0.16, 0.34], I 2 = 0%, and SMD, −0.03, 95% CI [−0.18, 0.13], p = 0.72, I 2 = 0%, respectively). Conclusions To sum up, probiotic supplement seems to be effective in ameliorating allergic rhinitis symptoms and improving the quality of life, but there is high heterogeneity in some results after subgroup analysis and clinicians should be cautious when recommending probiotics in treating allergic rhinitis. Systematic Review Registration https://www.crd.york.ac.uk/PROSPERO/, PROSPERO (CRD42021242645).


INTRODUCTION
Allergic rhinitis (AR) is characterized by a nasal sensitive inflammation, which is estimated to already affect 10%-40% of the worldwide population (1,2). Common symptoms of AR are nasal itching, sneezing, rhinorrhea, and nasal congestion. In addition, some patients experience symptoms of allergic rhinoconjunctivitis, such as watery or itchy or red eyes. Severe AR can affect the quality of life, sleep, and work performance (1).
In 1989, Strachan found that the number of siblings was inversely related to the prevalence of hay fever among peers in the UK. Then, he proposed the "Hygiene hypothesis" (3), that the changed intestinal microbiota due to the lack of contact with infectious sources, parasites, and symbiotic microorganisms affects the normal development of immune system. The "Hygiene hypothesis" extends to the "Old Friends" and the "Microflora hypothesis" (4,5). The "Microflora hypothesis" believes that a diverse gut microbiota plays an important role in shaping host immune development and that disruption or dysbiosis of the normal gut microbiota contributes to the development of immune disorders such as allergic diseases (6,7). Host-microbes symbiosis plays a cardinal role in maintaining health and immune homeostasis. Changes in the intestinal flora are considered to be one of the most important indicators of allergic diseases (8,9). Probiotics are live bacteria that colonize the gastrointestinal tract and they provide a health benefit to the host when administered in adequate amounts (10). Recent studies have shown that probiotics are non-pharmaceutical agents that can increase the production of systemic IFN, IL10, and IL12, improve the pre-Th1 immune response, and reduce Th2 cytokines (11), and thus have been proposed as modulators of the allergic response and advocated as therapeutic and preventive interventions for allergic disease (12,13).
Probiotics include the Lactobacillus group (L. rhamnosus GG, L. sporogenes, L. reuteri RC-14, L. plantarum 299v, L. acidophilus, and L. lactis), the Bifidobacterium group (B. bifidum, B. longum, and B. infantis), the Streptococcus group (S. thermophilus, S. lactis, and S. fecalis), and non-bacterial organisms (non-pathogenic yeast Saccharomyces boulardii). The most common probiotics are the Lactobacillus and Bifidobacterium groups (14). Many studies have attempted to assess the role of probiotics in the treatment of AR with inconsistent findings. While some have found a protective effect of probiotics on AR (15)(16)(17)(18), several others have found no association (19,20). Given that there have been further published studies, we undertook a systematic review and metaanalysis aiming to address the effect and safety of probiotics on AR, and meanwhile, we attempted to explore the possible causes of between-study heterogeneity via subgroup.

Study Registration
The protocol of this systematic review and meta-analysis has been registered on the PROSPERO platform with an assigned registration number CRD42021242645, based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocols statement guidelines. This research was conducted based on this protocol.

Database Search
We have performed a search in MEDLINE (PubMed), Embase, and the Cochrane Central Register of Controlled Trials. Additional studies will be sought by manually checking the references of included studies and relevant reviews. Searches will be restricted to publications appearing from inception to June 1, 2021. We used subject ("Rhinitis, Allergic", "Rhinitis, Allergic, Seasonal", "Rhinitis, Allergic, perennial", "prebiotics"," probiotics") and free words ("Seasonal Allergic Rhinitis", "Pollen Allergy" "Pollinosis", "Hay Fever", "allergic rhinitis", "Perennial Allergic Rhinitis", "prebiotics"," probiotics") to search in the databases aforementioned. The search strategy was as follows, taking PubMed as an example: (

Eligible Criteria
Studies were included if they met all of the following criteria (1): study design: experimental (randomized and quasi-randomized controlled trials) studies (2); study participants: participants with AR (3); intervention: the intervention group/s should receive probiotics supplementation in any dosage, or regimen as decided by the trialists of the respective trials (4); comparator(s)/control: the participants in the comparison group/s might receive a placebo or other drugs (5); if other drugs were used in the treatment group, they must also be used in the control group in the same way; and (6) language: articles published in the English language.
Articles were excluded if they were published in the form of conference abstract, case report, case series, letter to the editor, correspondence, editorial, narrative reviews, systematic reviews, and meta-analyses.

Study Selection and Data Extraction
Two investigators independently reviewed titles, abstracts, and full-text articles according to the aforementioned inclusion and exclusion criteria. Disagreement was resolved through discussion or a third investigator. The same two investigators extracted the following data from each selected study: literature characteristics (the first author's name, journal, year of publication, and study design); participant information (age and sample size); intervention information (intervention duration and comparison group components); outcome (AR and related adverse events); and conclusion.

Risk of Bias Assessment
The risk of bias assessment was conducted through The Cochrane Risk of Bias Tool Version 1 (21) in Review manager 5.3.4 software by CL and ML. Any disagreement was settled through consultation with the author SP.

Statistical Analyses
Statistical analyses were completed using Review Manager 5.3.4 software (RevMan; Version 5.3.4. Copenhagen, Denmark: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014). We chose the mean difference (MD) and standardized mean difference (SMD) for continuous outcomes. MD is the difference between the two means, which eliminates the influence of the absolute value between multiple studies. SMD can be simply understood as the quotient of the difference between the two means divided by the combined standard deviation, which not only eliminates the influence of the absolute value of multiple studies, but also eliminates the different effects of multiple study measurement units. Statistical heterogeneity was judged using the inconsistency index (I 2 ), and significant heterogeneity was reported if the I 2 is over 50%. The fixed-effect model was be used in this metaanalysis because larger sample studies will receive greater weight and provide greater contributions to pooled effects. Subgroup analyses were conducted to explore the source of heterogeneity. Publication bias assessment was conducted through funnel plots if more than 10 trials were included. Sensitivity analysis was used to explore the stability of the results. The Grading of Recommendations Assessment, Development and Evaluation (GRADE) Working Group was used to assess the evidence quality for outcomes across studies.

Database Search Results
The initial search was completed on June 1, 2021. We have identified 245 potentially relevant publications from PubMed, 580 from Embase, and 129 from The Cochrane Central Register of Controlled Trials. Endnote was used to eliminate duplicate publications, resulting in 97 records for review. After excluding publications that did not meet the inclusion or the exclusion criteria, we included 28 studies for systematic review and meta-Analysis. A flow diagram illustrating the exclusion of articles with specific reasons is shown in Figure 1 (PRISMA flowchart).

Risk of Bias Assessment
The risk of bias assessment is presented in Figures 2, 3. Most studies did not clearly show how to generate random sequences, nor did they clearly state whether association obfuscation was     performed. In terms of masking method, most of the studies have insufficient information to permit judgment of "Low risk" or "High risk". We assessed three trials having high risk of bias for different reasons. One of the trials did not report all the prespecified primary outcome indicators (30). The random allocation method in one of the studies was incorrect (The patients were randomized according to the birth date) (41). Since Nagata reported that participants were all female college students from the same university in the trial (33), it was marked as "high risk" in other bias.

Overall Analyses
Allergic Rhinitis Symptoms Score AR symptoms score included rhinoconjunctivitis total symptom score (RTSS) and total nasal symptom scores (TNSS). RTSS includes five individual AR symptoms (nasal congestion, sneezing, rhinorrhea, nasal pruritus, and eye itching) noted from 0 (no symptom) to 3 (severe symptom). TNSS were expressed as the sum of the scores for the four symptoms (nasal congestion, rhinorrhea, nasal itching, and sneezing) noted from 0 (no symptom) to 3 (severe symptom). Seven   Table 2).

Immunologic Parameters Total IgE
Nine trials reported the effect of probiotics on total IgE. After pooling nine estimates, there was no difference found in total IgE between the probiotic group and the control group (SMD, −0.03, 95% CI [−0.18, 0.13], p = 0.72, I 2 = 0%) ( Figure 6). Sensitivity analysis indicates that the result is stable (Supplementary Material 13). Subgroup analyses were conducted according to the classification of AR and combination of drugs. The results of subgroup analysis showed that the effect of probiotics on total IgE could not be affected by the classification of AR (PAR or SAR) or combined with other drugs (Supplementary Materials 7 and 8) ( Table 2).

Specific IgE
Specific IgE was evaluated in six studies. After pooling six estimates, there was no difference found in sIgE between the probiotic group and the control group (SMD, 0.09, 95% CI [−0.16, 0.34], p = 0.49, I 2 = 0%) (Figure 7). Sensitivity analysis indicates that the result is stable (Supplementary Material 13). Subgroup analyses were conducted according to the classification of AR and combination of drugs. The results of subgroup analysis showed that the effect of probiotics on sIgE could not be affected by the classification of AR (PAR or SAR) or combined with other drugs (Supplementary Materials 9 and 10) ( Table 2).

Th1/Th2 ratio
Four trials reported enough data to allow meta-analysis for the Th1/Th2 ratio. The results showed that the Th1/Th2 ratio was lower in the control group when the effect estimates from four trials were pooled (MD, −2.47, 95% CI [−3.27, −1.68], p < 0.00001, I 2 = 72%) (Figure 8). Sensitivity analysis indicates that the result is stable (Supplementary Material 13). Subgroup analyses were conducted according to the classification of AR. The results of subgroup analysis showed that the effect of probiotics on the Th1/ Th2 ratio could not be affected by the classification of AR (PAR or SAR) or treatment plan (monotherapy/combined) (Supplementary Materials 11 and 12) ( Table 2).

Adverse Events
Of the twenty-eight studies included, seventeen RCTs mentioned that no obvious adverse events were found during the research, while seven RCTs did not mention whether any adverse events occurred. Four RCTs have reported adverse events including diarrhea, abdominal pain, flatulence, and fever episodes. One study reported that loose stools and diarrhea were observed in the active and placebo groups, which had no significant differences in adverse events between the two groups (chisquare test, p < 0.4) (17). Another study showed that subjects with these adverse drug reactions (diarrhea, abdominal pain, and flatulence) recovered within a few days. In this study, it was found that one subject's adverse reaction was almost certainly related to the drug (15). One study reported slight abdominal pain in probiotic groups and all of the adverse events were spontaneously alleviated without drug treatment (41). One study revealed that abdominal symptoms (abdominal symptoms, diarrhea, and fever episodes) were reported in 56.5% versus 64.2% of children in intervention and control groups, respectively (p = 0.282) (26).

GRADE Evidence Quality Evaluation
The quality of evidence applied for each outcome is summarized in Table 3. The quality of evidence on the Allergic Rhinitis Symptoms Score, Rhinoconjunctivitis Quality of Life Questionnaire Score, Total IgE, Antigen-specific IgE, and Th1/Th2 ratio was rated as very low, very low, low, low, and very low, respectively ( Table 3).

DISCUSSION
In this study, the clinical evidence of probiotics in the treatment of AR was systemically collated and analyzed so as to provide a better guidance for clinical practice. Our results showed that probiotics supplementation for patients with AR can ameliorate AR symptoms and improve the quality of life. Probiotics supplementation can correct the Th1/Th2 balance. There was no  significant change in overall or antigen-specific IgE levels between probiotic-treated and placebo-treated subjects. The results of this study have significant heterogeneity, and the source of heterogeneity was explored by subgroup analysis. The results of subgroup analysis showed that probiotics can significantly relieve AR symptoms in patients with SAR. Subgroup analysis according to combination of drugs again found some evidence of a protective effect of probiotics (monotherapy) in relieving AR symptoms compared with placebo. Compared with antihistamines, probiotics combined with antihistamines (combination therapy) have no significant relief of AR symptoms. Subgroup analyses of these outcomes failed to find out the source of heterogeneity. The different doses, durations, and strains of probiotics may be the sources of heterogeneity. With regard to RQLQ score, the results of subgroup analysis according to combination of drugs showed that probiotics (single probiotic strain) compared with placebo can significantly improve the quality of life. Similarly, probiotics combined with antihistamines (combination therapy) compared with antihistamines showed a significant decrease in RQLQ scores, which means an improvement in the quality of life. As we all know, helper T cells play a key role in the adaptive immune response. Human T helper cells can be divided into two main subtypes, Th1 and Th2. The significant trend of immune response to Th2 lineage may lead to allergic diseases. Immunoglobin E (IgE)-mediated allergic inflammation is the main pathophysiological mechanism of AR and drives T helper 2 (Th2) cell polarized immune reactions (45). The balance Th1/Th2 is associated with AR. Th2 induces the activation of B cells and IgE class switching, which leads to B-cell differentiation into plasma cells that produce allergen-specific IgE. IgE enters the circulation and binds through its Cϵ3 domain to the high-affinity IgE receptor (FcϵRI) on the surface of mast cells and basophils (46). Activated mast cells and basophils release inflammatory mediators (e.g., histamine and leukotrienes) that cause symptoms such as nasal itching, sneezing, and runny nose. At the same time, these inflammatory mediators lead to a predominance of Th2 immune responses, further exacerbating inflammation. Therefore, the predominance of Th2 and its related cytokines correlates with the severity of AR. The Th1/Th2 ratio can reflect the effect of improving allergy symptoms by drugs to a certain degree.
Our meta-analysis demonstrated that probiotics supplementation can correct the Th1/Th2 balance, which indicates that probiotic supplementation can ameliorate AR by regulating the balance of Th1/Th2. However, only four of the included studies reported the Th1/Th2 ratio.
The purpose of most systematic reviews or meta-analyses is to explore the preventive effect of probiotic supplementation on allergic diseases (47)(48)(49)(50). There are less systematic reviews or meta-analyses to explore the therapeutic effect of probiotics on AR. A systematic review and meta-analysis of probiotics in the treatment of AR published in 2015 has shown that probiotics may be beneficial in improving symptoms and quality of life in patients with AR (51). One meta-analysis showed that probiotics have beneficial effects in the treatment of AR, especially with SAR and LP-33 strains (52). However, previous systematic reviews failed to explore the causes of heterogeneity as much as possible. Compared with previous systematic reviews and meta-analyses, our meta-analysis conducted subgroup analysis according to types of AR (PAR/ SAR) and treatment plan (single probiotic strain/mixed probiotic strains/probiotics combined with antihistamines; monotherapy/combined). We found that a single probiotic strain (LP-33) can significantly improve the quality of life of patients with AR from the meta-analysis of three studies. Two studies used mixed probiotic strains. One study demonstrated that a Bifidobacteria mixture (B. longum BB536, B. infantis M-63, and B. breve M-16 V) was able to significantly improve AR symptoms and quality of life in children with pollen-induced AR and intermittent asthma (39). Another study showed that probiotic NVP-1703 (a mixture of B. longum and L. plantarum) relieves AR symptoms by prompting Treg cells to release IL-10 (42). However, there was a high heterogeneity from metaanalysis of two studies, which may be related to the use of different probiotics. The different strains of probiotics, doses, and durations may be the sources of heterogeneity. To date, no serious adverse events have been observed for probiotic treatment; thus, it appears to be safe.
To sum up, probiotic supplement seems to be effective in ameliorating AR symptoms and improving the quality of life, but there is high heterogeneity in some results after subgroup analysis, and clinicians should be cautious when recommending probiotics in treating AR.
There are some limitations in this meta-analysis. First, the sample size of some included RCTs was small. Second, airborne pollen concentrations are associated with symptom severity and recovery in patients with SAR. The pollen concentrations varied due to different regions in different trials. This is a source of clinical heterogeneity.       *The basis for the assumed risk (e.g., the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI, Confidence interval. GRADE Working Group grades of evidence. High quality, Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality, Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality, Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality, We are very uncertain about the estimate. 1 In some studies, random sequence generation, allocation concealment, and blinding of participants and personnel are not described. 2 There is a significant heterogeneity (I 2 > 50%). 3 PICO is not exactly the same.