To our best knowledge, this is the first meta-analysis that investigated the core symptoms as well as behavioral and emotional problems of ASD. A previous meta-analysis, which only focused on the overall behavioral symptoms of ASD, failed to show significant improvement in subjects with ASD treated with probiotics compared to the control group [18]. By contrast, the results of the present meta-analysis, which included ten RCTs with 522 participants, demonstrated significant improvement in the overall behavioral symptoms in the probiotics group compared with the control group but without notable beneficial impact of probiotics on the core symptoms of ASD including social interaction, communication, and RRB. Our secondary analyses further showed significant improvement in the adapted behaviors in the probiotics groups compared to the controls as well as a non-significant trend of probiotics-related improvement in anxiety symptoms. On the other hand, no significant difference was noted in other secondary outcomes including irritability/aggression, hyperactivity/impulsivity, inattention, and parental stress between the probiotics and control groups. Overall, our study supported an association of probiotics use with improvement in the behavioral but not the core symptoms of ASD. Besides, the use of probiotics correlated with significantly improved adapted behaviors but only a trend of improvement in anxiety.
In contrast to our finding of significant improvement in the overall behavioral symptoms of ASD in the probiotics group compared to the controls, a previous meta-analysis did not show a significant difference between the two groups despite demonstration of a trend in favor of the former [18]. The inclusion of up to ten RCTs in our investigation compared to seven studies in the previous meta-analysis [18] may contribute to the difference in findings. Several mechanisms may explain the association between probiotics use and the observed alleviation of ASD symptoms [5]. One of the key hypotheses involves the gut-brain-axis involving a network of neuroendocrine pathways that enable a bidirectional communication between intestinal microbiome and the central nervous system [5]. Consistently, a previous study has identified the vagus nerve as the route of communication between intestinal microbiome and the central gamma-aminobutyric acid system [36]. In addition, given the known association between systemic inflammation and neurocognitive impairment [37, 38], the anti-inflammatory properties of certain probiotics [28, 29] may be beneficial to the maintenance of neurocognitive functions. Moreover, a prior experimental study using a mouse ASD model has demonstrated a correlation of an abnormal increase in intestinal mucosal permeability (e.g., dysbiosis) with the triggering of the systemic inflammation as well as abnormal neurotransmitter signaling in the brain [39]. Adopting the same ASD model, another study further demonstrated a normalization of such an increased intestinal mucosal permeability through the oral administration of the probiotic B. fragilis [8]. Nevertheless, although our results are supported by prior studies that suggested a possible modulating role of probiotics in GBA which has been reported to be associated with the behavioral symptoms of ASD, evidence derived from the current study is still not solid enough given the limited number of available trials.
Consistent with the results of a previous meta-analysis that showed significant improvements in the overall behavioral symptoms in subjects diagnosed with ASD treated with probiotic blends compared with controls [18], our subgroup analysis focusing on the use of single- versus multiple-strain probiotics demonstrated a significant improvement in the overall symptoms of ASD only in the multiple-strain probiotics group compared to the control group. In concert with our finding, a prior animal study suggested the merit of multiple-strain probiotics due to an increased opportunity of adhesion of beneficial microbiota to intestinal mucosa [19]. Besides, another meta-analysis reported the effectiveness of using multiple- instead of single-strain probiotics in the prevention of necrotizing enterocolitis and mortality in preterm infants [40]. Taken together, our findings and those from a previous meta-analysis [18] supported the use of multiple-strain probiotics for the alleviation of ASD symptoms. Nevertheless, the fact that only five studies were available for subgroup analysis suggests the need for further large-scale studies to address this issue.
Because most RCTs in our study used probiotics as dietary supplements except two that combined probiotics with other therapeutic approaches (i.e., probiotics plus bovine colostrum product or probiotics plus applied behavior analysis) [34, 35], we conducted subgroup analysis of studies using probiotics only as supplements that consistently showed greater improvement in the overall behavioral symptoms of ASD in the probiotics group than that in the control group. The results, therefore, suggested that probiotic supplementation without combining with other treatments may be effective in this setting, although more clinical investigations are required to validate our findings due to the limited number of RCTs and the small ES in our study.
Despite the significant overall behavioral improvement in subjects diagnosed with ASD treated with probiotics compared with the controls, no significant difference in improvement was noted in the three core symptoms of ASD (i.e., social functioning, communication, and RRB) between the two groups. There are several possible reasons for this observation. First, the overall improvement in behavioral symptoms may be attributed to the collective minor improvements in each core domain of behavioral problems. However, statistically significant differences may be obscured by the small number of RCTs available for analyzing each core behavioral symptom. Second, the lack of available information about changes in the core behavioral symptoms in two RCTs that provided findings in favor of probiotics in the treatment of the overall symptoms of ASD [11, 35] may render the improvement in core symptoms non-significant because of the absence of their probable positive contributions. Third, the observed behavioral improvement may stem from other associated symptoms of ASD such as mood or irritability. Taken together, our study could not provide robust evidence either to support or dismiss the effectiveness of probiotics for each core behavioral symptom of ASD due to the small numbers of available trials (i.e., a maximum of eight RCTs for each core behavioral symptom). Future studies are warranted to elucidate this issue.
Despite the lack of significant probiotics-associated improvement in the core behavioral symptoms of ASD, our results on secondary outcomes showed a significant improvement in adaptive behaviors assessed mainly by the vineland adaptive behavior scales in subjects treated with probiotics compared with controls. Consistently, prior RCTs have reported an association between the use of probiotics and cognitive function improvement [41, 42], which may be attributed to the systemic anti-inflammatory properties of certain probiotics [37, 38], that may be protective against neurocognitive impairment given the known negative effect of systemic inflammation on neurocognitive functions [28, 29]. Although we were unable to evaluate the result of cognitive function with standardized tools such as Wechsler Intelligence Scale for Children (WISC) due to a lack of available data, our finding of improved adaptive functions may imply a beneficial influence of probiotics on overall adaptation. However, the limited number of available RCTs (n = 3) suggests the need for further studies based on more objective cognitive assessment (e.g., WISC) to verify our findings.
Analyses of our secondary outcomes focusing on other associated problems of ASD showed an apparent trend of improvement in anxiety symptoms in individuals treated with probiotics compared with the controls despite a lack of statistical significance, while no difference was noted in other behavioral issues of ASD including irritability/aggression, inattention, hyperactivity/impulsivity, and parental stress between the two groups. Previous research has demonstrated a correlation between intestinal microbiome and mood regulation [43]. Compared with the intestinal microbiota composition in healthy individuals, prior investigations have also shown increased levels of the phyla Bacteroidetes, Proteobacteria, and Actinobacteria but a reduced level of Firmicutes in those diagnosed with major depressive disorders [44]. One proposed hypothesis is the mood-modulatory effect of probiotics through the microbial-gut-brain axis [45]. Despite the reported superiority of probiotics over placebos in anxiety relief from previous RCTs [46, 47], a recent meta-analysis showed effectiveness of probiotics only against depressive symptoms but not anxiety in patients with anxiety- or depression-related diagnoses [48]. Since one out of the three included RCTs used a measurement tool for both depression and anxiety [13] but the remaining two used tools for only anxiety [10, 14], the mixed results from both depression and anxiety may influence the specificity of our outcomes. Nevertheless, given our finding of a nearly significant effect and the positive results of a previous meta-analysis for mood symptoms [48], more investigations into the therapeutic efficacy of probiotics in improving mood symptoms specifically focusing on anxiety or depression in patients with ASD are warranted.
Several limitations in the current study need to be taken into consideration. First, notwithstanding our inclusion of up to ten trials and a total of 522 participants, the results were still not robust enough to provide solid evidence. Moreover, our results on some core or associated symptoms of ASD, which were derived from limited numbers of available trials (i.e., only two trials for inattention), require validation from future studies. Second, certain heterogeneities in treatment strategies such as the use of probiotics as supplements or part of combination therapy may be potential confounding factors that influence the therapeutic outcomes. Nevertheless, our consistent findings on subgroup analysis after excluding studies using probiotics as part of combination therapies indicated minimal influence of this confounder. Third, a number of other factors that may affect the therapeutic outcomes of probiotics, such as dietary habits or the use of other nutritional supplements, were unaccounted for due to a lack of relevant information for meta-regression or subgroup analysis. Third, although gastrointestinal symptoms are relatively common in individuals diagnosed with ASD [4], most of the included studies did not provide related information so that the possibility of improvements in behavioral symptoms secondary to an alleviation of gastrointestinal symptoms cannot be ruled out. Finally, two studies that did not use placebo control may be more susceptible to performance and detection bias [34, 35]. Nevertheless, our sensitivity test demonstrated consistent results after excluding those two studies.