The gut microbiota (GM) is a complex ecological community that influences host physiology and susceptibility to disease through its metabolic activities and interactions with many systems and functions of the host (1). The human microbiota is estimated to contain 10¹³-10¹⁴ microbial cells, with a 1:1 microbial to human cells ratio (2), and is located predominantly in the large intestine of the gastrointestinal (GI) tract (3).

The microbiota normally establishes a mutually beneficial cohabitation with the host: it protects against pathogens, extracts nutrients and energy from the diet, and contributes to the maintenance of the immune response (4). Alterations of the GM, caused either by environmental or host-related factors, can impair its ability to maintain good health and lead to dysbiosis. Dysbiosis, defined as an “imbalance” between GM and host, is characterized by a reduction in microbial diversity and/or an increase in proinflammatory species (5) and has been associated with a plethora of gastro-intestinal-extra-intestinal disorders, such as inflammatory bowel diseases (IBDs), obesity, cancer, malnutrition, and neurological disorders (5).

Several evidence have reported a bidirectional interaction between the GI tract and the central nervous system (CNS). Communication occurs through the autonomic nervous system, the enteric nervous system, the neuroendocrine and neuroimmune systems, and the GM itself, which modulates neurotransmitter production, influences brain development and myelination patterns, and produces metabolites able to impact behavior in mice (6, 7). Indeed, the concept of the brain-gut axis (GBA) has been extended to the microbiota-gut-brain axis (MGBA) (8), to account for the complex networks existing between the GM and the CNS. Although previous studies reported a link between GM colonization and the CNS (9), recent evidence showed that dynamic microbiota-host interaction may alter brain-gut signaling and increase the risk of neurodevelopmental disorders. The concept of critical parallel-interacting windows in neurodevelopment and GM colonization provides new ways to develop novel therapeutic interventions modulating the microbiota to fight neurodevelopmental deficits and brain disorders.

Several studies have reported GM variation in children with neurological impairment (NI). Lee et al. (10) reported that children with intractable epilepsy (IE) had a significantly less variegate GM than healthy controls; Huang C et al. (11) described distinct GM composition and functions in children with cerebral palsy and epilepsy (CPE). These shreds of evidence suggest that GM provides a therapeutic target for the treatment of neurodevelopmental disorders.

On the other hand, feeding and diet, together with gastrointestinal disorders linked to neurological disability, may in themselves contribute to the immaturity and dysfunction of the GM, an imbalance in nutrient/energy intake, and, ultimately, malnutrition.

Nutritional management and treatment of gastrointestinal symptoms in children with NI have become a challenge in the past decades. In these children, malnutrition can occur frequently due to inappropriate dietary energy intake, oral motor dysfunction, increased nutrient losses, increased basal metabolic rate, and physical exertion, leading to the need for artificial nutritional support to ensure growth, counteract morbidity and mortality, and improve overall health and quality of life for both patients and their families (12–14). The European Society of Gastroenterology, Hepatology, and Nutrition (ESPGHAN) position papers clarified the criteria for the diagnosis of oropharyngeal dysfunction, the assessment, and monitoring of nutritional status, the standardization of nutritional requirements, and the gastrointestinal management of children with NI (14). The assessment of nutritional status and continuous monitoring are mandatory to prevent overfeeding or underfeeding (12–14). Nutritional support is normally provided enterally [enteral nutrition (EN)] rather than parenterally, assuming competency of the gastrointestinal tract (12–14).

Long-term EN (LTEN), especially when used exclusively, can deeply affect the GM taxonomic composition and functions. In general terms, LTEN implies a profound change in the provision of nutrients to the gut and this, together with metabolic stress and the use of drugs, normally administered to these patients, is responsible for a marked dysbiosis (15, 16).

A number of reports have addressed changes in GM composition following the intervention of exclusive enteral nutrition (EEN) as the first-line therapy in the setting of IBDs in children (17). Scarce data are available for the effects of EN in other pediatric settings (18). For example, examinations of intestinal microbial diversity were conducted in children with NI performing blenderized enteral nutrition (BEN). A significant improvement in gut bacterial diversity following BEN was reported in a sample of twenty children with rapidly progressive degenerative disorders. The reduction of Proteobacteria in the feces showed the benefit of BEN in preserving a healthy gut microbiota (19). Although increased gut microbial diversity may improve gastrointestinal mobility, the use of BEN still raises concerns and is not currently the first choice (13, 14, 20).

In this study, using amplicon sequencing, we investigated the fecal microbiota composition in a pediatric population suffering from severe NI and under exclusive LTEN and compared it to sex- and age-matched controls. A correlation between GM and metabolic profile, focusing on insulin resistance and glycemic derangement, was also considered.

Inadequate nutritional status and disability are intimately linked (40). Malnutrition can occur in children with NI as a result of imbalance in nutrient/energy intake, oral motor dysfunction, increased loss of nutrients and basal metabolic rate, and physical inactivity (25, 41), necessitating long-term dependence on EN for providing the entire (or partial) caloric intake (15, 16, 25).

A 2021 ESPGHAN position paper aimed to standardize the nutritional and gastrointestinal management of children with NI (14). Initiation of EN is recommended in cases of inefficient, unsafe, and/or stressful oral energy and fluid intake, and gastrostomy placement is considered the primary way to provide enteral access. A standard EN with a standard polymeric formula (1 kcal/ml) suitable for age is a common choice. The ESPGHAN recommendation advises caution in the administration of home mixed diets due to concerns about their safety and nutritional adequacy to maintain an adequate nutritional status (14).

Diet is a strong modifier of the GM and its gene content, especially in children, whose gut bacterial communities are subject to rapid changes. Strong evidence links the microbial composition of the gut to situations of malnourishment in children, due to the pivotal role in extracting and metabolizing dietary ingredients (42, 43). Indeed, the composition and functional capacity of the gut microbiota appear to be altered in childhood undernutrition (44). In particular, an “undernourished microbiota” has been described as “immature” based on an altered diversity, a less efficient nutrient utilization, and strong compositional differences in specific bacterial taxa (e.g., enrichment in pathobionts and inflammogenic species, often aerotolerant and belonging to Proteobacteria), as compared with age-matched children exhibiting a consistently healthy growth (43, 45).

On the other hand, data on different GM compositions in children with neurological damage have been reported. Lee et al. (10) reported that children with intractable epilepsy (IE) had significantly greater diversity in GM than healthy controls. The species biomarkers for IE included the Enterococcus faecium group, the Bifidobacterium longum group, and Eggerthella lenta. Low levels of Bacteroidetes and high levels of Actinobacteria were also found in epileptic children compared with healthy controls. Huang et al. (11) published a case-control study showing significant differences in GM composition between children with cerebral palsy and epilepsy (CPE) as compared with a healthy group. In detail, significantly increased relative abundances in Bifidobacterium, Streptococcus, Akkermansia, Enterococcus, Prevotella, Veillonella, Rothia, and Clostridium (IV) and, on the other hand, significant reductions in Bacteroides, Faecalibacterium, Blautia, Ruminococcus, Roseburia, Anaerostipes, and Parasutterella characterized the CPE group.

In this study, we investigated, for the first time to the best of our knowledge, the consequences of LTEN in a neurologically impaired pediatric population, focusing on the GM.

In agreement with what was described above, the overall comparison of GM composition in NI children under LTEN and age- and sex-matched subjects revealed a clear clustering of the two groups, as a reflection of profound differences in the bacterial populations. In general, the differential taxonomic picture in NI children under LTEN seems to mirror a profound dysbiotic condition, in which protective taxa as Short-Chain Fatty Acids (SCFA) producers and anti-inflammatory taxa appear severely depleted (e.g., the Clostridiales families of Lachnospiraceae and Ruminococcaceae), while known pathobionts (Gammaproteobacteria and Klebsiella) or emerging pathogens (Synergistales and Cloacibacillus, Fusobacterium) are significantly enriched. In general terms, this picture is reminiscent of gut microbiota changes found in inflammatory disorders, as well as in malnourished children, as will be better detailed above.

Moving on to more details on the differential taxa, an interesting point is the strong depletion of the genus Faecalibacterium (family: Ruminococcaceae) whose only known species is Faecalibacterium prausnitzii. Normally, this is an abundant species, and one of the main butyrate producers, with protective and anti-inflammatory properties, and the capacity of enhancing the intestinal barrier function (46). Its abundance is reduced in different intestinal disorders (type 2 diabetes, colorectal cancer, and the two main IBDs: Crohn's disease (CD) and ulcerative colitis (UC), and its role in promoting gut health is increasingly recognized, to the point that it has been suggested to consider its monitoring as a biomarker to assist in gut diseases diagnostics and prognostics (46). Other SCFA producers belonging to Firmicutes, known to diminish in many inflammatory disorders among which pediatric IBD, appear depleted, e.g. Lachnospiraceae and Gemmiger (family: Ruminococcaceae), another genus of commensal bacteria linked to a healthy gut status (47). Concerning Proteobacteria, many concordant pieces of evidence link an unstable gut microbiota community enriched in this phylum (rich in aerotolerant and pro-inflammatory taxa, e.g., those belonging to Enterobacteriaceae) to a plethora of infective and inflammatory conditions, and undernutrition (48, 49). Their outgrowth is generally recognized as a sign of microbial dysbiosis in the gut (49). Another interesting point is the enrichment in the NI microbiota of emerging pathogens: (i) Cloacibacillus, belonging to the Synergistetes phylum (50) and considered an intestinal opportunistic pathogen; (ii) Fusobacterium, able to strongly activate host inflammatory responses and linked to several diseases, as well as a promoter of colorectal carcinogenesis (51).

The dysbiotic picture that emerges for these taxa is strongly reminiscent of what is repeatedly reported in the literature for children malnourished as a consequence of inadequate food intake, often coming from low-income Western African and Asian countries (42, 43): the strong depletion of several gut obligate anaerobes belonging to Firmicutes, Bacteriodetes, and Actinobacteria; and the enrichment in aerotolerant taxa, many of which Proteobacteria, more competitive and starvation resistant, and able to metabolize broader classes of substrates, such as amino acids (44, 45, 52). As for the link to our results, it is to be noted that the depletion of Faecalibacterium was found to be the most highly discriminatory bacterial taxonomic biomarker to distinguish malnourished children from age-matched controls exhibiting healthy growth (42). The “anaerobic depletion” also encompasses Bifidobacterium, whose reduction has been described as one of the first steps in gut microbiota alterations associated with severe acute malnutrition (45).

It is interesting to underline that, while most of these data come from pediatric populations in low-income countries, we retrieve comparable taxonomic biomarkers and dysbiosis pictures in a completely different context, regardless of geographic provenience, lifestyle, and therapeutic assessment.

Another important point is that data on the effects of EN on gut microbiota are scarce, and, at least for pediatric populations, generally limited to the use of exclusive EN as the first-line therapy in children neo-diagnosed with Crohn's disease and not always consistent among studies (53). Our article adds new results for a completely different pediatric setting.

Finally, it is interesting to cite a previous work investigating the effects of parenteral nutrition (PN) on gut microbiota structure (52). The authors retrieved similar changes in community structure, in terms of decreased relative abundance of Firmicutes and increased percentages of Bacteroidetes and Proteobacteria.

We are aware that our study has some limitations. First of all, a relatively small sample size could limit the power of the analysis, even if the differential clustering and bacterial taxonomic pictures of the two groups are supported by strong statistical evidence. Second, we considered a cohort of children with a severe grade of disability that could in itself contribute to the severity of dysbiosis; a comparison with a group of children with moderate disabilities would be useful in future studies. Finally, long-term anticonvulsant drugs may be associated with GM alterations. However, all the subjects were on anticonvulsant medications and this factor can be considered as a systematic error. Despite these limitations, the present study presents a first picture of the relevant modifications in GM taxonomic composition found in NI children and adolescents under LTEN. This could represent the premise of a broader process for improving the management of these children, as it implicitly suggests possible routes and targets to try to alleviate the observed dysbiosis, and improve the overall clinical conditions, in primis, the gastroenterological problems, which are often considerable (54). A relevant example could be F. prausnitzii, one of the most heavily depleted taxa, whose use as a probiotic species is currently under evaluation. Other two strategies that could potentially be explored are the supplementation with prebiotics able to enhance populations of SCFA producers, with a special focus on the anti-inflammatory butyrate, of which F. prausnitzii is a major producer, or giving the postbiotic butyrate directly.

Finally, further research is desirable, especially to establish whether dysbiosis may be a primary manifestation of a secondary consequence of chronic impaired health and malnutrition.

Our results suggest that LTEN can influence the GM taxonomic composition in neurologically impaired children. However, the role of bidirectional interaction between GI impairment/immaturity and the CNS in disabled children probably remains a key player in dysbiosis. Additional factors, such as drugs and physical inactivity, could not be excluded.

The physiologic, metabolic, and immunologic consequences of changes in GM and how they might contribute to associated morbidities and sequelae of malnutrition still need to be fully elucidated. New insights are presented here and addressed in patients that are not usually investigated in this sense.

Future interventions based on dietary modulations of the GM, e.g., through biotics-based interventions, to support a healthy microbiota in these children, should be investigated as a tool to help to combat their malnourishment and improve their general conditions, clinical outcomes, and quality of life, even if keeping in mind that restoration of health status may be incomplete. In this regard, longitudinal studies would certainly be useful to clarify the long-term efficacy of nutritional interventions. Additionally, interventional studies before and after a planned EN will be useful to define the specific role of EN on dysbiosis.

The data presented in this study have been deposited in the NCBI BioSample database, BioProject ID: PRJNA842554.

The studies involving human participants were reviewed and approved by ARNAS Civico-Di Cristina-Benfratelli, PA Register No. 354 Civico 2016. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin.

SP and VC participated in the study design and project management. SP, VC, EV, FC, GP, EB, CB, and GZ were responsible for the conceptualization and design of forms, data management and quality control, and writing and editing the manuscript. SP, FC, and CB were responsible for the statistical analysis and documentation. VC, GP, and EB were responsible for the recruitment of participants. SP, VC, EV, GP, CB, and GZ participated in the study supervision. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.