Comparison of the gut microbiome composition among individuals with acute or long-standing spinal cord injury vs. able-bodied controls

Abstract

Objective: Compare the gut microbiome composition among individuals with acute spinal cord injury (A-SCI), long-standing SCI (L-SCI), vs. able-bodied (AB) controls.

Design: Cross-sectional study.

Setting: The University of Alabama at Birmingham.

Participants: Seven adults with A-SCI (36 ± 12 years, 2F/5M, C4-T10, and American Spinal Injury Association Impairment Scale [AIS] A–D), 25 with L-SCI (46 ± 13 years, 6F/19M, C4-L1, and AIS A–D), and 25 AB controls (42 ± 13 years, 9F/16M).

Methods: Stool samples were collected after a median of 7 days and 18 years after injury in the A-SCI and L-SCI groups, respectively. Gut microbiome composition was analyzed using the 16S rRNA sequencing technique and QIIME software. The abundances of bacteria communities among groups were compared using the Kruskal–Wallis test adjusted for age.

Results: Several alpha diversity indices were different among groups (Chao1, Observed species, and Phylogenetic Diversity), but not others (Shannon and Simpson). Beta diversity differed among each pair of groups (P < 0.05). A number of microbial communities were differentially abundant among the groups (P < 0.05).

Conclusion: Our results revealed differences in the gut microbiome composition among groups. Compared to the AB controls, the SCI groups demonstrated microbiome profiles that shared features linked to metabolic syndrome, inflammation-related bowel disorders, depressive disorders, or antibiotics use, whereas the L-SCI group’s microbiome included features linked to reduced physical activity compared to the A-SCI and AB controls. Our results provided preliminary data and a scientific foundation for future studies investigating the impact of the gut microbiome composition on long-term health in individuals with SCI.

Introduction

The human gastrointestinal tract is colonized by 100 trillion bacteria, both beneficial and harmful, the balance of which is pivotal for maintaining human health.¹ The gut microbiome assists in a number of essential physiological functions, including, but not limited to, digestion, nutrient production and absorption, immune system development, and hormone production.² Considering its ability to produce and regulate numerous circulating compounds that influence the functions of distal organs and systems, in many respects, the gut microbiome resembles an endocrine organ.²

The composition and function of gut microbial communities are fluid, modulated by a plethora of environmental (diet, medication use, physical activity, etc.) and host genetic factors. Although the gut microbiome is a promising area of research in the able-bodied (AB) population, gut microbiome-related research in individuals with spinal cord injury (SCI) is scarce.^3,4 It was shown in animal models that at the onset of SCI, due to the partial or complete loss of central nervous system control over the gastrointestinal tract, neurological bowel dysfunction and altered colonic transit time occur, which disrupt gut microbiota composition (i.e. gut dysbiosis [GD]).⁵ Gut dysbiosis is commonly defined as the depletion of beneficial, nonpathogenic gut bacteria and coincidental increase of pathogenic bacteria. Notably, recent research suggests that SCI-induced GD could create a feedback loop that 1) hinders neurological recovery and 2) accelerates the development of secondary health conditions after SCI. For example, Kigerl et al.⁶ showed that antibiotics-induced GD impaired functional recovery and aggravated lesion pathology after subacute SCI in a mouse model.⁶ In addition, GD, as a result of SCI, has been shown to be associated with systemic inflammation and inflammation of the intestine,^6–8 increased intestinal permeability (leaky gut), and bacterial translocation from the gut into distal organs^6,9 in animal models. Leaked microbial products and associated systemic inflammation may collectively contribute to the development of urinary tract infections, impaired metabolic health, and other secondary conditions in SCI.^2,10,11 Lastly, the frequent use of antibiotics,¹² physical inactivity,¹³ and depression,¹⁴ commonly experienced by individuals in the chronic stages of SCI, may further disrupt the gut microbiome composition, progressively augmenting the negative influences of GD on metabolism, overall health, and quality of life.

Only limited research examined the gut microbiome composition in human SCI model and demonstrated the presence of GD among individuals with long-standing SCI (L-SCI) compared with AB controls. In addition, no human study has evaluated the gut microbiome composition among patients with acute SCI (A-SCI) in a clinical setting, where patients were prescribed antibiotics not only as treatment, but also as prophylactic measures. Patients with L-SCI often have indwelling or intermittent urinary catheters and are prone to have asymptomatic bacteriuria and urinary tract infections. As a result, they frequently receive antimicrobial therapy. SCI-induced GD coupled with intensive broad-spectrum antibiotic use may disturb normal gut microbiota including pathogenic bacterial colonization in the gut.¹⁵ To address this gap in current literature, we used 16S rRNA gene sequencing to evaluate our hypothesis that gut microbiome composition differs between individuals with A-SCI, L-SCI, and AB controls. We also aimed to explore the microbial communities that are different among these 3 groups.

Materials and methods

Participants recruitment

The study population included individuals with A-SCI or L-SCI and AB controls. There were seven adults (2F and 5M, 36 ± 12 years old) with A-SCI (injury levels C4-T10), 25 adults with L-SCI (6F and 19M, 46 ± 13 years old, injury levels C4-L1), and 25 AB controls (9F and 16M, 42 ± 13 years old) that were age-matched, generally healthy individuals with no known chronic conditions. Individuals with A-SCI were taken to acute care after injury; neurological examinations were performed within 72 h of injury (after the initial trauma evaluation and resuscitation); and stool and blood specimens were collected at a median of 7 days (range: 4–11 days) after injury. To limit the acute impact of antibiotics on gut microbiome composition, stool samples were collected at least 4 weeks after any antibiotics use among participants with L-SCI and AB controls. A 4-week period was chosen as research showed that the taxonomic composition of the community closely resembled its pretreatment state by 4 weeks after the end of antibiotics treatment¹⁶ and has been used in other human research studies.^3,17 However, the avoidance of antibiotics for those with A-SCI was not feasible.

For the A-SCI group, all neurological tests were performed by the same American Spinal Injury Association (ASIA) International Standards Training e-Learning Program-certified physical medicine and rehabilitation clinician. Each participant’s level of injury was documented using the ASIA Impairment Scale [AIS]. The sacral sparing definition was used to define the completeness of injury.¹⁸ Individuals with A-SCI were originally enrolled in a completed SCI clinical trial (NCT03509571, removed for blinding purposes).¹⁹ Individuals with L-SCI were identified from a list of individuals who are enrolled in an ongoing SCI Model Systems (SCIMS) clinical trial (NCT03204240) at the University of Alabama at Birmingham (UAB)-SCIMS and currently reside in the Greater Birmingham area. The level of injury was retrieved from their medical records. Able-bodied participants were recruited throughout the university and the surrounding communities as previously reported.²⁰ All studies were approved by the UAB Institutional Review Board. All participants provided written informed consent after receiving a thorough explanation of study procedures and risks and the opportunity to ask questions. Monetary compensation was provided to all participants.

Microbiome analysis

Stool collection and microbiome analysis were performed using established protocols.²¹ Briefly, stools samples were collected in Para-Pak vials (Meridian Biosciences, Inc; Cincinnati, OH) during acute care at the hospital (A-SCI group). For the L-SCI group, participants or their caregivers were provided with a stool collection kit (a Para-Pak vial for sample preservation, a stool collection container, and a 2-gallon Ziploc bag for disposal of collection container). Stool collection was performed at participants’ homes by participants or their caregivers during their regular bowel program. The Para-Pak vials containing stool samples from participants were picked up by FedEx on the day of collection and shipped back (at ambient temperature) in a FedEx Clinical Pak package to our facility using FedEx Standard Overnight service (arriving by 3:00 pm the next day). The Para-Pak contains a non-nutritive solution that preserves the microbial composition in the stool. Upon delivery, each sample was diluted to 0.1 mg/mL in Cary-Blair medium for a total volume of 20 mL with 10% glycerol (by volume). Aliquots of 5 mL were dispensed into cryovial tubes and stored at −80°C until DNA extraction.

Stool bacterial DNA was extracted using a Zymo Research Fecal DNA isolation kit (Zymo Research; Irvine, CA) per the manufacturer’s instructions. Polymerase chain reaction (PCR) was then used to amplify the V4 region of the 16S rRNA gene, and the PCR products were separated on an agarose gel using electrophoresis, excised from the gel, and subsequently purified with a QIAquick Gel Extraction Kit (Qiagen; Germantown, MD). The Illumina MiSeq DNA sequencing platform was used.

Bioinformatics analysis

Microbiome analyses, including quality control, sequence clustering, and phylogenic diversity analyses were performed using the Quantitative Insights into Microbial Ecology (QIIME) bioinformatics software and QWRAP 4 program as previously described.^21,22 Briefly, the quality of the raw data was assessed using FASTQC with low-quality data filtered out using the FASTX toolset. A combination of tools within the QIIME suite was utilized for Amplicon Sequence Variants (ASVs) clustering, taxa assignment (SILVA 16S rRNA gene database)²³, and as necessary, alignment and phylogenetic inference using PyNAST and FastTree. These procedures allowed us to quantitatively assess the microbiome population down to the genus, and frequently species, level. For comparative analyses, we determined several metrics for alpha and beta diversity.²¹ The following alpha diversity indices were calculated as measures of diversity within each group: 1) Chao1, an estimate of the species richness; 2) Observed species, a measure of unique ASVs in the sample; 3) Phylogenetic Diversity; 4) Shannon’s index, a measure of both richness and evenness; and 5) Simpson’s index, a measure of both richness and evenness, but less sensitive to the presence of rare species when compared to Shannon’s index. Beta diversity, a measure of between-sample similarity, was calculated using a non-phylogeny based (Bray–Curtis dissimilarity), as well as, phylogeny-based (unweighted UniFrac and weighted UniFrac) methods. The unweighted matrix depends on the presence and absence of ASVs between samples and their phylogenic distances, whereas the weighted matrix further accounts for the abundance of the bacteria. Principal coordinate analysis (PCoA) was performed to visualize the beta diversity distance matrices.²¹ Permutational multivariate analysis of variance testing was used to assess the differences in beta diversity between each pair of the 3 groups.

The abundances of bacterial communities among groups were compared in SAS 9.4 (SAS Institute, Inc., Cary, NC). Specifically, in order to adjust for age, we applied ordinary least squares regression to the ASV abundances. Utilizing the residuals (the ASV abundance not explained by age), we performed the Kruskal–Wallis test to compare the abundances among the 3 groups. A false discovery rate (FDR) of 0.1 was applied. If an ASV met an FDR of 0.1, we conducted post-hoc pairwise comparisons using the Neyemini (1963)²⁴ method cited by Wilcoxon and Wilcoxon (1964)²⁵ and illustrated in Zar (2010).²⁶ For pairwise comparisons, the threshold of statistical significance was defined as P < 0.05.

Results

Participant characteristics

Participant’s characteristics, including age, sex, level and completeness of injury, antibiotics usage, and time of sample collection, are shown in Table 1. Approximately 30% (2 of 7) of participants in the A-SCI group, 24% (6 of 25) of participants in the L-SCI, and 36% (9 of 25) in the AB control group were females. Age did not differ among groups (AB: 42 ± 13, A-SCI: 36 ± 12, L-SCI: 46 ± 13 y, P = 0.16). Six of the seven participants with A-SCI were prescribed various types and doses of antibiotics due to the complications associated with their injuries, such as pneumonia and urinary tract infections (information about the type and dose of antibiotics prescribed are shown in Supplementary Table 1). Stool samples were collected after a median of 7 days (min 4, max 11) and 18 years (min 3, max 53) after the onset of injury from participants with A-SCI and L-SCI group, respectively.

Table 1. Participants’ characteristics.

Group*	AB Controls	A-SCI	L-SCI
Total, n	25	7	25
Age, y	42 ± 13	36 ± 12	46 ± 13
Sex (Female/Male)	9F/16M	2F/5M	6F/19M
Level of Injury$, n	N/A	C: 5 T: 2	C: 6 T: 17 L: 2
Severity of Injury, n	N/A	AIS A: 2 AIS B: 1 AIS C: 3 AIS D: 1	AIS A: 17 AIS B: 2 AIS C: 2 AIS D: 4
Time of stool collection post-injury	N/A	7 days; min 4, max 11	18 years; min 3, max 53
Antibiotics Use	0	n=6	0

*AB, able-bodied; A-SCI, acute spinal cord injury; L-SCI, long-standing spinal cord injury; AIS: American Spinal Injury Association Impairment Scale.

^$C, cervical; T, thoracic; L, lumbar level;

Microbial analyses—overall microbiome composition, beta-diversity, and alpha-diversity

Bray–Curtis dissimilarity, unweighted, and weighted UniFrac beta diversity differed significantly between each pair of the 3 groups studied (P < 0.05 for all) with the exception of unweighted UniFrac index between the A-SCI and L-SCI groups (P = 0.11). The PCoA showed a clear separation between each pair of the 3 groups based on Bray–Curtis dissimilarity (Figure 1). With or without adjusting for age, alpha diversity indices such as Chao1 and Observed species, are higher in the SCI groups than controls, with the A-SCI group being highest (P < 0.05). However, Phylogenetic Diversity (PD_whole_tree), Shannon, and Simpson indices, the latter two of which take richness and evenness into consideration, were not different among groups (Table 2), suggesting that the A-SCI group had more unique bacteria communities, yet these communities are not well-represented (i.e. low abundances).

Figure 1.

The principal coordinate analysis of the gut microbiota of participants using the Bray–Curtis index (Red: able-bodied, Black: L-SCI, Blue: A-SCI groups).

Table 2. Comparison of Alpha diversity among groups.

	AB	A-SCI	L-SCI	P#
Chao1	273.4 ± 13.9a	447.9 ± 25.7b	345.8 ± 13.6c	0.045
Observed species	215.6 ± 13.6a	402.7 ± 25.2b	289.5 ± 13.3c	0.046
Phylogenetic Diversity (PD_whole_tree)	17.8 ± 0.7	23.5 ± 1.3	17.7 ± 0.7	0.25
Shannon	4.1 ± 0.2	5.1 ± 0.4	4.7 ± 0.2	0.30
Simpson	0.82 ± 0.03	0.93 ± 0.05	0.89 ± 0.03	0.26

Data are least-square means and their standard error estimated by the ANCOVA model.

^a,b,c: Columns with different letters are significantly different at a significance level of P < 0.05.

^#The significance of group effect upon adjusting for age.

Differences in microbial abundances among AB, A-SCI, and L-SCI groups

Further analysis of ASVs identified several bacteria communities that were differentially abundant among groups (Table 3). Specifically, compared with the AB control group, both the SCI groups had higher abundances of the families Erysipelotrichaceae, Acidaminococcaceae, Rikenellaceae, and genera Lachnoclostridium and Eisenbergiella of the Lachnospiraceae family, Alistipes of the Rikenellaceae family, Oscillibacter and Anaerotruncus of the Ruminococcaceae family. Furthermore, the A-SCI and L-SCI differed from the AB control group uniquely. The L-SCI group had higher abundances of the order Clostridiales, families Lachnospiraceae and Eggerthellaceae, and lower abundances of genus Campylobacter of the Campylobacteraceae family, order Bacillales than the AB controls, and the A-SCI group have higher Desulfovibrionaceae family than the AB control. The A-SCI group had higher abundances of genus Sutterella of the Burkholderiaceae family, Marinifilaceae family, and its genus Odoribacter than the AB control and L-SCI groups. Lastly, L-SCI had lower Burkholderiaceae family than the AB and A-SCI groups.

Table 3. Bacteria communities with differential abundance among groups categorized by their putative roles in the literature.

Bacteria	AB control group#	A-SCI group#	L-SCI group#	A-SCI vs AB*	L-SCI vs AB*	A-SCI vs L-SCI*
SCI Animal Model6
O_clostridiales	29.7%	33.7%	52.9%		L-SCI > AB
Metabolic Health 27–30
F_Acidaminococcaceae	0.3%	2.0%	0.8%	A-SCI > AB	L-SCI > AB
F_Acidaminococcaceae; G_Phascolarctobacterium	0.1%	1.1%	0.5%	A-SCI > AB	L-SCI > AB
F_lachnospiraceae	14.9%	17.2%	35.7%		L-SCI > AB
Antibiotics Use 31,32
F_Erysipelotrichaceae	0.4%	1.1%	0.8%	A-SCI > AB	L-SCI > AB
F_Lachnospiraceae; G_Lachnoclostridium	0.5%	4.1%	1.7%	A-SCI > AB	L-SCI > AB
Intestinal Inflammation32,35
F_Erysipelotrichaceae	0.4%	1.1%	0.8%	A-SCI > AB	L-SCI > AB
F_Burkholderiaceae; G_Sutterella	0.1%	1.9%	0.2%	A-SCI > AB		A-SCI > L-SCI
Depressive Disorders33,34
F_Rikenellaceae; G_Alistipes	0.4%	4.3%	2.1%	A-SCI > AB	L-SCI > AB
F_Ruminococcaceae; G_Oscillibacter	0.1%	0.5%	0.4%	A-SCI > AB	L-SCI > AB
Physical Activity36
F_Burkholderiaceae	3.6%	2.1%	0.6%		L-SCI < AB	A-SCI > L-SCI
Roles Unknown in Literature
F_Desulfovibrionaceae	0.1%	1.1%	0.5%	A-SCI > AB
G_Eggerthellaceae	0.0%	0.0%	0.1%		L-SCI > AB
F_Marinifilaceae	0.1%	1.5%	0.2%	A-SCI > AB		A-SCI > L-SCI
F_Marinifilaceae; G_Odoribacter	0.1%	1.3%	0.2%	A-SCI > AB		A-SCI > L-SCI
F_Campylobacteraceae; G_Campylobacter	0.2%	0.0%	0.0%		L-SCI < AB
O_Bacillales	0.7%	0.1%	0.1%		L-SCI < AB
F_Lachnospiraceae; G_Eisenbergiella	0.0%	1.3%	0.7%	A-SCI > AB	L-SCI > AB
F_Ruminococcaceae; G_Eubacterium coprostanoligenes group; S_unidentified	0.0%	0.7%	0.7%	A-SCI > AB	L-SCI > AB
F_Ruminococcaceae; G_Anaerotruncus	0.0%	0.0%	0.2%	A-SCI > AB	L-SCI > AB
F_Acidaminococcaceae; G_Phascolarctobacterium	0.1%	1.1%	0.5%	A-SCI > AB	L-SCI > AB
F_Erysipelotrichaceae; G_Erysipelatoclostridium	0.0%	0.2%	0.0%	A-SCI > AB

^#AB, able-bodied; A-SCI, acute spinal cord injury; L-SCI, long-standing spinal cord injury;

*Significantly different groups

Discussion

Recent research has revealed the complex and intricate interaction between the gut microbiome and the local gastrointestinal tract as well as distal organs, which is pivotal for host health. Evidence from several animal SCI models showed that SCI induces GD, increases gut permeability, promotes intestinal and systemic inflammation, and bacteria translocation.^6,7,9 The resulting microbial imbalance may impair neuro-recovery and predispose patients with SCI to secondary complications. Furthermore, additional external factors, such as extreme reduction in physical activity, increased psychological stress, and antibiotic usage may further exacerbate GD in the chronic stages of SCI. In this study, we performed a comprehensive 16s rRNA gut microbiome analysis and showed significant alterations of the gut microbiome among age-matched individuals with A-SCI, L-SCI, and AB controls, as demonstrated by differences in both the alpha- and beta-diversity indices. Our results provide the first evidence for GD in the A-SCI population and supported findings from previous studies in the L-SCI group. Specifically, we observed differential abundances of several microbial communities in the A-SCI and L-SCI groups that are linked to impaired metabolic health (families Acidaminococcaceae and Lachnospiraceae),^27–30 antibiotics use (genus Lachnoclostridium and family Erysipelotrichaceae),^31,32 depression (genera Alistipes and Oscillibacter),^33,34 and intestinal inflammations (family Erysipelotrichaceae and genus Sutterella)^32,35 compared to AB controls. In addition, we showed that L-SCI share microbiome features linked to reduced physical activity (family Burkholderiaceae) compared to A-SCI and AB controls.³⁶

Consistent with previous work in animal SCI models,⁶ we observed a higher abundance of Clostriadiales order in the L-SCI group than the AB control group. Within this order, Lachnospiraceae, one of the most prevalent families, was more abundant in the L-SCI group than in the A-SCI and AB control groups. A higher abundance of the Lachnospiraceae family is associated with elevated serum glucose and lipid concentrations in older populations.²⁸ Although our results are not equipped to delineate the relationship between GD and metabolic health, a causal effect of the Lachnospiraceae family on host glucose metabolism has been documented where the colonization of germ-free obese mice with Lachnospiraceae induced hyperglycemia.²⁸ Additionally, Acidaminococcaceae family and its genus Phascolarctobacterium, which are also linked to impaired glucose or lipid metabolism,^29,30 had higher abundances in the SCI groups than their AB controls. These data support the hypothesis that SCI-induced GD may be a previously unknown factor contributing to impaired metabolic health among individuals with SCI.

Previous studies using acute/subacute animal SCI models documented intestinal and systemic inflammation after SCI.^6,7 Similarly, we observed higher abundances of bacteria communities linked to gut inflammation-related disorders of the GI tract in the SCI groups. Specifically, both SCI groups had elevated Erysipelotrichaceae family, which was shown to be higher in inflammatory bowel disease and colorectal cancer.³² In addition, The Erysipelotrichaceae family has members that are immunogenic and associated with increased circulating tumor necrosis factor in the host.³⁷ In our study, compared to both the L-SCI and control groups, the A-SCI group had a greater abundance of genus Sutterella, which was shown to be higher in patients with inflammatory bowel disease.³⁵ Given the well-recognized role of GD in the development of inflammation in the gut,³⁸ our data suggest a potentially exacerbated inflammatory state in the gut of individuals with SCI compared with healthy AB controls.

There are many obstacles for individuals with SCI to be physically active, which leads to a sedentary lifestyle³⁹ and increased risk for cardio-metabolic diseases.⁴⁰ Research support that adopting an exercise regime modifies the gut microbiome, whereas the changes are reversed upon discontinuation of exercise.⁴¹ In our study, compared to the A-SCI and AB groups, the L-SCI group had lower abundances of the Burkholderiaceae family, which was found to be increased with exercise. It is reasonable to consider that the extreme physical inactivity in the L-SCI group might have contributed to such differences in microbiome composition.³⁶

Individuals with SCI often experience depressive disorders, which is partly a result of external stressors, such as extreme lifestyle changes. Expansive research on the microbiota-gut-brain axis implicated GD in the pathophysiology of depressive disorders,^42,43 where gut microbiome can affect the brain directly and indirectly by activating the vagus nerve, modulating the immune system, and producing neurotransmitters.⁴⁴ In our study, gut microbial communities, such as the Alistipes genus of the Rikenellaceae family and the Oscillibacter genus of the Ruminococcaceae family were higher in the A-SCI and L-SCI than the AB control groups. These microbial communities were found to be elevated in depressive disorders,^33,34 though their functions are not well understood. Thus, future studies assessing the relationship between the gut microbiome and depressive disorders in SCI may offer additional insights.

It is well-documented that the use of antibiotics causes rapid (within 4 days), as well as, persistent changes⁴⁵ in gut microbiome composition. Six out of seven participants in the A-SCI group were prescribed broad-spectrum antibiotics to prevent and/or treat infections. Although no antibiotics use was reported for the L-SCI group within 4 weeks prior to study enrollment, frequent prior use of antibiotics in this group may be common due to recurrent urinary tract infections.⁴⁶ In support of this idea, we found that microbial communities that flourish after antibiotics use, such as the Lachnoclostridium genus in the Lachnospiraceae family³¹ and Erysipelotrichaceae family,³² were more abundant in both A-SCI and L-SCI groups. Furthermore, given that antibiotics use is an integral part of standard care in A-SCI, our study is the first to document early changes in the gut microbiome that may be a result of both SCI and heavy antibiotics use. Further research is warranted to understand the impact of early changes in microbiome composition on functional recovery and long-term health.

Our study has several limitations. We did not have access to data on biomarkers of metabolic health and inflammation for all participants; consequently, whether the changes in microbiome composition coincided with worsened metabolic health or inflammatory status in the SCI groups is unknown.⁴ Similarly, our analysis is limited to gut microbiome compositional data with the use of 16s rRNA. The functional output of the gut microbiome may be predicted using programs like the phylogenetic investigation of communities by reconstruction of unobserved states (PICRUst). However, its predictions are limited due to the relatively low taxonomy resolution of 16s rRNA. As a result, PICRUst may not capture the important pathophysiological pathways carried out by different strains of bacteria.⁴⁷ Our future plans include performing metagenomics analysis, which offers higher taxonomical resolution as well as functional characterization of the gut microbiome than 16s rRNA sequencing, reaching the species and strain levels.⁴⁸ In addition, we could not delineate the effects of a number of potential confounders, including but not limited to diet, exercise, previous antibiotics use among L-SCI and AB controls, body composition, and the type of neurogenic bowel disorders (upper motor neuron [UMN] versus lower motor neuron [LMN])³ on gut microbiome composition. Although a previous study revealed different microbiome composition between UMN and LMN bowel disorders,³ statistically controlling for bowel program among SCI groups in our study was not possible as many patients in the A-SCI group still did not establish regimented bowel programs and lacked a diagnosis of the underlying pathophysiology (UMN versus LMN) for bowel dysfunction.

Conclusions

Our findings revealed a significant difference in gut microbiome composition among individuals with A-SCI, L-SCI, and AB controls. Compared to the AB controls, microbiome composition in the SCI groups share features linked to metabolic syndrome, inflammation-related bowel disorders, depressive disorders, or antibiotics use. Compared to A-SCI and AB control groups, individuals with L-SCI share features linked to physical inactivity. Our findings in A-SCI provide the first evidence for GD in this population. Whether SCI or antibiotics use alone or SCI combined with antibiotics use induce GD still remain unknown; however, our findings are expected to highlight a number of new avenues in acute and chronic SCI that could be explored in future microbiome studies. Particularly, future investigations are needed to assess the involvement of gut microbiome composition in neurorecovery potential in A-SCI and the development of chronic disorders among individuals with L-SCI.

Funding Statement

This work was supported by National Institute on Disability, Independent Living, and Rehabilitation Research: [Grant Number H133N110008/90SIS5019-03]; National Institute of Health: [Grant Number P30AR050948]; National Institute of Health: [Grant Number UL1TR001417].

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available as they contain information that could compromise the privacy of research participants.