Exercise pretreatment intervention alleviates intestinal dysbiosis and improves intestinal dysfunction in stroke rats

Abstract

Background

Gastrointestinal dysfunction is a common and detrimental complication following stroke, closely linked to poor outcomes. Emerging evidence implicates gut microbiota dysbiosis and intestinal barrier disruption in this process. This study investigated whether exercise preconditioning (EP) could protect against post-stroke gastrointestinal impairment and elucidated the underlying mechanisms involving the gut microbiota–intestine axis.

Methods

A meta-analysis was conducted to estimate the global incidence of post-stroke constipation. In vivo, Sprague–Dawley rats underwent 2 weeks of treadmill-based EP prior to transient middle cerebral artery occlusion (tMCAO). Neurological deficits, fecal properties, gut microbiota composition (16S rRNA sequencing), intestinal permeability, and mucosal inflammation were systematically evaluated.

Results

Meta-analysis revealed a 40% pooled incidence of post-stroke constipation. In the tMCAO model, EP significantly improved neurological scores and alleviated constipation, as evidenced by increased fecal water content and motility. EP restored gut microbiota diversity, enriched beneficial taxa (e.g., Lactobacillus), and reduced opportunistic pathogens. Furthermore, EP attenuated intestinal barrier disruption—marked by upregulated tight junction proteins (ZO-1, occludin)—and suppressed colonic inflammation (reduced TNF-α and IL-6).

Conclusion

Exercise preconditioning mitigates post-stroke gastrointestinal dysfunction, which is associated with preserving gut microbiota homeostasis, reinforcing the intestinal epithelial barrier, and dampening local inflammation. These findings highlight EP as a promising non-pharmacological strategy to improve systemic resilience and outcomes after stroke through gut-targeted mechanisms.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12876-025-04554-y.

Introduction

Stroke represents a significant global health concern. The underlying pathology of an ischaemic stroke is the formation of an intravascular thrombus, which results in the necrosis of brain tissue and focal neuronal deficits [1]. Clinically, ischemic stroke can result in significant neurological impairment, including paralysis, language and cognitive deficits, and loss of self-care abilities. Nevertheless, There is an increasing body of evidence indicating that, in addition to neurological deficits, stroke patients frequently present with a range of non-neurological complications. These include gastrointestinal disorders, immunosuppression and bacterial infections, and even sepsis, which can occur at any stage of the recovery process, from the hyper-acute to the chronic phase [2]. The study by Jia Fu et al. concluded that gastrointestinal complications are the most common cause of increased morbidity and mortality in patients [3]. Several studies have demonstrated the occurrence of gastrointestinal disturbances, including dysphagia, gastrointestinal bleeding and constipation, in patients who have suffered a stroke [4]. The study revealed that the second most prevalent symptom observed in patients with brainstem infarction was constipation due to paralytic bowel obstruction, with an overall incidence of 45% [5, 6]. Consequently, gastrointestinal complications in patients who have suffered a stroke have a considerable impact on the patient's overall health and significantly impede the recovery of the brain after a stroke.

A substantial amount of attention has been dedicated to the prevalence of post-stroke gastrointestinal complications in clinical practice, particularly given the association between post-stroke constipation and an increased risk of secondary stroke episodes, as well as the potential complications it may pose for stroke patients [7]. Changes in intestinal flora resulting from a stroke may be a significant contributing factor to gastrointestinal. Following the onset of ischemic stroke (IS), increased gut microbiota richness, reduced levels of beneficial metabolites such as short-chain fatty acids (SCFAs), and elevated abundance of Gram-negative bacterial lipopolysaccharides (LPS) trigger an inflammatory response characterized by upregulated expression of pro-inflammatory factors. The cytotoxic effect of inflammatory substances results in damage to the microvilli of the intestinal epithelium [2]. This leads to a breakdown in the intestinal epithelial barrier and a leaky gut, which can result in gastrointestinal complications such as intestinal dyskinesia and intestinal paralysis. A compromised intestinal barrier allows for the permeation of inflammatory cytokines, bacteria, and toxic intestinal metabolites across the damaged intestinal epithelial barrier, subsequently entering the circulatory system. This phenomenon exacerbates gastrointestinal dysfunction and may even precipitate entheogenic sepsis [2]. Dietary interventions, probiotics, microbial metabolites, faecal microbial transplants and traditional Chinese medicine, when employed alone or in combination, have been demonstrated to prevent and improve gastrointestinal complications following a stroke [8]. However, stroke patients represent a heterogeneous population, and the aforementioned interventions have notable clinical limitations. Preventive strategies implemented before stroke onset may therefore be more effective than treatments initiated after the event.

Exercise pretreatment (EP) is an effective non-pharmacological intervention that promotes ischemic tolerance and provides neuroprotection in high-risk individuals. It has been demonstrated that EP can be an effective method for inducing ischemic tolerance, exerting neuro-protection and reducing post-stroke brain injury. This is achieved by providing multiple training sessions before the onset of stroke. In comparison to alternative preconditioning techniques, this method offers the distinct advantages of simplicity and patient acceptance [9]. On the other hand, exercise as an intervention can effectively alleviate gastric dyskinesia [10]. Furthermore, recent animal studies have shown that EP not only significantly improved intestinal dysbiosis but also attenuated the severity of HF + DSS-induced colitis in mice [11]. Whether EP intervention can play a positive role in stroke complications, especially gastrointestinal dysfunction, and thus improve the prognosis of stroke is precisely what we are interested in. We first confirmed the occurrence of post-stroke constipation by meta-analysis and observed the effect of EP on the intestinal mucosal barrier in post-stroke rats on MCAO rats.

Materials and methods

Meta-analysis

This study included observational studies (cohort and cross-sectional) that met the following criteria: 1) the study population comprised definitively diagnosed stroke patients; and 2) constipation was diagnosed using the Rome III criteria for functional gastrointestinal disorders. Intervention studies articles with incomplete data or unclear definitions of constipation were excluded. Relevant databases including PubMed Embase CochraneLibrary, Web of Science, China Knowledge Network, Wanfang Database, and Wipo Database, were searched from their inception to September 2024. Search terms combined subject headings and free-text words related to stroke and constipation (see Table S1 for detailed search strategy, See Appendix). Literature screening and data extraction were independently conducted by two investigators. Disagreements were resolved through discussion or consultation with a third party. Extracted data included study title, authors, publication year, investigation year, study location, stroke sample size and subtypes, stroke stage, constipation prevalence, patient gender, and age. The quality of cohort studies was assessed using the Newcastle–Ottawa Scale (NOS; Table S2, See Appendix), evaluating selection, comparability, and outcome determination, with a maximum score of 9. Higher scores indicated better study quality. For cross-sectional studies, the AHRQ scale (Table S3, See Appendix) was used to evaluate various aspects comprehensively, ensuring robust quality assessment. Heterogeneity across studies was assessed using Cochran’s Q test and the I² statistic. An I² value of 0% indicates no observed heterogeneity, while values of 25%, 50%, and 75% are considered to represent low, moderate, and high heterogeneity, respectively.

Animal experimental design

Fifteen male Sprague–Dawley rats (5 weeks old) were obtained from Shanghai Jihui Laboratory Animal Care Co., Ltd. (Shanghai, China) and housed in the Animal Laboratory Facility of the Shanghai Public Health Clinical Center under controlled conditions: a 12-h light/dark cycle, temperature (22.5 ± 2.5 °C), and humidity (50 ± 5%). Rats were randomly assigned to three groups (n= 5 per group): sham-operated (Sham) middle cerebral artery occlusion (MCAO) and exercise preconditioning plus MCAO (EP + MCAO). Fourteen days prior to surgery the EP + MCAO group underwent daily treadmill exercise at 15 m/min for 30 min starting at 14:00. A 3-day acclimatization period preceded formal intervention with incremental increases in speed and duration (from 6 m/min for 10 min increasing by 2 m/min and 10 min daily) to minimize stress. Neurological deficits were evaluated at 24 h and 5 days post-surgery using theLonga scoring system. Following assessment, all animals were euthanized for tissue collection. At the designated endpoint, rats were deeply anesthetized with 4 % isoflurane in oxygen. Euthanasia was then performed via intraperitoneal injection of a saturated sodium pentobarbital solution (150 mg/kg). Death was confirmed by the absence of heartbeat, respiration, and corneal reflex for at least 5 min. This protocol ensures a rapid and humane death, consistent with institutional ethical guidelines. All procedures were performed in accordance with institutional ethical guidelines and approved by the Animal Welfare and Ethics Committee of the Shanghai Public Health Clinical Center (approval number: 2022-A004-01).

Transient middle cerebral artery occlusion

Focal cerebral ischemia was induced via transient middle cerebral artery occlusion (tMCAO) as previously described [12]. Briefly, rats were anesthetized with isoflurane, and the right common carotid artery (CCA), internal carotid artery (ICA), and external carotid artery (ECA) were exposed by blunt dissection. The distal end of the ECA was ligated, and a silicone-coated nylon monofilament (embolic suture) was inserted into the ICA via the ECA stump and advanced to occlude the origin of the middle cerebral artery. After 90 min of ischemia, the filament was carefully withdrawn to allow reperfusion, and the proximal ECA was ligated. Sham-operated animals underwent identical surgical procedures without filament insertion. Throughout surgery, rectal temperature was monitored and maintained at 37.0 ± 0.5 °C using a feedback-controlled heating pad to prevent hyper- or hypothermia.

Neurological function assessment

Neurological deficits were evaluated in all animals on postoperative days 1 and 5 using the 5-point Longa scoring system to assess model success and functional recovery [13]. Scores were assigned as follows: 0 = no deficit; 1 = failure to fully extend the contralateral forelimb; 2 = circling to the right; 3 = lateral drift to the right when walking; and 4 = inability to walk spontaneously, accompanied by depressed consciousness. Assessments were performed by an investigator blinded to group allocation.

Fecal sample collection, storage, and parameter analysis

Fecal samples were collected after behavioral testing. Rats were individually placed in clean cages lined with sterile filter paper, and freshly excreted feces were immediately collected using sterile tweezers, avoiding contamination from fur or debris. Samples were transferred to pre-labeled cryovials, rapidly frozen in liquid nitrogen, and stored at − 80 °C until analysis. Fresh fecal weight was measured daily from samples collected at the same time across all groups. Dry weight was determined after drying at 60 °C for 24 h. Fecal water content (%) was calculated as: (wet weight − dry weight)/wet weight × 100% [14].

Histopathological staining (H&E and crystal violet)

Brain and rectal tissues were fixed in 4% paraformaldehyde, embedded in paraffin, and sectioned using a microtome (not a cryostat, unless frozen sections were specifically used—see note below). Tissue sections were stained with hematoxylin and eosin (H&E) or crystal violet (CV) for histopathological evaluation. H&E staining was used to assess general tissue morphology and cellular architecture, while CV staining was employed to visualize neuronal integrity, particularly in brain sections. All stained sections were examined under a light microscope for qualitative and semi-quantitative analysis.

Brain water content measurement

Cerebral edema was assessed by measuring brain water content using the wet-dry weight method [15]. Immediately after euthanasia, the brain was rapidly excised and the right hemisphere (or ischemic region) was weighed to obtain the wet weight. The tissue was then dried in an oven at 110 °C for 24 h to a constant weight, followed by measurement of the dry weight. Brain water content (%) was calculated as: (wet weight − dry weight)/wet weight × 100%.

16S rRNA gene sequencing

Gut microbiota composition was analyzed by 16S rRNA gene sequencing. Genomic DNA was extracted using CTAB/SDS and the V4–V5 region (515F–907R) was amplified with barcoded primers. PCR products were purified (GeneJET Gel Extraction Kit Thermo Scientific) and libraries were constructed using the NEB Next® Ultra™ DNA Library Prep Kit (NEB USA).Library quality was assessed using the Qubit 2.0 Fluorometer and Agilent Bioanalyzer 2100. Paired-end sequencing (250–300 bp) was performed on an Illumina MiSeq platform. Bioinformatic analysis included read assembly (FLASH), OTU clustering and taxonomic annotation (UPARSE), and diversity analysis (QIIME). Principal Coordinate Analysis (PCoA) and UPGMA clustering were used to assess community structure. Statistical comparisons between groups were performed using ANOSIM, MRPP, Metastats, and LEfSe to identify taxonomic differences and their biological relevance.

Immunofluorescence

Paraffin-embedded tissue sections were deparaffinized in xylene and rehydrated through a graded ethanol series. After PBS washing, sections were blocked with 5% bovine serum albumin (BSA) for 30 min, then incubated overnight at 4 °C with primary antibodies against ZO-1 (ab53444, Abcam, 1:200), occludin (ab300136, Abcam, 1:200), TNF-α (ab6671, Abcam, 1:200), and IL-6 (ab9324, Abcam, 1:200). Following three PBS washes, sections were incubated with fluorescently labeled secondary antibodies and counterstained with DAPI (4, 6-diamidino-2-phenylindole). Fluorescence images were acquired using a confocal laser scanning microscope (Leica, Germany).

Statistical analysis

Data are expressed as mean ± standard deviation (SD). Statistical analyses were performed using SPSS 25.0 or GraphPad Prism 9.0. Normality was assessed using the Shapiro–Wilk test. For comparisons among three groups, one-way ANOVA was used followed by post hoc tests: the Least Significant Difference (LSD) test for equal variances, or Dunnett’s T3 test for unequal variances. Non-parametric data were analyzed using the Kruskal–Wallis test with Dunn’s post hoc test. For microbiota data, group differences in alpha and beta diversity were assessed using Kruskal–Wallis and PERMANOVA (Adonis), respectively. Statistical significance was set at P < 0.05.

Results

Literature search and constipation incidence analysis

A comprehensive search was conducted, resulting in the evaluation of 272 articles, from which 15 studies comprising a total of 35,864 patients were selected for inclusion (Figure S1). The included studies (Table 1) were published between 2000 and 2021, encompassing research sites across 8 countries and regions. Among these, 12 studies were conducted in Asia [16–26], 1 in Europe [27], 1 in Oceania [28], and 1 in South America [29]. Sample sizes ranged from 36 to 31,953 cases. The primary diagnostic criteria for constipation included the Rome III criteria and the Bristol Stool Trait Scale. The synthesis of results from the 15 studies revealed significant variability in the incidence of post-stroke constipation, with rates ranging from 0.04 [28]to 0.79 [24]. A forest plot demonstrated an overall incidence of post-stroke constipation of 40% (95% CI: 20%, 61%), with substantial heterogeneity across studies (I² = 95.8%, P < 0.001), as shown in Fig. 1A. The weights for each study ranged from 6.47 to 6.74, indicating the relative contribution of each study to the overall estimate. Subgroup analysis further investigated the prevalence of constipation in patients with ischemic stroke or hemorrhagic stroke. The results indicated that the prevalence of constipation in patients with ischemic stroke was 40%, with a 95% confidence interval of 29% to 51% (I² = 61.2%, P < 0.001) (Fig. 1B). For hemorrhagic stroke patients, the prevalence was 66%, with a 95% confidence interval of 48% to 84% (I² = 91.3%, P = 0.003) (Fig. 1C). Additionally, the prevalence of constipation at various stages of stroke was examined. In the acute phase, the prevalence of constipation among stroke patients was 39%, with a 95% confidence interval ranging from 14 to 64% (I² = 99.8%, P < 0.001) (Fig. 1D). In contrast, during the recovery phase, this prevalence increased to 52%, with a 95% confidence interval ranging from 9 to 95% (I² = 97.8%%, P < 0.001) (Fig. 1E). This suggests that the incidence of constipation is significantly higher in the recovery phase compared to the acute phase, potentially due to prolonged disease duration allowing for continued factors that trigger constipation.

Table 1.

Characteristics of included studies examining the incidence of constipation in stroke patients

No	Author	Year	Survey data	Location	Male/female (n)	Age, years	Total samples (n) (stroke subtype)	Stage of stroke	New-Onset Samples of constipation (%)
1	Yun Sun	2022	2019–2021	China	144/78	65.42 ± 13.37	222(IS 222)	acute	128(58)
2	F Bracci	2007	NA	Italy	43/47	68(27–95)	90	Recovery	27(30)
3	S.-F. Lim	2015	NA	Singapore	33/22	61.2 ± 9.7	55	Acute	18(33)
4	CJ Lin	2013	2010.8–2011.8	Taiwan, China	94/61	60.3	155(IS 124; HS 31)	Acute	30(19)
5	Cheng, Jiafei	2020	2017.11–2018.6	China	27/47	63.55 ± 11.05	71	Acute	34(48)
6	Chen, Jianbing	2014	2008.1–2014.1	China	362/138	65.2 ± 8.6	500(IS 378; HS 95)		186(37) (IS 125; HS 49)
7	V. S. Doshi	2003	2001.1–2001.6	Singapore	65/75	65.7	140(IS 99; HS 37)	Acute	32(23)
8	Yongjing Su	2009	2003.11–2004.11	China	92/62	65.61 ± 14.53	154(IS 122; HS 32)	Acute	85(55)) (IS 63; HS 22)
9	VerityE.Bond, B	2023	2000–2020	Australia	17,018/14,935	73.8 (13.8)	31,953(IS 22417; HS 5521)	Acute	1406(4)
10	Youn-I Choi	2015	2008–2009	South Korea	16/20	NA	36	Acute	21(58)
11	Koçaýli	2014	NA	Turkey	53/59	62.5 ± 13.0	112(IS 83; HS 29)	Recovery	83(74)
12	Tânia	2014	2009.12–2010.5	Brazil	49/49	58.13 ± 12.64	98(IS 87; HS 11)		28(29)
13	I Joon Han	2023	2018.1–2019.12	South Korea	164/92	67	256(IS 256)	Acute	81(32)
14	BIAN Jing	2017	2015.2–2016.9	China	183/91	62.35 ± 13.30	274(IS 274)	Acute	74(27)
15	Wei Chen	2021	2010.10–2018.12	China	1229/519	58 ± 14	1748(HS 1748)	Acute	1340(77)

HS Hemorrhagic stroke, IS Ischemic stroke, NA Information not provided

Fig. 1.

Incidence of constipation in stroke patients. Note: A Forest plot of the incidence of constipation after stroke in all studies. B Forest plot of the incidence of constipation after ischemic stroke. C Forest plot of the incidence of constipation after hemorrhagic stroke. D Forest plot of the incidence of constipation in acute phase stroke patients. E Forest plot of the incidence of constipation in recovering stroke patients

Effects of EP intervention on brain damage

To evaluate the impact of exercise pretreatment (EP) intervention on brain damage, HE staining was performed on brain sections from rats in each group. The results demonstrated significant necrosis, edema, and swelling of neuronal cells in the ischemic semi-dark area, pale staining, and nuclear condensation in the cortex, hippocampus, and striatum of rats in the model group. In contrast, the intracerebral injury in the EP group was notably alleviated compared to that of the model group (Fig. 2A-C). The status of cerebral infarction in rats from each group was observed using CV staining, and the volume of cerebral infarction was analyzed using image analysis software (Huai et al., 2013). The results showed that the volume of cerebral infarction in the EP group was significantly lower than that in the model group (Fig. 2D-G). Additionally, the ratio of dry weight to wet weight of brain tissue indicated the degree of brain edema after ischemia–reperfusion injury. The brain water content in both the model group and the EP group was significantly higher than that of the sham group, but the EP group exhibited a significantly lower brain water content compared to the model group (Fig. 2H). Motor function was objectively assessed usingLonga scores at 24 h and 5 days post-surgery. The results demonstrated that theLonga scores of rats in both the model and EP groups were significantly higher than those of the sham group at 24 h and 5 days after modeling. However, the scores of the EP group were significantly lower than those of the model group at the corresponding time points (Fig. 2I).

Fig. 2.

Effects of EP intervention on brain and intestinal functions. Note: (A, B, C) HE staining of brain tissue of each group; (D, E, F) crystal violet staining of brain tissues of each group; (G) Cerebral infarct volume in each group of rats; (H) Brain water content of each group; (I) comparison ofLonga scores between 24 h after modelling and 5 days after modelling in each group; (J) faecal water content of rats in each group. n = 5 per group. Error bars represent standard deviation. *P < 0.05, **P < 0.01, ***P < 0.001 vs. Sham; #P < 0.05, ##P < 0.01, ###P < 0.001 vs. tMCAO. *P < 0.05 and**P < 0.01 vs. Sham group; #P < 0.05 and ##P < 0.01 vs. EP group

Effects of EP intervention on intestinal function

To further evaluate the impact of EP intervention on intestinal function, the fecal water content of rats in the three groups was quantified. The results demonstrated a significant reduction in fecal water content in ischemic stroke rats compared to the sham group. Notably, the EP intervention effectively restored fecal water content in ischemic stroke rats (Fig. 2J). Fecal water content is a critical indicator of intestinal function, reflecting the degree of intestinal dysmotility and constipation. In this study, ischemic stroke led to a marked decrease in fecal water content, indicative of impaired intestinal transit. The EP intervention significantly mitigated this impairment, suggesting its potential role in improving intestinal motility and alleviating post-stroke constipation. Moreover, the restoration of normal fecal water content by EP intervention aligns with the hypothesis that pre-exercise conditioning can enhance gastrointestinal health and reduce the risk of complications associated with stroke-induced gastrointestinal dysfunction. These findings provide evidence for the beneficial effects of EP intervention on maintaining intestinal homeostasis and preventing the onset of constipation in stroke patients.

Effects of EP intervention on gut microbiota diversity and composition

To evaluate the impact of EP intervention on gut microbiota, 16S rRNA sequencing was performed on fecal samples from the three groups. Although no significant differences in Shannon index were observed, the Chao1 index revealed a notable discrepancy between the MCAO and EP groups, indicating altered microbial richness (Fig. 3A–B). The Bacteroidetes/Firmicutes ratio was significantly increased in the model group compared to the sham group, while the EP group exhibited a marked reduction in this ratio (Fig. 3C). β-diversity analysis using PCoA and non-metric multidimensional scaling (NMDS) clearly separated the three groups (Fig. 3D–E), with an NMDS stress value of 0.097, confirming robust group discrimination. At the phylum level, the model group showed decreased Firmicutes and increased Bacteroidota and Desulfobacterota abundances compared to the sham group—changes that were reversed by EP intervention (Fig. 4A–B). At the family level, EP restored stroke-induced reductions in Lactobacillaceae, Lachnospiraceae, Clostridia_UCG-014_norank, and Oscillospiraceae, and attenuated the increase in Bacteroidaceae (Fig. 4C–D). LEfSe analysis identified significant enrichment of taxa including p_Bacteroidota, c_Bacteroidia, o_Bacteroidales, f_Bacteroidaceae, and g_Bacteroides in the model group—alterations that were reversed by EP. Notably, EP also increased the abundance of p_Proteobacteria and c_Gammaproteobacteria (Fig. 4E–F). At the genus level, while Bacteroides was elevated in the model group, Akkermansia and UBA1819 were significantly increased compared to the sham group; EP intervention prevented these dysbiosis-related shifts (Fig. 4G).

Fig. 3.

EP regulates gut microbiota. Note: A, B Chao1 index, Shannon index in α-diversity. C Ratio of the phylum Bacteroidetes in Firmicutes. D β-diversity PCoA analysis. E β-diversity NMDS analysis. PERMANOVA (Adonis) was used to assess differences in beta diversity. Stress value for NMDS is 0.097. n = 5 per group. Error bars represent standard deviation. *P < 0.05, **P < 0.01, ***P < 0.001 vs. Sham; #P < 0.05, ##P < 0.01, ###P < 0.001 vs. tMCAO

Fig. 4.

Effect of EP intervention on the composition of intestinal flora in stroke rats. Note: A Structure of the phylum level microbiota of different taxa. B Community bar graph analysis of the microbiota at the phylum level in mice. C Microbial community structure at the family level in different groups. D The bar graph analysis of microbiota at the family level. E, F Linear discriminant analysis (LDA) effect size (LEfSe) analysis of microbiota in different groups. G Relative abundance of genus-level differential bacteria. Statistical significance was determined using LEfSe with LDA score > 2.0 and P < 0.05. n = 5 per group. Error bars represent standard deviation. *P < 0.05, **P < 0.01, ***P < 0.001 vs. Sham; #P < 0.05, ##P < 0.01, ###P < 0.001 vs. tMCAO

Effects of EP intervention on intestinal mucosal barrier and inflammatory response

HE staining of rectal sections revealed that the colonic mucosa in the sham group exhibited intact, well-organized epithelial architecture with normal muscularis mucosae and goblet cells. In contrast, the model group displayed severe mucosal disruption, epithelial detachment, thinning of the muscularis mucosae, and increased goblet cell proliferation. EP intervention markedly attenuated these structural abnormalities, preserving mucosal integrity (Fig. 5A). Immunofluorescence analysis showed that the tight junction proteins ZO-1 and occludin were predominantly localized at epithelial cell junctions in the rectal mucosa. Their expression was significantly reduced in the model group compared to the sham group, indicating compromised barrier function. EP intervention effectively restored the expression levels of both proteins (Fig. 5B–E). Plasma levels of LPS, DAO, and D-lactate (D-LA), markers of intestinal permeability, were significantly elevated in the model group, confirming increased gut barrier leakage. EP treatment significantly reduced the levels of all three markers, suggesting improved barrier integrity (Fig. 5F–H). In the rectal intestinal wall, immunofluorescence revealed a marked increase in the expression of pro-inflammatory cytokines IL-6 and TNF-α in the model group. EP intervention significantly reduced the expression of both cytokines (Fig. 6A–D). Notably, plasma levels of PGC-1α and FNDC5, myokines associated with exercise-induced protection, were significantly decreased in the model group. EP intervention effectively restored their circulating levels (Fig. 6E–F).

Fig. 5.

Effect of EP intervention on the intestinal mucosal barrier in tMCAO rats. Note: (A) HE staining of rat intestinal tissue of each group(20x); (B, D) Relative protein levels of ZO-1(Green) detected by immunofluorescence(100x). Scale bar: 50 um; (C, E) Relative protein levels of Occludin(Green) detected by immunofluorescence(100x).Scale bar: 50 um; (F, G, H) plasma LPS, DAO, D-LA levels detected by Elisa. Data were analyzed using one-way ANOVA followed by LSD post hoc test. n = 5 per group. Error bars represent standard deviation. *P < 0.05, **P < 0.01, ***P < 0.001 vs. Sham; #P < 0.05, ##P < 0.01, ###P < 0.001 vs. tMCAO

Fig. 6.

Effects of EP intervention on IL-6 and TNF-α expression in the rectal intestinal wall of stroke rats. Note: (A, C) Relative protein levels of IL-6(Green) were detected by immunofluorescence (100x). Scale bar: 50 um; (B, D) Relative protein levels of TNF-α(Green) detected by immunofluorescence(100x). Scale bar: 50 um. E, F plasma PGC-1α, FNDC5 levels detected by Elisa. Data were analyzed using one-way ANOVA followed by LSD post hoc test. n = 5 per group. Error bars represent standard deviation. *P < 0.05, **P < 0.01, ***P < 0.001 vs. Sham; #P < 0.05, ##P < 0.01, ###P < 0.001 vs. tMCAO

Discussion

Stroke not only results in significant neurological deficits but is also accompanied by a range of non-neurological complications, among which gastrointestinal dysfunction is particularly prevalent. Constipation represents a significant gastrointestinal complication following a stroke [20, 30]. In this study, a meta-analysis of 15 clinical studies from 2000 to 2021 revealed a 40% incidence of post-stroke constipation. Our animal studies also showed a significant reduction in faecal water content in rats after stroke, which is consistent with clinical reports. In addition, the meta-analysis of this study found that the incidence of constipation varied at different stages after stroke. The incidence of constipation reached 52% in stroke patients in the recovery phase, which was significantly higher than 39% in the acute phase. Clinical studies have also shown that gastrointestinal dysfunction after stroke, especially constipation, not only affects the recovery of stroke patients but also increases the risk of stroke recurrence [31]. Therefore, the treatment of post-stroke constipation has been taken very seriously in clinical practice. However, the current symptomatic treatment is not very effective in controlling and preventing the occurrence of post-stroke constipation [32].

Many studies have shown that EP not only reduces the risk of stroke but also improves the incidence of post-stroke constipation [33, 34]. To further confirm the preventive effect of EP on poststroke constipation and to preliminarily investigate its mechanism, we performed EP intervention in a group of rats to be modelled for stroke. The results showed that EP treatment before stroke not only reduced oedema and ischemia–reperfusion injury in the brain tissue on the injured side of the rats but also significantly increased the faecal water content during the post-stroke recovery period in rats. Since previous literature suggests that EP modulates the distribution and abundance of intestinal flora [35]. Therefore, the present study also investigated the mechanism of EP on the improvement of intestinal dysfunction in rats after stroke from the perspective of intestinal flora and intestinal mucosal barrier.

The results of this study showed that the EP intervention resulted in an increase in the α-diversity of intestinal flora in tMCAO rats, as well as a significant distinction from the model group in terms of β-diversity. Furthermore, the EP group resulted in a notable reversal of the compositional alterations in the flora, as evidenced by a reduction in the number of thick-walled phyla and an increase in the number of anaplasmosis phyla at the phylum level. Furthermore, our study demonstrated that EP not only enhanced the diversity and abundance of the gut flora but also reinforced the integrity of the intestinal barrier by suppressing the inflammatory response. This is consistent with previous research indicating that exercise can alleviate gastric dyskinesia and promote gut health. These results demonstrate that EP effectively ameliorates gastrointestinal injury in rats following ischemic stroke.

It has been demonstrated in previous studies that the repair of the intestinal mucosal barrier can be achieved by means of an improvement in the composition of the gut flora [36]. Given that the intestinal flora maintains a parallel symbiotic relationship with the intestinal mucosal barrier. A disruption to the intestinal flora can result in a breakdown of the intestinal mucosal barrier [37]. It should be noted that the sample size in the animal experiment was relatively small (n = 5 per group), which is consistent with common practice in preclinical tMCAO studies but may limit statistical power. Therefore, while the observed effects were consistent and significant, these findings should be interpreted cautiously and validated in larger cohorts in future studies. It should be noted that our animal experiments assessed outcomes at a 5-day post-stroke endpoint, which primarily captures the acute phase of gastrointestinal dysfunction and early responses to exercise pre-conditioning. While this timepoint aligns with the peak of gut barrier disruption and inflammatory activation observed in tMCAO models, it does not reflect longer-term recovery trajectories or the sustainability of exercise-induced benefits. We are currently extending our follow-up period to 28 days in ongoing studies to better characterize the chronic effects of exercise on post-stroke gut–brain axis remodeling. Furthermore, constipation was assessed indirectly via fecal water content, which, while commonly used as a surrogate marker of intestinal transit and hydration status, does not directly measure gastrointestinal motility. More direct functional assays—such as charcoal meal or carmine red transit tests—will be incorporated in our ongoing and future studies to provide a more comprehensive evaluation of post-stroke gut motility.

ZO-1 and occludin are integral components of the intestinal mucosal barrier. In instances where the intestinal mucosal barrier is compromised, the expression of ZO-1 and Occludin is observed to be diminished [38]. The findings of this study indicate that stroke results in damage to the rectal mucosal structures, including necrosis and detachment of rectal mucosal epithelial cells, as well as thinning of the muscular layer of the mucosa. However, there was a very significant repair of the intestinal mucosal structure in the rats of the EP group. Furthermore, the expression of ZO-1 and Occludin in the intestinal mucosa of stroke rats was also markedly reduced, whereas the expression of ZO-1 and Occludin in the intestinal mucosa of rats in the EP group was significantly elevated. These findings suggest that the EP intervention exerts a protective effect on the intestinal mucosa, preserves the integrity of the intestinal barrier, and reduces intestinal permeability.

In addition, after stroke onset, increased gut microbiota richness decreased beneficial metabolites and increased Gram-negative bacterial lipopolysaccharides triggered an inflammatory response with increased expression of inflammatory factors. The cytotoxic effects of inflammatory substances resulted in damage to intestinal epithelial microvilli and decreased expression of intestinal epithelial tight junction proteins. This leads to disruption of the intestinal epithelial barrier and leaky gut, which can lead to gastrointestinal complications such as intestinal motility disorders and intestinal paralysis [2, 7]. This further impairs recovery from stroke. In this study, we also observed a significant increase in the expression of TNF-α and IL-6 in the intestines of stroke rats, whereas the expression of TNF-α and IL-6 was downregulated in the intestinal mucosa of rats in the EP group.

EP intervention is a beneficial and effective rehabilitation method and a healthy lifestyle for the prevention of cardiovascular and cerebrovascular diseases [39]. It is easily accessible to patients and offers multiple potential benefits. Gastrointestinal complications, particularly after stroke, often impede recovery and increase the risk of stroke recurrence [40], yet current clinical management remains suboptimal in addressing these risks. Our findings demonstrate that EP not only promotes brain tissue repair but also significantly improves post-stroke intestinal dysfunction, which is closely associated with the modulation of gut microbiota and restoration of the intestinal mucosal barrier. However, while our meta-analysis reports a pooled incidence of post-stroke constipation of 40%, substantial heterogeneity exists across studies (ranging from 4 to 79%). For instance, the confidence intervals for subgroup analyses are broad (e.g., 9% to 95% for the recovery phase), highlighting variability in study populations and methods. This underscores the need for more standardized diagnostic criteria and consistent patient populations in future research. The single pooled estimate should thus be interpreted with caution due to this variability. Moreover, although microbial changes correlate strongly with functional recovery, causality cannot be established by the current study. The precise molecular mechanisms by which EP-regulated microbiota contribute to barrier repair and functional restoration remain to be elucidated. Although our study identifies an association between exercise and gut microbiota modulation, the underlying molecular mechanisms—particularly the putative role of exercise-induced myokines such as those in the PGC-1α/FNDC5 pathway—remain unresolved. This mechanistic gap represents a key limitation of our work and underscores the need for future studies employing targeted in vivo or in vitro approaches to dissect the muscle–gut axis in exercise-mediated microbial regulation. To this end, we are currently conducting fecal microbiota transplantation experiments to determine whether 'exercise-modulated' microbiota—particularly taxa such as Lactobacillus and Akkermansia—are functionally involved in mediating these protective effects. This will be the focus of our future research. It should be noted, however, that this study used only male rats, and given known sex differences in stroke and gut physiology, future studies must include females to assess the generalizability of these findings.

Conclusions

This study demonstrates that EP effectively mitigates post-stroke gastrointestinal dysfunction through a gut microbiota–intestine axis–mediated mechanism. Meta-analysis revealed a high prevalence of constipation after stroke (40%), underscoring the clinical significance of gastrointestinal complications. In tMCAO rats, EP not only reduced cerebral infarction and neurological deficits but also improved intestinal motility and alleviated constipation. Mechanistically, EP restored gut microbial homeostasis by enhancing overall diversity, modulating the Bacteroidetes/Firmicutes ratio, and enriching beneficial taxa such as Lactobacillaceae, Lachnospiraceae, and Clostridia_UCG-014_norank. Concurrently, EP reinforced the intestinal mucosal barrier, as evidenced by upregulated tight junction proteins (ZO-1, occludin) and reduced plasma levels of LPS, DAO, and D-lactate. This barrier protection was associated with suppressed intestinal inflammation, reflected in decreased expression of TNF-α and IL-6. Collectively, these findings indicate that EP acts via a multi-target, gut-centered pathway to prevent post-stroke gastrointestinal complications. EP may therefore represent a promising non-pharmacological strategy to improve systemic resilience, enhance recovery, and reducemortality after stroke.

Abbreviations

AHRQ

Agency for healthcare research and quality

CV

Crystal violet

DAO

Diamine oxidase

D-LA

D-lactate

EP

Exercise preconditioning

FNDC5

Fibronectin type III domain containing 5

H&E

Hematoxylin and eosin

i.p.

Intraperitoneal

IL-6

Interleukin-6

IS

Ischemic stroke

LEfSe

Linear discriminant analysis effect size

LPS

Lipopolysaccharide

MCAO

Middle cerebral artery occlusion

NMDS

Non-metric multidimensional scaling

NOS

Newcastle-Ottawa scale

OTU

Operational Taxonomic Unit

PCoA

Principal coordinate analysis

PGC-1α

Peroxisome proliferator-activated receptor gamma coactivator 1-alpha

SCFAs

Short-chain fatty acids

TNF-α

Tumor necrosis factor-alpha

Tmcao

Transient middle cerebral artery occlusion

UPGMA

Unweighted pair group method with arithmetic mean

ZO-1

Zonula occludens-1

Authors’ contributions

Ayipaxiaguli·Kasimu: Participated in study design, conducted the experiments, and was responsible for data analysis, data interpretation and writing the initial draft of the manuscript. Ying Huang: participated in study design, data analysis, and data interpretation. Bing Li designed the study and edited the paper. Dangshou Wang, Yanqi Jiang, Yu Liu, and Yuting Song recorded and analyzed the data. Yunpeng Zhang and Xinghua Chen made important revisions and edits to the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (grant numbers: 82174522) and the Backup Discipline Platform Construction Project of Jinshan Hospital, Fudan University (HBXK-2022–2).

Data availability

All data generated or analysed during this study are included in this published article.

Declarations

Ethics approval and consent to participate

All procedures were performed in accordance with institutional ethical guidelines and approved by the Animal Welfare and Ethics Committee of the Shanghai Public Health Clinical Center (approval number: 2022-A004-01).

Consent for publication

Not applicable. This study did not involve human participants, human data, or human tissues. Therefore, consent for publication is not required.

Competing interests

The authors declare no competing interests.