Constipation, affecting over 10% of the global population, represents a prevalent manifestation of gastrointestinal dysfunction and a critical worldwide epidemiological challenge (1). It is characterized by unsatisfactory defecation as a result of infrequent stools, difficult stool passage, or a combination of both mechanisms (2). According to the cause, constipation may be classified as primary or secondary. Secondary constipation arises from diverse etiological factors, including anatomical abnormalities (anorectal and colonic pathologies), dietary influences, pharmacological agents (notably opioids), and other underlying conditions (3). Resolution of secondary constipation—whether acute or chronic—requires addressing the primary etiology, potentially necessitating stool disimpaction, discontinuation of causative medications, or correction of structural colonic pathology. In contrast, primary chronic constipation represents a symptom-based disorder attributable to dysregulation of colonic motility, incoordination of anorectal neuromuscular function, and impaired brain–gut axis signaling (4). Based on distinct pathophysiological mechanisms, constipation is subcategorized into three subtypes: normal-transit constipation, slow-transit constipation, and rectal evacuation disorders, including dyssynergic and inadequate defecatory propulsion (5–7). Constipation may increase the risk of multiple diseases including Parkinson’s disease, multiple sclerosis, ischemic heart disease, depression or anxiety, and so on (8). Even among the diseases of gut-brain interactions, only constipation was significantly associated with mortality, while other diseases, including irritable bowel syndrome, chronic diarrhea, dyspepsia and abdominal pain, did not reduce survival rates (9). Stroke, as the most common severe neurological disorder, is frequently complicated by gastrointestinal dysmotility, with clinical studies reporting a constipation prevalence of 29–79% among affected individuals (10–12).

Although post-stroke constipation is a well-recognized complication, emerging clinical studies are now investigating the reverse association: whether constipation independently elevates stroke risk through mechanisms such as gut dysbiosis, gut-brain axis dysregulation, or shared vascular risk profiles (13–17). A previous meta-analysis (including 8 studies) indicated that constipation was associated with a higher risk of stroke. However, the study data included in this meta-analysis were limited to May 2024, and its subgroup analyses were limited to ischemic and mixed stroke subtypes, excluding hemorrhagic stroke (18). Some new studies involving relationship between constipation and risk of stroke were published from then on (14, 15). To obtain a more comprehensive estimate of the putative influence of the constipation on stroke, we performed a meta-analysis of published studies to determine the association between constipation and risk of stroke.

Our updated meta-analysis of 13 published studies confirmed a modest but statistically significant positive association between constipation and risk of stroke. Our results show that patients with constipation have a 1.23 times higher risk of stroke than those without constipation. Given the distinct pathophysiological mechanisms underlying hemorrhagic and ischemic stroke subtypes, we conducted a subgroup analysis. The disease-specific risk profile revealed a statistically significant association between constipation and ischemic stroke (OR = 1.39, 95% CI: 1.19–1.60), whereas no significant relationship was observed in hemorrhagic stroke (OR = 1.03, 95% CI 0.80–1.26). Interestingly, our subgroup analysis revealed a null association between constipation and stroke among women (OR = 1.00, 95%CI 0.92–1.07).

Stroke is the second most common cause of death and the leading cause of disability globally (29). Although significant progress has been made in the prevention and treatment of stroke in the past few decades, the burden will increase significantly due to factors such as population aging, especially in developing countries (29). Strokes manifest as two distinct pathophysiological entities - ischemic and hemorrhagic stroke - that while exhibiting divergent etiological pathways, share overlapping modifiable risk profiles. Both subtypes demonstrate strong epidemiological associations with hypertension, diabetes, smoking, hyperlipidemia, obesity, etc (30). Our results suggest that constipation is associated with the development of stroke. These findings indicate that the pathological characteristics of stroke may be accelerated by constipation, or the two may share etiological pathways.

The observed association between constipation and stroke may be mediated through multiple interrelated biological pathways. Firstly, the gut-brain axis, which serves as a pivotal bidirectional communication axis between the gut and the brain, may plays an important role in the connection between constipation and stroke. Constipation manifests as infrequent defecation, passage of hardened or lumpy stools, straining during elimination, sensation of incomplete evacuation, perception of anorectal obstruction/blockage, and necessity for manual maneuvers to enable defecation (31). Extended colon transit time during constipation may cause alteration of the intestinal microbiota (21). The gut microbiota has a key role in bidirectional interactions in the gut-brain axis, through modulation of central nervous system, immune cells, neuroendocrine cells, and peripheral neurons (32). In the gut-brain axis, migration of intestinal immune cells to the central nervous system may contribute to the pathogeneses of neurological and neurodegenerative diseases, such as stroke (33). Gut microbiota also play a pivotal role in Trimethylamine N-oxide (TMAO) production, augmented macrophage cholesterol accumulation and foam cell formation, all of which promote atherosclerosis progression (34). In addition, the prolonged colonic transit time in constipation may further promote the translocation of proinflammatory cytokines caused by gut bacteria and result in increased inflammatory responses and oxidative stress. So, patients with constipation may sustain a systemic low-grade inflammation state, which accelerates the development of atherosclerosis (35, 36). Secondly, chronic constipation might lead to subsequent persistent increases in blood pressure, through its induction of psychological stress, increased water absorption, and gut dysbiosis (16, 37). Constipation can also cause straining during defecation, during which patients may breathe in a strained manner similar to the Valsalva maneuver (38). This may induce transient blood pressure elevation, potentially triggering stroke occurrence. Third, serotonin (5-hydroxytryptamine, 5-HT) signaling represents a plausible mediating pathway. 5-HT is a key neurotransmitter in the brain-gut axis and is involved in several functions of the gastrointestinal tract including peristaltic reflexes (39). Current evidence remains equivocal regarding the expression of 5-HT increases or decreases in patients with constipation. Some studies suggest that patients with constipation have enhanced synthesis and release of 5-HT (40, 41). And 5-HT, as a vasoconstrictor that facilitates thrombus formation, is associated with the development of atherosclerotic plaques and elevated risks of atherosclerotic cardiovascular disease (42, 43). But a study conducted by Dunlop et al. indicates that relatively low post-prandial plasma 5-HT levels have been detected in patients with constipation-predominant irritable bowel syndrome (44). In generally, acute vascular constriction by 5-HT is usually shared by 5-HT_1B and 5-HT_2A receptors, except in intracranial arteries which constrict only through 5-HT_1B receptors (45). Studies have confirmed that 5-HT-induced vascular constrictions are mediated through 5-HT_1B receptors at low 5-HT concentrations (46–48). The serotonin pathway’s implications warrant further investigation.

Noteworthily, our meta-analysis demonstrated a significant correlation between constipation and elevated ischemic stroke risk, whereas no such association for hemorrhagic stroke. We hypothesize that the observed differences may be attributable to the distinct pathogenic mechanisms underlying the two stroke subtypes. The predominant mechanism underlying hemorrhagic stroke is hypertensive small-vessel disease, characterized by the formation of lipohyalinotic microaneurysms that eventually rupture (49). However, the majority of ischemic strokes are thromboembolic in nature, with common sources of embolism including large artery atherosclerosis and cardiac conditions, particularly atrial fibrillation (50). As previously established, chronic constipation contributes to the occurrence and development of atherosclerosis through mechanisms involving gut microbiota dysbiosis and additional biological pathways. Furthermore, Zhang et al. demonstrated that gut microbiota dysbiosis contributes to age-related atrial fibrillation via the lipopolysaccharide- and glucose-mediated activation of the NLRP3-inflammasome pathway (51). A two-sample Mendelian randomization analysis has provided additional evidence supporting a causal association between constipation and atrial fibrillation risk (52). In conclusion, constipation serves as a risk factor for ischemic stroke through multiple pathophysiological mechanisms.

Subgroup analyses demonstrated that constipation was not observed to increase stroke risk in both subgroups diagnosed with constipation according to stool traits and frequency criteria and self-reported criteria. Across three investigations employing stool traits and defecation frequency as diagnostic criteria, constipation conferred a 1.31-fold elevated stroke risk relative to non-constipated counterparts, albeit without statistical significance (8, 17, 24). We believe the underlying reason is that the number of participants was limited in these three studies, whose pooled analytical weight contributed minimally (6.87%) to the aggregate meta-analysis estimate. The two investigations employing self-reported criteria enrolled substantially larger cohorts, collectively contributing 27.68% to the aggregate meta-analytic weight (14, 21). Salmoirago-Blotcher et al. collected information on constipation by self-administered questionnaire. In this study, constipation was defined as “difficulty having bowel movements” over the previous 4 weeks, was rated using a scale ranging from none (did not occur), (mild did not interfere with usual activities), moderate (interfered somewhat with usual activities), or severe (symptom was so bothersome that usual activities could not be performed) (21). Park et al. defined constipation using the total number of prescribed laxatives ≥180 during the 1-year baseline period (14). Self-reported constipation exhibits lower diagnostic accuracy than symptom-based criteria (e.g., defecation frequency or Rome classifications), resulting in inflated prevalence estimates that may attenuate constipation-stroke risk associations.

Interestingly, our meta-analysis results show that constipation is not statistically significantly associated with stroke risk among female participants, and the reasons behind this are worth exploring. We hypothesized that pathophysiological mechanisms underlying the predominant subtype of constipation in female patients play a key important role in this phenomenon. As we know, the clinical spectrum of constipation encompasses three distinct phenotypes: (1) dyssynergic defecation disorder characterized by impaired rectal evacuation, (2) colonic dysmotility manifesting as slow-transit constipation, and (3) normal transit constipation presenting without discernible defecatory dysfunction or delayed colonic transit (53). Epidemiologic data consistently identify female sex as a predisposing factor for chronic constipation development (54). Compared with men, women are more likely to have findings suggestive of a functional defecatory disorder (55). And among women, dyssynergic defecation is the most prevalent subtype and slow transit constipation without dyssynergic defecation is uncommon (56). We postulate that, reduced fecal colonic residence time in dyssynergic defecation patients may confer relative protection against gut microbiota dysbiosis, compared to slow transit constipation cohorts, potentially mediated by diminished microbial fermentation duration. Thus, the reduced probability of gut microbiota dysbiosis may attenuate the association between constipation and stroke risk in women.

Our meta-analysis confirms the association between constipation and stroke risk, which raises interesting questions regarding the potential impact of therapeutic interventions on modifying stroke risk. The Danish population-based study revealed that, compared to the general population cohort, high-intensity laxative use (≥2 prescriptions) was associated with a 3.2-fold increased risk of ischemic stroke, low-intensity users(0–1 prescriptions) was 1.39 (28). Similarly, the US veteran cohort found that, compared to patients without constipation, patients using ≥2 types of laxatives faced a 21% higher stroke risk, patients using one faced a 19% (23). The observed associations potentially arise from two non-exclusive pathways: (1) intensified pharmacotherapy for constipation correlates with greater baseline disease severity, and (2) laxative regimens may inadvertently contribute to stroke pathogenesis through enhanced serotonin bioavailability or laxative-induced dehydration (23). Therapeutic optimization remains paramount, as first-line medical interventions (e.g., fiber supplementation, laxatives) often fail in refractory cases, particularly among patients with functional defecation disorders (e.g., dyssynergia). Management efficacy fundamentally depends on underlying etiology, with an example that biofeedback therapy is the preferred treatment for constipated patients with dyssynergia (57). Future prospective studies should evaluate whether etiology-targeted therapies—including intestinal microbiome modulation via probiotics, biofeedback protocols, or prokinetic agents—mitigate stroke incidence in constipated cohorts.

In interpreting the results, some limitations of this meta-analysis should be considered. First of all, substantial heterogeneity was observed in the present study, which may be related to gender, study type, region, stroke subtype, constipation definition, follow-up time, and so on. We observed a decrease in heterogeneity in subgroup analyses based on sex, region, constipation definition, and study type, which confirmed our hypothesis. We did not observe an association between constipation and stroke risk in studies from Asia and the America, which may be related to confounding factors such as stroke type and gender. However, due to the limited number of studies and the difficulty of obtaining detailed data, we did not perform further subgroup analysis. Notably, in the meta-analysis based on these cohort studies, constipation was still significantly associated with stroke risks, suggesting that our results are relatively reliable. Second, due to the small number of studies included in the subgroup analyses based on stroke subtypes, more studies are needed to determine the reliability of the association between constipation and different stroke subtypes. Third, this is a study-level meta-analysis, not an individual patient meta-analysis, which limits the ability to adjust for confounders at the participant level. Fourthly, our restricted database retrieval (not including Scopus/Embase) and exclusion of some types of literature (conference abstracts/non-English publications) may introduce selection bias through omitted eligible studies. Finally, sex-based subgroup studies are limited and we did not analyze constipation and stroke risks in men, which may inspire future research explorations.

The quality of the studies was assessed via means of the NOS, and all studies included in the analysis had scores ≥6 points. Therefore, the resultant risk estimate resulting from this analysis may also be deemed as fairly stringent estimate. It’s worth noting that these risk estimates demonstrate greater susceptibility to underestimation than overestimation of the true constipation-stroke association magnitude, attributable to preferential adoption of conservative constipation definitions. This is substantiated by subgroup analyses wherein strictly-defined constipation cohorts (ICD criteria/ stool traits and frequency criteria groups) demonstrated elevated stroke risk relative to loosely-defined populations (frequency criteria/ self-reported criteria groups). In future studies, as the number of studies permits, we recommend using universal definitions of constipation whenever possible, such as meeting the Rome IV criteria, to better determine the association between constipation and stroke risk.

Our meta-analysis demonstrated a significant positive association between constipation and stroke risk. Disease-specific stratification revealed elevated ischemic stroke risk but non-significant hemorrhagic stroke association. Notably, subgroup analyses showed complete risk attenuation in female populations. These findings underscore constipation as a modifiable risk factor in ischemic stroke management. Future studies on the risk of stroke in patients with constipation, and the explanatory factors for this association, are warranted in order to confirm and expand these results.