Constipation and risk of death and cardiovascular events in patients on hemodialysis
Article information
Abstract
Background
Constipation is a common gastrointestinal disorder and is often accompanied by alteration in the gut microbiota. Recently, several studies have disclosed its association with an increased risk of cardiovascular disease and mortality in the general population. Despite the high prevalence of constipation, data on the clinical impact of constipation in patients with chronic kidney disease are limited. We aimed to explore the prevalence of constipation and its association with cardiovascular disease in chronic kidney disease using a nationally representative cohort of hemodialysis patients.
Methods
This study used hemodialysis quality assessment and health insurance claims data from patients undergoing maintenance hemodialysis in South Korea. Chronic constipation was defined using the total number of laxatives prescribed during the 1-year baseline period. The primary outcome was a composite of acute ischemic stroke, hemorrhagic stroke, myocardial infarction, or all-cause death. Secondary outcomes were the individual components of the primary outcome.
Results
Among 35,230 patients on hemodialysis, 9,133 (25.9%) were identified as having constipation. During a median follow-up of 5.4 years, patients with constipation had a 15% higher incidence of the composite outcome, 16% higher incidence of ischemic stroke, and 14% higher all-cause mortality, after multivariate adjustment.
Conclusion
Chronic constipation requiring laxatives was associated with a higher risk of the composite outcome of cardiovascular events or all-cause death in patients on hemodialysis. Further studies are needed to confirm whether constipation is an independent predictor or a possible causal factor of cardiovascular disease.
Introduction
Constipation is a common gastrointestinal disorder and is often accompanied by alteration in the gut microbiota [1]. Although constipation is generally perceived as a benign and trivial condition, its chronic symptoms affect the patients’ quality of life and may place a considerable health economic burden [2,3]. Recently, several studies have demonstrated that constipation is independently associated with adverse outcomes, including chronic kidney disease (CKD) progression, cardiovascular disease, and mortality [4–7]. In a cohort of 45,112 Japanese individuals, Honkura et al. [8] found that those with constipation, assessed using a self-administered questionnaire, had a higher risk of cardiovascular mortality. Recently, in a cohort of United States veterans with an estimated glomerular filtration rate ≥60 mL/min/1.73 m2 (n = 3,359,653), those with constipation defined by the diagnostic code or laxative use had a higher incidence of coronary heart disease and ischemic stroke and also experienced a higher all-cause mortality [6]. Subsequently, in a nationwide cohort of the Danish population, constipation assessed using diagnostic codes was associated with an increased risk of myocardial infarction, ischemic stroke, hemorrhagic stroke, peripheral artery disease, and heart failure in the short term, and with ischemic stroke and venous thromboembolism beyond 5 years of follow-up [7]. In another large-scale prospective cohort study conducted in the United Kingdom (n = 502,229), regular use of laxatives was associated with a higher risk of dementia [9].
The prevalence of constipation is thought to be higher in CKD patients than in the general population, presumably due in part to dietary restrictions (e.g., limited fiber and fluid intake), comorbidities (e.g., diabetes mellitus), chronic medications (e.g., potassium-binding resin, phosphate binder), and altered gut microbiome. In a review by Murtagh et al. [10], constipation was identified as the third most common symptom, after fatigue and pruritus, with a prevalence of 57% in end-stage kidney disease (ESKD) patients. However, despite its high prevalence and possible connection with adverse clinical outcomes, we have a limited understanding of constipation, and there are very few studies in patients with ESKD.
In this study, we aimed to investigate the current status of laxative prescriptions in dialysis patients and their association with cardiovascular disease and mortality using a large nationally representative cohort of Korean hemodialysis (HD) patients.
Methods
Data source and study population
This study included outpatients undergoing maintenance HD in South Korea. We used the HD quality assessment data collected by the Korean Health Insurance Review and Assessment Service (HIRA). Public medical insurance is mandatory in Korea, and HIRA is an independent agency that reviews and assesses healthcare benefit costs. HIRA periodically conducts HD quality assessments in all HD facilities that claim HD fees to manage the quality of dialysis institutions. For the HD quality assessment, patients with ESKD who underwent HD at least twice a week were enrolled. Those admitted to the hospital at least once, did not undergo follow-up, or died during the assessment period were excluded.
The study population was based on the 2015 HD Quality Assessment (October 1 to December 31, 2015). We excluded patients aged <20 years and those undergoing dialysis for <3 months in this study. During the HD quality assessment, HIRA collected clinical data of each patient, such as dialysis vintage, height, body weight, blood pressure, and laboratory data, including plasma hemoglobin, serum albumin, phosphorus, total calcium, iron saturation, and ferritin levels. We also obtained data on comorbidities, medical procedures, and drug prescriptions from the HIRA claims database.
This study was approved by the Institutional Review Board (IRB) of Myongji Hospital in Goyang, South Korea (No. MJH2023-11030). Owing to the retrospective and de-identified nature of the study, the IRB waived the need for written consent from patients. The ethical guidelines of the Declaration of Helsinki were followed throughout the study.
Comorbidities
Chronic constipation has been defined as “unsatisfactory defecation for at least 3 months” [11]. In the present study, we defined chronic constipation using the total number of prescribed laxatives ≥180 during the 1-year baseline period (January 1 to December 31, 2015), considering many patients take laxatives more than two tables (packs) per day and chronic constipation does not usually end at 3 months. Laxatives collected from the HIRA claims database are listed in Supplementary Table 1 (available online). Preexisting comorbidities were identified based on the relevant International Classification of Diseases, 10th Revision (ICD-10) codes, with more than two diagnoses during admission or at outpatient clinics, during the baseline period (Supplementary Table 2, available online).
Outcomes
The primary outcome was a composite of acute ischemic stroke, hemorrhagic stroke, myocardial infarction, or all-cause death. Secondary outcomes were the individual components of the primary outcome in patients without such diagnoses. Each event was defined using the relevant ICD-10 diagnostic codes, accompanied by the related procedure codes, from January 1, 2016, to June 30, 2021 (Supplementary Table 2, available online). Specifically, ischemic stroke was defined by hospitalization diagnostic codes I63–I64 (ICD-10) with concomitant brain imaging [12,13], whereas hemorrhagic stroke was defined as hospitalization diagnostic codes I60–I61 (ICD-10) [7,14] with claims for brain imaging. Myocardial infarction was defined using cases with both hospitalization diagnostic codes I21–I23 (ICD-10) and procedure codes such as coronary angiography, stent insertion, and bypass surgery [15,16].
Statistical analysis
All statistical analyses were performed using the R software (version 4.0; R Foundation for Statistical Computing). Data are presented as mean with standard deviation or absolute numbers with percentages. Comparisons between the groups were performed using an independent t test or chi-square test, as appropriate. Univariate and multivariate Cox proportional hazards regression models were used to identify risk factors for clinical endpoints, including the presence of constipation. The following variables were included in the adjusted model. Model 1 was adjusted for age, sex (male), and dialysis vintage. Model 2 was adjusted for medical comorbidities (diabetes mellitus, hypertension, ischemic heart disease [IHD], congestive heart failure [CHF], cerebrovascular disease [CVD], atrial fibrillation or flutter [Af], and malignancy), in addition to the factors included in model 1. Model 3 analysis was performed on variables with p-values of <0.05 in univariate analysis. All tests were two-tailed, with p-values of <0.05 indicating statistical significance.
Results
Prevalence and characteristics of patients with constipation
In total, 35,230 patients were included in this study (Fig. 1). The mean age was 60.1 years; 20,716 patients (58.8%) were male, and the mean dialysis vintage was 5.6 years. Supplementary Table 3 (available online) shows the prevalence of constipation according to various criteria, ranging from 20.0% to 30.7% in the HD population. In this study, 25.9% (n = 9,133) of the patients on HD were identified as having constipation (25.4% in males and 26.7% in females). Among the 9,133 patients prescribed laxatives, 3,382 (37%) were prescribed more than two types of laxatives during the study period. The most commonly prescribed laxatives were lactulose/lactitol, followed by bulk-forming agents and magnesium oxide/hydroxide (Supplementary Table 4, available online).

Study cohort.
HD, hemodialysis; HS, hemorrhagic stroke; IS, ischemic stroke; MI, myocardial infarction.
The baseline characteristics categorized according to constipation status are shown in Table 1. Patients with constipation were more likely to be older, have a lower dialysis vintage, and have a higher body mass index. They had a higher prevalence of comorbidities and consumed higher amounts of oral iron supplements, potassium binders, and opioids. The use of statin, aspirin, angiotensin-converting enzyme inhibitor or angiotensin receptor blocker, and beta-blocker was also more common in patients with constipation.
Clinical outcomes
Primary outcome
After a median follow-up of 5.4 years, a total of 13,470 composite events were recorded (crude rate, 324.0 per 1,000 person-years; 95% confidence interval [CI], 319.0–329.5). Compared to patients without constipation, those with constipation had a higher risk of the primary outcome (hazard ratio [HR], 1.54; 95% CI, 1.48–1.5; p < 0.001) (Supplementary Table 5, available online). Although this association was attenuated after multivariate adjustment, the risk of primary outcome remained higher in patients with constipation than in those without constipation (adjusted HR, 1.22; 95% CI, 1.17–1.26 in model 2 and adjusted HR, 1.15; 95% CI, 1.11–1.20 in model 3) (Table 2).
All-cause death
A total of 12,118 patients died during the follow-up period (crude rate, 344.0 per 1,000 person-years; 95% CI, 339.0–348.9). Patients with constipation had a higher incidence of all-cause mortality than those without (HR, 1.61; 95% CI, 1.55–1.67; p < 0.001) (Supplementary Table 5, available online). Even after multivariate adjustment, the risk of death remained higher in patients with constipation (adjusted HR, 1.23; 95% CI, 1.18–1.28 in model 2 and adjusted HR, 1.14; 95% CI, 1.10–1.20 in model 3) (Table 2).
Ischemic stroke
A total of 1,193 patients experienced incident ischemic stroke (crude rate, 33.3 per 1,000 patient-years; 95% CI, 30.9–35.7). The crude incidence of ischemic stroke was higher in patients with constipation than those without (HR, 1.54; 95% CI, 1.36–1.73; p < 0.001) (Supplementary Table 5, available online). The risk of incident ischemic stroke remained higher in patients with constipation after multivariate adjustment (adjusted HR, 1.26; 95% CI, 1.11–1.42 in model 2 and adjusted HR, 1.16; 95% CI, 1.02–1.32 in model 3) (Table 2).
Hemorrhagic stroke
In total, 863 hemorrhagic strokes occurred, and the crude rate was 24.5 per 1,000 person-years (95% CI, 22.6–26.5). Constipation was not significantly associated with hemorrhagic stroke in HD patients (Table 2).
Myocardial infarction
During the follow-up period, 1,172 patients experienced myocardial infarction (crude rate, 33.9 per 1,000 person-years; 95% CI, 31.6–36.2). The crude cumulative incidence of myocardial infarction was higher in patients with constipation than in those without constipation (adjusted HR, 1.34; 95% CI, 1.18–1.32) in the univariate analysis. However, the risk was not significant after multivariate adjustment in models 2 and 3 (p = 0.07 in model 2 and p = 0.52 in model 3) (Table 2).
Discussion
In this nationwide cohort of 35,230 patients undergoing HD, we described the current status of laxative prescriptions and demonstrated an association of constipation with adverse clinical outcomes. Overall, 25.9% of HD patients had constipation (25.4% of males and 26.7% of females), and the most commonly prescribed laxatives were lactulose/lactitol. Additionally, after adjusting for potential confounders, patients with constipation had a higher risk of the primary composite outcome of stroke, myocardial infarction, or all-cause death than those without constipation.
Chronic constipation is defined as “unsatisfactory defecation characterized by infrequent stool, difficult stool passage, or both at least for the previous 3 months” [11]. The Rome criteria composed of constipation-related symptoms [17] or the Bristol Stool Form Scale based on visual inspection of feces [18], are often used to estimate constipation [19]. The reported prevalence of constipation varies across studies, with females and older individuals being more affected [20]. In a systematic review by Mugie et al. [21], the prevalence of constipation in the general population ranged from 0.7% to 79%, with a median of 16%. Diverse methodologies for data collection were used in the included studies, a self-administered questionnaire (n = 44) comprising the majority of the 68 studies, followed by face-to-face interviews, telephone surveys, and chart reviews [21]. Therefore, the wide variability in prevalence is probably due to different definition methods as well as the different cultures and diets of each study population [21].
The prevalence of constipation is assumed to be higher in CKD patients than in the general population [22]. Recently, Sumida et al. [23] examined the use of laxatives in patients with advanced CKD transitioning to ESKD in a retrospective cohort of United States veterans. They found that the proportion of patients prescribed laxatives increased as the disease progressed to ESKD, peaking at 37.1%. Zhang et al. [24] reported that the frequency of constipation was 71.7% in HD (n = 478) and 14.2% in peritoneal dialysis patients (n = 127) using the Rome III criteria [25]. Recently, in a large cohort of 12,217 HD patients in the Japan-Dialysis Outcomes and Practice (J-DOPPS) study, Honda et al. [26] demonstrated a prevalence of 30.5% when constipation was defined based on laxative prescriptions. In their study, stimulants such as bisacodyl and sennoside were the most common laxatives, followed by selective serotonin 5-HT4 receptor agonists and hyperosmotic agents. Notably, laxative use was associated with a greater risk of mortality, and similar results were observed regardless of the type of laxatives. In our study, oral stimulants (e.g., bisacodyl and sennoside) and stool softeners (e.g., docusate) were not included in the analysis because they are not covered by insurance in Korea.
In terms of baseline characteristics, patients with constipation were older and had a higher prevalence of comorbidities, including diabetes mellitus, hypertension, IHD, CHF, CVD, Af, and malignancy, which is in accordance with previous studies [6,26]. Patients with constipation were also taking more prescribed drugs including iron supplements, potassium-binding resins, and opioids, which are well-known constipation-causing drugs. Contrary to our expectations, phosphate binders were prescribed less frequently in the constipation group. Phosphate binders included calcium carbonate/acetate, sevelamer, lanthanum, and aluminum hydroxide in this study, and it has been reported that lanthanum and sevelamer could induce both constipation and diarrhea [27]. It is also possible that patients without constipation may have lower gastrointestinal discomfort and more food intake, resulting in the use of more phosphate binders because their phosphorus levels were greater than those of patients with constipation (5.02 ± 1.33 vs. 4.73 ± 1.30, p < 0.001).
Most importantly, constipation was associated with a significantly increased risk of the composite outcome of cardiovascular events or death after adjusting for potential confounders. In addition, constipation was associated with a higher risk of ischemic stroke and all-cause mortality. To the best of our knowledge, only one cohort study has examined the relationship between constipation and clinical outcomes in the CKD population [11]. In their prospective cohort of 12,217 patients on HD enrolled in the J-DOPPS study, constipation requiring laxatives was associated with an increased risk of all-cause mortality within 3 years. Our results are consistent with earlier reports that showed a significant association with ischemic stroke [6,7], all-cause mortality [6], and cardiovascular mortality [8] in the general population. Several studies have also reported a significant association with incident myocardial infarction [7,28] and coronary heart disease [6]. However, in the present study, the association between constipation and myocardial infarction was not significant after multivariate adjustment. The main pathophysiology of both coronary heart disease and ischemic stroke is based on atherosclerosis. However, some differences exist. Rupture or erosion of vulnerable plaques in the coronary arteries, resulting in occlusion, is the major cause of myocardial infarction. Ischemic stroke, on the other hand, shows a more heterogeneous etiology and more than 20% of cases are caused by large-artery atherosclerosis [29,30] compared with myocardial infarction. Moreover, although both entities have similar risk factors, some differences are observed in the magnitude of associations. Hypercholesterolemia [31,32], diabetes mellitus [31], and current smoking [32] were more strongly associated with coronary heart disease than with ischemic stroke in the previous studies. This may reflect the distinct biology of different vascular beds and partially explain the inconsistent results for each outcome. Although we cannot conclude the association with each cardiovascular event at present, it should be also noted that each study used different methods for defining constipation and outcomes as well as different study populations.
There are several speculative explanations for the mechanisms by which constipation increases cardiovascular risk and mortality. Constipation is characterized by infrequent bowel movements, and increased colonic transit time in constipation may cause alteration of the intestinal microbiota. Recently, the contributory role of gut dysbiosis in the development of immune dysfunction, oxidative stress, and cardiovascular disease has been observed consistently [33–35]. Given that constipation is one of the clinical forms of gut dysbiosis [1,36], it seems plausible that constipation partially contributes to the development of atherosclerosis by inducing low-grade chronic inflammation via bacterial endotoxins and/or altered gut metabolites. In particular, emerging evidence has revealed that gut dysbiosis is one of the main causes of chronic inflammation in CKD, which leads to the progression of cardiovascular disease and influences patient morbidity and mortality [37]. Second, constipation leads to straining of the stool and breathing, similar to the Valsalva maneuver, which is associated with transient increases in blood pressure and can sometimes conceivably trigger a cardiovascular event [38]. Moreover, constipation could contribute to decreased dietary intake, impaired intestinal absorption, and subsequent risk of malnutrition. Although constipation remained a significant risk factor after adjusting for hypoalbuminemia, it is still plausible that constipation can lead to nutritional deficiency and adverse outcomes.
This study has several limitations. First, our study was observational in nature, which does not provide causality and cannot eliminate the possibility of residual confounding factors. In addition, the outcomes were not controlled for the use of iron supplements, potassium binders, phosphate binders, and opioids which showed differences in baseline characteristics between the two groups. Second, laxative prescriptions were used to define constipation, and information on actual stool patterns or subjective symptoms was unavailable. Previous clinical studies assessed constipation using self-administered questionnaires [5], diagnostic codes [6,7,28], or laxative use [6,26]. Each method has its advantages and disadvantages. In the general population, only a few patients with constipation would seek medical care. Thus, defining constipation based on laxative prescriptions is likely to miss a considerable number of patients with constipation, leading to information bias. However, dialysis patients in Korea receive a specialized health benefit that pays only 10% of the cost of all insurance-covered drugs. Hence, patients undergoing HD in Korea can easily request drug prescriptions, leading to a low probability of missing patients with constipation. Third, the HIRA claims data contain information on insurance-covered drugs only; therefore, we could not include information on non-covered laxatives, such as docusates, sennosides, and oral bisacodyl. However, owing to their low cost and frequent contact with physicians, HD patients usually take non-reimbursed or over-the-counter laxatives as additive medications when insurance-covered drugs are insufficient for symptom relief. Therefore, we speculate that the proportion of patients who used only non-reimbursed laxatives and were misclassified into the non-constipation group would be low. Fourth, this study evaluated the laxative use of patients during the baseline period of 1 year and did not follow changes in constipation status. A laxative prescription may have been initiated or stopped after the baseline period. Research assessing changes in constipation over time and related outcomes would be valuable, providing more insight into the impact of constipation on clinical outcomes. Finally, cause-specific mortality data were unavailable for this cohort.
Despite these limitations, we believe that this study is important because it addresses the potentially deleterious effects of constipation in patients undergoing HD. Only one previous study suggested a link between constipation and mortality in dialysis patients [26]. To the best of our knowledge, this is the first study to evaluate the association between constipation and cardiovascular diseases in a dialysis population. Second, this study used a national cohort from Korea that included almost all outpatients on HD with nearly complete follow-up, unlike previous studies that enrolled the general population with only veterans (mostly males), females, or the elderly [5,6,28]. Third, we restricted the definition of incident cardiovascular events to cases with both diagnostic codes and relevant procedure codes such as brain imaging and coronary procedures, to decrease false positives, unlike previous studies that used only diagnostic codes to define primary endpoints [7,28]. In addition, most studies did not report on or effectively control for confounding factors such as biochemistry [6–8,28] and various medications [8,11,28]. Our study revealed that constipation may be a risk factor for cardiovascular disease and mortality even after adjusting for numerous factors.
In conclusion, chronic constipation requiring laxative use was associated with an increased risk of composite outcome of cardiovascular events and all-cause mortality in patients undergoing HD. Further studies are needed to confirm whether constipation is an independent predictor or possible causal factor of cardiovascular disease and to elucidate the mechanisms underlying these associations.
Supplementary Materials
Supplementary data are available at Kidney Research and Clinical Practice online (https://doi.org/10.23876/j.krcp.24.174).
Notes
Conflicts of interest
All authors have no conflicts of interest to declare.
Acknowledgments
We really appreciate to have opportunity to participate in Joint Project on Quality Assessment Research in 2022.
Data sharing statement
The data presented in this study are available from the corresponding author upon reasonable request.
Authors’ contributions
Conceptualization: All authors
Formal analysis: SCP
Methodology: HMC, YKL, YEK, DHK
Supervision: HMC, YKL
Writing–original draft: JJ
Writing–review & editing: HMC, SIB, DJO
All authors read and approved the final manuscript.