Abstract
Purpose
Hemorrhoidal disease (HD) and varicose veins (VV) are both common diseases. The aim was to investigate the association between them.
Methods
A nationwide cohort study was conducted using the Taiwan National Health Insurance Research Database (NHIRD) from 2005 to 2021, which included information of over 31 million beneficiaries. To assess the risk of developing VV among HD patients, records of 832,310 patients with HD diagnoses (the HD cohort) and those of propensity score-matched patients without HD diagnoses (the non-HD cohort) were identified and compared. Hazard ratios (HRs) and 95% confidence intervals (CIs) derived from the Cox proportional hazards model were used to estimate the association risks. Similarly, an opposite direction was approached using records of 112,027 patients with VV diagnoses (the VV cohort) to assess the risk of developing HD among VV patients.
Results
HD was associated with an increased risk of VV (adjusted HR, 1.52; 95% CI: 1.47–1.57). Similarly, VV were associated with an increased risk of HD (adjusted HR, 1.50; 95% CI: 1.45–1.55). The associations between the two diseases were evident in both sexes and all age groups. Patients with comorbid HD and VV had higher incidences of mitral valve regurgitation (P < 0.001), hernia (P < 0.001), varicocele (P = 0.008), and mortality (P = 0.006).
Conclusion
This study revealed a significant bidirectional association between HD and VV. It is recommended that physicians be mindful of either condition in patients already diagnosed with the other, as their coexistence could imply the potential for additional associated connective tissue diseases and adverse long-term outcomes.
Supplementary Information
The online version contains supplementary material available at 10.1007/s00384-025-04889-8.
Keywords: Hemorrhoidal disease, Mitral valve regurgitation, National health programs, Varicose veins
Introduction
Hemorrhoidal disease (HD) and varicose veins (VV), the most prevalent form of chronic venous disease (CVD), vary widely in occurrence: VV affects < 1% to 73% of females and 2% to 56% of males, while HD prevalence ranges from 4.4% to 36%, influenced by diagnostic criteria [1–5]. Both conditions cause significant discomfort and impact quality of life, leading to considerable social and economic costs [6–8].
HD and VV share similar risk factors. For VV, it is well known that age, family history, obesity, constipation, female sex, and standing occupation are significant risk factors [2], while for HD, constipation, diarrhea, low-fiber diet, pregnancy, age, family history, and any activities increasing intraabdominal pressure are reported [9]. The CHORUS study (Chronic venous and HemORrhoidal diseases evalUation and Scientific research) found 51.2% of HD patients showed chronic venous disease (CVD) symptoms or signs [10]. Moreover, the coexistence of the two diseases is associated with either HD severity or recurrence. The importance of examination for CVD in HD patients is emphasized, especially when shared risk factors are present.
Observational studies confirm the association between HD and VV, with HD surgery patients showing a higher incidence of VV, hinting at chronic constipation as a shared risk factor [11]. Similarly, patients with HD often report poor VV symptom scores on the validated Venous Insufficiency Epidemiological and Economic Study-Quality of Life/Symptoms (VEINES-QoL/Sym) questionnaire [12, 13]. Furthermore, research also shows a positive correlation between varicocele and other venous diseases including HD and VV, suggesting a common pathogenesis of venous dilatation [14].
The association between HD and VV is further implied by a common medical treatment modality. Micronized purified flavonoid fraction (MPFF), also known as Daflon, has been found effective in treating both diseases with different recommended dosages [15, 16]. Although several studies have been conducted to delineate the associations between HD and VV [10–12, 14], the evidence is still considered limited, and no studies to date have examined HD and VV in a Taiwanese population. Moreover, previous research has primarily examined the interrelationship between HD and VV through the lens of HD. Notably, there appears to be a gap in the literature regarding studies conducted from the perspective of VV. To address this shortfall, our study aimed to rigorously assess this association from both perspectives, thereby offering a more comprehensive understanding of the bidirectional risks between HD and VV within a Taiwanese population.
Therefore, the aims of the study were to examine the association between HD and the risk of VV in a Taiwanese population-based cohort and to explore the reciprocal relationship from the viewpoint of VV. We hypothesized that individuals with HD were at an increased risk of developing VV compared to those without it, and conversely, that individuals with VV were at an increased risk of developing HD.
Materials and methods
Data collection
This retrospective cohort study was conducted nationwide and relied on medical claims data from the Taiwan National Health Insurance Research Database (NHIRD). The National Health Insurance (NHI) system in Taiwan covers almost the entire population, representing about 23 million individuals or 99% of the total residents. The NHIRD offers a vast array of healthcare data, encompassing insured persons’ demographics, outpatient services, hospital stays, disease classifications, and medication records. It employs diagnostic classifications in accordance with the International Classification of Diseases, 9th and 10th Revisions, Clinical Modification (ICD-9-CM and ICD-10-CM). The database supports extensive epidemiological studies [17–20] and validates diagnostic, prescription, and hospitalization records [21–23]. Personal identifiers in the data are encrypted to ensure anonymity. This study’s data spanned from January 1, 2005, to December 31, 2021, involving over 31 million beneficiaries, and were sourced from the NHIRD. As the data, provided by the Health and Welfare Data Science Center (HWDC), contained only anonymized details of insured persons, the need for patient consent was obviated. Ethical approval was obtained from the Fu Jen Catholic University’s Institutional Review Board (IRB No. C110199).
Study cohorts
Patients with a primary HD diagnosis, identified by ICD-9-CM codes 455.xx and ICD-10-CM codes K64.xx, between January 1, 2005, and December 31, 2021, formed the HD cohort. They were compared with a control group (non-HD cohort) of individuals without an HD diagnosis. For accurate HD identification, patients needed at least three outpatient diagnosis records or one from an emergency/inpatient department. The index date was the first HD diagnosis for the HD cohort, with a corresponding medical consultation date for the non-HD cohort. The study included 866,867 newly diagnosed HD patients within the NHIRD during the specified period. VV diagnosis was categorized under ICD-9-CM 454.xx and ICD-10-CM I83.xx codes. Exclusions were made for patients aged < 20 years (N = 11,660), unknown sex (N = 15,294), pre-enrollment VV diagnosis (N = 7,803), pre-enrollment VV surgery (N = 1,880), or unknown identity (N < 3), resulting in 832,701 patients in the HD cohort.
To reduce the effects of confounding variables, propensity score matching was applied to the remaining participants. This matching considered demographic factors (age, sex) and medical comorbidities including hypertension, hyperlipidemia, diabetes mellitus, chronic obstructive pulmonary disease (COPD), cancer, atrial fibrillation, heart failure, ischemic heart disease, chronic kidney disease, and stroke. These are common factors used in propensity score matching that significantly impact disease risk, treatment responses, and health outcomes. In addition, constipation and diarrhea were included as confounding factors because both conditions can increase intraabdominal pressure. This may exacerbate venous pressure and affect venous return from the pelvic floor or lower limb venous system [2, 9]. Seven other confounding factors, including varicocele, mitral valve regurgitation, pelvic congestion syndrome, tricuspid valve insufficiency, aortic dissection, aortic aneurysm, and hernia, were also considered. These are associated with connective tissue weakness [14, 24–27], which might influence the formation of both HD and VV. Therefore, they were included as confounding variables. The ICD-9-CM and ICD-10-CM codes for all selected medical comorbidities are listed in the Appendix. This matching strategy enabled the selection of one control for each case, leading to a final total of 832,310 patients for both the HD and non-HD cohorts. The process of patient selection is illustrated in the left portion of Fig. 1.
Fig. 1.
The flow chart for the whole study. Left side is for the HD and the non-HD cohorts. Patients were followed until VV diagnosis, death, or December 31, 2021, whichever occurred first. Right side is for the VV and the non-VV cohorts. Patients were followed until HD diagnosis, death, or December 31, 2021, whichever occurred first. HD: hemorrhoidal disease; NHIRD: National Health Insurance Research Database; VV: varicose veins
Afterwards, we aimed to examine the incidence of HD from the viewpoint of VV. The VV cohort was constructed likewise, and compared to a control group, which consisted of patients without a VV diagnosis, with HD (ICD-9-CM 455.xx and ICD-10-CM K64.xx) being the outcome of interest. The exclusion criteria and matching methodology as previously described were replicated. The process for patient enrollment is depicted in the right side of Fig. 1. The HD cohort and the VV cohort were distinct from each other. In the HD cohort, any cases with a prior diagnosis of VV or VV surgery were excluded. This criterion ensured that all subsequent VV diagnoses occurred after HD diagnosis. Conversely, in the VV cohort, cases were excluded if HD diagnosis or surgery preceded the diagnosis of VV, ensuring that all subsequent HD diagnoses occurred after VV diagnosis. The different sequences of disease acquisition in each cohort confirmed their distinctness and negated any concern regarding patient overlap. The exclusion criteria are detailed in Fig. 1.
Statistical analysis
Chi-square and t-tests were employed to assess the variations in categorical and continuous variable distributions across the cohorts. Cox proportional hazards models was utilized to evaluate the risk between the two diseases, adjusting for age, sex, and comorbid conditions like hypertension, diabetes mellitus, hyperlipidemia, ischemic heart disease, COPD, chronic kidney disease, atrial fibrillation, heart failure, stroke, cancer, constipation, diarrhea, tricuspid valve insufficiency, aortic pathologies (including aortic dissection and aneurysm), and hernia. The Kaplan–Meier method, accompanied by the log-rank test, was applied to determine the variance in cumulative risks between the cohorts.
A further analysis was performed within the subgroups of VV presence between the HD versus the non-HD cohort. To evaluate the impact of HD, we focused on five outcomes: mitral valve regurgitation, hernia, varicocele, aortic pathologies, and mortality. The Chi-square test and the Cox proportional hazards model were utilized. The primary HD diagnosis date was set as the index date for HD patients (the HD cohort), with corresponding dates aligned for the non-HD patients. A 2-tailed P value of less than 0.05 was considered statistically significant. These analyses were executed using the SAS 9.4 software (SAS Institute, Cary, NC, USA).
Data availability statement
Access to these data is subject to restrictions. The data, sourced from the National Health Insurance database, can be provided by the authors upon receiving authorization from the National Health Insurance Administration of Taiwan.
Results
Patient characteristics of the HD and non-HD cohorts
Table 1 details the demographics and clinical profiles of both cohorts. Post-propensity score matching showed no significant differences in age (P = 0.622), sex (P = 0.928), or comorbidities including diabetes (P = 0.791), hyperlipidemia (P = 0.068), COPD (P = 0.658), cancer (P = 0.114), and constipation (P = 0.770) between the HD and non-HD cohorts. The majority were male (55.1%) with an average age of around 52 years in both cohorts.
Table 1.
Demographics of HD and non-HD cohorts after propensity score matching
| HD cohort | Non-HD cohort | ||
|---|---|---|---|
| Variable | No. = 832,310 | No. = 832,310 | P-value |
| Follow-up duration, mean ± SD (years) | 6.22 (4.27) | 5.56 (3.83) | < 0.001 |
| Age, mean ± SD (years) | 52.36 (16.43) | 52.30 (16.32) | 0.010 |
| Age, | 0.622 | ||
| 20–39 | 206,975 (24.9) | 207,301 (24.9) | |
| 40–59 | 343,489 (41.3) | 342,876 (41.2) | |
| ≥ 60 | 281,846 (33.9) | 282,133 (33.9) | |
| Women, No. (%) | 373,723 (44.9) | 373,782 (44.9) | 0.928 |
| Men, No. (%) | 458,587 (55.1) | 458,528 (55.1) | 0.928 |
| Comorbidities, No. (%) | |||
| Hypertension | 312,887 (37.6) | 314,328 (37.8) | 0.021 |
| Diabetes mellitus | 176,605 (21.2) | 176,746 (21.2) | 0.791 |
| Hyperlipidemia | 316,351 (38.0) | 317,495 (38.1) | 0.068 |
| COPD | 247,414 (29.7) | 247,676 (29.8) | 0.658 |
| Ischemic heart disease | 175,328 (21.1) | 174,112 (20.9) | 0.021 |
| Chronic kidney disease | 48,945 (5.9) | 47,301 (5.7) | < 0.001 |
| Atrial fibrillation | 18,252 (2.2) | 16,873 (2.0) | < 0.001 |
| Heart failure | 42,548 (5.1) | 40,803 (4.9) | < 0.001 |
| Stroke | 49,063 (5.9) | 47,636 (5.7) | < 0.001 |
| Cancer | 95,933 (11.5) | 96,587 (11.6) | 0.114 |
| Constipation | 308,737 (37.1) | 308,554 (37.1) | 0.770 |
| Diarrhea | 37,201 (4.5) | 36,412 (4.4) | 0.003 |
| Varicocele | 4,272 (0.5) | 3,879 (0.5) | < 0.001 |
| Mitral valve regurgitation | 57,258 (6.9) | 54,758 (6.6) | < 0.001 |
| Pelvic congestion syndrome | 3,411 (0.4) | 3,216 (0.4) | 0.017 |
| Tricuspid valve insufficiency | 5,291 (0.6) | 4,360 (0.5) | < 0.001 |
| Aortic pathologies | 2,327 (0.3) | 1,673(0.2) | < 0.001 |
| Hernia | 9,473 (1.1) | 8,714 (1.0) | < 0.001 |
| History of VV surgery, No. (%) | 0 (0.0) | 0 (0.0) | N/A |
COPD, chronic obstructive pulmonary disease; HD, hemorrhoidal disease; N/A, not applicable; No., numbers; SD, standard deviation
Comparison of VV incidence in HD and non-HD cohorts
The HD cohort experienced 10,023 VV cases over 5,176,205 person-years (incidence rate: 1.94 per 1000 person-years), compared to 5,680 cases over 4,630,148 person-years in the non-HD cohort (incidence rate: 1.23 per 1000 person-years). In Fig. 2, Kaplan–Meier curves showed the HD cohort had a higher VV risk, confirmed by a significant log-rank test result (P < 0.001). Adjusted hazards ratio (HR) analysis revealed a 1.52-fold higher VV risk in HD patients (95% confidence interval (CI), 1.47–1.57), consistent across both sexes and all age groups, as detailed in Table 2.
Fig. 2.
Kaplan–Meier curves of the cumulative incidence of varicose veins in the hemorrhoidal disease cohort and the non-hemorrhoidal disease cohort. HD: hemorrhoidal disease; VV: varicose veins
Table 2.
The Cox proportional hazards model analysis of the association between HD and risk of VV based on the whole group and age and sex subgroups
| Variables | HD cohort | Non-HD cohort | ||||||
|---|---|---|---|---|---|---|---|---|
| No. of VV cases | PYs | Ratea | No. of VV cases | PYs | Ratea | Crude HR (95% CI) | Adjusted HRb (95% CI) | |
| Total | 10023 | 5176205 | 1.94 | 5680 | 4630148 | 1.23 | 1.55 (1.50–1.60) | 1.52 (1.47–1.57) |
| Age, years | ||||||||
| 20–39 | 1194 | 1351044 | 0.88 | 685 | 1247451 | 0.55 | 1.56 (1.42–1.72) | 1.54 (1.41–1.70) |
| 40–59 | 4125 | 2197639 | 1.88 | 2305 | 1950793 | 1.18 | 1.53 (1.46–1.62) | 1.51 (1.44–1.59) |
| ≥ 60 | 4704 | 1627522 | 2.89 | 2690 | 1431904 | 1.88 | 1.53 (1.46–1.60) | 1.52 (1.45–1.60) |
| P for trend | < 0.001 | < 0.001 | ||||||
| Sex | ||||||||
| Women | 5865 | 2273663 | 2.58 | 3524 | 2055983 | 1.71 | 1.48 (1.42–1.54) | 1.45 (1.39–1.52) |
| Men | 4158 | 2902542 | 1.43 | 2156 | 2574165 | 0.84 | 1.67 (1.59–1.76) | 1.63 (1.55–1.72) |
a: Incidence rate was 1 per 1,000 person-years
b: Hazard ratios for the whole group analysis were adjusted for age, sex, and comorbidities including hypertension, diabetes mellitus, hyperlipidemia, ischemic heart disease, chronic obstructive pulmonary disease, chronic kidney disease, atrial fibrillation, heart failure, stroke, cancer, constipation, diarrhea, tricuspid valve insufficiency, aortic dissection, aortic aneurysm, and hernia; hazard ratios for age and sex subgroup analyses were adjusted for comorbidities including hypertension, diabetes mellitus, hyperlipidemia, ischemic heart disease, chronic obstructive pulmonary disease, chronic kidney disease, atrial fibrillation, heart failure, stroke, cancer, constipation, diarrhea, tricuspid valve insufficiency, aortic dissection, aortic aneurysm, and hernia, and were mutually adjusted for age and sex
CI, confidence interval; HD, hemorrhoidal disease; HR, hazard ratio; No., numbers; PYs , person-years; VV, varicose veins
Patient characteristics of the VV and non-VV cohorts
Demographics and clinical profiles are shown in Table 3. Post-propensity score matching showed no significant differences in age (P = 0.428) or comorbidities including diabetes (P = 0.871), hyperlipidemia (P = 0.053), COPD (P = 0.199), ischemic heart disease (P = 0.271), cancer (P = 0.470), and constipation (P = 0.524) between the VV and non-VV cohorts. The majority were female (approximately 65%) with an average age of around 60 years in both cohorts.
Table 3.
Demographics of VV and non-VV cohorts after propensity score matching
| VV cohort | Non-VV cohort | ||
|---|---|---|---|
| Variable | No. = 112,027 | No. = 112,027 | P-value |
| Follow-up duration, mean ± SD (years) | 6.32 (4.07) | 5.97 (3.83) | < 0.001 |
| Age, mean ± SD (years) | 60.11 (14.45) | 59.94 (14.15) | 0.004 |
| Age, | 0.428 | ||
| 20–39 | 10451 (9.3) | 10301 (9.2) | |
| 40–59 | 41673 (37.2) | 41892 (37.4) | |
| ≥ 60 | 59903 (53.5) | 59834 (53.4) | |
| Women, No. (%) | 72347 (64.6) | 73309 (65.4) | < 0.001 |
| Men, No. (%) | 39680 (35.4) | 38718 (34.6) | < 0.001 |
| Comorbidities, No. (%) | |||
| Hypertension | 53636 (47.9) | 54288 (48.5) | 0.006 |
| Diabetes mellitus | 29918 (26.7) | 29953 (26.7) | 0.871 |
| Hyperlipidemia | 48889 (43.6) | 49345 (44.0) | 0.053 |
| COPD | 35490 (31.7) | 35774 (31.9) | 0.199 |
| Ischemic heart disease | 29133 (26.0) | 29363 (26.2) | 0.271 |
| Chronic kidney disease | 6484 (5.8) | 6190 (5.5) | 0.007 |
| Atrial fibrillation | 4004 (3.6) | 3440 (3.1) | < 0.001 |
| Heart failure | 9200 (8.2) | 8437 (7.5) | < 0.001 |
| Stroke | 7115 (6.4) | 6768 (6.0) | 0.002 |
| Cancer | 11602 (10.4) | 11497 (10.3) | 0.470 |
| Constipation | 29481 (26.3) | 29615 (26.4) | 0.524 |
| Diarrhea | 4017 (3.6) | 3717 (3.3) | 0.001 |
| Varicocele | 402 (0.4) | 326 (0.3) | 0.005 |
| Mitral valve regurgitation | 9255 (8.3) | 8755 (7.8) | < 0.001 |
| Pelvic congestion syndrome | 443 (0.4) | 424 (0.4) | 0.540 |
| Tricuspid valve insufficiency | 1114 (1.0) | 947 (0.8) | < 0.001 |
| Aortic pathologies | 417 (0.4) | 319 (0.3) | < 0.001 |
| Hernia | 1252 (1.1) | 1070 (1.0) | < 0.001 |
| History of HD surgery, No. (%) | 0 (0) | 0 (0) | N/A |
COPD, chronic obstructive pulmonary disease; N/A, not applicable; No., numbers; SD, standard deviation; VV, varicose veins
Comparison of HD incidence in VV and non-VV cohorts
The VV cohort reported 10,024 new HD cases over 707,681 person-years (incidence rate: 14.16 per 1000 person-years), versus 6,296 cases over 669,355 person-years in the non-VV group (incidence rate: 9.41 per 1000 person-years). In Fig. 3, Kaplan–Meier analysis indicated a higher HD incidence in the VV cohort, supported by a significant log-rank test result (P < 0.001). Adjusted hazards ratio (HR) analysis revealed VV patients had a 1.50 times higher HD risk (95% CI, 1.45–1.55), consistent across both sexes and all age groups, as detailed in Table 4.
Fig. 3.
Kaplan–Meier curves of the cumulative incidence of hemorrhoidal disease in the varicose veins cohort and the non-varicose veins cohort. HD: hemorrhoidal disease; VV: varicose veins
Table 4.
The Cox proportional hazards model analysis of the association between VV and risk of HD based on the whole group and age and sex subgroups
| Variable | VV cohort | Non-VV cohort | ||||||
|---|---|---|---|---|---|---|---|---|
| No. of HD cases | PYs | Ratea | No. of HD cases | PYs | Ratea | Crude HR (95% CI) | Adjusted HRb (95% CI) | |
| Total | 10024 | 707681 | 14.16 | 6296 | 669355 | 9.41 | 1.49 (1.44–1.54) | 1.50 (1.45–1.55) |
| Age, years | ||||||||
| 20–39 | 905 | 66209 | 13.67 | 636 | 66973 | 9.50 | 1.45 (1.31–1.60) | 1.44 (1.30–1.60) |
| 40–59 | 3810 | 279801 | 13.62 | 2560 | 263816 | 9.70 | 1.38 (1.31–1.45) | 1.39 (1.32–1.46) |
| ≥ 60 | 5309 | 361671 | 14.68 | 3100 | 338566 | 9.16 | 1.59 (1.52–1.66) | 1.59 (1.52–1.66) |
| P for trend | < 0.001 | < 0.001 | ||||||
| Sex | ||||||||
| Women | 6224 | 457142 | 13.62 | 4060 | 439646 | 9.23 | 1.46 (1.40–1.52) | 1.48 (1.42–1.54) |
| Men | 3800 | 250539 | 15.17 | 2236 | 229709 | 9.73 | 1.54 (1.46–1.63) | 1.55 (1.47–1.63) |
a: Incidence rate was 1 per 1,000 person-years
b: Hazard ratios for the whole group analysis were adjusted for age, sex, and comorbidities including hypertension, diabetes mellitus, hyperlipidemia, ischemic heart disease, chronic obstructive pulmonary disease, chronic kidney disease, atrial fibrillation, heart failure, stroke, cancer, constipation, diarrhea, tricuspid valve insufficiency, aortic dissection, aortic aneurysm, and hernia; hazard ratios for age and sex subgroup analyses were adjusted for comorbidities including hypertension, diabetes mellitus, hyperlipidemia, ischemic heart disease, chronic obstructive pulmonary disease, chronic kidney disease, atrial fibrillation, heart failure, stroke, cancer, constipation, diarrhea, tricuspid valve insufficiency, aortic dissection, aortic aneurysm, and hernia, and were mutually adjusted for age and sex
CI, confidence interval; HD, hemorrhoidal disease; HR, hazard ratio; No., numbers; PYs, person-years; VV, varicose veins
Impact of HD on connective tissue-associated diseases and mortality with VV presence
In subgroups of VV development, further analysis based on the presence or absence of HD showed significantly higher incidences of connective tissue-associated diseases such as mitral valve regurgitation (MR), hernia, and varicocele in the HD cohort (MR: 1.9% vs. 1.3%, P < 0.001; hernia: 0.8% vs. 0.4%, P < 0.001; varicocele: 0.1% vs. 0.0%, P = 0.008). Mortality rates were also higher in the HD group (1.2% vs. 1.0%, P = 0.006), but differences in aorta-related pathology were not statistically significant (0.2% vs. 0.2%, P = 0.857). The details are listed in Table 5.
Table 5.
Five outcome variables in the subgroup of VV development based on the presence or absence of HD
| Outcome variables | HD | P-value (HR) | Crude HR | |
|---|---|---|---|---|
| YES (No. = 10023) |
NO (No. = 5680) |
(95% CI) | ||
| MR, No. (%) | < 0.001 | |||
| YES | 195 (1.9) | 74 (1.3) | 1.66 (1.44–1.91) | |
| NO | 9828 (98.1) | 5606 (98.7) | Reference | |
| Hernia, No. (%) | < 0.001 | |||
| YES | 77 (0.8) | 25 (0.4) | 1.95 (1.55–2.43) | |
| NO | 9946 (99.2) | 5655 (99.6) | Reference | |
| Varicocele, No. (%) | 0.008 | |||
| YES | 6 (0.1) | < 3 (0.0)a | 2.97 (1.34–6.62) | |
| NO | 10,017 (99.9) | Reference | ||
| Aortic pathologies, No. (%) | 0.857 | |||
| YES | 18 (0.2) | 10 (0.2) | 1.04 (0.66–1.66) | |
| NO | 10005 (99.8) | 5670 (99.8) | Reference | |
| Mortality, No. (%) | 0.006 | |||
| YES | 118 (1.2) | 54 (1.0) | 1.29 (1.08–1.55) | |
| NO | 9905 (98.8) | 5626 (99.0) | Reference | |
a: The Health and Welfare Data Science Center (HWDC) stipulates that when the number of patients is less than 3, the actual number cannot be disclosed in order to protect patient privacy. CI, confidence interval; HD, hemorrhoidal disease; HR, hazard ratio; MR, mitral valve regurgitation; No., numbers; VV, varicose veins
Discussion
This nationwide longitudinal study in a Taiwanese population found a significant bidirectional association between HD and VV. In addition, the significant association was evident in both sexes and all age groups. These phenomena are noteworthy and will be discussed in further detail.
The bidirectional association between either disease and risks of development of the other
From HD toward VV, according to Fig. 1 and Table 1, male patients were predominantly diagnosed with HD, aligning with previous findings [10, 11]. Furthermore, Table 2 shows that the association between HD and VV risk increases with age (P for trend < 0.001), highlighting the growing impact of HD on VV incidence with age, a trend supported by other epidemiological studies [2, 28].
A similar observation regarding the VV population was found when viewed from VV toward HD. Figure 1 and Table 3 indicate a higher diagnosis rate among females, consistent with the sex distribution found in other studies [2, 28]. Table 4 demonstrates that the association between VV and HD risk also strengthens with age (P for trend < 0.001), suggesting that VV’s influence on HD incidence intensifies as age advances.
The HD incidence was higher across all ages compared to VV, with rates of 14.16 vs. 1.94 per 1000 patient-years (Tables 2 and 4), suggesting a higher prevalence of HD in our study. While both diseases are common, HD appears more prevalent. Pata et al. noted that over 50% of individuals older than 50 years experience symptomatic hemorrhoids at least once [9]. Conversely, an international survey reported a 20–30% prevalence for chronic venous disorders, with VV being the most common form [1, 28]. A study in Taiwan found a 24.2% prevalence of VV [29]. The discrepancy in prevalence between the two diseases is also evident in our analysis, with 832,310 patients in the HD cohort compared to 112,027 in the VV cohort (Tables 1 and 3).
The explanations for the bidirectional association
The bidirectional association between the two diseases can be explained from several perspectives. First, the two diseases share several risk factors, such as age, obesity, family history, constipation, and activities increasing intraabdominal pressure [2, 9]. Interestingly, HD is more prevalent in male patients, while VV more often occur in female patients [2, 10, 11, 30]. Although a similar sex distribution was observed in our HD cohort, it would be prudent to consider the possibility that female patients in general may not be as willing as male patients to seek medical attention for HD as for VV. A study in Türkiye from Cuglan et al. did demonstrate a female preponderance in an HD cohort [12]. Therefore, the dissimilarity between the two diseases in terms of sex distribution may require further examination.
Second, the two diseases may share certain mechanisms anatomically. VV are considered as a manifestation of abnormal venous drainage, such as venous valve incompetency or proximal vein obstruction [6]. The most common cause is great or small saphenous vein reflux. HD, on the other hand, is more complex but commonly attributed to the sliding anal cushion theory, involving arterial inflow, venous drainage, and connective tissue weakness [31, 32]. Although the mechanisms of the two diseases are not exactly the same, both diseases are influenced by venous valve incompetency affecting venous drainage, primarily from the saphenous veins in VV and the internal hemorrhoidal venous plexus in HD [33].
Third, the two diseases have certain similarities at the molecular level. Specifically, vascular endothelial growth factor (VEGF) and matrix metalloproteinase-9 (MMP-9) are both overexpressed in both diseases. Regarding VEGF, Han et al. found higher VEGF levels in hemorrhoidal than in normal anal tissue [34], echoed by Chung et al.’s findings in hemorrhoidal stroma [35]. Shadrina et al. identified VEGF-A as crucial in VV via genome-wide association study (GWAS) [36], while another study noted elevated angiogenesis markers, including VEGF, in 43 VV tissue samples, particularly in recurrent cases [37].
Regarding MMP- 9, part of the matrix metalloproteinases family, it is crucial in breaking down the extracellular matrix, aiding angiogenesis, tissue repair, and morphogenesis [38]. Kisli et al. identified that serum MMPs including MMP-9 were activated in patients with hemorrhoids [38]. In addition, MMP- 9 was also found to be overexpressed in hemorrhoidal tissues compared to normal anal cushions in another study [34]. In terms of chronic venous disease (CVD), MMP-9 was elevated in venous wall remodeling six months after acute deep vein thrombosis (DVT) [39]. A systemic review also found MMPs, including MMP-9, elevated in varicose veins [40]. Moreover, genetic polymorphisms of MMP-9 have been identified in VV patients [41].
Fourth, the two diseases also have similarities at the genetic level. From the genetics perspective, Salnikova et al. found genetic overlaps between HD and VV, with both linked to MYH9 single nucleotide polymorphism (SNP) rs735854, suggesting a shared angiogenic and capillary integrity role [25, 42, 43]. Given their vascular anomaly connection [33], a common etiology for HD and VV might exist, highlighting MYH9 as a potential shared genetic factor.
The concept of shared genetic susceptibility has been applied in several diseases involving connective tissue weakness, such as pelvic organ prolapse, hemorrhoids, hernia, varicocele, varicose veins and pelvic congestion syndrome [25, 27, 44–47]. HD and VV are therefore grouped in the same disease category, connective tissue disorders. A GWAS study conducted in Taiwan also corroborated the findings that the ACOT-11 gene, which accounts for a well-known connective tissue weakness disorder called Loeys-Dietz Syndrome, was found to be significantly variated in VV patients [30]. The associations through the concept of connective tissue weakness may extend to other pathologies, such as aortic wall pathologies manifested as aortic aneurysm or dissection [26, 48] and mitral valve prolapse manifested as mitral valve regurgitation (MR) [24].
The impact of coexistence of HD and VV
Further analysis of VV subgroups within the HD and non-HD cohorts revealed significantly higher incidences of mitral valve regurgitation (MR), hernia, and varicocele in the HD cohort’s VV subgroup. Aortic pathologies exhibited a nonsignificant trend toward higher incidence (18 vs. 10; P = 0.857), as shown in Table 5, potentially due to small sample sizes. These findings suggest a positive association among these diseases, including at least VV, HD, MR, hernia, and varicocele. A common link of connective tissue weakness may be considered. Moreover, a slight increase in mortality risk was observed in patients with both HD and VV, emphasizing the importance of physician awareness regarding the potential for worsened outcomes due to connective tissue impairment in patients with both diseases.
Summary and clinical implications
Our study contributes significantly to the current body of knowledge regarding the association of these two common diseases. Based on the above findings, HD and VV are mutually associated with increased risks of each other and potentially elevate mortality due to related diseases such as MR, hernia, varicocele, or aortic diseases, likely through connective tissue weakness. The findings provide valuable insights for healthcare professionals in the prevention, early detection, and management of these diseases. For example, clinicians should be aware of the bidirectional risk between VV and HD during consultations, for either disease increases the risk of developing the other. Clinically, it would be prudent to advise a patient with an HD diagnosis to be mindful of the development of VV. For instance, patients with HD should be interviewed about symptoms and signs related to VV. A duplex ultrasound examination may be considered to assess the severity of venous insufficiency. They should also be encouraged to wear compressive stockings more often than patients without them as a preventative measure, for HD entails a higher risk of developing VV in the future. On the other hand, patients with VV diagnoses should be asked about any symptoms and signs associated with HD. They should also pay more attention to their dietary habits, since a low-fiber diet may increase the risk of developing HD in the future.
Moreover, the coexistence of both diseases may further imply weakness in the patient’s connective tissue composition, which is manifested in other body parts. As such, a further survey regarding the associated connective tissue disorders would be recommended, as this cohort of patients may portend relatively poor long-term outcomes. For instance, further arrangement of cardiac ultrasound examinations could be advised in patients having both HD and VV in order to early detect any potential valvular or aortic pathologies. If aortic pathology is detected, a computed tomography scan of the aorta and a stringent pressure control strategy should be implemented to better manage the disease and reduce the risk of associated aortic injury.
Limitations of the study
This study has at least three limitations. First, the dataset did not allow for assessing disease severity, limiting analyses on its impact on incidence, which warrants future investigation. Second, reliance on NHIRD records might lead to underreporting, especially for HD, where cultural stigma could inhibit disclosure, potentially causing underestimation of disease incidence and prevalence, particularly for HD compared to VV. Third, as the data pertained solely to the Taiwanese population, the findings might not apply universally across different ethnic groups.
Conclusion
This nationwide population-based cohort study demonstrated that varicose veins and hemorrhoidal disease are associated in a bidirectional manner. The coexistence of both diseases also increases risks of other connective tissue-associated diseases and mortality risks. Mindfulness of both diseases when either one is encountered is recommended during clinical consultation.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
The authors express sincere thanks to the Artificial Intelligence Development Center at Fu Jen Catholic University for the invaluable support and contributions.
Author contribution
M.L., Y.K., and Y.Y. contributed to the manuscript conception, writing, methodology, and formal analysis. M.C. and B.S. contributed to the supervision and administrative support. S.H., P.L., and T.C. contributed to the manuscript editing, and they provided a substantial intellectual contribution. Y.H. and W.C. critically revised the manuscript. All authors have read and approved the final manuscript for submission.
Funding
No funding was received to assist with the preparation of this manuscript.
Data availability
No datasets were generated or analysed during the current study.
Declarations
Ethical approval
Ethical approval was obtained from the Fu Jen Catholic University’s Institutional Review Board (IRB No. C110199).
Informed consent
Patient informed consent was waived due to the nature of the study involving only anonymized data, as provided by the Health and Welfare Data Science Center (HWDC).
Competing interests
The authors declare no competing interests.
Financial or non-financial interests
The authors have no relevant financial or non-financial interests to disclose.
Disclosure
This manuscript has not been published and is not under consideration for publication elsewhere.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Yan-Jiun Huang and Wei-Che Chiu contributed equally to this article as co-corresponding authors.
Contributor Information
Yan-Jiun Huang, Email: colorectalman@yahoo.com.tw.
Wei-Che Chiu, Email: ppk11642@gmail.com.
References
- 1.Bharath V, Kahn SR, Lazo-Langner A (2014) Genetic polymorphisms of vein wall remodeling in chronic venous disease: a narrative and systematic review. Blood 124(8):1242–1250 [DOI] [PubMed] [Google Scholar]
- 2.Beebe-Dimmer JL, Pfeifer JR, Engle JS, Schottenfeld D (2005) The epidemiology of chronic venous insufficiency and varicose veins. Ann Epidemiol 15(3):175–184 [DOI] [PubMed] [Google Scholar]
- 3.Gazet JC, Redding W, Rickett JW (1970) The prevalence of haemorrhoids. A preliminary survey. Proc R Soc Med 63(1):78–80 [PMC free article] [PubMed] [Google Scholar]
- 4.Johanson JF, Sonnenberg A (1990) The prevalence of hemorrhoids and chronic constipation. An epidemiologic study. Gastroenterology 98(2):380–386 [DOI] [PubMed] [Google Scholar]
- 5.Loder PB, Kamm MA, Nicholls RJ, Phillips RK (1994) Haemorrhoids: pathology, pathophysiology and aetiology. Br J Surg 81(7):946–954 [DOI] [PubMed] [Google Scholar]
- 6.Bergan JJ, Schmid-Schonbein GW, Smith PD, Nicolaides AN, Boisseau MR, Eklof B (2006) Chronic venous disease. N Engl J Med 355(5):488–498 [DOI] [PubMed] [Google Scholar]
- 7.Kaidar-Person O, Person B, Wexner SD (2007) Hemorrhoidal disease: a comprehensive review. J Am Coll Surg 204(1):102–117 [DOI] [PubMed] [Google Scholar]
- 8.Cosman BC (2019) Piles of money: “Hemorrhoids” are a billion-dollar industry. Am J Gastroenterol 114(5):716–717 [DOI] [PubMed] [Google Scholar]
- 9.Pata F, Sgro A, Ferrara F, Vigorita V, Gallo G, Pellino G (2021) Anatomy, physiology and pathophysiology of haemorrhoids. Rev Recent Clin Trials 16(1):75–80 [DOI] [PubMed] [Google Scholar]
- 10.Godeberge P, Sheikh P, Zagriadskii E et al (2020) Hemorrhoidal disease and chronic venous insufficiency: concomitance or coincidence; results of the CHORUS study (Chronic venous and HemORrhoidal diseases evalUation and Scientific research). J Gastroenterol Hepatol 35(4):577–585 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Ekici U, Kartal A, Ferhatoglu MF (2019) Association between hemorrhoids and lower extremity chronic venous insufficiency. Cureus 11(4):e4502 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Cuglan B, Ozturk S, Ozcan F et al (2022) Assessment of venous leg symptoms in patients with hemorrhoidal disease (VEIN-HEMORRHOID study). Phlebology 37(1):42–47 [DOI] [PubMed] [Google Scholar]
- 13.Kutlu A, Yilmaz E, Cecen D, Eser E, Ozbakkaloglu A (2011) The Turkish validity and reliability of the venous insufficiency epidemiological and economic study-quality of life/symptoms scales. Angiology 62(4):329–337 [DOI] [PubMed] [Google Scholar]
- 14.Aslan R, Erbin A, Celik S et al (2019) Evaluation of hemorrhoidal disease and lower extremity venous insufficiency in primary adult varicocele: a prospective controlled study. Phlebology 34(9):621–626 [DOI] [PubMed] [Google Scholar]
- 15.Godeberge P, Sheikh P, Lohsiriwat V, Jalife A, Shelygin Y (2021) Micronized purified flavonoid fraction in the treatment of hemorrhoidal disease. J Comp Eff Res 10(10):801–813 [DOI] [PubMed] [Google Scholar]
- 16.Lyseng-Williamson KA, Perry CM (2003) Micronised purified flavonoid fraction: a review of its use in chronic venous insufficiency, venous ulcers and haemorrhoids. Drugs 63(1):71–100 [DOI] [PubMed] [Google Scholar]
- 17.Wu CY, Chen YJ, Ho HJ et al (2012) Association between nucleoside analogues and risk of hepatitis B virus-related hepatocellular carcinoma recurrence following liver resection. JAMA 308(18):1906–1914 [DOI] [PubMed] [Google Scholar]
- 18.Chien HC, Kao Yang YH, Bai JP (2016) Trastuzumab-related cardiotoxic effects in taiwanese women: a nationwide cohort study. JAMA Oncol 2(10):1317–1325 [DOI] [PubMed] [Google Scholar]
- 19.Chang SH, Chou IJ, Yeh YH et al (2017) Association between use of non-vitamin K oral anticoagulants with and without concurrent medications and risk of major bleeding in nonvalvular atrial fibrillation. JAMA 318(13):1250–1259 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Chang SL, Huang YL, Lee MC et al (2018) Association of varicose veins with incident venous thromboembolism and peripheral artery disease. JAMA 319(8):807–817 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Cheng CL, Kao YH, Lin SJ, Lee CH, Lai ML (2011) Validation of the National Health Insurance Research Database with ischemic stroke cases in Taiwan. Pharmacoepidemiol Drug Saf 20(3):236–242 [DOI] [PubMed] [Google Scholar]
- 22.Chen CC, Chen LS, Yen MF, Chen HH, Liou HH (2012) Geographic variation in the age- and gender-specific prevalence and incidence of epilepsy: analysis of Taiwanese National Health Insurance-based data. Epilepsia 53(2):283–290 [DOI] [PubMed] [Google Scholar]
- 23.Kao WH, Hong JH, See LC et al (2018) Validity of cancer diagnosis in the National Health Insurance database compared with the linked National Cancer Registry in Taiwan. Pharmacoepidemiol Drug Saf 27(10):1060–1066 [DOI] [PubMed] [Google Scholar]
- 24.Lee ML, Chou WY, Kao YW et al (2023) Varicose veins are associated with an increased risk of mitral valve regurgitation: a nationwide population-based cohort study. Int Angiol 42(4):352–361 [DOI] [PubMed] [Google Scholar]
- 25.Salnikova LE, Khadzhieva MB, Kolobkov DS (2016) Biological findings from the PheWAS catalog: focus on connective tissue-related disorders (pelvic floor dysfunction, abdominal hernia, varicose veins and hemorrhoids). Hum Genet 135(7):779–795 [DOI] [PubMed] [Google Scholar]
- 26.Yetkin E, Ozturk S (2018) Dilating vascular diseases: pathophysiology and clinical aspects. Int J Vasc Med 2018:9024278 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Yetkin E, Ileri M (2016) Dilating venous disease: pathophysiology and a systematic aspect to different vascular territories. Med Hypotheses 91:73–76 [DOI] [PubMed] [Google Scholar]
- 28.Rabe E, Regnier C, Goron F, Salmat G, Pannier F (2020) The prevalence, disease characteristics and treatment of chronic venous disease: an international web-based survey. J Comp Eff Res 9(17):1205–1218 [DOI] [PubMed] [Google Scholar]
- 29.Chen CL, Guo HR (2014) Varicose veins in hairdressers and associated risk factors: a cross-sectional study. BMC Public Health 14:885 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Lee ML, Liang C, Chuang CH et al (2022) A genome-wide association study for varicose veins. Phlebology 37(4):267–278 [DOI] [PubMed] [Google Scholar]
- 31.Margetis N (2019) Pathophysiology of internal hemorrhoids. Ann Gastroenterol 32(3):264–272 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Lohsiriwat V (2012) Hemorrhoids: from basic pathophysiology to clinical management. World J Gastroenterol 18(17):2009–2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Holdstock JM, Dos Santos SJ, Harrison CC, Price BA, Whiteley MS (2015) Haemorrhoids are associated with internal iliac vein reflux in up to one-third of women presenting with varicose veins associated with pelvic vein reflux. Phlebology 30(2):133–139 [DOI] [PubMed] [Google Scholar]
- 34.Han W, Wang ZJ, Zhao B et al (2005) Pathologic change of elastic fibers with difference of microvessel density and expression of angiogenesis-related proteins in internal hemorrhoid tissues. Zhonghua Wei Chang Wai Ke Za Zhi 8(1):56–59 [PubMed] [Google Scholar]
- 35.Chung YC, Hou YC, Pan AC (2004) Endoglin (CD105) expression in the development of haemorrhoids. Eur J Clin Invest 34(2):107–112 [DOI] [PubMed] [Google Scholar]
- 36.Shadrina AS, Smetanina MA, Sokolova EA et al (2018) Allele rs2010963 C of the VEGFA gene is associated with the decreased risk of primary varicose veins in ethnic Russians. Phlebology 33(1):27–35 [DOI] [PubMed] [Google Scholar]
- 37.Sanchez FSL, Martinez JAC, Mendez-Garcia L, Garcia-Cenador MB, Pericacho M (2022) Endoglin and other angiogenesis markers in recurrent varicose veins. J Pers Med 12(4):528 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Kisli E, Kemik A, Sumer A, Kemik O (2013) Matrix metalloproteinases in pathogenesis of hemorrhoidal disease. Am Surg 79(11):1181–1184 [PubMed] [Google Scholar]
- 39.Deatrick KB, Elfline M, Baker N et al (2011) Postthrombotic vein wall remodeling: preliminary observations. J Vasc Surg 53(1):139–146 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Lim CS, Davies AH (2009) Pathogenesis of primary varicose veins. Br J Surg 96(11):1231–1242 [DOI] [PubMed] [Google Scholar]
- 41.Xu HM, Zhao Y, Zhang XM, Zhu T, Fu WG (2011) Polymorphisms in MMP-9 and TIMP-2 in Chinese patients with varicose veins. J Surg Res 168(1):e143–e148 [DOI] [PubMed] [Google Scholar]
- 42.Huang Y, Shi H, Zhou H, Song X, Yuan S, Luo Y (2006) The angiogenic function of nucleolin is mediated by vascular endothelial growth factor and nonmuscle myosin. Blood 107(9):3564–3571 [DOI] [PubMed] [Google Scholar]
- 43.Arrondel C, Vodovar N, Knebelmann B et al (2002) Expression of the nonmuscle myosin heavy chain IIA in the human kidney and screening for MYH9 mutations in Epstein and Fechtner syndromes. J Am Soc Nephrol 13(1):65–74 [DOI] [PubMed] [Google Scholar]
- 44.Allen-Brady K, Cannon-Albright L, Farnham JM et al (2011) Identification of six loci associated with pelvic organ prolapse using genome-wide association analysis. Obstet Gynecol 118(6):1345–1353 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Allen-Brady K, Cannon-Albright LA, Farnham JM, Norton PA (2015) Evidence for pelvic organ prolapse predisposition genes on chromosomes 10 and 17. Am J Obstet Gynecol 212(6):771.e1-771.e7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Allen-Brady K, Chua JWF, Cuffolo R, Koch M, Sorrentino F, Cartwright R (2022) Systematic review and meta-analysis of genetic association studies of pelvic organ prolapse. Int Urogynecol J 33(1):67–82 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Allen-Brady K, Norton PA, Farnham JM, Teerlink C, Cannon-Albright LA (2009) Significant linkage evidence for a predisposition gene for pelvic floor disorders on chromosome 9q21. Am J Hum Genet 84(5):678–682 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Yetkin E (2015) Hemorrhoid, internal iliac vein reflux and peripheral varicose vein: affecting each other or affected vessels? Phlebology 30(2):145 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Access to these data is subject to restrictions. The data, sourced from the National Health Insurance database, can be provided by the authors upon receiving authorization from the National Health Insurance Administration of Taiwan.
No datasets were generated or analysed during the current study.