International Journal of Colorectal Disease

, Volume 27, Issue 2, pp 215–220

The prevalence of hemorrhoids in adults


    • Department of SurgeryMedical University of Vienna
  • Friedrich Anton Weiser
    • Endoscopic Center
  • Katrin Schwameis
    • Department of SurgeryMedical University of Vienna
  • Thomas Riss
    • Hartmannspital Wien
  • Martina Mittlböck
    • Center for Medical Statistics, Informatics and Intelligent SystemsMedical University of Vienna
  • Gottfried Steiner
    • Endoscopic Center
  • Anton Stift
    • Department of SurgeryMedical University of Vienna
Original Article

DOI: 10.1007/s00384-011-1316-3

Cite this article as:
Riss, S., Weiser, F.A., Schwameis, K. et al. Int J Colorectal Dis (2012) 27: 215. doi:10.1007/s00384-011-1316-3



Exact data on the prevalence of hemorrhoids are rare. Therefore, we designed a study to investigate the prevalence of hemorrhoids and associated risk factors in an adult general population.


Between 2008 and 2009, consecutive patients were included in a prospective study. They attended the Austrian national wide health care program for colorectal cancer screening at four medical institutions. A flexible colonoscopy and detailed examination were conducted in all patients. Hemorrhoids were defined according to a standardized grading system. Independent variables included baseline characteristics, sociodemographic data, and health status. Potential risk factors were calculated by univariate and multivariate analysis.


Of 976 participants, 380 patients (38.93%) suffered from hemorrhoids. In 277 patients (72.89%), hemorrhoids were classified as grade I, in 70 patients (18.42%) as grade II, in 31 patients (8.16%) as grade III, and in 2 patients (0.53%) as grade IV. One hundred seventy patients (44.74%) complained about symptoms associated with hemorrhoids, whereas 210 patients (55.26%) reported no symptoms. In the univariate and multivariate analysis, body mass index (BMI) had a significant effect on the occurrence of hemorrhoids with p = 0.0391 and p = 0.0282, respectively. Even when correcting for other potential risk factors, an increase in the BMI of one increased the risk of hemorrhoids by 3.5%.


Hemorrhoids occur frequently in the adult general population. Notably, a considerable number of people with hemorrhoids do not complain about symptoms. In addition, a high BMI can be regarded as an independent risk factor for hemorrhoids.


HemorrhoidsEpidemiologyRisk factorsPrevalence


Hemorrhoidal disease is a common anorectal disorder. Although the vast majority of patients can be successfully treated with conservative medication and ointments, some patients need to undergo surgery during their life [1].

Main symptoms associated with hemorrhoids are bleeding, itching, soiling, and pain [2, 3]. A considerable number of patients complaining about these symptoms assume to suffer from hemorrhoids. Notably, only a precise anamnesis and a careful physical examination can make an accurate diagnosis, thus avoiding frequent and early misinterpretations [4].

Additionally, internal hemorrhoids, which do not prolapse, can only be found during endoscopy [5]. According to the definition by Banov et al., they are usually classified as grade I hemorrhoids [6]. However, it should be taken into consideration that some of these patients do not have anal complaints at all.

Very few attempts have been made to assess the prevalence of hemorrhoids. All studies have potential shortcomings, thus resulting in a wide range of prevalence rates. One study reviewed the records of patients seen in a colorectal department during a certain time period [7]. The authors revealed a prevalence rate of 86% for symptomatic and asymptomatic hemorrhoids. In contrast, Johanson and Sonnenberg reported an overall prevalence rate of 4.4% in the USA [8]. Anyway, due to the theoretical study design, hemorrhoids were not classified according to their symptoms and extensions.

The most effective way to perform an epidemiological study on hemorrhoids is to include a large sample of the general population, irrespective of concomitant symptoms. Furthermore, it is mandatory not only to question symptoms correlated with hemorrhoids but also to examine the anorectal region directly.

Therefore, the current investigation was designed to determine the prevalence of hemorrhoids in a large group of adults in the general population by direct questioning and anorectal examination. Moreover, we intended to identify risk factors associated with the occurrence of hemorrhoids.

Materials and methods


Between 2008 and 2009, consecutive patients, attending the national wide health care program for colorectal cancer screening, were included in a prospective study. The investigation was conducted in four medical institutions in Austria that provided the possibility to perform screening colonoscopies for colorectal cancer. According to the current recommendations in Austria, all patients from the age of 50 years were encouraged to participate in the national wide screening program. Patients with known risk factors, who usually have their screening colonoscopy in a younger age, were included as well. The study was approved by the local ethics committee, and written informed consent was given by all participants.


Flexible colonoscopies were performed by qualified endoscopic and colorectal surgeons as described previously [9]. All investigators were experienced in diagnosing and treating hemorrhoids. Patients were examined in the left lateral position. After anal inspection and digital examination, patients were asked to press with the intention to provoke potential hemorrhoidal prolapse. Thereafter, complete flexible colonoscopy was conducted whenever possible. In the rectal ampulla, a retroflexion maneuver was performed to visualize internal hemorrhoidal cushions.

A detailed standardized questionnaire was filled out by the examiner for each patient. All diagnostic findings during endoscopy were recorded.


The presence of hemorrhoids was documented and classified according to the international classification recommended by Banov et al.: Internal hemorrhoids that do not prolapse and appear as a bulge into the lumen of the anal canal with or without bleeding are classified as first degree. Second-degree hemorrhoids prolapse and reduce spontaneously. Third-degree hemorrhoids require digital reduction of prolapsed tissue. Fourth-degree hemorrhoids are not reducible [6].

Symptoms related to the presence of hemorrhoids were assessed and divided into bleeding, itching, soiling, and pain. Hemorrhoids were defined as symptomatic if patients reported one or more of those four symptoms. Contrary, patients with hemorrhoids, who did not complain about associated symptoms, were defined as asymptomatic hemorrhoids. The necessity of current treatment of hemorrhoids was documented too. Potential risk factors for the development of hemorrhoids (grades I to IV) were evaluated and classified as follows: age, sex, body mass index (BMI), pregnancy, birth type, educational status, professional status, liver cirrhosis, diabetes mellitus, and previous pelvic floor operations (exclusive operations for hemorrhoidal disease).

Statistical analysis

Continuous data are shown as mean and standard deviation (±SD) in case of normally distributed data and as median, minimum, and maximum otherwise. Differences between two groups were tested with t test for normally distributed data and with the nonparametric Wilcoxon rank sum test otherwise. Prevalence rates for age groups are given together with 95% confidence intervals (CI). Categorical data are described with absolute and relative frequencies. Differences of categorical data between groups are tested with the chi-square test or with Fisher’s exact test in case of small expected cell frequencies. In case of ordinal data, the trend version of the chi-square test was calculated. Additionally, associations between the covariates and the occurrence of hemorrhoids were modeled and described with odd ratios and 95% confidence intervals in a multiple logistic regression model.

All p values are two-sided and p ≤ 0.05 was considered significant. All calculations were performed with the statistical analysis software SAS (SAS Institute Inc., version 9.2, Cary, NC, USA).


The prevalence of hemorrhoids according to age’s distribution is outlined in Table 1. Three hundred eighty patients (38.93%) of 976 participants presented with hemorrhoids. In 277 patients (72.89%), hemorrhoids were classified as grade I, in 70 patients (18.42%) as grade II, in 31 patients (8.16%) as grade III, and in 2 patients (0.53%) as grade IV.
Table 1

The prevalence of hemorrhoids with regard to age


Prevalence of hemorrhoids

Hemorrhoids present/total

Prevalence (%)

95% CI for prevalence













































One hundred seventy patients (44.74%) of all hemorrhoid patients complained about symptoms associated with hemorrhoids, whereas 210 patients (55.26%) reported no symptoms (Table 2). The symptom of bleeding increased significantly with a higher grade of hemorrhoids (p = 0.0019). Interestingly, itching declined in those patients with a higher grade of hemorrhoids (p = 0.0267). The occurrence of symptomatic hemorrhoids showed no correlation with a higher grade of hemorrhoids (p = 0.5287; Table 3). Eighty-three of 380 patients (21.84%) reported to undergo conservative treatment for hemorrhoids, whereas 297 patients (78.16%) had no current medical therapy.
Table 2

Symptoms stratified by degree of hemorrhoids


Grade I (n = 277)

Grade II (n = 70)

Grade III (n = 31)

Grade IV (n = 2)

Total (n = 380)

p value

Any symptoms

125 (45.13)

26 (37.14)

17 (54.84)

2 (100)

170 (44.74)



65 (23.47)

18 (25.71)

15 (48.39)

2 (100)

100 (26.32)



74 (26.71)

13 (18.57)

4 (12.90)


91 (23.95)



11 (3.97)

4 (5.71)



15 (3.95)



30 (10.83)

5 (7.14)

1 (3.23)

1 (50)

37 (9.74)


Values are given as absolute numbers and frequencies (in percent). The tests are based on the exact chi-square test with an ordered trend in grades of hemorrhoids

Table 3

Symptomatic versus asymptomatic hemorrhoids stratified by degree of hemorrhoids


Grade I

Grade II

Grade III

Grade IV


p value

Symptomatic hemorrhoids

125 (45.13)

26 (37.14)

17 (54.84)

2 (100)

170 (44.74)


Asymptomatic hemorrhoids

152 (54.87)

44 (62.86)

14 (45.16)


210 (55.26)

Values are given as absolute numbers and frequencies (in percent)

Findings during colonoscopy were documented and compared with the presence of hemorrhoids. Notably, perianal skin tags and colonic polyps (polyps <6 mm and polyps >6 mm) showed a significant association with hemorrhoids (p = 0.0347 and 0.0287). Thirty-six patients (9.47%) with hemorrhoids (n = 380) had skin tags in contrast to 35 patients (5.87%) without hemorrhoids (n = 596).

Diverticulosis was found in 130 patients (34.21%) with hemorrhoids compared with 171 patients (28.69%) without hemorrhoids. This difference showed a tendency but was not significant (p = 0.0687). Colitis, colonic tumor, and other anal pathologies showed no significant association.

Potential risk factors for the occurrence of hemorrhoids were analyzed and listed in Table 4. Univariate analysis showed that only BMI had a significant correlation with an increased risk for hemorrhoids (p = 0.0391). The median BMI in the hemorrhoid group was 27.04 (range, 19.03–46.06) compared with 26.33 (range, 17.36–61.27) in the group without hemorrhoids, respectively. In a multiple logistic regression analysis, including the variables of Table 4 except pregnancy and birth type, BMI was confirmed to have an independent significant association with an increase in the risk for disease of 1.035 for a BMI increase of one (p = 0.0282, odds ratio = 1.035; 95% CI, 1.004–1.067). Additionally, we analyzed potential predictive factors for symptomatic hemorrhoids in patients with hemorrhoids. Although the difference was small, age was found to be a significant risk factor in the univariate analysis (symptomatic patients, 60.79 years (±8.97), asymptomatic patients, 62.94 (±8.60); p = 0.0179) but lost significance after adjusting for other covariates.
Table 4

Risk factors for developing hemorrhoids

Risk factors



p value





488 (50)

199 (40.78)

289 (59.22)



488 (50)

181 (37.09)

307 (62.91)


61.68 (±9.21)

61.98 (±8.82)

61.49 (±9.46)



26.54 (17.36–61.27)

27.04 (19.03–46.06)

26.33 (17.36–61.27)


Family status


60 (6.32)

19 (31.67)

41 (68.33)



696 (73.34)

272 (39.08)

424 (60.92)


147 (15.49)

63 (42.86)

84 (57.14)


46 (4.85)

16 (34.78)

30 (65.22)

Educational status

Grade school

565 (59.29)

212 (37.52)

353 (62.48)


High school

236 (24.76)

101 (42.80)

135 (57.20)


152 (15.95)

57 (37.50)

95 (61.50)

Professional status


357 (37.03)


232 (64.99)



29 (3.01)

13 (44.83)

16 (55.17)


539 (55.91)

222 (41.19)

317 (58.81)

At home

39 (4.05)

17 (43.59)

22 (56.41)



395 (80.94)

164 (41.52)

231 (58.48)



93 (19.06)

35 (37.63)

58 (62.37)

Birth type

No birth

93 (19.46)

35 (37.63)

58 (62.37)


Vaginal delivery

348 (72.80)

149 (42.82)

199 (57.18)

Cesarean delivery

22 (4.60)

9 (40.91)

13 (59.09)


15 (3.14)

4 (26.67)

11 (73.33)

Liver cirrhosis


18 (1.84)

7 (38.89)

11 (61.11)



958 (98.16)

373 (38.94)

585 (61.06)

Diabetes mellitus


103 (10.55)

46 (44.66)

57 (55.34)



873 (89.45)

334 (38.26)

539 (61.74)

Previous pelvic floor operations


233 (23.87)

102 (43.78)

131 (56.22)



743 (76.13)

278 (37.42)

465 (62.58)

Categorical variables are described as absolute numbers with percentages. Column percentages are given for total and row percentages otherwise. Continuous variables are described as means (±SD) for normally distributed data or with median (minimum–maximum) otherwise. p values refer either to chi-square test, t test, or Wilcoxon rank sum test, respectively

We compared the study patients with the demographic data (gender and education) of the “Statistics Austria.” Gender was equally distributed showing no significant difference (study group: female 50.0%‚ male 50.0%; general population: female 51.79%, male 48.21%; p = 0.3990). The education status was divided into grade school, high school, and college. We aimed to categorize the data of the “Statistics Austria” according to these groups: study group—grade school 59.29%, high school 24.76%, and college 15.95%; general population—62.86%, 26.94%, and 10.20%, respectively, p < 0.0001. This is not surprising as people with a higher education tend to participate in screening programs more frequently. Furthermore, as we included only adult patients, the probability of a completed higher education level is certainly increased.


The present epidemiological investigation was conducted to determine the prevalence of hemorrhoids in adults and to define associated risk factors. We revealed an overall prevalence of 39% for grades I to IV hemorrhoids classified according to the international classification of hemorrhoids in the current adult population. Notably, only 17% of patients complained about symptoms related to hemorrhoids, whereas 22% of patients reported not to have any problems.

The majority of patients in the asymptomatic group represented grade I hemorrhoids, thus merely a bulge in the anal canal caused by hemorrhoidal tissue was visible. To the author’s knowledge, there are no data available whether primary asymptomatic grade I hemorrhoids will become symptomatic in the future. It might also be possible that these hemorrhoids represent normal endoscopic findings and will never lead to typical hemorrhoidal complaints. However, further longitudinal studies need to clarify that.

Epidemiological data are of great importance as they reflect the burden of a disease. There is a considerable paucity of studies that intended to investigate the prevalence of hemorrhoids. Haas et al. assessed the prevalence of hemorrhoids by reviewing patients seen in a colon and rectal surgical clinic [7]. Patients underwent rectal examination and anoscopy. The overall prevalence rate of hemorrhoids was found to be 86%. In contrast to our results, 102 patients with grade I hemorrhoids reported symptoms compared to 42 patients without. However, it should be taken into consideration that the study population consisted of a highly selected group of patients; thus, selection bias cannot be ruled out.

On the other hand, Johanson and Sonnenberg observed a significant lower prevalence rate of 4.4% [8]. As mentioned above, their analysis was based on four different data sources: the National Health Interview Survey, the National Hospital Discharge Survey, the National Disease and Therapeutic Index, and the Morbidity Statistics from General Practice. Altogether, ten million people were studied. Nevertheless, these results should be interpreted with caution as the accuracy of such data depends on adequate recordkeeping. Moreover, the exact diagnosis of hemorrhoids requires a detailed examination; thus, the reliability of these prevalence rates can be questioned. A classification of hemorrhoids and a differentiation between symptomatic and asymptomatic hemorrhoids were missing as well.

In the present investigation, we also have correlated the presence of hemorrhoids with endoscopic findings. A significant association was found in regard to perianal skin tags and polyps. Generally, anal skin tags can occur as a primary lesion or secondary as a result of anal fissures or anorectal operations. It also might be possible that chronic perianal irritation caused by hemorrhoids can lead to anal skin tags. On the other hand, we cannot give a reasonable explanation why polyps were correlated with hemorrhoids. This finding probably happened by chance.

A high BMI was found to be the only independent risk factor for hemorrhoids. Notably, although it was supported by multiple logistic regression analysis, the difference of the BMI between patients with and without hemorrhoids was low. Thus, its clinical relevance can be put into question. Interestingly, we could not detect a correlation between hemorrhoids and women with a history of childbirth. Furthermore, the type of birth showed no influence on the development of hemorrhoids as well; however, the group of cesarean delivery was rather small. This finding is of quite interest as symptomatic hemorrhoids are common disorders during pregnancy [10]. However, Thompson et al. reported that hemorrhoids occurred frequently in pregnant women but showed resolution between 8 and 24 weeks postpartum [11].

Another study reported an increased prevalence of hemorrhoids in patients with a higher socioeconomic status [8]. In contrast, in the present investigation, the educational status had no impact on the prevalence rate.

We have to address few limitations of the present study. Although the colorectal cancer screening program is conducted nationwide in Austria and is recommended by all doctors and supported by companies, patients are free to choose whether to participate or not. Consequently, we are well aware of the fact that although we included a large number of patients, it does not represent a random sample of the general population. For that reason, we compared our demographic data with that of the Austrian population provided by Statistics Austria. There was only a slight difference with regard to the educational status of included patients. As mentioned above, this is not surprising as people with a higher education tend to participate in screening programs more frequently. In addition, as we included only adult patients, the probability of a completed higher education level is certainly increased.

In addition, it is under debate whether colonoscopy is a useful tool for diagnosing internal hemorrhoids [5, 1214]. Kelly et al. compared anoscopy with fiber endoscopy in identifying anorectal lesions [15]. The authors found a higher detection rate in the anoscopy group. However, other authors stated that colonoscopy with rectal retroflexion is a valuable technique to evaluate internal hemorrhoids [12, 13].


Hemorrhoids occur frequently in the adult general population. Notably, a considerable number of affected patients do not complain about typical hemorrhoidal symptoms. In addition, a high BMI can be regarded as an independent risk factor for hemorrhoids.

Competing interest

There were no conflicts of interest, sources of financial support, corporate involvement, patent holdings, etc. involved in the research and preparation of this manuscript.

Copyright information

© Springer-Verlag 2011