The burden of constipation on quality of life: results of a multinational survey


  • *

    IPSOS GmbH, Hamburg, Germany

Dr. O. Petrini, Istituto Cantonale di Microbiologia, Via Mirasole 22A, CH-6500 Bellinzona, Switzerland.


Background The impact of constipation on quality of life (QoL) may vary in different cultural or national settings.

Aim We studied QoL in a multinational survey to compare different social and demographic groups with and without constipation (defined according to Rome III criteria) and to detect country-specific differences among the groups studied.

Methods Health-related QoL (HRQoL) was assessed with the Short Form 36 (SF-36) questionnaire in 2870 subjects in France, Germany, Italy, UK, South Korea, Brazil and USA.

Results Respondents were mainly middle-aged, married or living together and part- or full-time employed. General health status, measured by the SF-36 questionnaire, was significantly worse in the constipated vs. non-constipated populations. Results were comparable in all countries. QoL scores correlated negatively with age. Constipated women reported more impaired HRQoL than constipated men. Brazilians were most affected by constipation as to their social functioning (35.8 constipated vs. 51.3 non-constipated) and general health perception (29.4 constipated vs. 54.4 non-constipated).

Conclusions There are significant differences in HRQoL between constipated and non-constipated individuals and a significant, negative correlation between the number of symptoms and complaints and SF-36 scores. The study detected a correlation of constipation with QoL and the influence of social and demographic factors on HRQoL in constipated people.


Clinically, constipation is defined according to the Rome II and III criteria as a complex of at least two symptoms including infrequent bowel movements, typically <3 per week over at least 3 months during the past half year, difficulty during defecation, straining during >25% of bowel movements, a subjective sensation of hard stools or incomplete bowel evacuation.1, 2

Approximately, 15% of the adult population in western countries suffers from constipation. The prevalence in North America ranges between 4% and 28%, being two to three times more frequent in women than in men, independently of the method of assessment used.3, 4 Constipation has been estimated to affect 5% of the male and 15% of the female population.5 Episodes become more frequent with increasing age, this trend being continuous in women and more marked in men older than 60 years.4, 5 Epidemiological studies suggest that the presence of constipation may be related to socioeconomic factors and is more prevalent in low compared with high-income groups in the same country. Subjects with higher education are less often affected than individuals with lower education and urban dwellers less so than those living in rural areas. Nutrition, physical activity or hormonal changes, on the other hand, apparently do not play a major role.4–9

Laxatives (bulk forming agents, osmotic laxatives and drugs influencing motility and/or secretion) are the treatment of choice for constipation, although changes in life style, nutrition and physical exercise are often included in the physician’s recommendations. It is assumed that 25–50% of the patients suffering from constipation take laxatives, but most of them use over-the-counter remedies, and only 10% consult a doctor.1

The importance of health-related quality of life (HRQoL) in assessing patient’s well-being and the impact of health disorders on socioeconomic factors such as work and social activities has been recognized.10 Relatively few studies have analysed HRQoL in constipation, although it has been previously demonstrated that HRQoL is lower in constipated than in non-constipated individuals, and treatment improves HRQoL.11

Health-related quality of life has been studied in a number of functional gastrointestinal (GI) disorders, including faecal incontinence,12, 13 functional constipation14 and irritable bowel syndrome (IBS),15 as well as after colectomy16 or ileal-pouch anal anastomosis.17 It has also been used to detect changes in well-being of patients with chronic constipation after laxative treatment.18

The choice of an adequate HRQoL instrument is often crucial to detect specific aspects of the well-being of patients. While disease-specific HRQoL instruments may address accurately the burden of a disease, they may fail to detect more general changes in a patient’s everyday functioning. Ideally, the combination of a disease-specific tool and a more general QoL instrument such as the Short Form 36 (SF-36) would be a valuable option, but this approach adds complexity to the design of the research, as the time required to answer questions may become unjustifiably long. It may be of interest, therefore, to use a generic instrument such as the SF-36 that is detailed enough to capture different facets of HRQoL but does not require too much time and effort for its completion.

The SF-36 has been widely used in clinical research and its validity and reliability have been demonstrated in a large number of studies.19–22 It is validated in several languages and therefore is well suited for use in large, multinational surveys. It has also been already used successfully to detect health-related changes in QoL in the presence of functional GI disorders23 and has helped to assess the outcome of subtotal colectomy for slow-transit constipation.16

Constipation-related HRQoL has so far been studied in selected patient populations and to our knowledge only one Canadian study has investigated the burden of functional GI disorders in a random sample.23 We decided, therefore, to study constipation-related HRQoL in a large, multinational survey to better understand differences in varying social and demographic groups, with and without constipation and to detect country-specific differences.

Materials and methods


The data analysed in this study were collected during a survey carried out between June and October 2005 in four countries in Europe (France, Germany, Italy and United Kingdom) and in South Korea, Brazil and USA. Interviews were conducted in 27 major towns in France, 10 major towns in USA and nationwide in Germany, Italy and UK. In Brazil, they were limited to Greater Sao Paulo and Greater Rio de Janeiro and in South Korea to Seoul, Incheon and Kyungi.

The survey was divided into two phases: constipated people were included in the first, whereas non-constipated subjects were interviewed during the second. Care was taken to include approximately 200 individuals from each country in each of the two phases.

An international market research company* performed the survey. In the USA, France and Italy, computer-assisted personal interviews were used, while in Brazil, Germany, UK and South Korea answers were recorded on paper forms by the interviewers. Previous research has shown that the SF-36 demonstrates high internal consistency regardless of the mode of administration and data collection.24

Sampling procedures

The survey was carried out anonymously and no medical intervention was planned, therefore no Ethics Committee approval was necessary and no signed informed consent was collected. The subjects were chosen randomly in each country. In the first phase of the study, 1435 subjects who claimed to suffer from constipation at least once every 3 months during the last 2 years were included in the survey. The interviewer assessed the self-reported constipation using a questionnaire that relied upon the Rome III criteria.2 The second phase interviewed an equally large group of healthy, age- and sex-matched individuals with no constipation. The gender split (70% women/30% men) and laxative usage rate (40%) were the same in all countries and were chosen a priori to reflect published prevalence data on constipation.25

Sample management

All the data were tracked and collected in a database by the Market Research Company, checked for accuracy of input and reliability and then exported in SAS format for all subsequent statistical calculations. SF-36 scores were computed and standardized as described by Spritzer26 using a T-score transformation to transform scores to a mean of 50 and standard deviations of 10 in the general US population. This produces the same mean and standard deviation for all eight scales and for the Physical and Mental Component Scores (MCS).27 The use of the same mean and standard deviation has the advantage to permit comparison of mean scores across populations. Standard scoring algorithms have been shown to be very reliable and allow interpreting differences across scales in the SF-36 profile and comparison of results across different samples, including populations from different countries for which normative data are not available.28

Questionnaires used

We used two different questionnaires. One included general questions on demographics, lifestyle and health aspects and definition of constipation, symptoms experienced and treatments. This questionnaire was validated internally for the definition of constipation to comply with the Rome III criteria for functional constipation.29 Subjects self-reporting constipation met at least two of the Rome III criteria.29

Health-related quality of life was assessed with the SF-36 questionnaire. The questionnaire contains 36 items to evaluate to what extent an individual’s health influences his/her physical, mental and social functioning.30, 31 The questionnaire has shown its usefulness in a variety of other GI disorders.23

Statistical analysis

The aim of the analysis was to explore the factors that could explain the variability of the level of mental and physical health and the differences in the health status between constipated and non-constipated subjects as well as between countries. The model was used in a descriptive way to assess the relationship between general health (GH) status and constipation, overall and by countries. Data were analysed using the SAS system 8.2 (SAS Institute Inc., Cary, NC, USA) multiple linear regression and the anova one-way procedures (Proc REG and Proc one-way anova). The significance level was set to α = 0.05 (two-tailed). The Pearson’s product moment correlation was used to identify the correlation between the respondents’ SF-36 scores and the number of symptoms and complaints reported in the interviews.



The sample size and the distribution of the respondents in the countries are shown in Table 1. Both samples (constipated and not constipated) were identically sized in all countries and the differences in sample size among countries were negligible. In both samples, the respondents were mainly middle-aged subjects (approximately 50% between 30- and 60-year old), 60% were married or living with a partner, >80% lived in a household with more than one subject and approximately 50% were employed either part-time or full-time (Table 2). Our previous study showed that approximately 80% of the population studied complied with Rome II and III criteria.29

Table 1.   Distribution of respondents by country
Country Constipated (n = 1435)Non- constipated (n = 1435)Percentage of total surveyed population (%)
nFemale (%)nFemale (%)
South Korea200702007013.9
Table 2.   Demographics of the population studied
 Constipated n (%)Non-constipated n (%)
Total sample14351435
 Male427 (29.8)427 (29.8)
 Female1008 (70.2)1008 (70.2)
 <16 years
 16–29 years276 (19.2)276 (19.2)
 30–59 years711 (49.5)711 (49.5)
 60–74 years394 (27.5)394 (27.5)
 75 years and older 54 (3.8) 54 (4)
Marital status
 Single/living alone264 (18.4)261 (18.2)
 Living with a partner/ married835 (68.2)885 (61.7)
 Widowed/divorced307 (21.4)252 (17.6)
 Single/living with parents 28 (2) 37 (2.6)
Household size
 1 subject266 (18.5)260 (18.1)
 2 subjects432 (30.1)413 (28.8)
 3 subjects281 (19.6)307 (21.4)
 4 subjects298 (20.8)277 (19.3)
 5 subjects and more 158 (11)178 (12.4)
 Part-time546 (38.1)532 (37.1)
 Full-time192 (13.4)176 (12.3)
 Unemployed696 (48.5)727 (50.7)
 No answer1 (0)


Tables 3–5 and Figures 1 and 2 outline the results of the univariate analyses for all SF-36 factors. Results shown are mean scores for SF-36 factors. Physical functioning (PF), role physical (RP), bodily pain (BP) and general health (GH) are combined to yield the Physical Component Score (PCS). Vitality (V), social functioning (SF), role emotional (RE) and mental health (MH) are psychological factors that, combined, result in the MCS.

Table 3.   Descriptive statistics for the SF-36 main scales according to gender, age and employment status
  1. PF, physical functioning; RP, role physical; BP, bodily pain; GH, general health; PCS, Physical Component Score; V, Vitality; SF, social functioning; RE, role emotional; MH, mental health; MCS, Mental Component Score. NA, not applicable; SF-36, Short Form 36.

  2. Abbreviations for SF-36 scores see Materials and methods.

  3. * Only constipated individuals; † Statistically significant differences between populations (i.e. constipated and not constipated subjects); ‡ Statistically significant differences due to the variable considered; § Constipated; ¶ Non-constipated.

  4. Italics: differences within the scale not significant.

Age class†‡
 16–29 years§27653.252.754.244.252.753.246.850.947.647.7
 16–29 years¶27654.353.756.752.455.855.452.751.349.950.3
 30–59 years§71151.350.651.844.
 30–59 years¶71152.451.154.049.953.252.450.449.348.448.5
 60–74 years§39444.54447.342.744.847.745.744.745.846.3
 60–74 years¶39445.745.348.944.6465047.745.648.548.5
 75 years and older§5431.435.139.137.534.441.540.537.941.643.4
 75 years and older¶5437.937.945.440.
Consultation of a physician*†‡
Table 4.   Descriptive statistics for the SF-36 main scales according to country
  1. PF, physical functioning; RP, role physical; BP, bodily pain; GH, general health; PCS, Physical Component Score; V, vitality; SF, social functioning; RE, role emotional; MH, mental health; MCS, Mental Component Score; SF-36, Short Form 36.

  2. Abbreviations for SF-36 scores see Materials and methods.

  3. * Statistically significant differences between populations (i.e. constipated and not constipated subjects); † statistically significant differences due to the variable considered; ‡ Constipated; § Non-constipated.

  4. Italics: differences within the scale not significant.

 South Korea‡20051.453.358.245.153.651.452.152.146.348.6
 South Korea§20053.154.157.347.154.252.853.751.949.450.3
Table 5.   Descriptive statistics for the SF-36 main scales according to marital status and household type
  1. PF, physical functioning; RP, role physical; BP, bodily pain; GH, general health; PCS, Physical Component Score; V, vitality; SF, social functioning; RE, role emotional; MH, mental health; MCS, Mental Component Score; SF-36, Short Form 36.

  2. Abbreviations for SF-36 scores see Materials and methods.

  3. * Statistically significant differences between samples (i.e. constipated and not constipated subjects); † Statistically significant differences due to the variable considered; ‡ Constipated; § Non-constipated.

  4. Italics: differences within the scale not significant.

Marital status*†
 Living with a partner/married‡83550.049.251.344.850.
 Living with a partner/married§88551.350.453.448.952.052.250.449.049.049.0
 Single, living with parents‡2842.346.843.930.950.537.246.44244.341.5
 Single, living with parents§375254.354.854.858.652.15352.153.454.2
Household Size (*; exception: SF = NS)†
 1 subject‡26645.54547.84347.344.944.244.644.946.1
 1 subject§26046.94650.346.35048.245.747.747.847.8
 2 subjects‡43248.347.149.444.548.746.746.545.846.248.3
 2 subjects§41349.347.951.847.351.249.748.248.849.249.5
 3 subjects‡28149.949.551.6435145.248.246.946.749.6
 3 subjects§3075251.254.549.65350.34948.248.553.2
 4 subjects‡29852.152.353.744.351.547.550.645.546.752.4
 4 subjects§27752.251.854.349.752.35149.448.848.853.3
 5+ subjects‡15849.650.450.640.850.544.549.245.846.449
 5+ subjects§17852.551.853.851.253.950.249.949.649.653.4
Figure 1.

 Mean physical scores (error bars: S.E.M.) of the SF-36 Health Survey in constipated and non-constipated subjects. All differences are significant (P < 0.01 for PF, BP, GH and PCS and P < 0.05 for RP). PCS, Physical Component Score; PF, physical functioning; RP, role physical; BP, bodily pain; GH, general health; SF-36, Short Form 36.

Figure 2.

 Mean mental scores (error bars: S.E.M.) of the SF-36 Health Survey in constipated and non-constipated subjects. All differences are significant (P < 0.01 for V, SF, MH and MCS and P < 0.05 for RE). MCS: Mental Component Score; V, vitality; SF, social functioning; MH, mental health; RE, role emotional; SF-36, Short Form 36.

Overall, the GH status, as measured by the aggregate scores of the Physical Component and Mental Component, was significantly worse in the constipated than in the non-constipated population. The subscales of both components also demonstrated significantly worse HRQoL in constipated compared with non-constipated individuals (Figures 1 and 2).

Table 3 shows significant differences between constipated and non-constipated subjects in all subscales after adjusting for the effect of gender. They were large (>5%) for GH perception, RE, SF and MH, in particular for the latter two, for which almost four points of difference in the standardized scores can be seen between the two samples. Differences were significant (but <5%) for PF, RP and BP. Females reported more impaired HRQoL than males in the PCS (constipated males, 50.7 and females, 48.4; non-constipated males, 52.4 and females, 50.7; P < 0.001). Similar findings were reported for the MCS (constipated males, 47.5 and females, 45.6; non-constipated males, 50.2 and females, 48.1; P < 0.001).

In all age groups, non-constipated individuals had a better HRQoL than constipated ones (Table 3). Quality of life scores correlated negatively with age. The physical components of SF-36 were, as expected, very low in the elderly (constipated, over 74-years old: 34.4; non-constipated, over 75-years old: 39.2; P < 0.001). The size of this subset (n = 54), however, is very small. Constipation did not seem to lower significantly the quality of life in young subjects with regard to the physical components, but the social and mental components revealed a significant association between constipation and psychological well-being (MCS, 16–29 years: 47.7 vs. 50.3; SF: 46.8 vs. 52.7; P < 0.001 in both cases).

As our data were standardized to reflect normative scores of the US population, the overall physical and MCSs were very close to 50 in both samples (constipated and non-constipated people) in the USA. The scores for the constipated subjects were, however, constantly lower than those of the non-constipated ones (Table 4). Constipation seems to have a major effect in Brazil with regards to SF (35.8 vs. 51.3, P < 0.001) and GH perception (29.4 vs. 54.4, P < 0.001). A similar effect, albeit to a lesser extent, was seen in French, German and UK respondents, whereas HRQoL of Italian and South Korean people did not seem to be particularly heavily influenced by constipation.

In general, constipated subjects consulting a physician had statistically significantly lower HRQoL scores than non-consulters (Table 3) in all subscales of the SF-36 questionnaire.

Employment played a major role in HRQoL (Table 3). The HRQoL scores of unemployed or retired respondents were constantly and significantly lower than those of the other subgroups. Constipation, however, also influenced significantly the GH perception and the MH score of HRQoL of those employed full-time or part-time.

Health-related quality of life scores of constipated singles living alone were more affected with regard to GH perception (46.2 vs. 50.2) and SF (47.3 vs. 50.7) than those living with a partner, while the MH (46.4 vs. 49) and GH perception (45.2 vs. 48.4) scores were more reduced in the married subgroup (Table 4). Constipated singles living with their parents reported a large impairment of their quality of life, both in the PCS (41.5 vs. 54.2) and the Social Component scale (44.3 vs. 53.4), with very low values for SF (37.2 vs. 52.1) and MH score (42 vs. 52.1). The size of the household, on the other hand, did not seem to influence greatly HRQoL, although those living in large households experienced more problems with regard to MH and GH perception, and these were aggravated by constipation symptoms (MH score for households with more than five people: 45.8 vs. 49.6; health perception score for households with more than five people: 40.8 vs. 51.2).


Functional GI disorders, particularly constipation, are widespread problems in the countries investigated. A previous survey has indicated that a significant percentage of the population considered themselves to be constipated, with prevalence values ranging from 5% in Germany to 18% in the USA.32 This study has shown that the majority (approximately 80%) of the respondents who consider themselves to be constipated fulfilled both Rome II and III criteria.29

In our survey, we interviewed 2870 subjects in seven countries and demonstrated that constipation is negatively correlated with quality of life of individuals to a similar extent in all countries. The QoL impairment, at first sight, was apparently modest, but comparable with that caused by other disorders. We observed changes ranging from 2–20%. For instance, the impairment reported by constipated singles living with their parents in the GH and Vitality scores is very close to that reported by patients with chronic obstructive pulmonary disease (COPD).10 Mallon et al.33 have described similar values for acne patients. Overall, the impairment observed in our study is comparable with that observed in patients reporting a single limiting long-standing illness in the Oxford Health and Lifestyle Survey,33 in patients suffering from gastro-oesophageal reflux34 or in women with a history of hypertension, diabetes, heart disease or depression.35

Previous studies have reported that the treatment of constipation relieves HRQoL.11, 36 Although this suggests that constipation is a possible cause of HRQoL deterioration, we have no data to support causality. The possibility that in selected groups of subjects (e.g. singles living with parents) HRQoL deterioration caused by psychological or physical distress leads to constipation cannot be excluded and needs to be studied in future investigations.

Using the SF-36 QoL questionnaire, we were able to detect significant differences in HRQoL between constipated and non-constipated individuals. This is an indication of the strong correlation between constipation and quality of life and confirms previous observations in IBS patients, whose SF-36 scores were correlated with the bowel symptom scores reported by patients and gastroenterologists.37

A number of studies have indicated that functional constipation adversely affects QoL. Generic instruments such as the Psychological General Well-Being Index, combined with the GI Symptom Rating Scale, have been used to study QoL in women with chronic constipation18 and in patients with different types of functional constipation.14 Using the SF-36, Mason et al. demonstrated that QoL was significantly impaired in patients with intractable constipation and that the SF-36 scores were improved after behavioural treatment.36 Overall, SF-36 is a generic QoL instrument that has been proven to be sensitive even to small changes in physical and psychological well-being: Thaler et al.38 have studied QoL changes in patients undergoing total abdominal colectomy for the treatment of colonic inertia and have reported scores very similar to those observed in our survey.

Chang39 reported a decrease of HRQoL in patients with functional GI diseases, in particular those suffering from IBS or functional dyspepsia. Dennison et al.11 expanded this to other forms of GI disorders, including constipation. We have observed that constipated individuals consistently report low scores in the psychological components of the SF-36 questionnaire. Young subjects seem to be particularly affected in their MCS and SF. It seems, therefore, that constipation affects psychological more than physical well-being. PCS Scale observed by Stahl et al.10 in COPD stage 0 patients were decidedly lower – although not substantially – than those we observed in constipated individuals who, in turn, had lower values of their MCS Scale.

Demographic and social factors have long been recognized as relevant to the presence or development of constipation. Women report constipation more often than men do,40 but few data have shown that this is correlated with a corresponding decrease in HRQoL. We have observed that constipated women are generally more likely to report impaired HRQoL than men, in particular with regards to SF, MH and GH perception. We were also able to show that people living under presumed psychological stress such as unemployment or retirement report lower HRQoL than working or married individuals. Similar data have been reported for men suffering from prostate cancer.41

The reduction of HRQoL seen in the healthy elderly is most likely correlated to their reduced physical well-being and this would explain the low SF-36 scores reported by the non-constipated elderly population. It has already been shown that constipation further reduces quality of life in the elderly42 and our study has confirmed this.

No previous studies have analysed national differences. Constipation affected well-being, regardless of geographic location. We detected differences between constipated and non-constipated people in all countries studied that were significant even after adjusting for other important variables. Statistically significant differences among countries were detected only for BP, GH, Vitality, SF and MH.

Our study has some limitations. Interviewing constipated and non-constipated individuals in separate surveys may have introduced some sampling bias and a sample size of 200 individuals in each country may not have been sufficient to detect differences among countries. One study, on the other hand, has shown that approximately 150 patients would be needed in a two-arm trial on upper respiratory tract infections to detect a significant change in HRQoL with a power of 0.9.43 We also cannot conclusively assess the influence of co-morbidities, such as IBS or anxiety on the SF-36 scores. The associated use of another scale would have been most helpful, but we are not aware of another validated scale that could have been used in all countries. Further, the separation of the two populations in the survey did not allow a reliable statistical model to detect the influence of each covariate on HRQoL. The a priori gender split may have introduced some bias in the data. The population surveyed was composed predominantly of urban dwellers: the results of the investigation may not be directly applicable to rural people.

Constipation apparently reduces several components of HRQoL, particularly on its psychological component. It should not be dismissed as trivial, as the changes observed are comparable with those reported for other disease states.33–35, 44 The improvement of quality of life is therefore an important goal in the management of constipation.


Declaration of personal interests: A. Wald, M. A. Kamm, S. Mueller-Lissner and C. Scarpignato have served as speakers, consultants or as advisory board members for Boehringer Ingelheim GmbH. I. Helfrich, C. Schuijt and J. Bubeck are employees of Boehringer Ingelheim, a company manufacturing laxatives. Declaration of funding interests: This study was funded in full by Boehringer Ingelheim GmbH.