AIM: Bowel habits are difficult to study, and most data on defecatory behaviour in the general population have been obtained on the basis of recalled interview. The objective assessment of this physiological function and its pathological aspects continues to pose a difficult challenge. The aim of this prospective study was to objectively assess the bowel habits and related aspects in a large sample drawn from the general population.

METHODS: Over a two-month period 488 subjects were prospectively recruited from the general population and asked to compile a daily diary on their bowel habits and associated signs and symptoms (the latter according to Rome II criteria). A total of 298 (61%) participants returned a correctly compiled record, so that data for more than 8000 patient-days were available for statistical analysis.

RESULTS: The average defecatory frequency was once per day (range of 0.25-3.25) and was similar between males and females. However, higher frequencies of straining at stool (P = 0.001), a feeling of incomplete emptying and/or difficult evacuation (P = 0.0001), and manual manoeuvres to facilitate defecation (P = 0.046) were reported by females as compared to males.

CONCLUSION: This study represents one of the first attempts to objectively and prospectively assess bowel habits in a sample of the general population over a relatively long period of time. The variables we analyzed are coherent with the criteria commonly used for the clinical assessment of functional constipation, and can provide a useful adjunt for a better evaluation of constipated patients.

Bowel habits are a difficult function to study objectively because of their highly private nature and negative associations. Therefore, it is not surprising that they represent one of the least understood aspects of human behaviour[1]. In the past, most knowledge of bowel habits was drawn from limited data on small groups of subjects (nurses, jail prisoners, elderly people, students)[2-5]. More recently, studies aimed generically at investigating functional gastrointestinal disorders[6-10] have yielded data on large numbers of subjects by means of telephone interviews or mailed questionnaires[11-14]. However, these studies and other reports have all been retrospective in nature and based on the subjects’ assertions regarding their recent bowel function[1,15,16]. More objective investigations have assessed small groups of subjects for limited periods of time (e.g. one week)[17].

Prospective studies on bowel habits conducted over an adequate period of time in the general population are still lacking. The aim of our prospective study was to objectively assess the frequency and characteristics of defecation in a sample of the general population over a longer period of time.

During a two-month period a questionnaire was consecutively distributed to 488 relatives or friends of patients attending the outpatient gastrointestinal clinic in six centres located in different regions of Italy (two in the north, two in the centre, and two in the south). A total of 259 women and 229 men received the form. To obtain the most objective possible data on bowel habits, the questionnaire took the form of a diary covering a period of 4 wk in which “yes-no” responses were to be given daily to six questions (Table 1). Drawing upon the Rome II criteria for functional constipation[18], data on the following symptoms and signs were recorded each day, namely number of bowel movements, straining during bowel movements, feeling of incomplete emptying and/or difficult evacuation, manual manoeuvres to facilitate defecation, lumpy or hard stools. In addition, the use of laxatives was recorded.

The questionnaires were anonymous, and the only personal information the participants were required to give was their age and sex. All subjects received an exhaustive explanation about the aim of the study and the structure of the questionnaire.

Each centre received approval from the local ethics commitee, the written consent of all subjects was obtained after they had been given a complete explanation of the aims of the study and the nature of the questionnaire, and the study was conducted in accordance with the Helsinki Declaration (Edinburgh revision, 2000).

For each subject, an overall score for each variable was computed as follows. The average number of bowel movements per day was obtained by taking the total number of defecations reported by the participants and divided by the total number of days in the study period (i.e., 28). The frequency in the use of laxatives was computed in the same way. The frequency of the four variables associated with defecation (straining, feeling of incomplete evacuation, need of manual help, lumpy/hard stools) was evaluated as the ratio between the total number of episodes recorded by the individual and the total number of bowel movements during the study period.

A total of 298 adult subjects (163 women (54.7%), mean age 42.5 ± 15.5 yr and 135 men (45.3%), mean age 42.4 ± 15.9 yr) returned the completed questionnaire. The mean response rate was 61.1 % (females 62.9 %; males 58.9 %, n.s.). Therefore, data for 8344 d were available for statistical analysis.

The distribution of the participating subjects by age and sex is shown in Table 2. Table 3 reports the frequency of defecation, expressed as the average number of evacuations per day, the frequency of pathological features and sensations at defecation, and the use of laxatives per day. Overall, the frequency of bowel movements averaged one per day (range 0.25-3.25), and was similar between males and females. No significant intra-personal variation in the parameters under examination was detected over the four-week period. Higher frequencies of straining at stool (P = 0.001), feeling of incomplete emptying and/or difficult evacuation (P = 0.0001), and manual manoeuvres to facilitate defecation (P = 0.046) were reported by females as compared to males.

Table 4 shows the correlations between pairs of defecatory variables. Bowel movement frequency was negatively correlated with other features of defecation and the use of laxatives. Straining, a sensation of incomplete/difficult evacuation, manual manoeuvres, lumpy/hard stools and the use of laxatives were positively correlated with each other.

Concerning the relative weights of the single variables, it might be noted that 15 (5%) subjects showed a low frequency (< 3/wk) of defecations, 35 (11.7%) straining during > ¼ defecations, 32 (10.7%) incomplete/difficult evacuation during > ¼ defecations, 2 (0.7%) manual manoeuvres during > ¼ defecations, and 18 (6%) lumpy/hard stools during ¼ defecations.

Most studies on bowel habits have been based on phone interviews and on the assumption that people would report accurately, but there has been good evidence that bowel movement frequency might be misreported[19,20]. Indeed, it is very difficult to remember and report accurately one’s bowel habits over recent months in a 20 min interview. Studies have shown marked discrepancies between recalled data and data that was recorded daily[19,21,22]. Moreover, people without a telephone or who were not at home when contact was attempted would be excluded from any given survey[23]. Another source of bias was the possibility that symptomatic individuals would be more keen to complete the survey process than asymptomatic subjects, which might lead to an overestimation of the frequency of symptoms.

Validated and universally accepted criteria are definitely needed if functional bowel disorders are to become a formally recognized disease entity by physicians, patients, and society[24].

In order to circumvent some of the methodological biases discussed above, for this study a questionnaire designed to elicit the most objective possible data on individual bowel habits was drawn up. With this instrument bowel movement frequency, and sensations and characteristics related to each bowel movement were prospectively recorded by nearly 300 subjects on a daily basis for 4 wk. Moreover, to obtain a geographically representative sample of our population, participants were recruited from different parts of the country.

It may be stressed that the 61% response rate could be considered relatively high, given the nature of the data being sought. Studies employing telephone interviews or mailed questionnaires have yielded a response rate ranging from 19% to 80%. We believe that our high response rate can be attributed to the simplicity of the questionnaire (which examined only six items) and its complete anonymity. A daily dial-in service might have been more reliable, but this is costly to implement and potentially dependent on the socio-cultural environment in which the study is conducted.

It must be pointed out that due to the recruiting procedure used, the individuals who participated in this study were not selected with respect to factors such as social status, education, occupation and, possibly, the prevalence and type of bowel habits reported. However, they were prospectively recruited and not selected based on the basis of factors such as the presence/absence of pathological symptoms. Nevertheless, in light of the high participation rate, it seems reasonable to postulate that the results of this study provide an acceptable approximation of the prevalence and type of symptoms in a general sample of Italian adults.

We are certain of our findings to be underlined. Firstly, there was a relatively large variation in bowel movement frequency, with an average of one per day, but a range of one evacuation every 4 d to about 3 bowel movements per day, with no differences in distribution between the sexes. It might also be noted that 5% of the participants reported less than 3 evacuations/week. Secondly, the number of subjects who reported abnormal features during > ¼ defecations was low, in particular, the incidence of manual manoeuvres to facilitate defecation was almost nil (0.7%). However, these variables showed a positive correlation with one another. Thirdly, the prevalence of defecation-related variables (except for the presence of lumpy/hard stools) was significantly different between the sexes, with a higher frequency in women, and interestingly, all of these are variables related with pelvic floor function. The use of laxatives was also rare (5% of the population sample), but much (P = 0.0001) more frequent among women.

The variables analysed in this study could be helpful in the clinical assessment of functional constipation. Our data furthermore suggest that different symptoms and signs should be attributed to different weights in the evaluation of constipation scores[25,26]. For instance, a value of less than one defecation per week or the use of manual manoeuvres to facilitate defecation could represent clinically important indications for the diagnosis of constipation.

In conclusion, this study represents one of the first attempts to prospectively assess bowel habits in a general population sample over a long period of time. Further studies in “normal” subjects will obviously be needed to confirm these observations.

The following researchers of the Italian Constipation Study (ICS) Group participated in the study: Bassotti G, Chistolini F, Morelli A (Perugia); Bellini M, Alduini P, Mammini C, Rappelli L, Costa F, Stasi C, Mumolo MG, Berni I, Giorgetti S, Marchi S (Pisa); Pucciani F, Iozzi L, Cianchi F, Cortesini C (Firenze); Bocchini R, Cimatti M, Fornasari L, Montaletti I, Pazzi P (Cesena/Forlì); Bove A, Balzano A (Napoli); Battaglia E, Dughera L, Emanuelli G (Torino); Bruzzi P (Genova).