Background
Severe constipation is a chronic condition with major morbidity and health care burden [
1,
2]. As symptoms of constipation are the first points of discussion between a patient and their clinician the ability to subtype patients based upon these symptoms is important because, ideally it should direct a logical and cost-effective investigative pathway without the need for expensive and inconvenient testing.
Two of the most commonly identified symptoms of constipation are infrequent and hard lumpy stools. Both of these characteristics are fundamental to major current consensus-derived definitions of constipation such as the Rome III criteria [
3] and the American College of Gastroenterology Chronic Constipation Task Force [
4]. The Rome III criteria also include straining, a feeling of incomplete evacuation and manual maneuvers to facilitate defecation. It is noteworthy that all these criteria are established in the absence of laxative use. Indeed the Rome criteria qualified by the statement that loose stools are rarely present without the use of laxatives [
3].
However, the ability to critically and realistically assess constipation symptoms in the absence of laxative use can be difficult in a chronically constipated population. Laxatives remain a primary treatment option in such patients [
5,
6] and in many instances patients may have been taking laxatives for many years. Therefore when asked at interview or in a questionnaire to detail their stooling characteristics in the absence of laxatives they have to do so from memory. Responses based purely on recall of stooling habits have been shown to be unreliable [
7‐
9].
The use of stool diaries avoids the problems associated with patients' recall ability [
9‐
16]. However, in systematic studies of constipation these diaries are usually kept with patients off laxatives [
8,
13]. As laxative usage can potentially alter stool frequency and consistency, the stereotypic classifier of infrequent hard stools in severe constipation may bear little or no resemblance to the reality of the day-to-day life in these patients.
This question is important to the clinician and the researcher, particularly when trying to quantify stooling habits and outcomes in clinical trials. There is a need to try and classify constipated patients according to subtype [
17,
18]. This is best achieved by physiological function tests of colonic transit and anorectal evacuation. These tests are costly, inconvenient and not always readily available. Currently, based on existing symptom measurement tools, prediction of the physiological measures is imperfect [
5]. It is unknown whether laxative-related perturbation of symptoms may contribute to this lack of predictive value of symptoms for measureable mechanical dysfunction.
There has never been a systematic evaluation of the potential impact of laxative usage on symptoms in patients with constipation. Utilising a 3-week stool diary completed by patients with slow transit constipation, our aim was to evaluate, on a day-to-day basis, this relationship and determine the specific impact of laxative use, if any, upon stool frequency and form, straining, feelings of complete evacuation in addition to reported instances of abdominal pain and bloating. Specifically we hypothesised that all of these symptoms are affected by laxative use in patients with slow transit constipation.
Discussion
This study indicates that frequent, loose stools, pain and bloating are all commonly reported in a patient group with confirmed slow colonic transit and constipation. Indeed three quarters of our patients fell within the defined 3/wk - 3/day normal range for stool frequency [
25,
26] and 78% described normal to loose stool. Straining at stool and a feeling of incomplete evacuation were also commonly reported. Over 80% of our patients regularly took laxatives, and our data indicate that laxative use has a significant impact upon the reported stool frequency and form. However despite increasing the stool frequency laxatives had little affect on a patient's ability to feel as though they had completely evacuated and they did not reduce the incidence of straining. Laxatives also had no apparent affect upon abdominal pain and bloating suggesting that pain and bloating are intrinsic to the constipation.
Laxatives remain the primary treatment option for severe constipation [
6] and in the US the annual expenditure on laxatives is estimated to exceed $US850 M [
27,
28]. Yet despite the common usage their potential impact upon the day-to-day constipation symptoms is often overlooked and this raises potential problems when attempting to critically assess the reality of symptoms in chronically constipation patients. In this study 91% of our patients reported having constipation for at least 5 years (62% >10 yrs) and laxative use for many had become part of every day life. Asking such patients to detail their stooling habits in the absence of laxative use is unlikely to provide realistic data. Indeed in our experience the classic response to the question "how often would you open your bowels a week without laxative use?" is "I'd never go". While such responses may represent the patient's perspective of their bowel function, several studies have shown that a patients recall ability of stool habits is inaccurate [
7‐
9].
These inaccuracies could potentially confound study findings. For example previous studies have suggested that infrequent stools may indicate delayed transit [
10,
12,
29] and some studies have presented supportive evidence for this [
30,
31], while others have not [
9,
13]. When looking at the methodology of these studies it is of interest to note that those that associated infrequent stools with delayed transit, based the patient's stool frequency upon answers provided in a questionnaire [
30,
31]. Thus the acquired data relied on the patients recall ability, presumably in the absence of laxatives, which is more likely to indicate a low weekly stool frequency. Those studies that found no correlation between stool frequency and delayed colonic transit assessed the stool frequency on data obtained in stool diaries [
9,
13]. While not directly assessed here our data would also suggest that neither stool form nor frequency provides an accurate indicator of slow colonic transit.
To overcome the impact of laxatives upon constipation symptoms, stool diaries could be maintained in the absence of laxatives use, however in our experience chronically constipated patients become dependent on laxatives and can be unwilling to stop using them for even short periods of time. Such patients can be excluded from studies assessing constipation symptoms [
8] but then these studies risk removing patients at the more severe end of constipation scale.
The potential for laxatives to alter stool frequency and form can also have implications for studies that use constipation symptoms in an attempt to define constipation subtypes. The mathematical techniques of factor analysis have been used in an attempt to determine whether certain symptoms groups can help to define subtypes of constipation. Studies applying these techniques have found conflicting results with Mertz et al [
32] identifying 3 constipation subtypes while Eltringham el at [
33] did not. In our own recent study, 4 subtypes were identified, however there was significant overlap amongst these subtypes [
17]. Given that the approach to constipation subtyping in our study relied entirely on answers provided in a questionnaire, it is important that patients understand how they are supposed to complete the questionnaire. For example we have had many patients ask if they should complete the questionnaire based on what they are like on or off laxatives. Therefore within any population of constipated patients the questionnaires may be assessing; i) a patient's estimate of what they believe they would be like without laxatives; ii) what they are actually like on laxatives; or iii) a combination of both situations.
While laxatives in this study increased stool frequency, it is important to note that on the majority of laxative affected days patients still had to strain to open their bowels and were left with a feeling of incomplete evacuation. As a consequence the patient still feels constipated. This highlights a potential problem associated with using stool frequency as a measure of constipation severity. To manoeuvrer around this many studies use 'spontaneous bowel motion' or 'complete spontaneous bowel motion' to define bowel habit [
4,
34]. However, once again in chronically constipated patients that have become dependant upon laxatives, the reality of obtaining several weeks of stool frequency data in the absence of any laxative use could be difficult. In such patients we would suggest that a more accurate measure of constipation severity could be the number of days per week in which a patient achieves a feeling of complete evacuation in light of their laxative use. In our patients, while many fit the 'normal' stool frequency criteria, the vast majority experienced a bowel movement associated with a feeling of complete evacuation on less than 3 days per week.
An important and often overlooked factor to emerge from our stool diaries is the prevalence of pain and bloating. Frequent, loose stools, abdominal pain and bloating are all symptoms commonly associated with irritable bowel syndrome, and the current Rome criteria for the diagnosis of constipation does not even mention the presence of pain or bloating [
3]. Only 8% of our patients reported an absence of these symptoms whereas >27% reported these symptoms as interfering or preventing normal daily activities. Previous studies have shown a large proportion of patients with constipation, including those with slow transit constipation, potentially met the Rome criteria for IBS [
17,
35,
36], thus highlighting the need to include pain and bloating in the assessment of constipation symptoms.
There are potential criticisms with our study. Straining and pain/bloating, reported by 89% and 92% of our patients, respectively are commonly reported by patients who have a disorder of rectal evacuation [
37,
38]. Given that tests of anorectal function were not used to exclude patients with co-existent disordered rectal evacuation, it is likely that such mixed disorders existed in some of our study population and this could potentially impact upon our results. Indeed the inclusion of such patients has recently been put forward as a potential confounder in clinical trials of prokinetics in constipation [
39]. However, this assertion still remains to be proven and as there is a large body of evidence that suggests there are no symptoms that are reliably predictive of an evacuation disorder [
35,
40‐
43], we feel that the potential inclusion of such patients should not detract from the overall symptomatic findings in our study.
Other potential criticisms of this study include the fact we did not measure the size or the weight of stools that were passed and we therefore we have no ability to determine what a patient considers to be a 'bowel motion'. Passing a small amount of hard or liquid stool may have constituted a bowel motion for some patients. Of course this factor remains true for almost all symptom based constipation studies. Finally while there is a risk of non-compliance with reporting the daily symptoms, especially over a three week period, we found that with these particular patients, who were all seeking help for the constipation symptoms, there was a willingness to comply and we are therefore confident that these data are an accurate representation of the symptoms in this population.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
PD & MJ; study concept and design, statistical analysis, draft and critical review of manuscript; LH data collation and entry; JK, DL & IC, patient selection and critical manuscript review. All authors all authors read and approved the final manuscript.