A total of 331 consecutive patients with the complaint of constipation (as per their own perception) were included in the prospective cross-sectional study from April 2012 to April 2015. This study hospital is a tertiary care hospital with a vast and predominantly general referral base and also has an open access system for patient consultation. Thus apart from tertiary care, it serves primary and secondary care also.

Patients were mostly recruited from OPD attendance, except some indoor patients admitted for other reasons who complained of constipation. Evaluation included: (i) a structured history based on questions pertaining to diagnosis of FC and IBS - C using the Rome III criteria [Table/Fig-1,2]; (ii) Evaluation for organic disease including history of alarm features, co-morbid conditions (Parkinson’s disease, vascular dementias, ischemic brain and heart diseases) and drug intake (like calcium, iron, calcium channel blockers, beta blockers, antidepressants and antipsychotics which are known to cause constipation), thorough physical examination and investigations including complete haemogram, blood biochemistry, fasting and postprandial blood sugar, thyroid function test, stool microscopy and occult blood test and colonoscopy.

To compare with Western criteria and also to determine the objective incidence in those having the subjective complaint, constipation was defined as any frequency less than 1/day (7/week) since normal bowel frequency in 90% of Indian population is 1-2/day [11].

Results were expressed as proportions (%). Continuous variables and discrete variables were compared with Student’s t-test and chi-square test respectively and two sided p <0.05 was considered significant. The study was approved by the hospital ethics committee and all patients consented to be included in the study.

The study results are shown in [Table/Fig-1,2 and 3]. All patients consumed mixed diet, none were pure vegetarians. 7/331 (2.1%) patients were diagnosed with cancer colon and all had alarm features like bleed per rectum, anaemia, weight loss. These were excluded from further analysis. Of the remaining 324 persons, 211 (65%) were above age of 60 years. 64 (19.8%) had normal bowel frequency i.e. at least 1 motion/day. Of them, 16 (25%) complained of straining and 60 (94%) of incomplete evacuation. FC [Table/Fig-1] was diagnosed in 224 (69.1%) patients overall by Western criteria while by Indian standard, it was in 224/260 (86.2%). The corresponding figures for IBS-C [Table/Fig-2] were 36 (11.1%) and 36/260 (13.8%).

BSS Type 1-3 stool was passed by 93.8% patients overall, 81.3% patients with normal stool frequency (by Indian standards), 100% in the FC group and 77.8% in the IBS-C group. Only 67.9% patients passed BSS type 1 and 2 stool. Most patients reported stool frequency of 3-4/week both in FC (50.9%) and IBS-C group (55.6%).

There was male preponderance in both IBS-C and FC group. BSS Type 1 and 2 was significantly more in patients of FC whereas Type 3 was more in those with normal bowel frequency. Except that more males had Type 4 stool and less than 4/week frequency than females, there was no gender difference for other stool types or frequencies in this cohort.

Co-morbidities were present in 10 (15.6%), 190 (84.8%), 3 (8.3%) of those with normal bowels, FC and IBS-C respectively. It was significantly more in FC group compared to the other two (p<0.05). Overall 122/324 (37.7%) regularly took drugs known to cause constipation.

This initial study of patients who have consulted doctors for their subjective feeling of constipation shows that subjective feeling is at variance with definition of constipation by stool frequency and form. The main difference with Rome III criteria of FC and IBS-C in Indian patients complaining of constipation also lies both in frequency and form of stool. Overall 64.6% of our patients had normal stool frequency and 32.1% normal stool form (BSS type 3,4) by Western criteria. Similar data was reported in a previous population based Indian study evaluating lower gastrointestinal symptoms (which included 637 controls and 398 cases from the present study area) where of 53% patients complaining of constipation, only 19% had true constipation by Western criteria and 64% had normal stool frequency [11]. Indian diet usually contains food with high fiber intake like chapatti, vegetables. Stool weight in Indians is also higher than West [16]. These factors lead to short colonic transit time and softer stool so that the frequency of defecation even in those with constipation is higher than the West. This fact is reflected in another recent Indian study evaluating bowel habit in the general population where the prevalence of constipation by stool frequency was very low (2.6%) by Western definition (<3/week) while it was 8.8% by Indian definition (<1/day). The corresponding figures for stool forms were 5.1% (BSS type1-2) and 13.3% (BSS type1-3), most patients pass type 4 stool [15]. Since this study was conducted in the state neighbouring that of the present study (about 450 km distance) with very similar dietary and sociocultural habits, both these studies can be considered as control group for comparing data. The main difference of the present study from these two studies done in the general population is that it is carried out only in patients whose constipation was bothersome enough for medical consultation and hence likely to reflect more objective data on the symptom. It is common knowledge that constipation increases with age and 65% of our patients were above 60 years of age. This is an indirect proof of the representation of our patient cohort. This may also be the reason why colon cancer appears slightly higher.

Stool form (e.g. harder BSS Type 1-3 stool) is also important in the subjective feeling of constipation. In the present study BSS Type 1-3 stool was passed by 93.8% subjects and endorses the Asian criteria for constipation. 51.5% of patients with constipation had stool frequency of 3-4/week hence a frequency of less than 5/week (rather than 7/week) appears more indicative of constipation in these subjects. In the recent MIIBS study [17], 19.7% and 40.5% patients could be classified into IBS-C based on stool frequency and form respectively, 77.6% had normal stool frequency. An interesting fact is that in defining IBS-C, patient number increased from 16 to 28 if BSS Type 1-3 is considered instead of Type 1-2 (as in Rome III criteria). A study from Israel [18] showed that just by changing the frequency requirement for symptoms from 25% to 10%, the prevalence of IBS increased with corresponding decrease in FC keeping overall prevalence of functional bowel disease constant. Feeling of incomplete evacuation was reported by 98.8% and may reflect the feel of constipation in Indian subjects. These factors should be considered while defining constipation in the Indian context.

Previous Indian studies [11,15,17] showed constipation to be more prevalent in elderly females and IBS in males. Our study showed male preponderance in both IBS-C and FC group. BSS Type 1 and 2 being the predominant stool form in patients of FC appears logical considering their advanced age compared to others. A recent hospital based Indian study evaluating constipation by Rome III criteria found FC in 58%, and IBS-C in19% with male preponderance [19], figures matching our results.

Last of all, co-morbidities and drugs also appear to have their share to contribute to constipation especially in the elderly (60 years and above, [Table/Fig-3]).

This initial conceptual study evaluating Indian patients consulting doctors for bothersome constipation shows that BSS types 1-3 and stool frequency of less than 5/week are good markers of constipation in this population. Feeling of incomplete evacuation should be given due weightage. Most constipated patients are elderly with secondary co-morbidities. All these factors need to be considered while defining constipation in India and conducting population based studies on the subject. Colon cancer rarely presents with constipation only.