- Email: [email protected]
Irritable bowel syndrome according to varying diagnostic criteria: are the new Rome II criteria unnecessarily restrictive for research and practice?
THE AMERICAN JOURNAL OF GASTROENTEROLOGY © 2000 by Am. Coll. of Gastroenterology Published by Elsevier Science Inc.
Vol. 95, No. 11, 2000 ISSN 0002-9270/00/$20.00 PII S0002-9270(00)01990-0
Irritable Bowel Syndrome According to Varying Diagnostic Criteria: Are the New Rome II Criteria Unnecessarily Restrictive for Research and Practice? Philip M. Boyce, M.D., F.R.A.N.Z.C.P., Natasha A. Koloski, B.A. (Hons.), and Nicholas J. Talley, M.D., Ph.D., F.A.C.G. Department of Psychological Medicineand Department of Medicine, University of Sydney, Nepean Hospital, Penrith, New South Wales, Australia
OBJECTIVES: It has been suggested that the variation in the prevalence of irritable bowel syndrome (IBS) may be due to the application of different diagnostic criteria. New criteria for IBS have been proposed (Rome II). It is unknown whether persons meeting different criteria for IBS have similar psychological and symptom features. The aim of this study was to measure the prevalence of IBS according to Manning and Rome definitions of IBS and to evaluate the clinical and psychological differences between diagnostic categories. METHODS: A total of 4500 randomly selected subjects, with equal numbers of male and female subjects aged ⱖ18 yr and representative of the Australian population, took part in this study. Subjects were mailed a questionnaire (response rate, 72%). Characteristics measured were gastrointestinal symptoms over the past 12 months, neuroticism and extroversion (Eysenck Personality Questionnaire), anxiety and depression (Delusions-Symptoms-States Inventory), mental and physical functioning (SF-12), and somatic distress (Sphere). RESULTS: The prevalence for IBS according to Manning, Rome I, and Rome II was 4.4% (95% confidence interval [CI] ⫽ 3.5–5.1%), 6.9% (CI 6.0 –7.8%), and 13.6% (CI 12.3–14.8%), respectively. Only 12 persons with Rome I did not also meet Rome II criteria; 196 persons with Manning criteria did not meet Rome II cut-offs. Having IBS regardless of which criteria were used was significantly associated with psychological morbidity, but psychological factors were not important in discriminating between diagnostic categories. However, pain and bowel habit severity independently discriminated between diagnostic groups. CONCLUSIONS: IBS is a relatively common disorder in the community. The new Rome II criteria may be unnecessarily restrictive in practice. (Am J Gastroenterol 2000;95: 3176 –3183. © 2000 by Am. Coll. of Gastroenterology)
INTRODUCTION The irritable bowel syndrome (IBS) is characterized by abdominal pain and disturbance of bowel habits in the absence of demonstrable structural pathology. It is the most
common source of referrals to gastroenterologists (1), and is very costly to the community (2). Community-based studies have reported prevalence rates for IBS ranging from 9% to 22% (3–5). This wide range in the prevalence of IBS may be due to the varying thresholds associated with different diagnostic criteria. The diagnosis of IBS relies upon clinical criteria alone, as there is no gold standard. The diagnosis is dependent upon identifying characteristic symptoms and then distinguishing IBS from structural bowel disorders. Manning et al. (5) first operationalized diagnostic criteria for IBS, by identifying the symptoms that differentiated between IBS and organic bowel disease (with a physician diagnosis being the “gold” standard). Abdominal pain relieved with defecation, looser and/or more frequent stools with the onset of pain and abdominal distension, were the discriminating symptoms and are known as the Manning criteria. Other less discriminating symptoms included; mucus per rectum and a feeling of incomplete rectal evacuation (5). Using these symptoms with abdominal pain criteria, the prevalence of IBS is high, generally around 15–20% (4, 6). Despite some concerns about the sensitivity of Manning criteria for accurately diagnosing IBS in male subjects (7, 8), these criteria have provided the foundation for the more recently accepted diagnostic classification of IBS, the Rome criteria. Subsequently, the Rome I criteria were developed by consensus (9). The prevalence rate for IBS, using the Rome I criteria (3), is lower, generally around 9 –12% (10, 11). The Rome criteria have recently been modified by relaxing the duration threshold and removing the duplication of stool frequency and consistency in Part 1 of the criteria (11). In 1999, the Rome II criteria were published, in part based on factor analytic studies (12); the revised Rome criteria (Rome II) include only the first part of the original criteria, but require the presence of two out of three symptoms relating abdominal pain to bowel symptoms (11). In research and clinical practice, the key issue is what defines a “case” of IBS as opposed to the normal variation in bowel symptoms. The difference between “cases” and noncases is likely to be dependent upon the diagnostic thresholds that are used, although this has not been carefully
AJG – October, 2000
studied. These thresholds can be set across a number of different parameters. First, the severity of the bowel symptoms (or the degree of impairment or disability) can be set as the threshold to distinguish between cases and noncases. Second, the threshold could be based purely on a symptom count. This may be done in two ways; all symptoms being given equal weight (a polythetic approach) or alternatively with a particular symptom or symptoms being regarded as essential (a monothetic approach) to the diagnosis and other symptoms being regarded as accessory symptoms. Third, a duration threshold can be used, in combination with a symptom count. Finally, cases could be distinguished from noncases on a combination of the parameters. The Manning criteria thus apply a simple polythetic approach, in that the diagnosis is dependent upon two or more symptoms being present with all symptoms having equal weight (5). The Rome I criteria use a mixed approach, with a strict duration threshold and a combination of essential and accessory symptoms. The new Rome II criteria relaxed the duration threshold, to allow for the waxing and waning of symptoms over a long period (11). None of the diagnostic systems has applied a severity dimension, or impairment or disability criteria, despite the fact that the severity of bowel symptoms can be objectively assessed (13). Separate studies using Rome I and Manning definitions have found that persons with IBS are more likely to be female (3), have a lower socioeconomic status (3), greater psychological morbidity including higher levels of anxiety, depression, and somatization disorders (14 –16) and experience a reduced quality of life (17–20) compared with persons without IBS or those who have IBS symptoms in the community but who do not seek health care for them. There are no data, however, on whether persons meeting these various diagnostic versions of IBS are similar in terms of sociodemographic, psychological, and symptom characteristics. We postulated that persons fulfilling a more restrictive threshold for IBS such as the Rome I criteria would exhibit greater psychological disturbances and more impaired quality of life because of the burden of coping with more symptoms of greater duration, compared with those persons meeting the more liberal Manning criteria. If true, then this would have important implications for clinical trials and criteria development in regard to determining the optimal symptom cut-off points for a diagnosis of IBS, as well as diagnosis in clinical practice. Therefore we aimed, in a population-based cross-sectional study, to 1) measure the prevalence of IBS according to the Manning criteria and for the first time variations of the Rome criteria and 2) evaluate the clinical and psychological differences, if any, between key diagnostic categories.
MATERIALS AND METHODS Subjects A random sample of 4500 persons aged ⱖ18 yr, with equal numbers of male and female subjects, were selected from
IBS and Varying Diagnostic Criteria
3177
the 1996 electoral roll for the local government area of Penrith (3.6% of the Sydney’s population). Penrith is homogeneous in terms of its sociodemographics and ethnic composition with the Australian population, except that its inhabitants are slightly younger and it has a slightly higher socioeconomic status (1991 census) (21). Of the original 4500 subjects mailed the questionnaire, 398 subjects had moved or were overseas, 11 subjects had died, 24 subjects were seriously ill, 75 subjects explicitly refused to participate and 1082 subjects failed to respond. A completed questionnaire was returned by 2910 subjects, giving an overall response rate of 72%. There were slightly more female responders (51.9%) to the questionnaire compared with nonresponders (47.0%). Procedure The mail out of a questionnaire to subjects was carried out in three batches of 500, 2000, and 2000. In order to ensure that non–English-speaking persons were not excluded from the study, an information sheet, printed in the common non-English languages used by the residents of Lindsay, was included with the questionnaire. Non–English-speaking persons, if they were interested in participating, were invited to contact the research team so that an interpreter could be organized for them. Nonrespondents were extensively followed-up over a period of 10 wk, with a replacement survey sent out at wk 4 and reminder letters sent out in wk 2, 5, 7, and 10. A telephone call was also made to subjects in wk 9. Those subjects in batch 1 received an additional replacement survey at wk 8 and a telephone call instead of a reminder letter was carried out in wk 7. Questionnaire The complete version of the 32-page questionnaire contained 216 items divided into sections, the first containing questions on a range of gastrointestinal symptoms relating to the following specific anatomic regions: esophagus, gastroduodenum, bowel, abdomen, biliary, and anorectum. The criteria for IBS in the different diagnostic systems are shown in Table 1. Also in the early sections of the questionnaire, we asked about the number of times health care was sought over the past 12 months for gastrointestinal symptoms. The latter sections of the questionnaire consisted of measures of physical and mental functioning, psychological distress, and personality. A shortened version of the Eysenck Personality Questionnaire (EPQ) (22) was used to assess the personality traits of neuroticism and extroversion. Psychological distress was measured using the DelusionsSymptoms-States Inventory (DSSI) (23) a measure of anxiety and depression. Physical and mental functioning was assessed using the SF-12 (24) and the Sphere (25) provided scores for the subscales of anxiety/depression, fatigue/pain, and somatic distress. In addition, we included items to assess the sociodemographic characteristics of our sample. These included questions relating to age, sex, country of birth, and educational level.
3178
Boyce et al.
Table 1. Diagnostic Criteria for the Irritable Bowel Syndrome According to Different Systems Manning criteria Two or more of the following symptoms: Pain eased after bowel movement Looser stools at onset of pain More frequent bowel movements at onset of pain Abdominal distension Mucus per rectum Feeling of incomplete emptying Rome I criteria At least 3 months of continuous or recurrent symptoms of: Abdominal pain or discomfort that is Relieved with defecation; and/or Associated with a change in frequency of stool; and/or Associated with a change in consistency of stool; and Two or more of the following, at least on one-fourth of occasions or days: Altered stool frequency (for research purposes “altered” may be defined as more than three bowel movements each day or less than three bowel movements each week); Altered stool form (lumpy/hard or loose/watery stool); Altered stool passage (straining, urgency, or feeling of incomplete evacuation); Passage of mucus; and/or Bloating or feeling of abdominal distension. Rome II criteria At least 12 weeks (which need not be consecutive) in the preceding 12 months, of abdominal discomfort or pain that has two of three features: Relieved with defecation; and/or Onset associated with a change in frequency of stool; and/or Onset associated with a change in form (appearance) of stool.
Diagnostic Criteria Computer algorithms were written to generate Manning, Rome I, and Rome II prevalence of IBS. The questions used to elicit the symptoms of IBS according to the different diagnostic systems (the same questions were used for each system) are reproduced in Appendix 1. Bowel habit severity was based on the mean score of the frequency (from 0 ⫽ “Not at all” to 5 ⫽ “Very often”) of occurrence of nine bowel symptoms over 3 months in the past year. Statistical Analyses The prevalence (%) for all the diagnostic versions of IBS were calculated with the corresponding 95% confidence intervals. The sociodemographic, psychological, symptom, and health care–seeking associations were analyzed in two ways. First, all the variables were analyzed for associations with each diagnostic category using t tests for dimensional scores and 2 tests for categorical variables. Comparisons were made with a “control” group of subjects who did not meet a diagnosis of IBS with any of the criteria but could have suffered from other functional gastrointestinal disorders. Second, we compared Rome I versus Manning, and Rome I versus Rome II criteria, on a range of variables using a logistic regression with the adjusted odds ratios and 95% confidence limits reported. The possibility of nonindependent data because of multiple observations on each subject
AJG – Vol. 95, No. 11, 2000
Table 2. Overlap of Subjects Identified by Applying Different Diagnostic Criteria
Diagnostic Criteria Met None at all Manning criteria Manning alone Manning & Rome I Manning & Rome II Manning & Rome I & Rome II Rome I criteria Rome I alone Rome I & Rome II Rome I & Manning Rome I & Rome II & Manning Rome II criteria Rome II alone Rome II & Rome I Rome II & Manning Rome II & Rome I & Manning
% of Subjects Meeting Index No. of Criteria and Subjects Alternative Criteria 2512 395 185 126 199 115 127 1 115 126 115 201 2 115 199 115
46.8 31.9 50.4 29.1 0.8 90.6 99.2 90.6 1 57.2 99 57.2
may have led to within-group correlation in residuals and consequent incorrect statistical inference (incorrect standard errors). Standard errors were adjusted by viewing the data as though subjects were cluster sampled and taking into account the calculation of the design effect. The degree of adjustment is proportional to the intrasubject correlation. All p values calculated were two tailed; the ␣ level of significance was set at 0.05.
RESULTS The mean age of the sample was 43.8 yr (95% CI: 43.2– 44.3); 51.3% were female (95% CI: 49.0 –53.5%). A total of 398 (13.7%, 95% CI 12.3–15.2%) of the participants met criteria for IBS on at least one of the diagnostic systems. Prevalence of IBS A total of 127 participants met Rome I criteria (4.4%, 95% CI: 3.6 –5.2%) over the previous year. By contrast, 395 participants fulfilled the Manning criteria (13.6%, 95% CI: 12.2–15.1%). The prevalence using the revised Rome criteria, Rome II, was 6.9% (95% CI: 5.9 – 8.0%). Overlap of IBS Diagnostic Categories There was considerable overlap applying the different definitions, as shown in Table 2. For example, 127 subjects met criteria for Rome I diagnosis and, of them, 115 (90.6%) also met the Rome II criteria. On the other hand, 46.8% of the subjects who met the Manning criteria did not meet either of the Rome criteria. Sociodemographic, Psychological, and Symptom Associations With IBS Diagnostic Criteria We then examined the mean scores on the psychological variables (neuroticism, anxiety, depression, and somatic distress), the SF-12 (mental and physical functioning) and
0.000 0.000 0.000 3.6 (1.6) 2.0 (1.1) 8.0 (5.7) 0.000 0.000 0.000 3.6 (1.7) 2.13 (1.13) 11.34 (4.17) 3.2 (1.8) 1.9 (1.0) 6.5 (5.9) 0.08 (0.6) 0.08 (0.4) 1.21 (3.2)
0.000 0.000 0.000
0.000 0.000 42.2 (10.2) 46.1 (10.1) 0.000 0.000 41.5 (10.3) 44.8 (10.5) 42.3 (10.3) 46.2 (10.2) 47.3 (8.75) 50.9 (8.16)
0.000 0.000
0.000 0.135 0.000 0.000 0.000 6.4 (3.0) 5.4 (3.0) 5.0 (3.9) 3.5 (4.1) 12.4 (9.8) 0.000 0.020 0.000 0.000 0.000 6.5 (3.1) 5.1 (2.8) 5.6 (4.3) 4.0 (4.5) 13.4 (9.1) 6.1 (3.3) 5.5 (3.1) 4.9 (4.1) 3.3 (4.2) 12.1 (10.0) 4.0 (3.2) 5.8 (3.0) 2.6 (2.9) 1.6 (2.7) 8.9 (12.4)
0.000 0.215 0.000 0.000 0.000
0.000 0.371 0.352 0.55 138 (68.7) 42.9 (14.2) 152 (77.2) 108 (54.5) 0.000 0.493 0.492 0.56 96 (75.6) 43.1 (14.27) 97 (78.2) 67 (53.2) 0.000 0.064 0.55 0.77 257 (65.4) 42.5 (13.8) 290 (75.3) 219 (56.3) 1232 (49.7) 44.0 (14.6) 1803 (74.4) 1365 (53.5)
Rome II (n ⫽ 201) Rome I vs Controls, p Rome I (n ⫽ 127) Manning vs Controls, p Manning (n ⫽ 395)
3179
Sociodemographics Female (n, %) Mean age (yr) Born in Australia (n, %) Year 12 or higher educational level (n, %) Psychological factors Mean neuroticism score (SD) Mean extroversion score (SD) Mean anxiety score (SD) Mean depression score (SD) Mean somatic distress score (SD) Quality of life Mean mental functioning score (SD) Mean physical functioning score (SD) Symptom factors Mean abdominal pain severity score (SD) Mean abdominal pain frequency score (SD) Mean bowel habit score (SD)
Differences Between Rome I and Rome II We repeated these analyses, this time comparing the Rome I (n ⫽ 127) and the new Rome II (n ⫽ 201) criteria. (Table 4). Being female, reporting higher levels of anxiety, depression, and somatic distress, having more impairment in physical functioning and increased bowel habit severity were all significantly associated with the Rome I criteria, the differences, although statistically significant, are small and not clinically relevant. A logistic regression revealed that those fulfilling Rome I criteria experienced greater impairment of their physical functioning (OR ⫽ 0.99, 95% CI: 0.98 –1.00)
Controls (n ⫽ 2512)
Differences Between Rome I and Manning Despite the considerable overlap among the diagnostic categories, we wished to determine whether Rome I and Manning— both representing the extremes in our diagnostic continuum— differed with respect to any sociodemographic, psychological, quality of life, and symptom characteristics. We first examined differences between Rome I (n ⫽ 127) and Manning criteria (n ⫽ 395). Subjects classified according to the Rome I criteria were significantly more likely to be female, to report higher levels of anxiety, depression, and somatic distress, to report greater abdominal pain frequency and severity, and to score higher on bowel habit severity compared with the Manning criteria on univariate analysis (Table 4). Although these differences were statistically significant, they were small and of limited clinical significance. We then entered the same variables into a logistic regression. Only two of the variables remained significant: participants fulfilling Rome I criteria had significantly more severe bowel problems (OR ⫽ 1.16, 95% CI 1.12–1.20), and were less introverted (OR ⫽ 0.95, 95% CI 0.91– 0.99) than those fulfilling Manning criteria. However, this model only explained 11.7% of the deviance.
Table 3. Sociodemographic, Psychological, and Symptomatic Associations With Each Set of IBS Diagnostic Criteria
symptoms (abdominal pain severity and frequency, and bowel habit severity) for the subjects meeting different sets of IBS criteria. These scores were then compared to the scores from the control subjects, who met no diagnostic criteria for IBS. The scores for subjects meeting each criteria set and the control subjects are shown in Table 3. For example, we first compared the 127 subjects diagnosed using the Rome I criteria with the 2512 control subjects. IBS was more common among female subjects, regardless of which criteria were used. The mean age in each diagnostic group was in the mid-40s, and the majority of persons with IBS were Australian-born and had completed high school. Age, ethnicity, and educational level were not significantly associated with a diagnosis of IBS. Having a diagnosis of IBS, regardless of which criteria set was used, was significantly associated with higher levels of neuroticism, anxiety, depression, and somatic distress, and with impaired physical and mental functioning on the SF-12. As expected, abdominal pain severity and frequency and bowel habit severity were also significantly associated with having an IBS diagnosis versus not having this diagnosis.
IBS and Varying Diagnostic Criteria
Rome II vs Controls, p
AJG – October, 2000
3180
Boyce et al.
AJG – Vol. 95, No. 11, 2000
Table 4. Predictors of IBS Diagnostic Group Membership
Sociodemographics Female (n, %) Mean age (SD), yr Born in Australia (n, %) Year 12 or higher educational level (n, %) Psychological factors Mean neuroticism score (SD) Mean extroversion score (SD) Mean anxiety score (SD) Mean depression score (SD) Mean somatic distress score (SD) Quality of life Mean mental functioning score (SD) Mean physical functioning score (SD) Symptom factors Mean abdominal pain severity score (SD) Mean abdominal pain frequency score (SD) Mean bowel habit score (SD)
Rome I vs Manning, Odds Ratio (95% CI)
Rome I vs Rome II, Odds Ratio (95% CI)
Manning only vs Rome criteria, Odds Ratio (95% CI)
0.61 (0.43–0.86) 0.99 (0.99–1.01) 0.91 (0.66–1.24) 1.13 (0.84–1.51)
0.71 (0.52–0.96) 1.00 (0.99–1.01) 0.93 (0.67–1.28) 1.12 (0.87–1.43)
1.32 (0.87–2.00) 1.00 (0.98–1.01) 1.07 (0.70–1.62) 0.97 (0.65–1.45)
1.04 (0.99–1.09) 0.96 (0.92–1.00) 1.04 (1.00–1.07) 1.04 (1.01–1.07) 1.06 (1.03–1.09)
1.01 (0.97–1.05) 0.97 (0.93–1.01) 1.01 (1.01–1.05) 1.02 (1.01–1.02) 1.05 (1.03–1.08)
0.96 (0.90–1.03) 1.02 (0.96–1.08) 0.99 (0.94–1.04) 0.99 (0.94–1.04) 0.97 (0.93–1.01)
0.99 (0.98–1.01) 0.99 (0.97–1.00)
0.99 (0.98–1.01) 0.99 (0.98–1.00)
1.00 (0.98–1.02) 1.00 (0.98–1.02)
1.19 (1.07–1.33) 1.37 (1.02–1.84) 1.16 (1.13–1.19)
1.01 (0.92–1.12) 1.13 (0.88–1.45) 1.14 (1.01–1.17)
0.75 (0.65–0.87) 0.64 (0.43–0.95) 0.90 (0.86–0.94)
and had more severe bowel problems (OR ⫽ 1.13, 95% CI: 1.10 –1.17) than those fulfilling Rome II criteria, accounting for 7.4% of the deviance. Differences Between Manning Only Versus Controls and Rome Criteria There was a considerable number of persons who only met Manning but not Rome criteria for IBS (n ⫽ 185). We compared the persons falling into this category with controls and those persons meeting Rome I and/or Rome II alone (n ⫽ 213). Table 4 shows that persons meeting Manning only criteria were more psychologically distressed (higher levels of neuroticism, anxiety, depression, and somatic distressed) and suffered more mental and physical dysfunction than controls. These differences were modest but did reach statistical significance. They also experienced significantly more severe abdominal pain and bowel problems than did controls. Those persons who met Rome definitions alone for IBS had significantly greater abdominal pain frequency, and more severe abdominal pain and bowel problems, than those persons only meeting Manning criteria. (Table 4). No other significant sociodemographic, psychological, or quality of life differences were found. According to logistic regression, those fulfilling Manning only criteria experienced less severe abdominal pain (OR ⫽ 0.79, 95% CI 0.67– 0.92) and bowel problems (OR ⫽ 0.91, 95% CI 0.87– 0.95) than those fulfilling Rome criteria, accounting for 7.4% of the deviance.
DISCUSSION Recently, the Rome II criteria have been developed by consensus (11), but further work is required to determine which set of criteria are the most valid and applicable. Here
we report for the first time the prevalence rates of IBS ascertained in a community-based study of functional bowel symptoms using three sets of diagnostic criteria. When different thresholds were applied, the prevalence rates differed, with the Rome I criteria giving a prevalence rate of 4.4%, the Rome II criteria a prevalence rate of 6.9%, and the Manning criteria a prevalence of 13.6%. Before reviewing the relevance of these findings, the strengths and limitations of this study will be considered. First, we sent a validated postal questionnaire to a random sample of the population, with a good response rate of ⬎70%. The evaluation of a large random sample in this study and the associated high response rate despite the use of a comprehensive questionnaire lends support for the reliability of our findings. We were unable to determine whether there was any systematic bias in our sample, as we could not compare those who responded to our survey with the nonresponders (all the information we had about them was their address). However, the demographic profile of our sample is very consistent with census data for the area and is likely to be representative. Second, we used a questionnaire to ascertain bowel symptoms. Although such an instrument may not have the same degree of validity as using a physician interview, the questionnaire, which includes a full range of bowel symptoms, has adequate face and concurrent validity (26). This instrument had the advantage of ensuring that participants considered all possible bowel symptoms in a nonthreatening manner, some of which may not have been elicited during the course of a physician interview. Epidemiological studies on the prevalence of IBS have been plagued by the lack of use of well validated assessments of symptoms and a lack of standardized diagnostic criteria for IBS. We found considerable variation in the
AJG – October, 2000
prevalence of IBS when different diagnostic criteria were used. Our data suggests that there is a normal distribution of lower bowel symptoms in the community and that applying various diagnostic criteria shifts the threshold for diagnosis. The diagnostic thresholds vary on two dimensions: duration and severity. Rome I demand persistence of symptoms and high severity, whereas the Manning criteria have a low threshold for severity and no duration criteria. Clearly, such differences will identify different subject groups. The variable prevalence of IBS using the Rome criteria may be due to different interpretations of the duration criteria being applied in the few studies that have been conducted (3, 27). We applied the criteria conservatively; thus, we required symptoms to be chronic or recurrent for at least three consecutive months in the prior 12-month period. Others have not required a similar duration (3). Moreover, some of these studies have not been population-based but have used volunteers (3). The present study has confirmed previous findings that IBS is associated with being female, having increased psychological morbidity and impaired mental and physical functioning (3, 17–20). Our results suggest for the first time that this applies regardless of which diagnostic criteria are used. A key question that emerges from this study is which diagnostic criteria are the most appropriate for identifying IBS. When different thresholds were used, the subjects identified as suffering from IBS, were not the same. Those identified using the rigorous Rome criteria applying the 3-month chronicity threshold had the more severe illness, as judged by independent symptom severity questions. They were also the group that demonstrated the highest levels of psychopathology, with modestly higher scores on depression and anxiety, and higher somatisation scores. Those participants identified as suffering from IBS using the Manning criteria had the least severe illness and demonstrated lower levels of psychopathology. The differences on the psychopathology measures were relatively small but still may influence treatment and etiological studies. We speculate that the chronic and persistent nature of the symptoms associated with the Rome I criteria contributed to the psychological distress identified within this population, as reflected by the severity of the bowel symptoms in this group. An alternative explanation is that the higher levels of psychological distress among this group perpetuate their bowel symptoms, thus leading to the chronicity. This issue can only be resolved through an appropriate longitudinal study, which is currently in progress. An important issue is whether the differences observed really matter. This depends, to a large extent, on the purposes for which the diagnostic criteria are being used. For the clinician diagnostic precision, particularly for irritable bowel syndrome, is a lower priority than excluding the possibility of a structural cause for the symptoms. The treatment for the symptoms would be no different if the patient were diagnosed using the Manning or Rome criteria.
IBS and Varying Diagnostic Criteria
3181
The situation is different in research, especially in clinical trials when diagnostic precision is essential. For clinical trials we need to be confident that the diagnosis is reliable across the recruitment sites (so that we can be confident that the same disorder is being treated). The same is the case for epidemiological and etiological studies where reliable diagnoses need to be made. We would argue that for these purposes the revised Rome II criteria are the most appropriate for two reasons. They identify a core group of persons suffering from irritable bowel syndrome who have the key symptoms associated with a chronic waxing and waning course. Second, in contrast with the Rome I criteria, they can be readily applied. The Rome I criteria are harder to assess because of the requirement of symptoms being present for the past 3 months, which may miss some subjects with irritable bowel syndrome and be potentially more difficult to reliably assess. In summary, we have demonstrated that irritable bowel syndrome, however defined, is a relatively common disorder associated with significant psychological morbidity and impairment in quality of life. Changing the diagnostic threshold only has a small effect on the characteristics of the population defined as suffering from irritable bowel syndrome. Therefore, there seems to be little advantage in applying the new criteria in clinical practice, although we would argue that the more restrictive Rome II criteria are the most relevant criteria to use in such research.
ACKNOWLEDGMENT Supported by a research grant from the National Health and Medical Research Council of Australia. Reprint requests and correspondence: Professor Philip Boyce, Department of Psychological Medicine, University of Sydney, Nepean Hospital, PO Box 63, Penrith NSW 2751, Australia. Received Jan. 26, 2000; accepted May 15, 2000.
REFERENCES 1. Everhart JE, Renault PF. Irritable bowel syndrome in officebased practice in the United States. Gastroenterology 1991; 100:998. 2. Longstreth GF. Irritable bowel syndrome—A multibilliondollar problem. Gastroenterology 1995;109:2029 –31. 3. Drossman DA, Li Z, Andruzzi E, et al. U.S. householder survey of functional gastrointestinal disorders: Prevalence, sociodemography and health impact. Dig Dis Sci 1993;38: 1569 – 80. 4. Jones R, Lydiard S. Irritable bowel syndrome in the general population. Br Med J 1991;304:87–90. 5. Manning AP, Thompson WG, Heaton KW, et al. Towards positive diagnosis of the irritable bowel. Br Med J 1978;2: 653– 4. 6. Talley NJ, Zinsmeister AR, Van Dyke C, et al. Epidemiology of colonic symptoms and the irritable bowel syndrome. Gastroenterology 1991;101:927–34. 7. Talley NJ, Phillips SF, Melton LJ, et al. Diagnostic value of
3182
8. 9. 10. 11. 12. 13. 14.
15.
16. 17.
Boyce et al.
the Manning criteria in irritable bowel syndrome. Gut 1990; 31:77– 81. Smith RC, Greenbaum DS, Vancouver JB, et al. Gender differences in Manning criteria in the irritable bowel syndrome. Gastroenterology 1991;100:591–5. Thompson WG, Dotevall G, Drossman DA. Irritable bowel syndrome: Guidelines for the diagnosis. Gastroenterol Int 1989;2:92–5. Talley NJ, Boyce P, Jones M. Identification of distinct upper and lower gastrointestinal symptom groupings in an urban population. Gut 1998;42:690 –5. Thompson WG, Longstreth GF, Drossman DA, et al. Functional bowel disorders and functional abdominal pain. Gut 1999;45(suppl 2):II43–7. Taub E, Cuevas JL, Cook EW, et al. Irritable bowel syndrome defined by factor analysis. Dig Dis Sci 1995;40:2647–55. Heaton KW, Radvan J, Cripps H, et al. Defecation frequency and timing, and stool form in the general population: A prospective study. Gut 1992;33:818 –24. Drossman DA, McKee DC, Sandler RS, et al. Psychosocial factors in the irritable bowel syndrome. A multivariate study of patients and non-patients with irritable bowel syndrome. Gastroenterology 1988;95:701– 8. Smith RC, Greenbaum DS, Vancouver JB, et al. Psychosocial factors are associated with health care seeking rather than diagnosis in irritable bowel syndrome. Gastroenterology 1990; 28:293–301. Whitehead WE, Bosmajian L, Zonderman AB, et al. Symptoms of psychologic distress associated with irritable bowel syndrome. Gastroenterology 1988;95:709 –14. Hahn BA, Kirchdoerfer LJ, Fullerton S, et al. Patient-perceived severity of irritable bowel syndrome in relation of symptoms, health resource utilization and quality of life. Aliment Pharmacol Ther 1997;11:553–9.
AJG – Vol. 95, No. 11, 2000
18. Whitehead WE, Burnett CK, Cook EW III. Impact of irritable bowel syndrome on quality of life. Dig Dis Sci 1996;41:2248 – 53. 19. O’Keefe E, Talley NJ, Zinsmeister AR, et al. Bowel disorders impair functional status and quality of life in the elderly. A population based study. J Gerentol Med Sci 1995;50A: M184 –9. 20. Jones RH. Clinical economics review—Gastrointestinal disease in primary care. Aliment Pharmcol Ther 1999;10:233–9. 21. Australian Bureau of Statistics. 1991 Census of population and housing, basic community profile: Area 6350 Penrith. ABS catalogue No. 2722.1. 1999. 22. Grayson DA. Latent trait analysis of the Eysenck Personality Questionnaire. J Psychiatry Res 1986;20:217–35. 23. Bedford A, Foulds GA. Validation of the Delusions-Symptoms-States Inventory. Br J Med Psychol 1977;50:163–71. 24. Ware JE, Kosinski M, Keller SD. SF-12: How to score the SF-12 physical and health summary scales. Boston: Health Institute, New England Medical Centre, 1999. 25. Hadzi-Pavlovic D, Hickie I, Ricci C. The SPHERE. Somatic and Psychological Health REport. Development and Initial Evaluation. Technical Report TR-97-002, School of Psychiatry, University of NSW and Academic Department of Psychiatry at St. George Hospital, 1997. 26. Talley NJ, Boyce PM, Owen BK, et al. Initial validation of a bowel symptom questionnaire and measurement of chronic gastrointestinal symptoms in Australians. Aust New Zealand J Med 1995;25:302– 8. 27. Talley NJ, Boyce PM, Jones M. Is the association between irritable bowel syndrome and abuse explained by neuroticism? A population based study. Gut 1998;42:47–53. (see Appendix on facing page)
AJG – October, 2000
IBS and Varying Diagnostic Criteria
3183
Appendix Specific questions from the survey used to diagnose IBS: 1. In the past 12 months, have you ever had any pain or discomfort in your abdomen, stomach, or tummy? (Yes/No) 2. Did this pain or discomfort in your abdomen, stomach, or tummy keep happening over a period of a month or more? (Yes/No) 3. Did this pain or discomfort in your abdomen, stomach, or tummy keep happening over a period of 3 months or more? (Yes/No) 4. During periods when you had this pain or discomfort in your abdomen, stomach, or tummy, how often would you say that:* ● the pain or discomfort was made better by having a bowel motion? ● you had more bowel motions than usual when the pain or discomfort began? ● you had less bowel motions than usual when the pain or discomfort began? ● you had looser bowel motions (stools) than usual when the pain or discomfort began? ● you had harder bowel motions (stools) than usual when the pain or discomfort began? 5. In the past 12 months have you ever had any of the following problems with your bowels? ● you had more than three bowel motions each day? (Yes/No) ● you had less than three bowel motions each day? (Yes/No) ● your bowel motions (stools) were very lumpy or hard? (Yes/No) ● your bowel motions (stools) were very loose or watery? (Yes/No) ● after finishing a bowel movement you felt that there was still bowel motion (stool) that needed to be passed? (Yes/No) ● you experienced an urgent need to have a bowel motion that made you rush to a toilet? (Yes/No) ● you needed to strain a lot to have a bowel motion? (Yes/No) ● you noticed mucus (white green slimy) material in your bowel motions (stools)? (Yes/No) 6. Did any of these problems with your bowels keep happening over a period of a month or more? (Yes/No) 7. Did any of these problems with your bowels keep happening over a period of 3 months or more? (Yes/No) 8. At times when you had problems with your bowels, how often would you say that:* ● you had more than three bowel motions each day? ● you had less than three bowel motions each day? ● your bowel motions (stools) were very lumpy or hard? ● your bowel motions (stools) were very loose or watery? ● after finishing a bowel movement you felt that there was still bowel motion (stool) that needed to be passed? ● you experienced an urgent need to have a bowel motion that made you rush to a toilet? ● you needed to strain a lot to have a bowel motion? ● you noticed mucus (white or green slimy) material in your bowel motions (stools)? 9. In the past 12 months, have you ever had the feeling that your abdomen, stomach, or tummy was swollen or bloated? (Yes/No) 10. Did that feeling that your abdomen, stomach, or tummy was swollen or bloated keep happening over a period of a month or more? (Yes/No) 11. Did that feeling that your abdomen, stomach, or tummy was swollen or bloated keep happening over a period of 3 months or more? (Yes/No) 12. During periods when you had the feeling that your abdomen, stomach, or tummy was swollen or bloated, how often did it occur?* * Options for scoring the answers to these questions were: Not at all. Sometimes (less than one quarter [25%] of the time). Often (more than one quarter [25%] of the time). Very often (more than half [50%] of the time). Almost always (more than three quarters [75%] of the time).
Vol. 95, No. 11, 2000 ISSN 0002-9270/00/$20.00 PII S0002-9270(00)01990-0
Irritable Bowel Syndrome According to Varying Diagnostic Criteria: Are the New Rome II Criteria Unnecessarily Restrictive for Research and Practice? Philip M. Boyce, M.D., F.R.A.N.Z.C.P., Natasha A. Koloski, B.A. (Hons.), and Nicholas J. Talley, M.D., Ph.D., F.A.C.G. Department of Psychological Medicineand Department of Medicine, University of Sydney, Nepean Hospital, Penrith, New South Wales, Australia
OBJECTIVES: It has been suggested that the variation in the prevalence of irritable bowel syndrome (IBS) may be due to the application of different diagnostic criteria. New criteria for IBS have been proposed (Rome II). It is unknown whether persons meeting different criteria for IBS have similar psychological and symptom features. The aim of this study was to measure the prevalence of IBS according to Manning and Rome definitions of IBS and to evaluate the clinical and psychological differences between diagnostic categories. METHODS: A total of 4500 randomly selected subjects, with equal numbers of male and female subjects aged ⱖ18 yr and representative of the Australian population, took part in this study. Subjects were mailed a questionnaire (response rate, 72%). Characteristics measured were gastrointestinal symptoms over the past 12 months, neuroticism and extroversion (Eysenck Personality Questionnaire), anxiety and depression (Delusions-Symptoms-States Inventory), mental and physical functioning (SF-12), and somatic distress (Sphere). RESULTS: The prevalence for IBS according to Manning, Rome I, and Rome II was 4.4% (95% confidence interval [CI] ⫽ 3.5–5.1%), 6.9% (CI 6.0 –7.8%), and 13.6% (CI 12.3–14.8%), respectively. Only 12 persons with Rome I did not also meet Rome II criteria; 196 persons with Manning criteria did not meet Rome II cut-offs. Having IBS regardless of which criteria were used was significantly associated with psychological morbidity, but psychological factors were not important in discriminating between diagnostic categories. However, pain and bowel habit severity independently discriminated between diagnostic groups. CONCLUSIONS: IBS is a relatively common disorder in the community. The new Rome II criteria may be unnecessarily restrictive in practice. (Am J Gastroenterol 2000;95: 3176 –3183. © 2000 by Am. Coll. of Gastroenterology)
INTRODUCTION The irritable bowel syndrome (IBS) is characterized by abdominal pain and disturbance of bowel habits in the absence of demonstrable structural pathology. It is the most
common source of referrals to gastroenterologists (1), and is very costly to the community (2). Community-based studies have reported prevalence rates for IBS ranging from 9% to 22% (3–5). This wide range in the prevalence of IBS may be due to the varying thresholds associated with different diagnostic criteria. The diagnosis of IBS relies upon clinical criteria alone, as there is no gold standard. The diagnosis is dependent upon identifying characteristic symptoms and then distinguishing IBS from structural bowel disorders. Manning et al. (5) first operationalized diagnostic criteria for IBS, by identifying the symptoms that differentiated between IBS and organic bowel disease (with a physician diagnosis being the “gold” standard). Abdominal pain relieved with defecation, looser and/or more frequent stools with the onset of pain and abdominal distension, were the discriminating symptoms and are known as the Manning criteria. Other less discriminating symptoms included; mucus per rectum and a feeling of incomplete rectal evacuation (5). Using these symptoms with abdominal pain criteria, the prevalence of IBS is high, generally around 15–20% (4, 6). Despite some concerns about the sensitivity of Manning criteria for accurately diagnosing IBS in male subjects (7, 8), these criteria have provided the foundation for the more recently accepted diagnostic classification of IBS, the Rome criteria. Subsequently, the Rome I criteria were developed by consensus (9). The prevalence rate for IBS, using the Rome I criteria (3), is lower, generally around 9 –12% (10, 11). The Rome criteria have recently been modified by relaxing the duration threshold and removing the duplication of stool frequency and consistency in Part 1 of the criteria (11). In 1999, the Rome II criteria were published, in part based on factor analytic studies (12); the revised Rome criteria (Rome II) include only the first part of the original criteria, but require the presence of two out of three symptoms relating abdominal pain to bowel symptoms (11). In research and clinical practice, the key issue is what defines a “case” of IBS as opposed to the normal variation in bowel symptoms. The difference between “cases” and noncases is likely to be dependent upon the diagnostic thresholds that are used, although this has not been carefully
AJG – October, 2000
studied. These thresholds can be set across a number of different parameters. First, the severity of the bowel symptoms (or the degree of impairment or disability) can be set as the threshold to distinguish between cases and noncases. Second, the threshold could be based purely on a symptom count. This may be done in two ways; all symptoms being given equal weight (a polythetic approach) or alternatively with a particular symptom or symptoms being regarded as essential (a monothetic approach) to the diagnosis and other symptoms being regarded as accessory symptoms. Third, a duration threshold can be used, in combination with a symptom count. Finally, cases could be distinguished from noncases on a combination of the parameters. The Manning criteria thus apply a simple polythetic approach, in that the diagnosis is dependent upon two or more symptoms being present with all symptoms having equal weight (5). The Rome I criteria use a mixed approach, with a strict duration threshold and a combination of essential and accessory symptoms. The new Rome II criteria relaxed the duration threshold, to allow for the waxing and waning of symptoms over a long period (11). None of the diagnostic systems has applied a severity dimension, or impairment or disability criteria, despite the fact that the severity of bowel symptoms can be objectively assessed (13). Separate studies using Rome I and Manning definitions have found that persons with IBS are more likely to be female (3), have a lower socioeconomic status (3), greater psychological morbidity including higher levels of anxiety, depression, and somatization disorders (14 –16) and experience a reduced quality of life (17–20) compared with persons without IBS or those who have IBS symptoms in the community but who do not seek health care for them. There are no data, however, on whether persons meeting these various diagnostic versions of IBS are similar in terms of sociodemographic, psychological, and symptom characteristics. We postulated that persons fulfilling a more restrictive threshold for IBS such as the Rome I criteria would exhibit greater psychological disturbances and more impaired quality of life because of the burden of coping with more symptoms of greater duration, compared with those persons meeting the more liberal Manning criteria. If true, then this would have important implications for clinical trials and criteria development in regard to determining the optimal symptom cut-off points for a diagnosis of IBS, as well as diagnosis in clinical practice. Therefore we aimed, in a population-based cross-sectional study, to 1) measure the prevalence of IBS according to the Manning criteria and for the first time variations of the Rome criteria and 2) evaluate the clinical and psychological differences, if any, between key diagnostic categories.
MATERIALS AND METHODS Subjects A random sample of 4500 persons aged ⱖ18 yr, with equal numbers of male and female subjects, were selected from
IBS and Varying Diagnostic Criteria
3177
the 1996 electoral roll for the local government area of Penrith (3.6% of the Sydney’s population). Penrith is homogeneous in terms of its sociodemographics and ethnic composition with the Australian population, except that its inhabitants are slightly younger and it has a slightly higher socioeconomic status (1991 census) (21). Of the original 4500 subjects mailed the questionnaire, 398 subjects had moved or were overseas, 11 subjects had died, 24 subjects were seriously ill, 75 subjects explicitly refused to participate and 1082 subjects failed to respond. A completed questionnaire was returned by 2910 subjects, giving an overall response rate of 72%. There were slightly more female responders (51.9%) to the questionnaire compared with nonresponders (47.0%). Procedure The mail out of a questionnaire to subjects was carried out in three batches of 500, 2000, and 2000. In order to ensure that non–English-speaking persons were not excluded from the study, an information sheet, printed in the common non-English languages used by the residents of Lindsay, was included with the questionnaire. Non–English-speaking persons, if they were interested in participating, were invited to contact the research team so that an interpreter could be organized for them. Nonrespondents were extensively followed-up over a period of 10 wk, with a replacement survey sent out at wk 4 and reminder letters sent out in wk 2, 5, 7, and 10. A telephone call was also made to subjects in wk 9. Those subjects in batch 1 received an additional replacement survey at wk 8 and a telephone call instead of a reminder letter was carried out in wk 7. Questionnaire The complete version of the 32-page questionnaire contained 216 items divided into sections, the first containing questions on a range of gastrointestinal symptoms relating to the following specific anatomic regions: esophagus, gastroduodenum, bowel, abdomen, biliary, and anorectum. The criteria for IBS in the different diagnostic systems are shown in Table 1. Also in the early sections of the questionnaire, we asked about the number of times health care was sought over the past 12 months for gastrointestinal symptoms. The latter sections of the questionnaire consisted of measures of physical and mental functioning, psychological distress, and personality. A shortened version of the Eysenck Personality Questionnaire (EPQ) (22) was used to assess the personality traits of neuroticism and extroversion. Psychological distress was measured using the DelusionsSymptoms-States Inventory (DSSI) (23) a measure of anxiety and depression. Physical and mental functioning was assessed using the SF-12 (24) and the Sphere (25) provided scores for the subscales of anxiety/depression, fatigue/pain, and somatic distress. In addition, we included items to assess the sociodemographic characteristics of our sample. These included questions relating to age, sex, country of birth, and educational level.
3178
Boyce et al.
Table 1. Diagnostic Criteria for the Irritable Bowel Syndrome According to Different Systems Manning criteria Two or more of the following symptoms: Pain eased after bowel movement Looser stools at onset of pain More frequent bowel movements at onset of pain Abdominal distension Mucus per rectum Feeling of incomplete emptying Rome I criteria At least 3 months of continuous or recurrent symptoms of: Abdominal pain or discomfort that is Relieved with defecation; and/or Associated with a change in frequency of stool; and/or Associated with a change in consistency of stool; and Two or more of the following, at least on one-fourth of occasions or days: Altered stool frequency (for research purposes “altered” may be defined as more than three bowel movements each day or less than three bowel movements each week); Altered stool form (lumpy/hard or loose/watery stool); Altered stool passage (straining, urgency, or feeling of incomplete evacuation); Passage of mucus; and/or Bloating or feeling of abdominal distension. Rome II criteria At least 12 weeks (which need not be consecutive) in the preceding 12 months, of abdominal discomfort or pain that has two of three features: Relieved with defecation; and/or Onset associated with a change in frequency of stool; and/or Onset associated with a change in form (appearance) of stool.
Diagnostic Criteria Computer algorithms were written to generate Manning, Rome I, and Rome II prevalence of IBS. The questions used to elicit the symptoms of IBS according to the different diagnostic systems (the same questions were used for each system) are reproduced in Appendix 1. Bowel habit severity was based on the mean score of the frequency (from 0 ⫽ “Not at all” to 5 ⫽ “Very often”) of occurrence of nine bowel symptoms over 3 months in the past year. Statistical Analyses The prevalence (%) for all the diagnostic versions of IBS were calculated with the corresponding 95% confidence intervals. The sociodemographic, psychological, symptom, and health care–seeking associations were analyzed in two ways. First, all the variables were analyzed for associations with each diagnostic category using t tests for dimensional scores and 2 tests for categorical variables. Comparisons were made with a “control” group of subjects who did not meet a diagnosis of IBS with any of the criteria but could have suffered from other functional gastrointestinal disorders. Second, we compared Rome I versus Manning, and Rome I versus Rome II criteria, on a range of variables using a logistic regression with the adjusted odds ratios and 95% confidence limits reported. The possibility of nonindependent data because of multiple observations on each subject
AJG – Vol. 95, No. 11, 2000
Table 2. Overlap of Subjects Identified by Applying Different Diagnostic Criteria
Diagnostic Criteria Met None at all Manning criteria Manning alone Manning & Rome I Manning & Rome II Manning & Rome I & Rome II Rome I criteria Rome I alone Rome I & Rome II Rome I & Manning Rome I & Rome II & Manning Rome II criteria Rome II alone Rome II & Rome I Rome II & Manning Rome II & Rome I & Manning
% of Subjects Meeting Index No. of Criteria and Subjects Alternative Criteria 2512 395 185 126 199 115 127 1 115 126 115 201 2 115 199 115
46.8 31.9 50.4 29.1 0.8 90.6 99.2 90.6 1 57.2 99 57.2
may have led to within-group correlation in residuals and consequent incorrect statistical inference (incorrect standard errors). Standard errors were adjusted by viewing the data as though subjects were cluster sampled and taking into account the calculation of the design effect. The degree of adjustment is proportional to the intrasubject correlation. All p values calculated were two tailed; the ␣ level of significance was set at 0.05.
RESULTS The mean age of the sample was 43.8 yr (95% CI: 43.2– 44.3); 51.3% were female (95% CI: 49.0 –53.5%). A total of 398 (13.7%, 95% CI 12.3–15.2%) of the participants met criteria for IBS on at least one of the diagnostic systems. Prevalence of IBS A total of 127 participants met Rome I criteria (4.4%, 95% CI: 3.6 –5.2%) over the previous year. By contrast, 395 participants fulfilled the Manning criteria (13.6%, 95% CI: 12.2–15.1%). The prevalence using the revised Rome criteria, Rome II, was 6.9% (95% CI: 5.9 – 8.0%). Overlap of IBS Diagnostic Categories There was considerable overlap applying the different definitions, as shown in Table 2. For example, 127 subjects met criteria for Rome I diagnosis and, of them, 115 (90.6%) also met the Rome II criteria. On the other hand, 46.8% of the subjects who met the Manning criteria did not meet either of the Rome criteria. Sociodemographic, Psychological, and Symptom Associations With IBS Diagnostic Criteria We then examined the mean scores on the psychological variables (neuroticism, anxiety, depression, and somatic distress), the SF-12 (mental and physical functioning) and
0.000 0.000 0.000 3.6 (1.6) 2.0 (1.1) 8.0 (5.7) 0.000 0.000 0.000 3.6 (1.7) 2.13 (1.13) 11.34 (4.17) 3.2 (1.8) 1.9 (1.0) 6.5 (5.9) 0.08 (0.6) 0.08 (0.4) 1.21 (3.2)
0.000 0.000 0.000
0.000 0.000 42.2 (10.2) 46.1 (10.1) 0.000 0.000 41.5 (10.3) 44.8 (10.5) 42.3 (10.3) 46.2 (10.2) 47.3 (8.75) 50.9 (8.16)
0.000 0.000
0.000 0.135 0.000 0.000 0.000 6.4 (3.0) 5.4 (3.0) 5.0 (3.9) 3.5 (4.1) 12.4 (9.8) 0.000 0.020 0.000 0.000 0.000 6.5 (3.1) 5.1 (2.8) 5.6 (4.3) 4.0 (4.5) 13.4 (9.1) 6.1 (3.3) 5.5 (3.1) 4.9 (4.1) 3.3 (4.2) 12.1 (10.0) 4.0 (3.2) 5.8 (3.0) 2.6 (2.9) 1.6 (2.7) 8.9 (12.4)
0.000 0.215 0.000 0.000 0.000
0.000 0.371 0.352 0.55 138 (68.7) 42.9 (14.2) 152 (77.2) 108 (54.5) 0.000 0.493 0.492 0.56 96 (75.6) 43.1 (14.27) 97 (78.2) 67 (53.2) 0.000 0.064 0.55 0.77 257 (65.4) 42.5 (13.8) 290 (75.3) 219 (56.3) 1232 (49.7) 44.0 (14.6) 1803 (74.4) 1365 (53.5)
Rome II (n ⫽ 201) Rome I vs Controls, p Rome I (n ⫽ 127) Manning vs Controls, p Manning (n ⫽ 395)
3179
Sociodemographics Female (n, %) Mean age (yr) Born in Australia (n, %) Year 12 or higher educational level (n, %) Psychological factors Mean neuroticism score (SD) Mean extroversion score (SD) Mean anxiety score (SD) Mean depression score (SD) Mean somatic distress score (SD) Quality of life Mean mental functioning score (SD) Mean physical functioning score (SD) Symptom factors Mean abdominal pain severity score (SD) Mean abdominal pain frequency score (SD) Mean bowel habit score (SD)
Differences Between Rome I and Rome II We repeated these analyses, this time comparing the Rome I (n ⫽ 127) and the new Rome II (n ⫽ 201) criteria. (Table 4). Being female, reporting higher levels of anxiety, depression, and somatic distress, having more impairment in physical functioning and increased bowel habit severity were all significantly associated with the Rome I criteria, the differences, although statistically significant, are small and not clinically relevant. A logistic regression revealed that those fulfilling Rome I criteria experienced greater impairment of their physical functioning (OR ⫽ 0.99, 95% CI: 0.98 –1.00)
Controls (n ⫽ 2512)
Differences Between Rome I and Manning Despite the considerable overlap among the diagnostic categories, we wished to determine whether Rome I and Manning— both representing the extremes in our diagnostic continuum— differed with respect to any sociodemographic, psychological, quality of life, and symptom characteristics. We first examined differences between Rome I (n ⫽ 127) and Manning criteria (n ⫽ 395). Subjects classified according to the Rome I criteria were significantly more likely to be female, to report higher levels of anxiety, depression, and somatic distress, to report greater abdominal pain frequency and severity, and to score higher on bowel habit severity compared with the Manning criteria on univariate analysis (Table 4). Although these differences were statistically significant, they were small and of limited clinical significance. We then entered the same variables into a logistic regression. Only two of the variables remained significant: participants fulfilling Rome I criteria had significantly more severe bowel problems (OR ⫽ 1.16, 95% CI 1.12–1.20), and were less introverted (OR ⫽ 0.95, 95% CI 0.91– 0.99) than those fulfilling Manning criteria. However, this model only explained 11.7% of the deviance.
Table 3. Sociodemographic, Psychological, and Symptomatic Associations With Each Set of IBS Diagnostic Criteria
symptoms (abdominal pain severity and frequency, and bowel habit severity) for the subjects meeting different sets of IBS criteria. These scores were then compared to the scores from the control subjects, who met no diagnostic criteria for IBS. The scores for subjects meeting each criteria set and the control subjects are shown in Table 3. For example, we first compared the 127 subjects diagnosed using the Rome I criteria with the 2512 control subjects. IBS was more common among female subjects, regardless of which criteria were used. The mean age in each diagnostic group was in the mid-40s, and the majority of persons with IBS were Australian-born and had completed high school. Age, ethnicity, and educational level were not significantly associated with a diagnosis of IBS. Having a diagnosis of IBS, regardless of which criteria set was used, was significantly associated with higher levels of neuroticism, anxiety, depression, and somatic distress, and with impaired physical and mental functioning on the SF-12. As expected, abdominal pain severity and frequency and bowel habit severity were also significantly associated with having an IBS diagnosis versus not having this diagnosis.
IBS and Varying Diagnostic Criteria
Rome II vs Controls, p
AJG – October, 2000
3180
Boyce et al.
AJG – Vol. 95, No. 11, 2000
Table 4. Predictors of IBS Diagnostic Group Membership
Sociodemographics Female (n, %) Mean age (SD), yr Born in Australia (n, %) Year 12 or higher educational level (n, %) Psychological factors Mean neuroticism score (SD) Mean extroversion score (SD) Mean anxiety score (SD) Mean depression score (SD) Mean somatic distress score (SD) Quality of life Mean mental functioning score (SD) Mean physical functioning score (SD) Symptom factors Mean abdominal pain severity score (SD) Mean abdominal pain frequency score (SD) Mean bowel habit score (SD)
Rome I vs Manning, Odds Ratio (95% CI)
Rome I vs Rome II, Odds Ratio (95% CI)
Manning only vs Rome criteria, Odds Ratio (95% CI)
0.61 (0.43–0.86) 0.99 (0.99–1.01) 0.91 (0.66–1.24) 1.13 (0.84–1.51)
0.71 (0.52–0.96) 1.00 (0.99–1.01) 0.93 (0.67–1.28) 1.12 (0.87–1.43)
1.32 (0.87–2.00) 1.00 (0.98–1.01) 1.07 (0.70–1.62) 0.97 (0.65–1.45)
1.04 (0.99–1.09) 0.96 (0.92–1.00) 1.04 (1.00–1.07) 1.04 (1.01–1.07) 1.06 (1.03–1.09)
1.01 (0.97–1.05) 0.97 (0.93–1.01) 1.01 (1.01–1.05) 1.02 (1.01–1.02) 1.05 (1.03–1.08)
0.96 (0.90–1.03) 1.02 (0.96–1.08) 0.99 (0.94–1.04) 0.99 (0.94–1.04) 0.97 (0.93–1.01)
0.99 (0.98–1.01) 0.99 (0.97–1.00)
0.99 (0.98–1.01) 0.99 (0.98–1.00)
1.00 (0.98–1.02) 1.00 (0.98–1.02)
1.19 (1.07–1.33) 1.37 (1.02–1.84) 1.16 (1.13–1.19)
1.01 (0.92–1.12) 1.13 (0.88–1.45) 1.14 (1.01–1.17)
0.75 (0.65–0.87) 0.64 (0.43–0.95) 0.90 (0.86–0.94)
and had more severe bowel problems (OR ⫽ 1.13, 95% CI: 1.10 –1.17) than those fulfilling Rome II criteria, accounting for 7.4% of the deviance. Differences Between Manning Only Versus Controls and Rome Criteria There was a considerable number of persons who only met Manning but not Rome criteria for IBS (n ⫽ 185). We compared the persons falling into this category with controls and those persons meeting Rome I and/or Rome II alone (n ⫽ 213). Table 4 shows that persons meeting Manning only criteria were more psychologically distressed (higher levels of neuroticism, anxiety, depression, and somatic distressed) and suffered more mental and physical dysfunction than controls. These differences were modest but did reach statistical significance. They also experienced significantly more severe abdominal pain and bowel problems than did controls. Those persons who met Rome definitions alone for IBS had significantly greater abdominal pain frequency, and more severe abdominal pain and bowel problems, than those persons only meeting Manning criteria. (Table 4). No other significant sociodemographic, psychological, or quality of life differences were found. According to logistic regression, those fulfilling Manning only criteria experienced less severe abdominal pain (OR ⫽ 0.79, 95% CI 0.67– 0.92) and bowel problems (OR ⫽ 0.91, 95% CI 0.87– 0.95) than those fulfilling Rome criteria, accounting for 7.4% of the deviance.
DISCUSSION Recently, the Rome II criteria have been developed by consensus (11), but further work is required to determine which set of criteria are the most valid and applicable. Here
we report for the first time the prevalence rates of IBS ascertained in a community-based study of functional bowel symptoms using three sets of diagnostic criteria. When different thresholds were applied, the prevalence rates differed, with the Rome I criteria giving a prevalence rate of 4.4%, the Rome II criteria a prevalence rate of 6.9%, and the Manning criteria a prevalence of 13.6%. Before reviewing the relevance of these findings, the strengths and limitations of this study will be considered. First, we sent a validated postal questionnaire to a random sample of the population, with a good response rate of ⬎70%. The evaluation of a large random sample in this study and the associated high response rate despite the use of a comprehensive questionnaire lends support for the reliability of our findings. We were unable to determine whether there was any systematic bias in our sample, as we could not compare those who responded to our survey with the nonresponders (all the information we had about them was their address). However, the demographic profile of our sample is very consistent with census data for the area and is likely to be representative. Second, we used a questionnaire to ascertain bowel symptoms. Although such an instrument may not have the same degree of validity as using a physician interview, the questionnaire, which includes a full range of bowel symptoms, has adequate face and concurrent validity (26). This instrument had the advantage of ensuring that participants considered all possible bowel symptoms in a nonthreatening manner, some of which may not have been elicited during the course of a physician interview. Epidemiological studies on the prevalence of IBS have been plagued by the lack of use of well validated assessments of symptoms and a lack of standardized diagnostic criteria for IBS. We found considerable variation in the
AJG – October, 2000
prevalence of IBS when different diagnostic criteria were used. Our data suggests that there is a normal distribution of lower bowel symptoms in the community and that applying various diagnostic criteria shifts the threshold for diagnosis. The diagnostic thresholds vary on two dimensions: duration and severity. Rome I demand persistence of symptoms and high severity, whereas the Manning criteria have a low threshold for severity and no duration criteria. Clearly, such differences will identify different subject groups. The variable prevalence of IBS using the Rome criteria may be due to different interpretations of the duration criteria being applied in the few studies that have been conducted (3, 27). We applied the criteria conservatively; thus, we required symptoms to be chronic or recurrent for at least three consecutive months in the prior 12-month period. Others have not required a similar duration (3). Moreover, some of these studies have not been population-based but have used volunteers (3). The present study has confirmed previous findings that IBS is associated with being female, having increased psychological morbidity and impaired mental and physical functioning (3, 17–20). Our results suggest for the first time that this applies regardless of which diagnostic criteria are used. A key question that emerges from this study is which diagnostic criteria are the most appropriate for identifying IBS. When different thresholds were used, the subjects identified as suffering from IBS, were not the same. Those identified using the rigorous Rome criteria applying the 3-month chronicity threshold had the more severe illness, as judged by independent symptom severity questions. They were also the group that demonstrated the highest levels of psychopathology, with modestly higher scores on depression and anxiety, and higher somatisation scores. Those participants identified as suffering from IBS using the Manning criteria had the least severe illness and demonstrated lower levels of psychopathology. The differences on the psychopathology measures were relatively small but still may influence treatment and etiological studies. We speculate that the chronic and persistent nature of the symptoms associated with the Rome I criteria contributed to the psychological distress identified within this population, as reflected by the severity of the bowel symptoms in this group. An alternative explanation is that the higher levels of psychological distress among this group perpetuate their bowel symptoms, thus leading to the chronicity. This issue can only be resolved through an appropriate longitudinal study, which is currently in progress. An important issue is whether the differences observed really matter. This depends, to a large extent, on the purposes for which the diagnostic criteria are being used. For the clinician diagnostic precision, particularly for irritable bowel syndrome, is a lower priority than excluding the possibility of a structural cause for the symptoms. The treatment for the symptoms would be no different if the patient were diagnosed using the Manning or Rome criteria.
IBS and Varying Diagnostic Criteria
3181
The situation is different in research, especially in clinical trials when diagnostic precision is essential. For clinical trials we need to be confident that the diagnosis is reliable across the recruitment sites (so that we can be confident that the same disorder is being treated). The same is the case for epidemiological and etiological studies where reliable diagnoses need to be made. We would argue that for these purposes the revised Rome II criteria are the most appropriate for two reasons. They identify a core group of persons suffering from irritable bowel syndrome who have the key symptoms associated with a chronic waxing and waning course. Second, in contrast with the Rome I criteria, they can be readily applied. The Rome I criteria are harder to assess because of the requirement of symptoms being present for the past 3 months, which may miss some subjects with irritable bowel syndrome and be potentially more difficult to reliably assess. In summary, we have demonstrated that irritable bowel syndrome, however defined, is a relatively common disorder associated with significant psychological morbidity and impairment in quality of life. Changing the diagnostic threshold only has a small effect on the characteristics of the population defined as suffering from irritable bowel syndrome. Therefore, there seems to be little advantage in applying the new criteria in clinical practice, although we would argue that the more restrictive Rome II criteria are the most relevant criteria to use in such research.
ACKNOWLEDGMENT Supported by a research grant from the National Health and Medical Research Council of Australia. Reprint requests and correspondence: Professor Philip Boyce, Department of Psychological Medicine, University of Sydney, Nepean Hospital, PO Box 63, Penrith NSW 2751, Australia. Received Jan. 26, 2000; accepted May 15, 2000.
REFERENCES 1. Everhart JE, Renault PF. Irritable bowel syndrome in officebased practice in the United States. Gastroenterology 1991; 100:998. 2. Longstreth GF. Irritable bowel syndrome—A multibilliondollar problem. Gastroenterology 1995;109:2029 –31. 3. Drossman DA, Li Z, Andruzzi E, et al. U.S. householder survey of functional gastrointestinal disorders: Prevalence, sociodemography and health impact. Dig Dis Sci 1993;38: 1569 – 80. 4. Jones R, Lydiard S. Irritable bowel syndrome in the general population. Br Med J 1991;304:87–90. 5. Manning AP, Thompson WG, Heaton KW, et al. Towards positive diagnosis of the irritable bowel. Br Med J 1978;2: 653– 4. 6. Talley NJ, Zinsmeister AR, Van Dyke C, et al. Epidemiology of colonic symptoms and the irritable bowel syndrome. Gastroenterology 1991;101:927–34. 7. Talley NJ, Phillips SF, Melton LJ, et al. Diagnostic value of
3182
8. 9. 10. 11. 12. 13. 14.
15.
16. 17.
Boyce et al.
the Manning criteria in irritable bowel syndrome. Gut 1990; 31:77– 81. Smith RC, Greenbaum DS, Vancouver JB, et al. Gender differences in Manning criteria in the irritable bowel syndrome. Gastroenterology 1991;100:591–5. Thompson WG, Dotevall G, Drossman DA. Irritable bowel syndrome: Guidelines for the diagnosis. Gastroenterol Int 1989;2:92–5. Talley NJ, Boyce P, Jones M. Identification of distinct upper and lower gastrointestinal symptom groupings in an urban population. Gut 1998;42:690 –5. Thompson WG, Longstreth GF, Drossman DA, et al. Functional bowel disorders and functional abdominal pain. Gut 1999;45(suppl 2):II43–7. Taub E, Cuevas JL, Cook EW, et al. Irritable bowel syndrome defined by factor analysis. Dig Dis Sci 1995;40:2647–55. Heaton KW, Radvan J, Cripps H, et al. Defecation frequency and timing, and stool form in the general population: A prospective study. Gut 1992;33:818 –24. Drossman DA, McKee DC, Sandler RS, et al. Psychosocial factors in the irritable bowel syndrome. A multivariate study of patients and non-patients with irritable bowel syndrome. Gastroenterology 1988;95:701– 8. Smith RC, Greenbaum DS, Vancouver JB, et al. Psychosocial factors are associated with health care seeking rather than diagnosis in irritable bowel syndrome. Gastroenterology 1990; 28:293–301. Whitehead WE, Bosmajian L, Zonderman AB, et al. Symptoms of psychologic distress associated with irritable bowel syndrome. Gastroenterology 1988;95:709 –14. Hahn BA, Kirchdoerfer LJ, Fullerton S, et al. Patient-perceived severity of irritable bowel syndrome in relation of symptoms, health resource utilization and quality of life. Aliment Pharmacol Ther 1997;11:553–9.
AJG – Vol. 95, No. 11, 2000
18. Whitehead WE, Burnett CK, Cook EW III. Impact of irritable bowel syndrome on quality of life. Dig Dis Sci 1996;41:2248 – 53. 19. O’Keefe E, Talley NJ, Zinsmeister AR, et al. Bowel disorders impair functional status and quality of life in the elderly. A population based study. J Gerentol Med Sci 1995;50A: M184 –9. 20. Jones RH. Clinical economics review—Gastrointestinal disease in primary care. Aliment Pharmcol Ther 1999;10:233–9. 21. Australian Bureau of Statistics. 1991 Census of population and housing, basic community profile: Area 6350 Penrith. ABS catalogue No. 2722.1. 1999. 22. Grayson DA. Latent trait analysis of the Eysenck Personality Questionnaire. J Psychiatry Res 1986;20:217–35. 23. Bedford A, Foulds GA. Validation of the Delusions-Symptoms-States Inventory. Br J Med Psychol 1977;50:163–71. 24. Ware JE, Kosinski M, Keller SD. SF-12: How to score the SF-12 physical and health summary scales. Boston: Health Institute, New England Medical Centre, 1999. 25. Hadzi-Pavlovic D, Hickie I, Ricci C. The SPHERE. Somatic and Psychological Health REport. Development and Initial Evaluation. Technical Report TR-97-002, School of Psychiatry, University of NSW and Academic Department of Psychiatry at St. George Hospital, 1997. 26. Talley NJ, Boyce PM, Owen BK, et al. Initial validation of a bowel symptom questionnaire and measurement of chronic gastrointestinal symptoms in Australians. Aust New Zealand J Med 1995;25:302– 8. 27. Talley NJ, Boyce PM, Jones M. Is the association between irritable bowel syndrome and abuse explained by neuroticism? A population based study. Gut 1998;42:47–53. (see Appendix on facing page)
AJG – October, 2000
IBS and Varying Diagnostic Criteria
3183
Appendix Specific questions from the survey used to diagnose IBS: 1. In the past 12 months, have you ever had any pain or discomfort in your abdomen, stomach, or tummy? (Yes/No) 2. Did this pain or discomfort in your abdomen, stomach, or tummy keep happening over a period of a month or more? (Yes/No) 3. Did this pain or discomfort in your abdomen, stomach, or tummy keep happening over a period of 3 months or more? (Yes/No) 4. During periods when you had this pain or discomfort in your abdomen, stomach, or tummy, how often would you say that:* ● the pain or discomfort was made better by having a bowel motion? ● you had more bowel motions than usual when the pain or discomfort began? ● you had less bowel motions than usual when the pain or discomfort began? ● you had looser bowel motions (stools) than usual when the pain or discomfort began? ● you had harder bowel motions (stools) than usual when the pain or discomfort began? 5. In the past 12 months have you ever had any of the following problems with your bowels? ● you had more than three bowel motions each day? (Yes/No) ● you had less than three bowel motions each day? (Yes/No) ● your bowel motions (stools) were very lumpy or hard? (Yes/No) ● your bowel motions (stools) were very loose or watery? (Yes/No) ● after finishing a bowel movement you felt that there was still bowel motion (stool) that needed to be passed? (Yes/No) ● you experienced an urgent need to have a bowel motion that made you rush to a toilet? (Yes/No) ● you needed to strain a lot to have a bowel motion? (Yes/No) ● you noticed mucus (white green slimy) material in your bowel motions (stools)? (Yes/No) 6. Did any of these problems with your bowels keep happening over a period of a month or more? (Yes/No) 7. Did any of these problems with your bowels keep happening over a period of 3 months or more? (Yes/No) 8. At times when you had problems with your bowels, how often would you say that:* ● you had more than three bowel motions each day? ● you had less than three bowel motions each day? ● your bowel motions (stools) were very lumpy or hard? ● your bowel motions (stools) were very loose or watery? ● after finishing a bowel movement you felt that there was still bowel motion (stool) that needed to be passed? ● you experienced an urgent need to have a bowel motion that made you rush to a toilet? ● you needed to strain a lot to have a bowel motion? ● you noticed mucus (white or green slimy) material in your bowel motions (stools)? 9. In the past 12 months, have you ever had the feeling that your abdomen, stomach, or tummy was swollen or bloated? (Yes/No) 10. Did that feeling that your abdomen, stomach, or tummy was swollen or bloated keep happening over a period of a month or more? (Yes/No) 11. Did that feeling that your abdomen, stomach, or tummy was swollen or bloated keep happening over a period of 3 months or more? (Yes/No) 12. During periods when you had the feeling that your abdomen, stomach, or tummy was swollen or bloated, how often did it occur?* * Options for scoring the answers to these questions were: Not at all. Sometimes (less than one quarter [25%] of the time). Often (more than one quarter [25%] of the time). Very often (more than half [50%] of the time). Almost always (more than three quarters [75%] of the time).