Williams, Rachel E. PhD*; Hartmann, Katherine E. MD, PhD*†; Sandler, Robert S. MD, MPH*‡; Miller, William C. MD, MPH*‡; Steege, John F. MD†
Irritable bowel syndrome (IBS) is defined by consensus-established criteria that include abdominal pain and altered stool characteristics. Depending on the classification criteria used, this syndrome affects 10% to 20% of the general population in developed countries1–3 and is the most common condition diagnosed by gastroenterologists, accounting for 10.4% of all digestive system diagnoses.4 Characteristics associated with IBS include age,3,5 female gender,3,6 depression and anxiety,4,6 history of abuse,7 increased physician visits and health care expenditures,6 reduced quality of life,8 endometriosis, and dyspareunia and dysmenorrhea.9,10
Chronic pelvic pain is defined by an extended duration of pain in the pelvis. The pain may originate from any organ system or pathology and may have multiple contributing factors.11 It affects 12% to 39% of women, with a higher prevalence found in studies using clinic populations than in studies of the general population.11,12 Women with chronic pelvic pain use 3 times more medications, have 4 times more nongynecologic operations, are 5 times more likely to have a hysterectomy,13 and have reduced quality of life14,15 compared with women without this pain.
In studies of the general population that included both health care seekers and non–health care seekers, IBS was reported in 39% of women with chronic pelvic pain10 and in 48% of women having diagnostic laparoscopy for chronic pelvic pain.16 Irritable bowel syndrome has been associated with painful menstruation, pain during intercourse, endometriosis, and history of pelvic inflammatory disease in women with chronic pelvic pain.10 The objective of our study was to evaluate whether there are unique characteristics associated with IBS within a population that has chronic pelvic pain.
MATERIALS AND METHODS
The population of this cross-sectional study was comprised of new patients who entered the Pelvic Pain Clinic at the University of North Carolina (UNC) between June 10, 1993, and December 11, 2000 (N = 987). Typically, women are referred to the Pelvic Pain Clinic at UNC by their gynecologists or primary care physicians within UNC Hospitals or from elsewhere in North Carolina. A few patients come from surrounding states or other countries. Most referrals occur because previous treatment failed, additional treatment was needed, or complex surgery was anticipated. Data collection was limited to women with self-reported pelvic pain for 6 months or longer, although it is possible for a patient to be referred for pelvic pain lasting less than 6 months. The Institutional Review Board approved use of these clinical data for research purposes.
The clinic personnel collected data systematically to aid in patient evaluation and treatment. Patients completed self-administered questionnaires before the initial visit to the clinic, including the Beck Depression Inventory,17 Symptom Checklist-90 (SCL-90),18 McGill Pain Questionnaire,19 a clinic-specific general information form, and life experiences survey. Physicians recorded diagnostic impressions and physical findings at the initial visit. The SCL-90 and life experiences survey, which includes mental health and history of abuse, were not administered after 1998.
Irritable bowel syndrome was classified by the Rome I criteria. In our study, the Rome I symptom questions were part of the general information questionnaire that was self-administered before the initial visit. To be classified as having IBS, the patient must have at least 3 months of continuous or recurrent symptoms of 1) abdominal pain or discomfort that is relieved with defecation or associated with a change in frequency of stool or associated with a change in consistency of stool or all three and 2) 2 or more of the following, at least one fourth of occasions or days: altered stool frequency (more than 3 bowel movements each day or less than 3 bowel movements each week), altered stool form (lumpy and hard or loose and watery stool), altered stool passage (straining, urgency, or feeling of incomplete evacuation), passage of mucus, and bloating or feeling of abdominal distension.6
At the initial visit, one of the small number of clinic physicians who specialize in chronic pelvic pain performed a complete medical history and physical examination. Prior pathology reports were generally available. The physician recorded clinical diagnostic impressions based on his or her best judgment. These diagnoses were grouped as “yes” if the physician indicated it was definite or probable that it existed in the patient and grouped as “no” if they indicated it was possible or did not apply to the patient. The locations of recorded pain included left side of pelvic area, right side of pelvic area, middle area of pelvis, low back pain, around opening of vagina, deep inside the vagina, high in the pelvis near the navel, and in the thighs. These locations were grouped to make “total number of pain sites,” because many women had more than one location of pain.
The characteristics that we evaluated included demographic characteristics, clinical diagnoses, history of abuse, depression, pain, and prior abdominal surgeries. Clinical diagnoses included muscular back pain, pelvic floor tension myalgia, endometriosis, pyriformis syndrome, vaginismus, pelvic congestion syndrome, myofascial syndrome, adhesions, adenomyosis, fibroids, urethral syndrome, pelvic relaxation problems, and vulvar vestibular syndrome. History of abuse included adult sexual abuse, child sexual abuse, rape at any age, adult physical abuse, and physical discipline during childhood. Depression was measured with the Beck Depression Inventory. The SCL-90 was also used to evaluate mental health. Pain measurements included number of pain sites, duration of pain, and McGill Pain Score. Prior abdominal surgeries included the prior lysis of adhesions, prior ablation or excision of endometriosis, oophorectomy (bilateral, unilateral, none), prior hysterectomy, prior diagnostic surgery (diagnostic laparoscopy or pain mapping), other prior gynecologic procedures (vulvar, vaginal, uterine, ovarian, uterine suspension, laparoscopic uterine nerve ablation), prior nongynecologic abdominal procedures (bladder, bowel, appendectomy, cholecystectomy, hernia), prior surgery by laparoscopy, prior surgery by laparotomy, prior surgery through vagina, total number of prior abdominal procedures (gynecologic, bladder, bowel, appendectomy, cholecystectomy, hernia), and total number of prior surgeries for “this pain.”
We performed univariate analyses to gain a better understanding of each characteristic and evaluated outliers for data entry errors and biologic plausibility. We used bivariate analyses to calculate odds ratios (ORs) and 95% confidence intervals (CIs) for the relationship between each characteristic and IBS.
Each characteristic was evaluated as a potential modifier of the relationship between every other characteristic and IBS. Effect modifiers were considered if they had a Breslow–Day test of homogeneity of P < .10 and stratum-specific odds ratios with greater than 100% difference or stratum-specific odds ratios in opposite directions (SAS 8.01, SAS Institute Inc., Cary, NC). No effect modifiers were identified, therefore, the final model contains only main effects. All variables associated with IBS in bivariate analyses with a P < .10 were evaluated in multivariable models. In addition, physical discipline in childhood, child sexual abuse, rape at any age, and race were included in the full model although they had a P > .10 because the literature supports a possible association with IBS. We used backward stepwise logistic regression with a Wald p-value of 0.10 to develop a final reduced model of the characteristics associated with IBS in this population (Stata 7, StataCorp LP, College Station, Texas).
Seventy-eight percent of 987 new referral patients to this pelvic pain clinic were younger than 40 years of age, 81% were white, 53% had a college or graduate education, 40% had an annual household income of $35,000 or greater, 52% worked full-time, 36% were not employed, and 59% were married. These demographic characteristics were similar by IBS status, although women with IBS were slightly older, more educated, more often separated or divorced or widowed, and working part-time or not employed (Table 1).
The prevalence of IBS in women with chronic pelvic pain was 34.5% (95% CI: 31.6, 37.5, n = 341). Characteristics that were associated with IBS in bivariate analyses (P < .10) were age, pelvic floor tension myalgia as assessed by tenderness to touch, muscular back pain, endometriosis, pyriformis syndrome by positive thigh rotation test, a working diagnosis of adenomyosis, history of adult physical abuse, history of adult sexual abuse, higher Beck Depression Inventory score, Symptom Checklist-90 scores, total number of pain sites, duration of pain, chronic pain syndrome, and McGill Pain Questionnaire score (Tables 1 and 2).
In the final reduced multivariable model, IBS remained associated with muscular back pain (OR = 5.37, 95% CI: 0.98, 29.29), age 40 years or older (OR = 1.98, 95% CI: 1.27, 3.11), Beck Depression Inventory score indicating depression present (10 or greater) (OR = 1.93, 95% CI: 1.24, 3.01), Symptom Checklist-90 global score in top 24% (OR = 1.77, 95% CI: 1.09, 2.86), 6 to 8 pain sites (OR = 1.67, 95% CI: 1.01, 2.78), and history of adult physical abuse (OR = 1.51, 95% CI: 1.01, 2.26) (Table 3).
We found that IBS is associated with specific characteristics of women with chronic pelvic pain, indicating they may not be manifestations of the same disorder. In addition, these characteristics may help us better define the diagnostic criteria of chronic pelvic pain by breaking out IBS as a partial cause of overall pain. In turn this could lead to a more timely diagnosis of other causes of the patient's abdominal and pelvic pain.
Thirty-five percent of women with pelvic pain in this clinic setting met Rome I criteria for IBS, which is similar to previous studies of women referred to a gynecologic clinic9 and women with chronic pelvic pain in a community.10 This prevalence of IBS is approximately twice as high as that reported in the general population of women in the United States (14% to 24%),8 yet little is understood about this association. We did not use the Rome II criteria because they were not available at the start of data collection and Rome I could be applied across the entire study period. Although IBS criteria have changed over time, the Rome I criteria has 97% agreement with Rome II when administered to a population-based sample.20 If IBS and chronic pelvic pain were manifestations of the same underlying process, we would not expect to find associations between IBS and specific characteristics of women with chronic pelvic pain.
Irritable bowel syndrome has an established association with depression, anxiety disorder,6 and history of abuse.7 We found these associations also exist in women with chronic pelvic pain. Although depression and history of abuse are associated with chronic pelvic pain,21 these factors were more common in women with IBS in this population. Antidepressants and psychotherapy have been used to treat bowel symptoms as well as treat depression.22 Hence, future research could evaluate whether active treatment of depression and abuse reduce symptoms of IBS and pelvic pain in this population.
We found that women aged 40 years or older are more likely to have IBS than younger women, which is similar to previously published literature on this association.3 Because 90% of the patients in our study were between 20 and 50 years of age, the category of women over 40 years old primarily reflects the experience of patients 40–50 years old and not those over 65 years old. Hence our results are similar to population-based estimates indicating the highest prevalence in those 45–65 years of age.3
We cannot confirm the previously reported relationship of IBS with endometriosis and dyspareunia.9,10 Endometriosis was associated with IBS with P < .1 in bivariate analyses, but the relationship did not remain when we accounted for other characteristics in the multivariate model. Previous studies of endometriosis and IBS only used bivariate analyses, therefore the association may have been confounded by other factors. The three classifications of dyspareunia (organic, functional, or mixed) were not associated individually or as a group with IBS in bivariate analyses. Women with muscular back pain and pain in many locations were more likely to have IBS. This is consistent with the conceptual expectation that a condition that influences the gastrointestinal tract broadly would have more diffuse manifestations. An extensive pain distribution and muscular back component should prompt evaluation of bowel symptoms in this population.
We did not have information about the percent of the diagnoses that were based on pathologic reports. It is possible that some misclassification occurred, but we have no reason to believe that access to prior pathologic reports should be associated with IBS status.
Other factors not abstracted but may be considered for future studies include: gravidity, family size (number of dependents), care giving burden, method of delivery, and insurance status. Sexually transmitted infections could also be considered. In our population, less than 1% of patients had a sexually transmitted infection at evaluation.
Although the women in this study represent the severe end of the spectrum of chronic pelvic pain, our findings illustrate that specific characteristics are associated with IBS in this population. Our study population has a higher proportion of whites and college educated women than the general United States population, which may be characteristics associated with health care seeking behavior.
We had a unique opportunity to use data that was systematically collected for eight years by four physicians specializing in pelvic pain in a clinic setting using a battery of validated questionnaires, including consensus-established IBS criteria. This small expert provider pool with one research nurse who entered the data gives the study consistency and reduces potential information bias.
We evaluated more characteristics and had a larger sample size than previously published studies of IBS in women with chronic pelvic pain. The size of our population gave us power to assess characteristics that may be too rare to evaluate with smaller sample sizes. We also had the ability to use multivariate modeling and control for the effects of multiple factors at once, unlike previous studies that reported only bivariate associations. Our objective was to evaluate which characteristics were associated with IBS, rather than assess a specific relationship that we established a priori, therefore we used stepwise logistic regression.
Our findings could help physicians attempt to effectively treat women with IBS and chronic pelvic pain. These syndromes affect a large proportion of women1–3,11 and are associated with increased health care use8,12 and decreased quality of life.8,13,14 Our study indicates that specific characteristics distinguish IBS in women with chronic pelvic pain, suggesting that physicians could diagnose and treat IBS in conjunction with treatment for chronic pelvic pain. Future research could focus on the effectiveness of diagnosing and treating IBS among this population. Furthermore, it is unknown whether there is a causal link between IBS and the characteristics that we identified. Prospective studies, if possible, would aid in better understanding the origin of these associations. These and related questions may help clinicians better understand how to alleviate pain in patients with pelvic pain and IBS.
1. Jones R, Lydeard S. Irritable bowel syndrome in the general population. BMJ 1992;304:87–90.
2. Talley N, Zinsmeister A, Van Dyke C, Melton L. Epidemiology of colonic symptoms and the irritable bowel syndrome. Gastroenterology 1991;101:927–34.
3. Sandler R. Epidemiology of irritable bowel syndrome in the United States. Gastroenterology 1990;99:409–15.
4. Everhart J, Renault P. Irritable bowel syndrome in office-based practice in the United States. Gastroenterology 1991;100:998–1005.
5. Thompson W. The irritable bowel. Gut 1984;25:305–20.
6. Drossman D, Whitehead W, Camilleri M. Irritable bowel syndrome: a technical review for practice guideline development. Gastroenterology 1997;112:2120–37.
7. Drossman D. Irritable bowel syndrome and sexual/physical abuse history. Eur J Gastroenterol Hepatol 1997;9:327–30.
8. Drossman D, Li Z, Andruzzi E, Temple R, Talley N, Thompson W, et al. U.S. householder survey of functional gastrointestinal disorders: prevalence, sociodemography, and health impact. Dig Dis Sci 1993;38:1569–80.
9. Prior A, Wilson K, Whorwell P, Faragher E. Irritable bowel syndrome in the gynecological clinic: survey of 798 new referrals. Dig Dis Sci 1989;34:1820–4.
10. Zondervan K, Yudkin P, Vessey M, Jenkinson C, Dawes M, Barlow D, et al. Chronic pelvic pain in the community—symptoms, investigations, and diagnoses. Am J Obstet Gynecol 2001;May:1149–55.
11. Steege J. Scope of problem. In: Steege J, Metzger D, Levy B, editors. Chronic pelvic pain: an integrated approach. Philadelphia (PA): WB Saunders; 1998. p. 2.
12. Jamieson DJ, Steege JF. The prevalence of dysmenorrhea, dyspareunia, pelvic pain, and irritable bowel syndrome in primary care practices. Obstet Gynecol 1996;87:55–8.
13. Reiter R, Gambone J. Demographic and historical variables in women with idiopathic chronic pelvic pain. Obstet Gynecol 1990;75:428–32.
14. Harrop-Griffiths J, Katon W, Walker E, Holm L, Russo J, Hickok L. The association between chronic pelvic pain, psychiatric diagnoses, and childhood sexual abuse. Obstet Gynecol 1988;71:589–94.
15. Mathias S, Kupperman M, Liberman R, Lipschutz R, Steege J. Chronic pelvic pain: prevalence, health-related quality of life, and economic correlates. Obstet Gynecol 1996;87:321–7.
16. Longstreth G, Preskill D, Youkeles L. Irritable bowel syndrome in women having diagnostic laparoscopy or hysterectomy. Dig Dis Sci 1990;35:1285–90.
17. Beck A, Ward C, Mendelson M, Mock J, Erbaugh J An inventory for measuring depression. Arch Gen Psychiatry 1961;4:561–71.
18. Derogatis L, Lipman R, Covi L. SCL-90: an outpatient psychiatric rating scale—preliminary report. Psychopharmacol Bull 1973;9:13–28.
19. Melzack R. The McGill Pain Questionnaire: major properties and scoring methods. Pain 1975;1:277–99.
20. Saito YA, Talley NJ, Melton L, Fett S, Zinsmeister AR, Locke GR. The effect of new diagnostic criteria for irritable bowel syndrome on community prevalence estimates. Neurogastroenterol Motil 2003;15:687–94.
21. Fry R, Crisp A, Beard R, McGuigan S. Psychosocial aspects of chronic pelvic pain, with special reference to sexual abuse: a study of 164 women. Postgrad Med J 1993;69:566–74.
22. Jailwala J, Imperiale TF, Kroenke K. Pharmacologic treatment of the irritable bowel syndrome: a systematic review of randomized, controlled trials. Ann Intern Med 2000;133:136–47.