|Home | About | Journals | Submit | Contact Us | Français|
The burden of functional GI disorders and their associations with psychological distress in women veterans is unclear.
To examine one-year prevalence of IBS and dyspepsia symptoms and their associations with anxiety, depression and PTSD among women veterans receiving primary care at a Veteran Affairs Medical Center Women’s Clinic.
IBS, dyspepsia and psychological distress were assessed using the validated self-administered Bowel Disorder Questionnaire, the Beck Depression and Anxiety Inventories, as well as the Mississippi Scale for Combat-Related posttraumatic stress disorder (PTSD) questionnaire.
We enrolled 248 women (84% participation rate). Ninety-three (38%) reported IBS and 51 (21%) dyspepsia symptoms. Women with IBS and dyspepsia reported higher mean scores of anxiety (IBS: 24 vs. 12, p<.0005 and dyspepsia: 26 vs. 12, p <.0005), depression (IBS: 22 vs. 11, p=.0005 and dyspepsia: 23 vs. 11, p <.0005), and PTSD (IBS: 87 vs. 69, P<.001 and dyspepsia: 86 vs. 69, p <.0005). Age- and ethnicity-adjusted logistic regression analyses showed a 3- to 46-fold increase in odds of IBS and dyspepsia among women with anxiety, depression, or PTSD.
Women veterans have high prevalence of IBS and dyspepsia symptoms, both of which are highly associated with presence of depression, anxiety and PTSD.
Irritable bowel syndrome (IBS) and functional dyspepsia are common functional gastrointestinal (GI) disorders. IBS and dyspepsia prevalence rates vary according to diagnostic criteria and target population (e.g., community, primary care, or specialty care patient populations), but tend to be more common in women than in men. In Western countries, the prevalence of IBS symptoms ranges from 5% to 27% in women and from 2% to 19% in men,1,2 while the prevalence of dyspepsia symptoms, most of which are thought to be functional in origin, ranges from 12% to 38% in women and 8% to 27% in men.3–5
A biopsychosocial model of illness has been used to explain connections between GI and emotional and cognitive functions, in which functional GI disorders are conceptualized as dysfunction of brain-gut interactions.1,6–11 In this model, stressful life events and psychological factors exacerbate IBS and dyspepsia symptoms,1,12,13 influence health care seeking behaviour,1,14,15 and impact clinical outcomes,1,6,12 such as improvement in symptom intensity.13
Psychological distress and trauma are commonly reported in patients with functional GI disorders. For example, significant associations have been reported between IBS and increased psychological distress,16–19 including depression,16,20–23 anxiety,20,23–27 and PTSD.28 Similar associations have been reported for dyspepsia in clinic-based studies.29,30 Sexual and physical abuse are also more commonly reported among patients with functional GI disorders (40% to 53%) compared with organic GI disorders (10% to 37%).31–33
Psychological distress as well as trauma is also commonly reported among women veterans.34–37 Previous research highlights the considerable physical and mental illness burden among women veterans who use VA medical facilities.38 Women veterans report high rates of military sexual assault; for example, the reported prevalence of rape during military service ranges from 11% to 48%.39 Due to women veterans’ increased risk of experiencing psychological distress, we expect an excess burden of IBS and dyspepsia among women veteran VA users. The Department of Veterans Affairs' Office of Policy and Planning estimates that in 2006 women comprised approximately 7% of the total veteran population (1,731,125 / 23,976,991) and by 2030 is projected to comprise 13% (2,002,971/15,155,603).40 Approximately 11% of all women veterans currently obtain some or all of their health care from the VA.39
Few studies have investigated prevalence of IBS or dyspepsia and their associations with common psychological disorders among women veterans.41,42 We therefore conducted this cross-sectional study to examine the prevalence of IBS and dyspepsia symptoms among women veterans at the Women’s Clinic at the Michael E. DeBakey Veterans’ Affairs (VA) Medical Center in Houston, Texas. We also examined the association between IBS and dyspepsia, with depression, anxiety and PTSD symptoms. To our knowledge, this is the first study that examines prevalence of IBS and dyspepsia in relation to commonly associated factors of psychological distress among women veterans.
The sampling frame for this study consisted of women veterans scheduled for outpatient primary care at the Michael E. DeBakey VA Women’s Clinic. Overall, the female patient population at this clinic is 44% white and 47% black, and 25% are between 18 and 40 years of age. We obtained daily lists of patients routinely scheduled for primary care clinic. We included only women veterans (not family members of veterans) 18 years of age and older and scheduled for the primary care clinic. All eligible women were approached and introduced to the study at the time of their arrival at the clinic and asked to complete the study questionnaires provided by a female researcher who was not part of the clinical staff. Participants were given the opportunity to complete questionnaires at the clinic or return completed questionnaires in a pre-paid envelope. We also manually reviewed the comprehensive and integrated electronic medical records of all participants in the study to identify any diagnosis or symptoms compatible with IBS for the one year time period before questionnaire administration as well as the one year time period after. Among those women veterans who fulfilled BDQ symptom-based criteria for IBS, we also examined their entire available medical record for either definitive or potential occurrence of any alternate organic explanations of these symptoms including inflammatory bowel disease, celiac sprue, or GI malignancy.
A sociodemographic questionnaire that included 11 items was used to ascertain age at time questionnaire was completed, race/ethnicity, marital status and educational attainment. In addition, we included the following widely used screening instruments:
The Bowel Disorder Questionnaire (BDQ) is a validated and reliable questionnaire that was used to determine presence of IBS and dyspepsia symptoms during the past year. The BDQ includes 59 items, 46 of which help identify and quantify the type, frequency and severity of gastrointestinal symptoms.43,44 In a sample of 395 subjects (50% women; mean age, 51 years) the mean kappa statistic indicated a high reliability for all BDQ items (kappa=0.78; range: 0.52–1.0) for retests conducted a second time (ranging from 24 hours to 7 weeks).43 The BDQ also discriminated functional bowel disease from healthy controls with a sensitivity of 83% and a specificity of 76%.43
The Beck Depression Inventory-2nd Edition (BDI-II) is a validated and reliable self-report questionnaire that was used to identify the existence and severity of depression symptoms during the past two weeks, based on diagnostic criteria in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. Each of its 21-items is rated on a 4-point Likert scale from 0 (strongly agree) to 3 (strongly disagree). The BDI-II manual reports high internal consistency among 500 psychiatric outpatients (coefficient alpha =0.92) and 120 college students (coefficient alpha =0.93).45 High internal consistency has also been reported in a sample of 340 adult primary care patients (correlations ranged from 0.54 to 0.74) 46. Among 220 African American primary care patients (52% women; ≥18 years of age) the BDI-II had high internal consistency (coefficient alpha =0.90), sensitivity (88%), specificity (84%), and diagnostic accuracy (85%). 47 Additionally, the BDI-II correlates well with psychological tests such as the Revised Hamilton Psychiatric Rating Scale for Depression (r=0.71), the Beck Hopelessness Scale (r=0.68), and the original BDI-IA (r=0.93).45 For our study, depression was computed as a total of all items and then categorized according to recommended cutoff values for total scores: 0 to 13 to indicate none or minimal range depression; 14 to 19 to indicate mild depression, 20 to 28 to indicate moderate depression, and 29 to 63 to indicate severe depression.45
The Beck Anxiety Inventory (BAI) is a validated and reliable self-report questionnaire that is used to assess common anxiety symptoms in the past week. Each of its 21 items is rated on a 4-point Likert scale ranging from 0 (strongly agree) to 4 (strongly disagree). Anxiety was computed as a total of all items and then categorized according to recommended cutoff values for total scores: 0 to 7 to indicate minimal anxiety; 8 to 15 to indicate mild anxiety, 16 to 25 to indicate moderate anxiety; and 26 to 63 to indicate severe anxiety.48,49 Beck et al48 analyzed reliability and validity among 160 psychiatric outpatients, and reported a high internal consistency reliability (coefficient alpha=0.92), as well as a high correlation (0.75) between one-week test-retest scores.
The Mississippi Scale for Combat-Related PTSD Questionnaire (M-PTSD) is a validated and reliable self-report questionnaire used to assess symptom severity of PTSD in combat veterans. Each of the 35 items is rated on a 5-point Likert scale ranging from 1 (not at all true or never) to 4 (extremely true or very frequently). The scale range of PTSD symptom severity (35 to 175) is derived by summing all items. In a sample of 362 male Vietnam-era veterans seeking care from Vietnam Veteran Outreach Centers, high internal consistency was reported for the entire scale (coefficient alpha=0.94).50 In a sample of 30 Vietnam combat veterans who were diagnosed with PTSD according to the DSM-III criteria, 30 non-combat Vietnam-era veterans receiving care for non-psychotic problems, and 32 Vietnam veterans with no history of psychological or psychiatric treatment, the PTSD group scored significantly higher on the Mississippi compared with the psychiatric group (P< .0001), and the psychiatric group scored marginally higher than the well adjusted group (P< .10).50 In this study we define PTSD using the recommended diagnostic cutoff value of 107 to indicate PTSD is present. At this cutoff level, the Mississippi scale had a previously reported sensitivity of 93%.50
We defined IBS and dyspepsia based on self-reported symptoms during the 12 months prior to clinical presentation using the BDQ. A full listing of symptom-based diagnostic criteria (based on modified Rome II criteria) is listed in table 1.
Sociodemographic characteristics and psychological distress were compared in different groups (either IBS or dyspepsia present versus neither IBS nor dyspepsia present) using χ2 statistics to test statistical significance for dichotomous and categorical variables and t tests for continuous variables. We examined socio-demographic variables including age (ages ≤ 44 years, ≥45 years); race/ethnicity (black, white, and Hispanic or other, including American Indian/Alaskan native ethnic groups); education (high school graduate or less, at least some college); and marital status (married, separated, divorced or widowed, and never married). Psychological distress variables were examined using previously presented cutoffs values for depression and anxiety scores (minimal, mild, moderate and severe) 45,49 and PTSD (present, absent).50 The Levene's test was used to test for equality of variances between groups and if the criterion was not met (P < .05) we computed the t tests for continuous variables assuming unequal variance. Linear-by-linear χ2 tests for trend for categorical data were computed to examine the relationship between IBS and dyspepsia across levels of depression and anxiety.
We examined the associations between IBS or dyspepsia (dependent variables), and psychological distress, including depression, anxiety, and PTSD symptoms (independent variables) using logistic regression models. Depression, anxiety and PTSD were each modeled separately as continuous as well as categorized variables described above. For continuous variables, the parametric estimates reflect the degree of risk change associated with every 1-point change in score. To account for possible confounding, we adjusted for age-groups (≤45 years, >45 years) and ethnicity-groups (white, non-white). Another model included all psychiatric disorders (depression, anxiety and PTSD) simultaneously as continuous variables to adjust for co-existing psychiatric symptomatology. All analyses were conducted using SPSS version 15 (SPSS Inc, Chicago, IL, USA).
The study was reviewed and approved by the Institutional Review Board for Human Subject Research at Baylor College of Medicine.
Between November 2005 and February 2006, we identified and approached 296 women veterans potentially eligible to participate. Of these, 283 consented to participate and 248 completed the questionnaires, producing an overall participation rate of 84% (248/296). Among all participants, 40% were 45 years of age or less. Most self-identified as either black (48%) or white (41%), 9% identified as Hispanic, and 2% as other ethnic groups. Twenty-eight percent were married and 54% were previously married (divorced, separated or widowed), and most (72%) had some college education or less.
Thirty-eight percent (93/248) of participants reported symptoms consistent with a potential diagnosis of IBS (based on self-reported symptoms using the BDQ), most with diarrhea features (92/93; 99%). Fifty-two percent (48/93) of IBS cases had constipation, almost all (47/48) also reporting presence of diarrhea symptoms. Twenty-one percent (51/248) of participants reported dyspepsia symptoms. Fifteen percent (36/248) reported both IBS symptoms and dyspepsia symptoms, while 57% (140/248) reported neither IBS nor dyspepsia symptoms.
Sensitivity analysis of the BDQ-derived case definition of IBS was also conducted. For all study participants, we reviewed medical records to identify presence of an IBS diagnosis or of two or more symptoms compatible with IBS anywhere in the medical record within a one-year time period before and also after questionnaire administration. Furthermore, the chart review of the 93 BDQ-derived IBS patients identified possible alternate organic explanations of their symptoms in only three patients (n=1 endometriosis, n=1 celiac sprue, and n=1 pelvic mass).
Among all participants, the mean BDI score for depressive symptoms was 16.2 (SD: 14.3), indicating mild depression symptoms. Half (53%) of women scored in the minimal depression range (0–13), while 14% scored in the mild (14–19), 13% moderate (20–28), and 21% in the severe range for depression (29–63). The mean BAI score for anxiety symptoms was 17.1 (SD: 13.7), indicating moderate anxiety symptoms overall. Nearly one third of women (29%) were categorized as minimal anxiety (0–7), while 23% scored in the mild (8–15), 24% moderate (16–25), and 23% in the severe (26–63) range for anxiety. The mean score for PTSD symptoms, as measured by the Mississippi Scale for Combat-Related PTSD Questionnaire, was 76.7 (SD: 25.0). Approximately 15% of all women screened positive for PTSD (M-PTSD score ≥ 107).
Table 2 presents sociodemographic differences between participants with IBS and dyspepsia and those with neither IBS nor dyspepsia symptoms. A larger proportion of younger women veterans reported IBS (46%) and dyspepsia (43%) symptoms than those without IBS or dyspepsia symptoms (37%), however, the differences were not statistically significant. There were also no significant differences in race/ethnicity, education, or marital status between those with IBS or dyspepsia and those with neither symptom.
Table 3 provides comparisons of self-reported psychiatric distress between participants with IBS or dyspepsia and participants with neither symptom. Women reporting IBS symptoms had significantly higher mean scores for depression (+10.9 points; P <.0005), as well higher mean scores for anxiety (+11.9 points; P <.0005), and PTSD (+17.4 points; P <.0005) compared with women with neither IBS nor dyspepsia symptoms. Women with dyspepsia also had significantly higher mean scores for depression (+11.6 points; P <.0005), anxiety (+13.8 points; P < .0005) and PTSD (+16.7 points; P < .0005), compared with women with neither IBS nor dyspepsia symptoms. The SPSS linear-by-linear association test of trends were significant for an association between anxiety and both IBS (P < .0005) and dyspepsia (P < .0005) indicating significant linear increase in the proportions of participants with IBS and dyspepsia as anxiety severity increased. Significant trends were also observed for depression with IBS (P < .0005), as well as dyspepsia (P < .0005).
Age- and sex-adjusted logistic regression models estimated that the odds of IBS or dyspepsia symptoms increased by at least 3-folds (between 3- and 9-fold) in women with mild or greater degree of depression as compared with women with none or minimal depression. The associations between anxiety and IBS or dyspepsia were even higher than these with depression and showed a clear severity-response relationship where greater severity of anxiety was associated with up to 16-fold increase in IBS and more than 40-fold increase with dyspepsia (Table 3).
Each one-point increase in PTSD score increased the odds of reporting for both IBS and dyspepsia symptoms by approximately 3% (Tables 3). The excess odds of a woman veteran with PTSD reporting IBS symptoms was four-fold greater than the odds of a woman veteran without PTSD. Similarly, the odds of a woman veteran with PTSD reporting dyspepsia symptoms was nearly five times greater than a woman veteran without PTSD. When adjusting for age, ethnicity and all psychiatric disorders (as continuous variables) simultaneously, only anxiety symptoms remained significantly associated with IBS (adjusted odds ratio=1.07; 95% CI, 1.03–1.12) and dyspepsia (adjusted odds ratio=1.08; 95% CI, 1.03–1.13).
To our knowledge, this is the first study to examine the prevalence of self-reported IBS and dyspepsia symptoms and their association with several indicators of psychological distress, in an exclusively female veteran population of VA users. We observed a high prevalence of IBS (38%) and dyspepsia (21%) symptoms. Separate logistic regression models revealed significant excess odds of IBS and dyspepsia among participants reporting depression, anxiety and PTSD symptoms, as compared to reporting neither symptom. The associations were stronger for anxiety, which was the only psychiatric disorder to remain significant after adjusting for all psychiatric disorders (depression, anxiety and PTSD) simultaneously.
The prevalence rate of IBS in this study is relatively higher than that previously reported among women in the general population in Western countries (5% to 27%).1,2 In contrast, the prevalence we ascertained for dyspepsia (21%) falls into the range of prevalence rates (12% to 38%) previously reported among women in Western countries,3–5 but is lower than the 44% reported by Dominitz and Provenzale among 107 women veteran VA users.41 However, direct comparison of our rates is not possible due to differences in sampling methods as well as definitions and number of criteria used to identify IBS and dyspepsia cases.2
Our study examined a large proportion of African American women (47% of the Women Clinic’s population and 48% of the study population). Thus, our data provided important information on this relatively unexamined segment of the population.51 Our results indicated no statistically significant differences of IBS and dyspepsia prevalence by race/ethnicity among women veteran VA users.
Whether the prevalence of functional gut disorders varies according to type or region of military service was not examined in this study. Previous evidence based on Persian Gulf War veterans, reported by Gray et al52 revealed an over three-fold increased odds of IBS among predominantly male U.S. military Persian Gulf War Navy Seabees (n=3,831) compared with Seabees deployed elsewhere (n=4,933) or with non-deployed Seabees (n=3,104). Similarly, Eisen et al53 reported a nearly two-fold significantly increased prevalence odds of dyspepsia among deployed (n=1,061) compared with non-deployed (n=1,128) predominantly male Gulf War veterans. Reasons for these differences are unclear.
Our results are consistent with positive associations found between anxiety and IBS20,23–26 and dyspepsia,29,30,54 as well as between depression and IBS and dyspepsia.16,20–24 In general, when modeled separately, depression had a significant effect on IBS and dyspepsia, but weaker effect than anxiety. Among IBS participants, we found a higher prevalence of moderate and severe levels of anxiety than depression. Similarly, those with dyspepsia had a higher prevalence of mild, moderate and severe levels of anxiety than depression. This discrepancy corresponds to findings by Sykes et al26, who reported a higher frequency of current (50%) and lifetime anxiety (58%) compared with current (5%) and lifetime (28%) major depression among treatment-seeking IBS patients. However, the comparatively weak association found in our study between IBS or dyspepsia and current depression may be a result of our limited sample size or conduct of the study within one VA primary care clinic. Whether IBS and dyspepsia are more strongly associated with anxiety than depression among the larger female veteran population merits further investigation; however, our findings suggest that anxiety may be a risk factor for the acquisition or impact of IBS on female veteran patients, and could have important clinical implications. For example, the high proportion of diarrhea sufferers suggests that this is a specific feature of IBS in this anxious female population.
While the overall PTSD prevalence rate of 15% in this study may be regarded as high, it was lower compared with results from a previous study conducted by Dobie et al42 among women veteran VA users at the VA Puget Sound Health Care System (22%). The discrepant finding may be due to methodological differences, for example, Dobie et al42 used a mailed survey, a different instrument to identify PTSD, and the study population was more white (73% compared with 41% in our study) and attained higher levels of formal education (84% attended college compared with 28% in our study). While the PTSD literature in women indicates that PTSD is associated with increased use of health care services, few studies have examined the direct impact of PTSD on female-specific health outcomes,55 including IBS and dyspepsia. We found that PTSD is associated with significantly increased odds of both IBS (OR=4.11) and dyspepsia (OR=4.85). Our results are consistent with those previously reported regarding IBS. 28,42 Irwin et al28 reported a high lifetime prevalence of PTSD (36%) in a small sample (n=50, including 40 women) of IBS patients and Dobie et al42 reported a nearly three-fold significant increased age-adjusted odds of IBS among women veterans who screened positive for PTSD compared with women who screened negative.
The definitions of IBS and dyspepsia used in our study were intended as the best possible approximations of the Rome II diagnostic criteria, which require symptoms of abdominal discomfort or pain to be present at least 12 weeks, not necessarily consecutive, in the preceding 12 months.6 We used particularly conservative definitions that require discomfort or pain of at least moderate severity and at least six times in the past year. While the BDQ was not specifically designed to measure Rome II criteria, an advantage of using the instrument to identify symptoms consistent with a potential clinical diagnosis of IBS and dyspepsia is its ease of administration and minimal burden on study participants. In addition, the BDQ is widely used within the literature. Therefore its use in our study helps maximize comparability with other studies that have similarly used the BDQ and adapted Rome II criteria to define and identify IBS56.
A particular strength of the current study is our parallel performance of a structured medical record review for all study participants. It provided information important to qualifying our study findings including: First, our BDQ-derived definition of IBS using adapted Rome II criteria yields a substantively higher prevalence of IBS than that documented in medical records. This is not an unexpected finding as it has been shown that the majority of persons with IBS do not seek medical attention for their symptoms57. However, our medical record review also demonstrated a very low false positive rate (7%) or very high specificity (93%) for the BDQ. Overall, the BDQ has good test performance characteristics including a good PPV or positive predictive value (i.e., 75% probability that a woman veteran that has a positive finding of a diagnosis or symptoms in her medical record will be BDQ-classified as having IBS) and also good NPV or negative predictive value (i.e., 71% probability that a woman veteran with a negative finding of symptoms or diagnoses in her medical record will also be classified by the BDQ as not having IBS). Overall, we believe both the good NPV and PPV support use of the BDQ as a screening instrument for IBS in our study population. Second, participants in the study did not differ significantly from non-participants in the prevalence of IBS or dyspepsia diagnoses or symptoms recorded in their medical records thus providing some reassurance against selection bias. We believe that our approach of targeting consecutive patients irrespective of their presenting complaints coupled with high participation rate contributed to this apparent lack of selection bias based on GI symptoms. Finally, additional assurance is provided by the fact that the vast majority of participants with IBS symptoms did not have an obvious organic explanation for their symptoms in their medical records.
Although this study adds substantially to sparse literature on IBS and dyspepsia among women veterans, it has potential limitations that also need to be discussed. First, as our study was conducted in a primary care clinic at a single VA, it is possible that our findings may not be generalizable to other women veterans who use the VA system. However, the large representative sample, including a broad range of age and ethnic groups is reassuring. Additionally, our cross-sectional study design precludes conclusions regarding a causal relationship between psychological distress such as anxiety and depression and risk of IBS or dyspepsia.. Prospective research that can assess additional risk factors, as well as timing and duration of the psychological factors in relationship to women veterans’ development of IBS and dyspepsia symptoms is necessary to better understand what etiologic and prognostic roles depression, anxiety and PTSD may play. Another limitation is that we did not collect information on combat history or traumas that occurred in relation to military service, most notably a history of military sexual trauma and of combat experience that could potentially modulate some of the associations we observed. However, given the strength, significance and consistency of the effects of these psychosocial variables on functional GI disorders demonstrated by this study, it is very unlikely that they could be explainable by confounding due to a history of selected military traumas, particularly as some of these traumas are expected to be very uncommon among women veterans (e.g., combat experience) and some would be expected to effect a substantial number of women veterans (e.g., a history of military sexual harassment, abuse or assault). We have subsequently begun collecting data to allow us to examine the effect selected military service-related traumas, particularly military sexual trauma, may play in risk for and outcomes with functional GI disorders.
In summary, our study used validated measures to determine the prevalence of IBS (38%) and dyspepsia (21%) among women veterans receiving primary care at a VA Women’s Clinic. We found significant associations between IBS and dyspepsia symptoms and increased presence of depression, anxiety and PTSD symptoms among these women veterans. These results indicate that screening for IBS and dyspepsia might be justified in women veterans with psychological distress and conversely, that screening for psychological distress among women veterans with IBS and dyspepsia might be justified. The effect of history of combat as well as specific trauma needs to be examined will be examined in future work based on a larger study population, in addition to the potential effect of comorbid presentation of functional GI disorders and psychological distress on patient outcomes (e.g., quality of life) or system outcomes (e.g., health resource utilization)..
Financial support: This material is based upon work supported in part by Health Services Research and Development (HSR&D) Service, Office of Research and Development, Department of Veterans Affairs. This study is partly funded by a grant from Novartis pharmaceuticals to Dr. El-Serag. Dr El-Serag is supported by National Institute of Health grant K-24 DK078154-03. Dr. Savas is supported by the National Research Service Award 5 T32 HP10031-09 funded by the Health Resources and Services. Dr. Cully is supported by a Career Development Award (CDA-2; 05-288) from the Veterans Health Administration, HSR&D Service.
Specific author contributions: Hashem B. El-Serag was responsible for the study design. Hashem B. El-Serag, Donna L. White and Lara S. Savas were responsible for the interpretation of results and manuscript. Data analysis was done by Lara S. Savas. All authors reviewed and revised the manuscript of the paper and approved its final version.
There are no conflicts of interest to disclose.