PDF
Citation
Print
INTRODUCTION
Irritable bowel syndrome (IBS) is a functional bowel disorder defined by the presence of chronic or recurring symptoms that include abdominal pain and discomfort (for Rome III criteria [1], but not for Rome IV criteria [2]), flatulence, bloating, and altered bowel habits [1]. IBS is a disorder of unknown etiology; there are no known biochemical, structural, or physiological abnormalities to characterize IBS. The prevalence of IBS in Western populations is estimated to be between 5% and 20% [3]. Sex ratio, frequencies of abnormal transit or defecation disorders, association to dyspepsia vary strongly according to different countries [3].
In order to test adult personality, the Minnesota Multiphasic Personality Inventory (MMPI) is the most widely used psychometric test [4]. The psychological profile of many patients with digestive disorders was previously assessed with this test [5-7]. The MMPI-2, a significant revision of the MMPI, standardized on a national sample of U.S. adults [8], introduced a wide variety of subscales to help clinicians in the interpretation of the results of the original clinical scales.
To understanding illness and human health in their entire contexts, the biopsychosocial approach [9] was proposed and used to analyze functional gastrointestinal disorders (FGIDs) such as IBS in the Rome III presentation [10]. In this hypothesis, IBS results from dysfunction of the enteric and central nervous system, which is manifested as dysmotility and/or visceral hypersensitivity and is modified by psychosocial processes. In this multicausal pathology, biological and psychological factors interact to determine the disease activity and experience of illness. IBS symptoms can be associated with psychological co-morbidities: anxiety disorders, depression, somatoform disorders, and phobic disorders [11]. This psychological profile explains an exaggerated intestinal pain and emotional distress [12].
Nevertheless, most of these studies do not use the Rome criteria [13], and MMPI was never used to characterize IBS subtypes as defined by the Rome III criteria. IBS is a clinically defined disorder without physiopathologic processes [14]. We hypothesize that different IBS subtypes are associated with different personality profiles. The present study aims to evaluate the psychological profile of IBS and IBS subtypes in a cohort of outpatients consulting for FGID in a tertiary center by using a normalized psychometric test, the MMPI-2.
METHODS
1. Subjects
Between September 2010 and December 2014, 1,293 outpatients were consecutively referred by gastroenterologists to our Center for Functional GI and Motility Disorders in the Gastroenterology Clinic of the Avicenne Hospital, a tertiary center for FGID management. A full clinical evaluation, including morphological (endoscopy or radiology) and the exclusion of metabolic, endocrinologic, and neurologic etiologies, failed to yield an organic cause for their complaint. The use of narcotics and previous surgery of the gastrointestinal tract were exclusion criteria. A single investigator (M.B.) independently confirmed the validity of the initial FGID diagnosis.
We ask the patients that have not initially filled in the MMPI questionnaire to do so at the second clinical visit. In toto, 608 patients (69% female), aged 44.5 ± 17.0 years (mean ± standard deviation [SD]), body mass index (BMI) 26.5 ± 6.0 kg/m², filled in the personality questionnaire and a standard Rome III adult diagnosis questionnaire.
2. Methods
Following the French legislation, this study was registered in the French National Agency for drug safety (decision number: 2015-A01661-48). We perform the present study according to the provisions of the Declaration of Helsinki. Informed written consent was obtained from each patient.
4. Clinical Questionnaire
Patients in the gastroenterologist’s office filled out a standard Rome III adult diagnosis questionnaire for FGIDs [15].
For functional bowel disorders, the diagnosis of IBS was based on the report of recurrent abdominal pain or discomfort at least 3 days per month in the last 3 months associated with 2 or more of the following: improvement with defecation, beginning associated with a change in frequency of stool, or with a change in form (appearance) of stool. Subtypes of IBS (IBS with constipation [IBS-C], IBS with diarrhea [IBS-D], mixed IBS [IBS-M], and unsubtyped IBS [IBS-U]) were defined according to the Rome III criteria [1]. In the absence of IBS, the diagnosis of other functional bowel disorders (bloating, constipation, diarrhea, and unspecified) was made. Finally, nonspecific bowel disorders were diagnosed by exclusion when bowel disorders were present but did not meet the criteria mentioned above.
Functional esophageal [16] (heartburn, chest pain, dysphagia, globus), gastroduodenal [17] (dyspepsia, postprandial distress syndrome, epigastric pain syndrome, aerophagia), anorectal [18] (fecal incontinence, anorectal pain, including levator ani syndrome proctalgia fugax, difficult defecation), disorders and abdominal pain [19] were diagnosed according to the Rome III criteria.
5. Psychological Profiles
We used the MMPI-2 to assess the psychological profile as previously described [20]. The individual raw MMPI-2 score was converted to a T-scale score (mean ± SD, 50 ± 10), which is age and sex-adjusted to the healthy controls [21]. Briefly, for each subject, the MMPI-2 is scored by 6 validity scales namely L, F, Fb, K, true response inconsistency (TRIN), and variable response inconsistency (VRIN), and 10 clinical scales namely 1-Hypochondriasis, 2-Depression, 3-Hysteria, 4-Psychopathic deviate, 5-Masculinity-Femininity, 6-Paranoia, 7-Psychasthenia, 8-Schizophrenia, 9-Hypomania, and 10-Social introversion. Sixty other items, also calculated, are described in the Supplementary Text.
6. Statistical Analysis
We used IBM SPSS software version 20.0 (IBM Corp., Armonk, NY, USA) to carry statistical analyses and express the results for quantitative parameters as mean ± SD. Differences among the groups were searched by one-way analysis of variance using Bonferroni correction for multiple comparisons with post hoc tests. Analysis of qualitative variables was used chi-square tests.
Logistic regression was used for data analysis that systematically included IBS as the dependent variable and as independent variables, BMI, and the validity and clinical MMPI-2 scales. For the analysis of the psychological characteristics of the IBS subtypes, 2 multinomial logistic regression models, adjusted for BMI, with the non-IBS and the unspecified IBS subtype used as reference groups, were created for each group. The backward selection was used for model selection during all multivariate logistic regression. Statistically significant variables (P< 0.05) remained in the adjusted model.
RESULTS
1. Patients Characteristics
The Rome questionnaire divided the 608 FGIDs patients into 235 IBS patients (39%; 77 for IBS-C, 68 for IBS-D, 54 for IBS-M, 36 for IBS-U) and 373 non-IBS patients (61%) (Table 1). IBS and non-IBS patients have similar age and sex ratio, but nonIBS patients have higher BMI than IBS patients (P< 0.001). Among the 4 IBS subtypes, IBS-C patients are more frequently of female sex than IBS-M patients (P= 0.003) and have lower BMI than IBS-D (P= 0.002) and IBS-U (P= 0.002) patients. The univariate analysis shows that IBS patients report a higher prevalence of chest pain (P< 0.001), dysphagia (P= 0.004), postprandial distress syndrome (P< 0.001), nonspecific dyspepsia (P< 0.001), aerophagia (P< 0.001), and levator ani syndrome (P= 0.004).
2. Psychological Characteristics of IBS Patients
IBS patients show higher symptom exaggeration than non-IBS patients (P< 0.001) and decreased test defensiveness score (P< 0.001) (Table 2, Supplementary Table 1). IBS patients have a significantly higher score for all clinical scales: Hypochondriasis (P< 0.001), Depression (P< 0.001), Hysteria (P< 0.001), Psychopathic deviate (P= 0.005), Masculinity-Femininity (P< 0.001), Paranoia (P< 0.001), Psychasthenia (P< 0.001), Schizophrenia (P< 0.001), Hypomania (P= 0.005) and Social introversion (P= 0.042). The results of the content, supplementary and Harris-Lingoes scales for IBS and non-IBS patients are shown in the Supplementary Text.
As shown in Fig. 1, the multivariate logistic regression found that IBS group is characterized by decreased BMI (P= 0.002; odds ratio [OR], 0.957; 95% confidence interval [CI], 0.930–0.984), decreased test defensiveness score (P= 0.001; OR, 0.972; 95% CI, 0.955–0.989) and increased scales for Hypochondriasis (P< 0.001; OR, 1.049; 95% CI, 1.035–1.064) and Masculinity-Femininity (P= 0.018; OR, 1.020; 95% CI, 1.003–1.037).
The frequency of abnormal clinical scales is shown in the Supplementary Tables 2 and 3 and analyzed in the Supplementary Text.
3. Demographics and Psychological Characteristics of IBS Subtypes Patients
By comparison with non-IBS patients (Table 3, Supplementary Table 4), IBS-C patients have a more frequently female sex (P< 0.01) and a lower BMI (P< 0.001) while IBS-M patients have lower BMI (P= 0.014).
Among the IBS patients, IBS subgroups show a significant difference for one validity scale, Infrequency scale with the highest value for the IBS-M group (P< 0.001 vs. non-IBS patients and P= 0.040 vs. IBS-C patients) and 4 clinical scales: Hypochondriasis, Depression, Psychasthenia, and Hypomania. The differences of the content, supplementary, and HarrisLingoes scales are shown in the Supplementary Table 3 and described in the Supplementary Text.
By comparison with non-IBS patients, IBS-C, IBS-D, and IBSM patients have increased Hypochondriasis (P< 0.001 for these 3 groups), increased Depression (P= 0.013 for IBS-C, P= 0.005 for IBS-D, and P< 0.001 for IBS-M), IBS-M patient have higher Psychasthenia (P< 0.001) and higher Hypomania (P= 0.004). No item was significantly different in the IBS-U group.
The multinomial logistic regression shows that, by comparison with the non-IBS patients (Fig. 2), IBS-C patients have lower BMI (P< 0.001; OR, 0.911; 95% CI, 0.869–0.954) and increased value for Hypochondriasis (P= 0.001; OR, 1.058; 95% CI, 1.022–1.095) and Masculinity-Femininity scale (P= 0.034; OR, 1.028; 95% CI, 1.002–1.055), the IBS-D patients have increased Hysteria score (P= 0.003; OR, 1.059; 95% CI, 1.020–1.099) and the IBS-M patients have decreased BMI (P= 0.005; OR, 0.922; 95% CI, 0.871–0.976), decreased test defensiveness score (P =0.036; OR, 0.946; 95% CI, 0.898–0.996) and increased Hypochondriasis scale (P= 0.008; OR, 1.058; 95% CI, 1.015–1.103). In contrast, no significant association was found with the IBS-U group.
By comparison to IBS-U patients, the multinomial logistic regression shows that IBS-C patients gave increased Depression score (P= 0.011; OR, 1.095; 95% CI, 1.021–1.174) and decreased BMI (P= 0.018; OR, 0.909; 95% CI, 0.839–0.983), the IBS-M patients have only decreased BMI (P= 0.030; OR, 0.909; 95% CI, 0.835–0.991) while the IBS-D patients were not significantly different.
The frequency of abnormal values is shown in Supplementary Tables 5 and 6 and analyzed in the Supplementary Text. Briefly, the multinomial logistic regression shows that, as compared to the non-IBS patients, only IBS-C and IBS-M patients are associated with an increased frequency of abnormal Hypochondriasis score: for IBS-C patients (P= 0.001; OR, 2.958; 95% CI, 1.603–5.458), and IBS-M patients (P= 0.003; OR, 3.245; 95% CI, 1.489–7.070).
DISCUSSION
This study confirms the existence of significant psychological characteristics in IBS patients. They have higher symptom exaggeration than non-IBS patients (P< 0.001), and higher score for all clinical scales. Also, psychological characteristics associated with IBS phenotypes, except IBS-U, were found. By comparison with the non-IBS patients, IBS-C patients have a higher score for Hypochondriasis and Masculinity-Femininity scale, IBS-D patients have a higher score for Hysteria, and the IBS-M patients have decreased test defensiveness score and increased Hypochondriasis scale. By comparison to IBS-U patients, IBS-C patients reported a higher score of Depression.
We used a French validated translation of the MMII 2. This 567 items-test usually takes between 1 and 2 hours to complete depending on the reading level. To avoid any bias, we excluded from the study patients unable to respond to the entire test (5% of FGIDs patients) and patients with organic diseases. Data are based on the other 608 patients.
The MMPI has been commonly used to assess medical patients (chronic pain [22], chronic pelvic pain [23], premenstrual symptoms [24]) and to explore patients with digestive disorders of known (inflammatory bowel disease), or of unknown origin: dyspepsia [25], nutcracker esophagus [26], slow transit constipation [7], or IBS [27-30].
In previous clinical studies, IBS patients were compared to healthy controls 28 or in patients with extra digestive pain [7]. The present study uses other FGIDs patients as the control group for 2 reasons. First, MMPI scales used are T scores with a known distribution in the healthy population (mean ± SD, 50 ± 10), according to the gender and age; second, other FGID patients suffer from another chronic digestive disorder. The main superiority of the MMPI on other psychological tests is the presence of validity scales to validate the questionnaires [31]. Nevertheless, the high correlation between the MMPI clinical scales limits the power of this test [32]. The improved version of the MMPI, the MMPI-3, which has a slightly different item pool, could correct this limitation.
The MMPI-2 uses 6 validity scales, but the Infrequency scale is significantly different between IBS patients and non-IBS patients. The F scale, defined by 60 items of the MMPI-2 test, is used to identify attempts at “faking good” or “faking bad.” High scores on this item characterize people that are trying to appear better or worse than they are. F score is higher in IBS patients than in non-IBS patients. Abnormal F score is found more often in IBS patients than non-IBS patients (n = 22 [17.6%] vs. n = 17 [7.7%]; P= 0.009), but there is no difference among the IBS subgroups (P= 0.670).
The other validity scales were not different between IBS and non-IBS patients (Table 2). The L scale was developed to detect patients that positively present themselves. Despite the mean L scores are higher than the normal range (45–55), this score was not different between IBS and non-IBS patients, nor between the different IBS phenotypes. L scores between 56 and 64, indicative of individuals who have a tendency to resort to denial mechanisms, or are more conforming than usual, was observed in 35% of IBS patients and 32% of non-IBS patients (P= 0.425).
The K scale measures defensiveness and assesses the willingness of the client to disclose personal information and to discuss his/her problems. The K scale is lower in IBS patients than in non-IBS patients, but it does not change with the IBS phenotypes.
In the present study, all clinical scales were different between IBS and non-IBS patients. Nevertheless, after multivariate analysis, only 2 scales remained significant: Hypochondriasis and Masculinity-Femininity.
The Supplementary Text contains the discussion of results on content, supplementary, and Harris-Lingoes scales.
Surprisingly, although all clinical scales were higher in IBS patients than in the other FGID patients, only 1 validity scale (F) and 2 clinical scales (Depression and Hypomania) varied significantly according to the IBS subtype. In a previous study using IBS subgroups clinically and not Rome III defined subgroups, it was found, contrarily to our study, only significant differences on scales K, F, and Social introversion [33]. Nevertheless, post hoc analysis revealed that for scale K, the IBS-D group scored higher than both the IBS-C and IBS-M groups. In contrast, while on scale Social introversion the pain-diarrhea group only scored higher than the pain-constipation group, but this study did not use the Rome III criteria, and did not include logistic analysis. In the present study, the different IBS phenotypes are characterized by specific psychological profiles as compared to non-IBS patients or IBS-U patients, despite the main differences are found between IBS and non-IBS FGID patients.
The IBS-C group has higher values for the scales K, Anxiety, and lower values for the scales Schizophrenia, and Masculine gender role than non-IBS patients. These patients have a defensive approach to the test and lack of confidence. As reported in previous studies, we found in the IBS-C group, a tendency to exaggerate physical symptoms [34] related to a psychological need to exaggerate and complain, and increased preoccupation about one’s health. The IBS-C group also has a higher frequency of patients with abnormal Hypochondriasis scale (see Supplementary Text).
As previously found in a limited group of Iranian IBS patients [35], IBS-D patients have a higher value for the Infrequency scale, in favor of overall psychopathology, resentment, or moodiness.
By comparison with the non-IBS group, IBS-M patients have lower values for the VRIN scale, the best measure of random, or inconsistent responding. They also have a lower Masculine gender role scale, as previously described in 70 non-subtyped male IBS patients [36], a lower Fears scale, lower Low self-esteem scale, and a larger Psychasthenia scale. In this group, a higher frequency of patients with abnormal Hypochondriasis scores was found (see Supplementary Text). Previously, it was found that IBS-M patients reported a higher number of somatic symptoms than IBS-C and mostly IBS-D patients [37]. These results could explain the results of the present study.
However, as compared to the non-IBS patients, only IBS-C and IBS-M patients are associated with an increased frequency of abnormal Hypochondriasis score. The frequency of abnormal Hypochondriasis score could be associated with the high frequency of overlapping disorders in IBS patients [38,39].
One major limitation of the present study is that the studied population is a population referred to a tertiary center focusing on FGIDs. Consequently, the expectation would be that there would be a higher level of co-morbid psychiatric distress. The present results could not be predictive of similar findings in a community population.
The cause-effect relationship between abnormal psychological traits and IBS symptoms remains controversial. Our data and other reports illustrate that abnormal psychological profiles do exist in IBS patients seeking medical help [40]. On the other hand, Drossman et al. [13] reported that psychological profiles in IBS nonpatients and subjects showing IBS symptoms but never seeking medical help were not different from the psychological profiles of healthy controls. From these observations, it was proposed that abnormal psychological behaviors are closely related to IBS patients seeking medical help. However, the control population of our study was composed of FGIDs patients with the same history of chronic functional digestive disease.
This study has some clinical implications. The present results tend to associate IBS to specific psychological profile rather than to a pattern of health care seeking pattern but also shows that abnormal scales are found in only a minority of IBS patients. In consequence, the use of antidepressant drugs must be proposed to only a limited percentage of patients as well as tricyclic or selective serotonin reuptake inhibitor antidepressants [41]. Besides, the difference classically performed between tricyclic antidepressants in case of diarrhea and abdominal pain, and selective serotonin reuptake inhibitor antidepressants for IBS patients with abdominal pain and constipation must be reevaluated [42], mainly in IBS-U patients. Thus, the tailored management of IBS patients must be the rule, including medical treatment of the predominant cardinal symptom [43]: constipation, diarrhea, bloating, and abdominal pain in combination with dietary and psychological interventions.
To conclude, the present study shows that IBS patients, as compared with non-IBS patients, have significant increases in all clinical MMPI scales. Furthermore, we found significant differences in the psychological profiles according to the different IBS phenotypes.
XML Download