| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Departments of Medicine and Psychiatry (D.A.D., J.L., Z.L., Y.J.B.H., T.C.T.), Division of Digestive Diseases, and UNC Center for Functional GI and Motility Disorders, University of North Carolina, Chapel Hill, NC; and Department of Psychology (F.K.), Duke University, Durham, NC.
Address reprint requests to: Douglas A. Drossman, MD, Division of Digestive Diseases, 726 Burnett-Womack Bldg., CB 7080, University of North Carolina, Chapel Hill, NC 27599-7080. Email: drossman{at}med.unc.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
OBJECTIVES: The primary objective was to study the effect of different coping strategies on the health outcome of women with GI disorders and how these coping strategies may modify the effects of education, GI disease type, neuroticism, and abuse severity on health outcome.
METHODS: We followed 174 patients in a referral GI clinic for 12 months to assess their health status as a derived variable of daily pain, bed disability days, psychological distress, daily dysfunction, number of visits to physicians, and number of surgeries and procedures. We obtained at baseline their GI diagnosis (functional vs. organic), neuroticism score (NEO Personality Inventory), sexual and/or physical abuse history, and scores on two coping questionnaires. Regressions analyses were used to determine the relative effect of the coping measures on health outcome and their modifying effects on education, GI disease type, neuroticism, and abuse severity.
RESULTS: A higher score on the Catastrophizing scale and a lower score on the Self-Perceived Ability to Decrease Symptoms scale (Coping Strategies Questionnaire ) predicted poor health outcome. Less education, a functional GI diagnosis, a higher neuroticism score, and greater abuse severity also contributed to poor health status. However, the effect of GI disease type and neuroticism on health outcome was significantly reduced by the coping measures.
CONCLUSIONS: Maladaptive coping (eg, catastrophizing) and decreased self-perceived ability to decrease symptoms may adversely affect health outcome and may modify the effect of GI disease type and neuroticism on health outcome.
Key Words: coping • health status • outcome • gastrointestinal disorders • psychosocial factors
Abbreviations: CSQ = Coping Strategies Questionnaire; GI =gastrointestinal; GSI = General Severity Index; IBD =inflammatory bowel disease; NEO scale = neuroticism extraversionopenness scale; SCL-90-R = Symptoms Checklist 90 Revised; SIP= Sickness Impact Profile; WCQ = Ways of CopingQuestionnaire.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
An important psychosocial predictor, not well studied in gastroenterology, relates to the patient’s coping strategies. Coping is defined as "efforts, both action-oriented and intrapsychic, to manage (ie, master, tolerate, minimize) environmental and internal demands and conflicts, which tax or exceed a person’s resources" (3). Coping may have adaptive or maladaptive effects on an individual’s health status (4).
Two standardized measures of coping commonly used to evaluate how individuals with medical conditions respond to their illness or painful symptoms are the WCQ and the CSQ. The WCQ, developed by Lazarus and Folkman (3), contains eight coping scales that are broadly classified into problem-focused coping and emotion-focused coping. Problem-focused coping involves trying to deal directly with the situational stressor by changing the stressor or oneself (eg, by seeking social support or reappraising the stressor). Emotion-focused coping involves managing the distressing emotions evoked by the situation or condition (eg, by praying or denial). In general, problem-focused coping is used in situations appraised by patients as changeable, and emotion-focused coping is used in situations not amenable to change.
The CSQ, developed by Keefe et al. (5, 6), focuses primarily on coping in response to painful conditions. It measures six different cognitive pain coping strategies and one behavioral strategy. In particular, one subscale, catastrophizing (eg, "I feel it’s never going to get any better" or "I feel my life isn’t worth living") correlates strongly with depression and has been shown in other clinical populations to have significant negative effects on health status (5–7).
The purpose of this study was to determine prospectively the relationship of baseline coping strategies on health outcome measured during a subsequent year. Furthermore, we were interested in the extent to which these coping strategies might mediate the effect of such variables known to affect health outcome, such as education, neuroticism, GI disease type, and history of abuse, in a female GI referral population. In addition, we included the two described coping measures to determine, for future investigation, which measure would be more effective in predicting health outcome. The study population for this study was confined to women because the focus was sexual abuse and functional GI disorders, both of which are more prevalent among women.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
All 239 subjects in the baseline study were followed for 1 year, during which time they received additional questionnaires to assess health outcome over time. At 3, 6, 9, and 12 months after the study visit, each subject filled out a healthcare utilization form. A total of 222 patients (93%) returned at least one follow-up questionnaire, 196 (82%) returned at least two, 170 patients (71%) returned at least three, and 138 (58%) returned all four. In addition, at 6 and 12 months, subjects completed the SIP (10) to assess daily functional status and quality of life, the SCL-90-R (11) to evaluate psychological distress, and a 2-week daily symptom diary. One hundred eighty-two patients (76%) returned at least one of these two mailings, and 141 patients (59%) returned both mailings. To be included in the study, subjects had to have completed at least two follow-up measurements for measures administered four times and at least one follow-up for those given twice. One hundred seventy-four subjects (72.8% response rate) met these requirements and thus made up our sample.
A comparison of respondents (N = 174) and nonrespondents (N = 65) of the follow-up mailing indicated that there were no differences in age, education, GI disease type, neuroticism scores, and abuse severity. The group of nonrespondents, however, had a higher proportion of nonwhites (27.7% vs. 12.1%, p = .004) than the group of respondents. There were no differences between these two groups with regard to baseline coping variables (defined below), except that nonrespondents had a higher mean score on the Escape-Blame scale (9.9 vs. 8.0, p = .02) and a higher score on the Positive Reappraisal scale (17.6 vs. 15.5, p = .03). Adjustments for differences in race were made in the multivariate data analysis.
Measurement
All predictor variables were assessed at baseline. The health outcome variable was measured over the 1-year follow-up period.
Demographics and physician’s diagnosis.
Patients were asked their date of birth, race (white or nonwhite), and years of formal education. A patient’s physician was asked to identify the patient’s main GI complaint; for patients with more than one GI diagnosis, the principal GI diagnosis was used. Missing or ambiguous diagnoses were completed by the first author’s (D.A.D) discussion with the physician or by chart review. The GI diagnoses were grouped into two major categories: functional, based on the Rome Diagnostic Classification System for Functional Gastrointestinal Disorders (12, 13), including globus sensation, rumination, functional chest pain, dysphagia, dyspepsia, abdominal bloating, constipation, diarrhea or abdominal pain, aerophagia, irritable bowel syndrome, incontinence, anorectal pain, or dyschezia/pelvic floor dyssynergia; and organic, including ulcerative colitis, Crohn’s disease, acid peptic disease, liver disease, pancreatic/biliary disease, and other organic diagnoses.
Neuroticism.
Because neuroticism, a personality trait, is associated with poor health status in this population (14, 15), we obtained a neuroticism score at baseline using the revised NEO Personality Inventory (16).
Sexual and physical abuse history.
Because sexual and physical abuse history is a strong predictor of poor health status (8, 17–19), we also studied the contributory affect of abuse along with coping on health outcome, using a previously developed structured interview (19, 20). Two female psychology graduate students, trained by the principal investigator and coinvestigators (D.A.D., J.L., T.C.T.) to administer and score the interview, conducted the interview during the study visit. From these data, we used a previously validated abuse severity scale to quantify the severity of the abuse experienced (9). The score ranges from 0 (none) to 6 (severe) and is the sum of 1) invasiveness of sexual abuse (0 = no invasive sexual abuse, 1 = touching/fondling, 2 = rape), 2) degree of injury during abuse (0 = no or minor injury, 2 = serious physical injury), and 3) number of life-threatening physical assaults (0 = none, 1 = one to three, 2 = four or more).
Coping.
Two coping questionnaires were administered during the study visit.
Ways of Coping Questionnaire.
The original WCQ developed by Lazarus and Folkman (3) is organized into eight coping scales to assess problem-focused and emotion-focused coping. Subjects are asked how they cope with a stressor. The instructions we used were as follows: "Let’s assume your GI symptoms worsen until they are almost unbearable, and you are no longer able to carry on your usual activities. You are informed that the problem may continue indefinitely. Please describe how you would respond to the situation." The WCQ was developed for a nonclinical population. In our study, the eight coping scales were found to have low reliability (Cronbach’s 
= 0.51–0.75) and had no effect in predicting health outcomes in a preliminary analysis of the data. Therefore, we performed factor analysis to determine whether other factors might better represent this population. In other words, we sought to develop a more condition-specific coping measure applicable to the gastroenterology population. We used principal component analysis to determine how the 50 items included in the eight ways of coping scales were clustered. Using varimax-rotated factor weights greater or close to 0.4 as criteria, 44 items were grouped into four factors, which explained 36.1% of the total variance in the matrix. On the basis of clinical features, the four factors were named "escape-blame," for example, blaming oneself and trying to escape or deny the problem (12 items, = 0.77); "avoid feelings," for example, avoid thinking or feelings about the problem (10 items, = 0.59); "positive problem solving," for example, make a plan and take action about the problem (11 items, = 0.77); and "positive reappraisal," for example, change or grow from the experience (11 items, = 0.80). The first two factors are emotion-focused, and the last two are problem-focused. We created scales by computing the mean score of items for each coping factor (0 = "would not apply," 3 = "would apply a great deal"). Cronbach’s values showed that the four revised scales had better internal consistency than the original eight.
Coping Strategies Questionnaire.
The CSQ is widely used to measure how an individual copes with chronic pain (21). The questionnaire contains seven scales of coping strategies. For the purpose of this study, we used only the Catastrophizing scale, which contains six items that indicate maladaptive negative cognitions (eg, "It’s terrible and I feel it’s never going to get any better" or "I worry all the time about whether it will end"). The items are scored on a seven-point Likert scale (0 = "never think or feel that," 6 = "always think or feel that"), and the scores are summed to range from 0 to 36.
In addition, we included two single-item questions used in the CSQ to rate overall coping strategy effectiveness: "Based on all the things you do to cope or deal with your symptoms, on an average day, how much control do you feel you have over it?" (amount of control over symptoms); and "Based on all the things you do to cope or deal with your [gastrointestinal] symptoms, on an average day, how much are you able to decrease it?" (self-perceived ability to decrease symptoms). Response scores ranged from 0 ("no control" or "inability to decrease it") to 6 ("complete control" or "can decrease it completely").
We chose these components because previous research found that amount of control over symptoms and self-perceived ability to decrease symptoms were positively loaded, and the catastrophizing subscale was negatively loaded on a single coping factor (22). This factor had strong predictive effects on health outcome among patients with painful conditions (6, 22–24). Test-retest reliability varied from 0.80 to 0.91 for catastrophizing, from 0.52 to 0.81 for amount of control over symptoms, and from 0.43 to 0.75 for self-perceived ability to decrease symptoms (21).
Health status.
Quarterly or semiannual mailings assessed health status over a 1-year period. This was done to improve the reliability of these measures and to examine how coping may affect subsequent health status. Mean scores for current pain, psychological distress, and daily dysfunction were calculated using data collected at 6 and 12 months, and sums for number of bed disability days, physician visits, and surgeries and procedures were calculated using data collected at 3, 6, 9, and 12 months. The number of bed disability days, physician visits, and surgeries and procedures for subjects with one or two missing data points was computed by substituting the mean of nonmissing values in calculating the sum for 1 year. Subjects missing more data points were excluded from the study.
To have a single outcome measure, we created a composite health status score, overall poor health, from six health status variables (current pain, bed disability days, psychological distress, daily dysfunction, number of physician visits, and number of surgeries and procedures) that are clinically consistent with the expected effects of chronic or severe GI illnesses. A linear T score was first collected for each of the six variables, and then the composite variable was derived as the mean of the six T scores. For all measures, 0 was considered optimum or good health status. The Cronbach’s value of the six variables (below) for the follow-up sample was 0.86, indicating good internal consistency.
The six health status measures used by our research group (8, 19, 25) are described below.
Current pain.
Subjects prospectively filled out daily diary cards for 2 weeks during months 6 and 12. The diary card asked the subjects to "Place a vertical mark that indicates the amount of pain you felt today" on a 100-mm visual analog scale (0 = "none," 100 = "very severe"). Scores from each of the 14 days were averaged.
Bed disability days.
The number of bed disability days was determined quarterly. Subjects were asked, "How many days during the past 3 months did you stay in bed (more than half of the day) because of illness? Include any days you stayed in the hospital." This question was adapted from the Epidemiological Catchment Area Survey for North Carolina (26).
Psychological distress.
Psychological distress was measured using the GSI of the SCL-90-R (11) at months 6 and 12. The SCL-90-R consists of 90 items that measure the following psychological symptoms: somatization, obsessive-compulsive neurosis, interpersonal sensitivity, depression, anxiety, hostility, paranoid ideation, phobic anxiety, and psychoticism. For this study, we used the GSI summary raw score (mean of 90 items); the score for each item ranged from 0 ("not at all") to 4 ("extremely").
Daily dysfunction.
Daily dysfunction was obtained using the overall score of the SIP (10) at 6 and 12 months. The SIP consists of 136 items that assess 12 different categories of behavioral dysfunction: work, eating behavior, recreation and pastimes, sleep and rest, alertness behavior, home management, emotional behavior, social interaction, body care and movement, mobility, ambulation, and communication.
Number of visits to physicians and number of surgeries and procedures.
Number of visits to physicians and number of surgeries and procedures were obtained by quarterly mailings. Each subject was asked to provide the number of visits to a physician and the number of surgical or endoscopic procedures received in the previous 3 months.
Statistical Analysis
Data were entered and analyzed using SAS software (SAS Institute, Cary, NC) in several steps. First, we obtained summary statistics of the sample’s demographics, GI diagnosis, neuroticism, abuse history, coping strategies, and health status. We computed correlations among coping variables and neuroticism and t test differences between patients with functional and organic disorders. Next, we ran three multiple regression models in the following hierarchical order to determine the effect on health status (overall poor health): 1) demographic variables (age and race); 2) background history variables (education, GI disease type, neuroticism, and abuse severity); and 3) the seven coping variables, including the four revised WCQ scales and the three CSQ measures. Incremental R2 values for the four background history variables and for the seven coping variables were computed in models 2 and 3, respectively, to determine the explanatory effects of these predictors and the extent to which the effects of the background history variables were modified by the seven coping variables. We ran collinearity statistics, and there was no obvious collinearity between predictors in our final model; a regression model run with only the significant predictor variables showed results identical to those of the full model. Results of the regression models are presented with standardized parameter estimates, p values pertaining to the estimates, R2 values of the models, and incremental R2 values of the four background history variables and the seven coping variables. All tests were two-tailed, and p < .05 was used to determine the significance level.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
|
The average scores of our sample on the catastrophizing (10.1) and self-perceived ability to decrease symptoms (3.1) measures were similar to those of patients studied in pain clinics (range, 8.4–13.8 and 2.6–3.0, respectively). The amount of control over symptoms score for our sample (3.4) was somewhat better than that of patients in pain clinics (range, 2.2–2.5) (21).
Health outcome.
The health outcome of the sample was generally poor. Subjects had an average daily pain score of 24 (on a scale of 0–100), and, during 1 year, they spent more than 27 days in bed because of illness, visited a physician approximately 14 times, and had nearly one surgical and/or diagnostic procedure. The mean score on the GSI of SCL-90-R was 0.7 (SD, 0.6), compared with 0.54 for ulcerative colitis patients in another study (27) and 0.38 in a national database of persons with IBD (28), and equaled approximately the 86th percentile of a national normative sample of women (11). The mean overall score on the SIP was 8.9 (SD, 9.8), compared with 7.4 for the national IBD database (28).
Correlational Analyses
The correlations among coping variables and neuroticism ranged from 0 to 0.64, as shown in Table 2. Neuroticism was most highly related to coping by trying to escape the problem or blaming oneself (r = .43) and catastrophizing (r = .54); these two coping scales (escape-blame and catastrophizing ) were highly correlated (r = .42). The two problem-focused coping scales, positive problem solving and positive reappraisal, were also highly related (r = .49). Catastrophizing was associated with feeling less in control over symptoms (r = .35) and less ability to decrease symptoms (r = .36). The latter two CSQ scales were highly correlated (r = 0.64). Note also that patients with functional disorders were slightly less likely to feel in control of symptoms (p = .001), to feel able to decrease symptoms (p = .002), and to use positive reappraisal (p = .02) and more likely to catastrophize (p = .001) than patients with organic disorders.
|
|
In model 3, we also added the seven coping variables, which explained an additional 13% of the variance in overall poor health. Three coping variables were statistically significant. Poor health was associated with more problem solving (p = .05), more catastrophizing (p = .04), and less self-perceived ability to decrease symptoms (p = .003). Nearly 41% of the total variance of overall poor health was explained by all the variables in the model.
When we compared models 3 and 2, we found that the effect of the four background history variables (education, GI disease type, neuroticism, and abuse severity) on health outcome was reduced by the coping variables. With the coping variables in the model, the power of these four background history variables in explaining overall poor health dropped by more than half (variance of overall poor health explained by these four variables changed from 27.7% in model 2 to 12.9% in model 3). GI disease type and neuroticism were no longer significant. Education remained marginally significant. Abuse remained highly predictive (p < .0001), although its effect on overall poor health was marginally reduced (standardized parameter estimate dropped from 0.39 to 0.34).
Because of a possible interaction effect between neuroticism and abuse severity, we tested the interaction between these two variables but did not find it to be significant. We also examined interactions of disease type and coping variables, and none were significant. In addition, we also ran our final model excluding psychological distress (GSI of SCL-90) from the health outcome measure to determine whether our results would still hold if based on physical health only. The results were identical with and without psychological distress in the health outcome measure.
In summary, the Positive Problem Solving scale of the revised WCQ and the Catastrophizing and Self-Perceived Ability to Decrease Symptoms scales of the CSQ were strong predictors of poor health outcome among female GI patients. These coping variables seemed to modify the effect of GI disease type and neuroticism on health outcome. In addition, abuse severity remained a strong predictor of adverse health outcome.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
Overall, the predictor variables selected (demographic factors, functional vs. organic GI diagnosis, neuroticism, abuse severity, and coping measures) explained 41% of the variance of health outcome. The outcome variable was derived from several measures recognized as clinically important in chronic or severe GI illness: pain, bed disability, psychologic distress, daily function, number of physician visits, and number of surgeries and procedures. Health status was primarily explained by the effects of abuse history, more positive problem solving, more catastrophizing, and less perceived ability to reduce symptoms. The coping measures may also explain the adverse consequences of having a functional GI disorder (8) and of scoring high on the neuroticism measure (15), because their effects were no longer significant after coping variables were included in the regression model.
Few previous gastroenterology studies have used standardized measures to assess the effects of coping on health status, and none of these examined health status prospectively. Using the WCQ (3), our group studied persons with IBD belonging to the Crohn’s and Colitis Foundation of America (28). The members dealt with the stress of their disease predominantly through problem-focused strategies: seeking social support, problem solving, and positive reappraisal. Additional analyses indicated that these strategies helped to reduce psychological distress and dysfunction in this population. Similarly, in a study using a different coping measure, patients with IBD were more likely to use problem-focused strategies than emotion-focused strategies and were satisfied with the way they were coping and their life in general (30). The results of these studies suggest that, at least for the IBD population, problem-focused strategies may be more adaptive in reducing the adverse psychosocial consequences of the illness and improving health status.
Contrary to our expectations, we found that in patients with worsening GI symptoms, higher positive problem-solving scores were related to poorer health. To explain this finding, we hypothesized that those who used more positive problem solving (eg, "Talk to someone to find out more about the situation" or "Talk to someone who could do something concrete about the problem") might seek more healthcare advice and treatment. This could increase the number of physician visits and possibly the number of surgeries and procedures, two components of the overall poor health outcome measure. To test this hypothesis, we repeated model 3, using a dependent variable from which the number of physician visits and number of surgeries and procedures were removed. We found that the positive problem-solving score was marginally changed (p = .06, NS). Perhaps patients who react to their symptoms with hypervigilance also tend to be vigilant about planning and problem solving when it comes to their illness.
The CSQ is a process-oriented measure that evaluates cognitions in relation to pain. In particular, the Catastrophizing scale includes negative self-statements and thoughts about the future in which the patient unrealistically assumes that the worst possible outcome will occur. This strategy occurs in patients with high levels of pain and functional disability (6) and, on the basis of studies of patients with chronic pain, seems to be a separate construct from depression, another condition also negatively associated with outcome (7). Furthermore, reductions in the catastrophizing score correlate with improvements in physical and psychological function (6).
The Catastrophizing scale has been studied primarily among patients with arthritis and musculoskeletal pain, in whom it is associated with poorer health status. Within gastroenterology, one study evaluated the relationship between somatic pain perception, sexual or physical abuse, psychiatric morbidity, and scores on the Catastrophizing scale of the CSQ among 50 women referred to a tertiary-care center with painful GI disorders (gastroesophageal reflux disease, functional chest pain, and irritable bowel syndrome) (31). Patients with a history of abuse had significantly lower pain threshold levels in response to finger-pressure stimuli, possibly because they had lower cognitive standards for judging stimuli as noxious (response bias). They also had poorer daily function, more psychiatric disorders, and higher catastrophizing scores than nonabused patients with these GI disorders. The authors proposed that a history of abuse might contribute to increased pain, psychologic distress, and hypervigilance for noxious stimuli. The painful symptoms may then lead to functional disability and greater healthcare seeking and may also alter personal beliefs and coping strategies in relation to the pain, thereby leading to catastrophizing.
Another purpose of the study was to determine, within a gastroenterology population, which of the two coping measures were more predictive of outcome. Therefore, we evaluated the WCQ, which was subjected to factor analysis and rescaled for use in a GI population, and subscales of the CSQ, which measures factors known to have the most significant effects on health status (5, 6, 24). We found that the strongest predictive effect was obtained with use of the CSQ, particularly the six-item Catastrophizing scale and the additional item relating to one’s self-perceived ability to decrease symptoms. These findings are similar to those obtained in patients with chronic pain and rheumatoid arthritis (5, 6, 24). In contrast, only one of the revised WCQ scales, Positive Problem Solving, had a marginal effect in predicting health outcome. These results lead us to recommend that future studies of coping in this population include use of the brief Catastrophizing scale of the CSQ and the self-perceived ability to decrease symptoms question.
The results of this study have important clinical implications. Toner (32) proposed that the way individuals think about their symptoms may have important effects on their health. As shown in this study, a patient who has a profoundly pessimistic view of her illness (ie, catastrophizing) or who perceives little ability to decrease the symptoms is likely to have a poorer outcome. This effect may be enhanced if the patient has a history of abuse, possibly because this type of experience is commonly associated with a perceived inability to effect positive consequences in the environment or because abuse leads to hypervigilance to bodily sensations with an overinterpretation of their significance. We note that abuse and maladaptive coping are the most important predictors of health outcome.
The findings also lend support to the use of psychological treatments like cognitive-behavioral treatment or other techniques that seek to modify these maladaptive cognitions. In one controlled study of cognitive-behavioral treatment for patients with irritable bowel syndrome (33), eight 2-hour group treatment sessions over a 3-month period led to an increase in the number of effective cognitive and behavioral coping strategies and a concurrent reduction in abdominal complaints. Furthermore, improvement continued at the 2-year follow-up examination. We believe that cognitive-behavioral treatment should be considered for patients in whom maladaptive cognitions and coping interfere with the patient’s adjustment to the GI illness, regardless of diagnosis.
Some limitations need to be considered in interpreting these data. First, because this study was performed at a referral center, the findings may not be generalizable to all GI patients seen in other clinical settings (eg, in primary care). The frequency of patients with psychosocial disturbances and a history of abuse is lower in primary-care and nonclinical settings compared with tertiary-care settings (34–36), so fewer patients in those settings may exhibit maladaptive coping strategies. In this study, we controlled for age, education, neuroticism, and abuse history, so the findings would apply at least to female patients seen in other referral centers. Furthermore, even if psychosocial disturbance is less frequent in nonreferral populations, when present, it should not minimize the effect of maladaptive coping or abuse history when present.
Second, we used a health status composite variable derived from six conceptually related variables, and this may be more difficult to interpret than looking at several outcome variables sequentially. However, realizing that measurement errors are inevitable, an increasing number of investigators, especially those in social or psychological fields, have used latent or composite variables derived from a group of measures of a similar nature. For example, the popular 36-item Short-Form Health Survey (SF-36) is a multiple-item instrument encompassing eight concepts of health (37). Another example is the Crohn’s Disease Activity Index (38). This composite measure, which is widely used in Crohn’s disease studies, consists of several disparate variables: stool frequency, abdominal pain, sense of well-being, disease-related complications, use of opiates, abdominal mass, hematocrit, and body weight. The advantages of using a composite measure include 1) reduction of the impact of measurement errors, because a composite variable generally has better distribution than the individual variables; and 2) reduction of Type I error resulting from multiple comparisons. The results of our study are preliminary, because we need more research to validate the health outcome composite measure.
Third, we cannot establish causal links among neuroticism, functional disorders, and coping because we do not have longitudinal data on these constructs. We posit that the coping measures explain the adverse health consequences of having a functional GI disorder and of scoring high on neuroticism. It may be that neuroticism and maladaptive coping increase the probability that one will develop a functional disorder and subsequent poor health. Additional research is needed to disentangle these relationships.
Fourth, the study did not involve men. It needs to be determined whether such psychosocial factors as maladaptive coping, abuse, or neuroticism are as common in men seen in GI practices as in women and whether these factors have similar effects on outcome in men. We believe that such a study is warranted.
Finally, the intensity of a 1-day study visit might burden some study subjects. To reduce respondent fatigue, we took the following measures: 1) provided breaks, including an hour-long lunch break, during the study period; 2) split a long session into shorter ones if needed; 3) alternated difficult sessions with easy ones; and 4) shortened visits by allowing a subject to return for another visit to complete the questionnaires if needed.
In conclusion, we found that among women referred for GI disorders, maladaptive coping (catastrophizing), decreased self-perceived ability to decrease symptoms, and a history of abuse significantly predict poor health outcome. We recommend that these factors be included in future studies. Finally, future studies will need to determine the degree to which modification of these effects through psychological interventions, such as cognitive-behavioral treatment, will improve medical and psychological symptoms and clinical outcome.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
Received for publication July 20, 1999.
Revision received October 1, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. M. Heitkemper and M. E. Jarrett Update on Irritable Bowel Syndrome and Gender Differences Nutr Clin Pract, June 1, 2008; 23(3): 275 - 283. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. A. Sansone, M. Butler, H. Dakroub, and M. Pole The Relationship Between Number of Types of Childhood Trauma, Number of Psychophysiological Disorders, and Length of Medical Disability Traumatology, June 1, 2006; 12(2): 143 - 147. [Abstract] [PDF]
|
|
|
|
 |
|
 C. L. Kwan, N. E. Diamant, G. Pope, K. Mikula, D. J. Mikulis, and K. D. Davis Abnormal forebrain activity in functional bowel disorder patients with chronic pain Neurology, October 25, 2005; 65(8): 1268 - 1277. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Creed, E. Guthrie, J. Ratcliffe, L. Fernandes, C. Rigby, B. Tomenson, N. Read, and D. G. Thompson Reported Sexual Abuse Predicts Impaired Functioning but a Good Response to Psychological Treatments in Patients With Severe Irritable Bowel Syndrome Psychosom Med, May 1, 2005; 67(3): 490 - 499. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Kendall-Tackett Exciting Discoveries on the Health Effects of Family Violence: Where We Are, Where We Need to Go J Interpers Violence, February 1, 2005; 20(2): 251 - 257. [Abstract] [PDF]
|
|
|
|
 |
|
 L. A. Martinez Self-Care for Stoma Surgery: Mastering Independent Stoma Self-Care Skills in an Elderly Woman Nurs Sci Q, January 1, 2005; 18(1): 66 - 69. [Abstract] [PDF]
|
|
|
|
|
|
J. M. Lackner, B. M. Quigley, and E. B. Blanchard Depression and Abdominal Pain in IBS Patients: The Mediating Role of Catastrophizing Psychosom Med, May 1, 2004; 66(3): 435 - 441. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
