SEARCH

SEARCH BY CITATION

Keywords:

  • inflammatory bowel disease;
  • Crohn's disease;
  • ulcerative colitis;
  • disease activity;
  • quality of life;
  • psychosocial issues

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES
  7. APPENDIX:

Background: The aim was to assess quality of life (QOL) and psychological functioning in inflammatory bowel disease (IBD) as related to patterns of disease activity over time.

Methods: Study participants were 388 recently diagnosed individuals from the population-based Manitoba IBD Cohort Study. They completed mail-out surveys at 6-month intervals and clinical interviews annually. Based on their 2-year pattern of self-reported disease activity, participants were assigned to 1 of 3 groups: consistently active, fluctuating, or consistently inactive disease. Disease type (Crohn's disease [CD] or ulcerative colitis [UC]) was confirmed through chart review. Change over time was modeled for measures of QOL and positive and negative psychological functioning using mixed-effects regression analyses.

Results: Half of the participants had fluctuating disease activity, while almost one-third of participants reported consistent active disease. Participants with the fluctuating activity pattern showed significant improvement in disease-specific QOL compared to participants with consistent activity. Perceived stress, health anxiety, and pain anxiety decreased while pain catastrophizing and mastery increased over time, although the amount of change was not significantly different among disease activity patterns. However, when the data were averaged over time there were significant differences among disease activity patterns on most outcomes. Significant effects of CD versus UC were observed only for the pain measures.

Conclusions: Change in IBD QOL is influenced by one's longitudinal profile of disease activity, but change in psychological functioning is not. Effects of disease activity on psychological functioning were modest, suggesting that disease has an impact even when patients are not experiencing active symptoms.

(Inflamm Bowel Dis 2008)

Inflammatory bowel disease (IBD) is a chronic illness characterized by periods of remission and disease exacerbation. As with many diseases, the clinical expression of IBD cannot be fully accounted for by pathophysiology. A biopsychosocial understanding of illness describes clinical outcome and disease exacerbation as influencing and strongly influenced by both biological and psychosocial factors.1, 2

Individuals with IBD have lower quality of life (QOL) than the general population, as well as diminished psychological functioning and well-being,3–6 Moreover, psychosocial outcomes tend to be poorer during periods of active than of inactive disease.7–11 However, there has been little investigation of how variation in disease activity over time is associated with psychological function and QOL, which should help to further clarify these relationships.

Blondel-Kucharski et al3 collected repeated measurements of QOL for patients with Crohn's disease (CD) over a 1-year period with assessments at 3-month intervals. There was significant improvement in QOL, which the authors attribute to lower rates of disease activity at study endpoint, although this association was not directly tested. Porcelli et al6 compared patients with CD and ulcerative colitis (UC) at just 2 timepoints in a 6-month period. Regardless of disease type, individuals for whom disease activity increased were more likely to report increased anxiety and depression, while individuals for whom disease activity decreased over time reported reduced psychological distress.

Using data from the population-based Manitoba IBD Cohort Study, this study examines the trajectory of change for several IBD outcomes across multiple occasions in a 2-year period, focusing on sustained periods of active and inactive disease and their relationship to improving or worsening QOL and psychological functioning. Based on previous research, we hypothesized that change in QOL and psychological functioning would be sensitive to disease activity, but that there would not be differences between individuals with different types of disease (i.e., CD versus UC). On the one hand, while active disease might negatively impact on QOL and psychological functioning, the ability to psychologically adapt to active disease could minimize its impact over time.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES
  7. APPENDIX:

The Manitoba IBD Cohort Study is a prospective longitudinal study to identify the determinants of disease outcomes across multiple domains. The study was initiated in 2002, drawing on subjects from the University of Manitoba IBD Research Registry, a population-based registry established in 1995. A description of the methods used to establish the Registry has been published previously,10 and is therefore not repeated here. The Cohort Study was approved by the University of Manitoba Health Research Ethics Board.

Individuals diagnosed within 7 years (to capture relatively early disease) and who were 18 years of age or older (N = 606) were identified from the Registry and invited to join the study. A total of 418 (69.0%) individuals agreed to participate. Following the initial contact, 19 individuals did not complete both the baseline survey and interview components, another 4 individuals withdrew, and 7 individuals were found to be ineligible, resulting in a final sample of 388 participants for longitudinal follow-up. Data were collected using standardized self-report instruments distributed by mail at 6-month intervals and interviews conducted by clinic staff at 12-month intervals. This study focuses on data collected between the baseline wave (i.e., month 0) and the 24-month wave, giving a total of 5 measurement occasions. During this observation period, 34 subjects dropped out of the study, an attrition rate of 8.8%.

Chart review by study staff was used to confirm patient report of disease type following the baseline wave of data collection. Self-report data on demographic characteristics (e.g., age, sex, ethnicity, marital status) were also collected at baseline.

Disease activity was based on patient report of IBD symptom persistence for the previous 6 months using a 6-point response format: “In the past 6 months my disease has been (a) constantly active, giving me symptoms every day (b) often active, giving me symptoms most days (c) sometimes active, giving me symptoms on some days (for instance 1–2 days/week) (d) occasionally active, giving me symptoms 1–2 days/month (e) rarely active, giving me symptoms on a few days in the past 6 months and (f) I was well in the past 6 months, what I consider a remission or absence of symptoms.” Using a previously published method,10 individuals were classified as having active or inactive disease at each measurement occasion. Active disease was defined as experiencing symptoms constantly to occasionally (i.e., response categories a through d), and inactive disease was defined as experiencing infrequent symptoms or feeling well (response categories e and f). Study participants were then assigned to 1 of 3 mutually exclusive groups based on their pattern of disease activity across the measurement occasions: the consistently active disease group reported active disease at all measurement occasions, the consistently inactive disease group reported inactive disease at all measurement occasions, and the fluctuating disease activity group was composed of the remaining individuals. Standardized clinical indices of disease activity, the Harvey–Bradshaw for CD12 and Powell–Tuck for UC,13 were positively correlated with our 6-point measure of disease activity (Spearman's ρ = 0.40 [P < 0.0001] for CD; ρ = 0.50 [P < 0.0001] for UC at the baseline measurement occasion). The correlations are modest in value, likely due to differences in the reference time period for the clinical indices (2 weeks) and the disease activity self-report measure (6 months).

Study participants provided responses for multiple instruments designed to measure QOL, negative psychological functioning, and positive psychological functioning. A disease-specific QOL measure, the Inflammatory Bowel Disease Questionnaire14 (IBDQ), was used in addition to a general QOL measure, the Medical Outcomes Study 36-item Short Form Questionnaire15 (SF-36). Five aspects of negative psychological functioning were considered: distress, perceived stress, health anxiety, pain anxiety, and pain catastrophizing. Three aspects of positive psychological functioning were also investigated: social support, well-being, and mastery. The instruments used to measure these constructs are described in Table 1.

Table 1. Description of Quality of Life (QOL) and Psychological Functioning Instruments
ConstructInstrumentPurposeNumber of Items/Range/Interpretation
Disease-specific QOLInflammatory Bowel Disease Questionnaire (IBDQ)14A well-known quality of life instrument for inflammatory bowel diseases33 that measures 4 domains: gastrointestinal symptoms, systemic problems, emotional dysfunction, social difficulties32 items/0 to 224/ higher scores correspond to better QOL
General QOLSF-3615A standard health assessment instrument that measures quality of life in 8 domains: physical health, mental health, emotional role-functioning, physical role-functioning social health, general health perceptions, vitality, and pain36 items/0 to 100 on separate physical health and mental health component scores/higher scores correspond to better QOL
DistressKessler Psychological Distress Scale34Measures global psychological distress10 items/10 to 50/scores less than 16 correspond to a reduced risk for an anxiety or depressive disorder
StressCohen Perceived Stress Scale35, 36Examines the role of stress in disease14 items/0 to 56/higher scores correspond to greater perceived stress
Health anxietyHealth Anxiety Questionnaire (HAQ)37Measures health concerns and the individual's somatic focus21 items/0 to 63/higher scores indicate greater health anxiety
Pain anxietyPain Anxiety Symptom Scale (PASS)38Captures 4 dimensions of pain-related anxiety: cognitive, fear, escape/avoidance, physiological40 items/0 to 200/higher scores indicate greater pain anxiety
Pain catastrophizingCoping Strategies Questionnaire– Pain Catastrophizing Subscale39Measures maladaptive response to pain that involves the tendency to exaggerate the threat of the pain and to negatively evaluate one's own ability to manage the pain8 items/0 to 28/higher scores indicate greater pain catastrophizing
Social supportMultidimensional Scale of Perceived Social Support40Assesses degree of support from family and friends12 items/12 to 84 (often rescaled to range from 1 to 7)/higher scores correspond to greater support
Well-beingPsychological Well-Being Manifestations Scale41Measures an individual's degree of positive sense of self and day-to-day functioning25 items/25 to 125/higher scores correspond to better psychological well-being
MasteryMastery Scale42Measures an individual's perceived control or efficacy7 items/7 to 28/higher scores correspond to greater mastery

In summary, longitudinal data were collected on 11 outcomes. However, not all measures were collected at each occasion because of the potential for participant fatigue due to the large number of response items. The instruments to measure QOL, perceived stress, and social support were administered at all 5 measurement occasions, while the instruments to measure health anxiety, well-being, and mastery were administered on 3 occasions: 0, 12, and 24 months. The remaining instruments were each administered on 2 occasions: distress was measured at 0 and 12 months, and pain anxiety and catastrophizing were measured at 0 and 24 months.

Trend plots of the data were used to assess the assumption of a linear trajectory of change in the study outcomes and identify potential outliers/influential observations. Measures of skewness and kurtosis were computed to assess potential departures from a multivariate normal distribution. Correlations were computed to assess the pattern of association among the repeated measurements. Frequency distributions and logistic regression revealed that subject attrition was not associated with age group, disease type, or sex; we assume that the missing observations are missing at random, or ignorable.16

Linear mixed-effects regression models, also known as hierarchical or multilevel models,7–19 were used to test the trend in each of the outcomes over time (measured in months) and to model the data as a function of the fixed effects of pattern of disease activity and disease type, as well as random (i.e., subject-specific) effects. The confounding covariates of age and sex were included as fixed effects in all models. Mixed-effects models have been recommended for analyzing longitudinal QOL outcomes, in part because they can accommodate data that are unbalanced due to missing observations.20 Regression parameters were estimated using maximum likelihood estimation.

The model initially fitted for each outcome included 2-way interactions of time × disease type, time × disease activity pattern, and disease type × disease activity pattern. To ensure a parsimonious final model, interaction effects that were nonsignificant were excluded. A random intercept and a random slope for time were initially included in each model; the random slope for time was only retained in the final model when it resulted in a significant improvement in model fit and where model convergence was achieved. Model fit was assessed using the Aikake Information Criterion (AIC), a penalized log-likelihood measure.21 An independence covariance structure was fit to the model residuals because goodness of fit (as measured by the AIC) did not improve substantially for any of the outcomes when a model for correlated residuals (e.g., first-order autoregressive or compound symmetric structure) was adopted. All analyses were conducted using the MIXED procedure in SAS v. 9.1 (Cary, NC).22

The output from the mixed-effect models includes F tests of main and interaction effects as well as regression parameter estimates (i.e., b̂) for the fixed effects, and their associated standard errors, t-statistics, and P-values. Tests of statistical significance were conducted using α = 0.05. The intraclass correlation (ICC), the proportion of variation in the outcome variable(s) explained by the random effect(s), is also reported.18 The ICC ranges in value from 0 to 1, with higher values indicating a greater proportion of total variation due to subjects.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES
  7. APPENDIX:

For the 388 study participants the baseline chart review confirmed that 187 (48.2%) had CD and 169 (43.6%) had UC. Another 18 (4.6%) were identified as having indeterminate colitis and 14 (3.6%) did not have definite IBD. The latter 2 groups of individuals were excluded, leaving 356 participants for subsequent analyses. More than half of the study participants (59%) were women. They were predominantly Caucasian (91%), with small numbers having self-described backgrounds such as East Indian, Hispanic, or Metis (of Native American and European ancestry). The age range was from 18 to 83 years. For participants with CD the mean age was 38.5 years (SD = 14.6), while for those with UC the mean age was 43.0 years (SD = 14.7). Average duration of disease at baseline (in years) was 4.4 (SD = 2.1) for participants with CD and 4.3 (SD = 2.1) for those with UC.

The number and percent of study participants with self-reported active and inactive disease at each measurement occasion is reported in Table 2 for CD and UC. Overall, 109 (30.6%) participants were classified as having a consistently active disease pattern during the 24-month observation period, 60 (16.9%) were classified as having a consistently inactive disease pattern, and 174 (48.9%) were classified as having a fluctuating disease activity pattern. The remaining 13 (3.7%) could not be classified into 1 of these 3 groups because of missing data on disease activity, and were excluded from subsequent analyses. Descriptive statistics for the QOL and psychological functioning measures are reported in Table 3 for these 3 disease activity patterns at the baseline and 24-month measurement occasions; separate results are provided for CD and UC. In addition, the Appendix reports the correlations among each of the outcome measures at the baseline measurement occasion; there was a moderate negative correlation between the IBDQ and the negative psychological functioning measures of distress, perceived stress, health anxiety, and pain anxiety, but a weak negative association between the IBDQ and pain catastrophizing. Associations of the IBDQ with the positive psychological functioning measures were positive and weak to moderate in strength. A similar pattern was evident for the SF-36 mental health component, but the physical health component was only weakly associated with the positive and negative psychological functioning measures.

Table 2. Frequency (%) of Active and Inactive Disease at Each Measurement Occasion
 CD (n = 187)UC (n = 169)
ActiveInactiveActiveInactive
  1. Row totals may not add to N = 356 because of missing responses.

Baseline (0 months)135 (72.2)48 (25.7)109 (64.5)56 (33.1)
6 months118 (63.1)49 (26.2)96 (56.8)51 (30.2)
12 months117 (62.6)56 (30.0)98 (58.0)58 (34.3)
18 months107 (57.2)51 (27.3)86 (50.9)66 (39.1)
24 months103 (55.1)62 (33.2)81 (47.9)69 (40.8)
Table 3. Means (SDs) for Quality of Life and Psychological Functioning Measures at Baseline and 24 Months for Study Participants Classified by Pattern of Disease Activity
 CD (n = 187)UC (n = 169)
Consistently Active DiseaseFluctuating Disease ActivityConsistently Inactive DiseaseConsistently Active DiseaseFluctuating Disease ActivityConsistently Inactive Disease
  • a

    Distress was not collected at the 24-month wave. The possible range of scores for each measure are as follows: IBDQ: 0 to 224; SF-36 Physical Health: 0 to 100; SF-36 Mental Health: 0 to 100; Distress: 10 to 50; Perceived Stress: 0 to 56; Health Anxiety: 0 to 63; Pain Anxiety: 0 to 200; Catastrophizing: 0 to 28; Social Support: 1 to 7; Well-Being: 25 to 125; Mastery: 7 to 28.

Quality of Life      
IBDQ      
 0 months152.9 (26.7)165.4 (31.5)184.2 (23.6)160.0 (29.4)168.5 (31.1)189.4 (20.8)
 24 months155.0 (31.5)180.2 (23.4)193.3 (20.4)163.0 (27.5)180.6 (23.2)198.7 (16.9)
SF-36 Mental Health      
 0 months41.9 (5.9)43.2 (5.8)45.6 (5.1)43.2 (7.1)43.6 (5.9)43.9 (5.5)
 24 months41.1 (5.9)44.2 (4.5)44.7 (4.3)43.8 (3.4)42.8 (5.1)44.4 (5.5)
SF-36 Physical Health      
 0 months41.9 (5.5)43.4 (4.2)42.7 (6.3)42.4 (5.3)42.3 (4.7)44.0 (3.6)
 24 months43.0 (5.4)44.3 (3.7)41.8 (3.3)41.7 (4.4)42.6 (4.5)43.5 (2.4)
Negative Psychological Functioning      
Distressa      
 0 months20.7 (6.4)18.4 (5.9)16.0 (4.8)17.8 (6.0)17.4 (5.5)15.4 (4.0)
 12 months21.2 (6.9)17.9 (6.4)16.4 (4.6)17.4 (5.1)17.1 (5.3)15.4 (4.4)
Perceived Stress      
 0 months25.1 (8.1)22.9 (8.8)20.1 (7.0)23.2 (7.5)22.3 (7.6)17.6 (6.9)
 24 months23.8 (9.3)21.2 (7.7)17.0 (8.2)20.3 (7.7)20.6 (8.4)15.2 (7.7)
Health Anxiety      
 0 months17.7 (9.5)16.7 (10.9)12.2 (9.1)14.1 (7.7)14.1 (9.4)9.0 (5.4)
 24 months14.6 (9.6)10.5 (7.3)8.1 (8.5)12.9 (8.7)10.6 (8.2)6.2 (5.2)
Pain Anxiety      
 0 months79.8 (27.0)74.4 (30.6)72.0 (31.1)67.6 (30.7)67.8 (27.8)61.2 (26.1)
 24 months74.3 (30.8)64.9 (27.9)59.7 (29.2)56.8 (27.8)63.7 (30.3)41.5 (22.2)
Catastrophizing      
 0 month13.2 (4.6)12.7 (4.0)12.2 (5.0)11.1 (4.6)12.4 (3.8)12.0 (5.8)
 24 month14.7 (5.3)14.1 (5.1)14.0 (5.4)12.5 (5.8)13.1 (4.9)12.8 (4.4)
Positive Psychological Functioning      
Social Support      
 0 months5.4 (1.2)5.6 (1.3)5.7 (1.1)5.4 (1.1)5.4 (1.2)5.8 (0.9)
 24 months5.4 (1.5)5.6 (1.3)5.9 (1.0)5.3 (1.5)5.5 (1.3)5.8 (1.3)
Well-Being      
 0 months73.2 (17.2)76.9 (15.4)82.1 (14.1)74.6 (18.0)78.7 (16.6)85.8 (13.2)
 24 months71.4 (18.9)78.5 (14.8)84.1 (11.6)75.5 (15.5)77.6 (16.6)85.0 (15.8)
Mastery      
 0 months18.0 (5.0)19.2 (4.4)19.3 (4.2)19.2 (3.4)19.9 (4.9)21.9 (3.7)
 24 months18.6 (5.0)20.1 (4.0)19.3 (4.4)19.8 (4.3)19.9 (5.6)22.3 (4.2)

Assessments of QOL

Mixed-effects regression models for the QOL measures were conducted first. All models included a random intercept; retention of a random slope for time did not result in improved model fit as evaluated by the AIC.

For the IBDQ there was a statistically significant disease activity pattern × time interaction (F[2,1277] = 6.65, P = 0.0013). No other interaction terms were statistically significant. Figure 1 illustrates the trajectory of change in the unadjusted means for the 3 patterns of disease activity. The regression coefficients for the interaction are reported in Table 4. Compared to study participants with consistently active disease, those with consistently inactive disease showed no significant change in IBDQ scores over time (P = 0.1183), while those with fluctuating disease activity showed a small but significant improvement in IBDQ scores (P = 0.0003). There was no significant main effect of disease type for the IBDQ (F[1, 1277] = 0.60, P = 0.4391).

thumbnail image

Figure 1. Profile plot of means and 95% confidence intervals for the IBD quality of life measure.

Download figure to PowerPoint

Table 4. Regression Parameter Estimates from Mixed-effects Regression Models for Quality of Life Measures
Model EffectIBDQSF-36 Mental HealthSF-36 Physical Health
SESESE
  • All parameter estimates are adjusted for age and sex; SE, standard error; ICC, intraclass correlation; n/a indicates time × disease activity interaction was not included in the SF-36 mental and physical health models.

  • *

    A parameter estimate that is significant at α = 0.05; a positive value of b̂ for the time effect indicates that the average score increased over time while a negative value indicates that the average score decreased over time; a positive value of b̂ for disease activity pattern or disease type indicates that the average score was higher than the reference group average score, while a negative value indicates that the average score was lower than the reference group average score.

Time0.080.090.010.010.0040.01
Disease Activity Pattern      
 Consistently inactive31.89*4.041.62*0.600.480.50
 Fluctuating10.49*3.050.340.450.520.38
 Consistently activeRefRefRef
Disease Type      
 Crohn's disease−1.942.51−0.470.41−0.050.35
 Ulcerative colitisRefRefRef
Time × Disease Activity Pattern      
 Time × consistently inactive0.240.15n/an/a
 Time × fluctuating0.42*0.11n/an/a
 Time × consistently activeRefn/an/a
ICC0.580.340.34

None of the interaction effects were statistically significant for the SF-36 mental health or physical health components. As well, the main effect of time was not significant for either component (mental health: F[1, 1197] = 0.24, P = 0.6210; physical health: F[1, 1197] = 0.18, P = 0.6714). There was a significant main effect of disease activity pattern for the mental health component (F[2, 1197] = 3.85, P = 0.0216), but not for the physical health component (F[2, 1197] = 1.01, P = 0.3661). For the former (see Table 4), individuals with the pattern of consistently inactive disease had a significantly higher score than individuals with consistently active disease (P = 0.0067). There was no significant main effect of disease type on either the mental health (F[1, 1197] = 1.32, P = 0.8927) or physical health (F[1, 1197] = 0.02, P = 0.8811) components.

The ICC values for the QOL models are also reported in Table 4. For the IBDQ, slightly more than half of the total variation could be explained by the random intercept, which indicates that there was substantial subject-specific variation in the data. The ICC was much smaller for both SF-36 components.

Assessments of Negative Psychological Functioning

None of the 2-way interactions were statistically significant in the models for the negative psychological functioning measures; therefore, only main effects were retained. A random intercept was included in all models, but a random slope for time was not retained because it did not result in improved model fit.

A significant main effect of time was observed for perceived stress (F[1, 1260] = 25.79, P < 0.0001), health anxiety (F[1, 649] = 93.63, P < 0.0001), pain anxiety (F[1, 656] = 59.84, P < 0.0001), and pain catastrophizing (F[1, 651] = 23.52, P < 0.0001), but not for distress (F[1, 537] = 1.40, P = 0.2381). Examination of the regression coefficients revealed that there was a significant decrease in perceived stress, health anxiety, and pain anxiety, and a significant increase in pain catastrophizing over time (Table 5).

Table 5. Regression Parameter Estimates from Mixed-effects Regression Models for Negative Psychological Functioning Measures
Model EffectDistressStressHealth AnxietyPain AnxietyCatastrophizing
SESESESESE
  • All parameter estimates are adjusted for age and sex; SE, standard error; ICC, intraclass correlation.

  • *

    A parameter estimate that is significant at α = 0.05; a positive value of b̂ for the time effect indicates that the average score increased over time while a negative value indicates that the average score decreased over time; a positive value of b̂ for disease activity pattern or disease type indicates that the average score was higher than the reference group average score, while a negative value indicates that the average score was lower than the reference group average score.

Time−0.010.01−0.08*0.02−0.15*0.02−0.34*0.040.05*0.01
Disease Activity Pattern          
 Consistently inactive−3.80*0.84−5.08*1.08−5.78*1.27−7.624.21−0.170.64
 Fluctuating−1.93*0.63−1.510.81−1.550.96−2.563.180.020.49
 Consistently activeRefRefRefRefRef
Disease Type          
 Crohn's disease0.760.581.190.751.130.887.27*2.920.95*0.45
 Ulcerative colitisRefRefRefRefRef
ICC0.740.580.690.760.57

A significant main effect of disease activity pattern was observed for distress (F[2, 537] = 10.77; P < 0.0001), perceived stress (F[2, 1260] = 11.13, P < 0.0001), and health anxiety (F[2, 649] = 10.46, P < 0.0001), but not for pain anxiety (F[2, 656] = 1.65, P = 0.1947) or catastrophizing (F[2, 651] = 0.05, P = 0.9483). As the results in Table 5 reveal, study participants with the consistently inactive disease pattern had a significantly lower mean score than respondents with a consistently active disease pattern for distress, perceived stress, and health anxiety (P < 0.0001) for all measures. As well for distress, the fluctuating disease activity group had a significantly lower mean score than the consistently active group (P = 0.0002).

There was a significant main effect of disease type on pain anxiety (F[1, 656] = 6.21, P = 0.0129), and catastrophizing (F[1, 656] = 4.53, P = 0.0337). Scores were significantly higher for CD than for UC participants on both measures (Table 5). ICC values were highest for distress and pain anxiety. However, the ICC exceeded 0.50 for all measures, indicating that more than half of the total variation in the data could be attributed to the subject-specific effect.

Assessments of Positive Psychological Functioning

For the 3 measures of positive psychological functioning, none of the 2-way interactions were statistically significant; therefore, the final models retained only main effects. A random intercept was included in each model, but a random slope for time was not retained because it did not result in improved model fit.

There was a significant main effect of time for mastery (F[1, 652] = 5.46, P = 0.0197), but not for social support (F[1, 1278] = 0.08, P = 0.7837) or psychological well-being (F[1, 653] = 0.06, P = 0.8123). An increasing trend over time was evident for mastery (Table 6), although the magnitude of the effect was small.

Table 6. Regression Parameter Estimates from Mixed-effects Models for Positive Psychological Functioning Measures
Model EffectSocial SupportWell-BeingMastery
SESESE
  • All parameter estimates are adjusted for age and sex; SE, standard error; ICC, intraclass correlation.

  • *

    A parameter estimate that is significant at α = 0.05; a positive value of b̂ for the time effect indicates that the average score increased over time while a negative value indicates that the average score decreased over time; a positive value of b̂ for disease activity pattern or disease type indicates that the average score was higher than the reference group average score, while a negative value indicates that the average score was lower than the reference group average score.

Time0.0010.002−0.00.030.02*0.01
Disease Activity Pattern      
 Consistently inactive0.410.1810.37*2.222.41*0.63
 Fluctuating0.230.135.21*1.670.930.48
 Consistently activeRefRefRef
Disease Type      
 Crohn's disease0.080.120.061.53−0.380.44
 Ulcerative colitisRefRefRef
ICC0.630.560.66

A significant main effect of disease activity pattern was observed for well-being (F[1, 653] = 11.51, P < 0.0001) and mastery (F[2, 652] = 7.27, P = 0.0008), but not for social support (F[2, 1278] = 2.96, P = 0.0519). As Table 6 reveals, compared to the consistently active disease group the consistently inactive disease group reported significantly higher scores for well-being (P < 0.0001) and mastery (P = 0.0002). As well, participants with a fluctuating pattern of disease activity had significantly higher scores for well-being compared to those with a consistently active disease pattern (P = 0.0019).

There was no significant main effect of disease type on any of the measures of positive psychological functioning: social support (F[1, 1278] = 2.79, P = 0.0950), well-being (F[1, 653] = 0.01, P = 0.9679), mastery (F[1, 652] = 0.73, P = 0.3919). ICC values for the measures of positive psychological functioning ranged from 0.56 to 0.66. These values indicate that for all measures, more than half of the total variation was due to the subject-specific effect.

For CD subjects there was a significant effect of disease location (small bowel versus colonic or perineal) on IBDQ and mastery scores, and a significant effect of disease behavior (fistulizing and stricturing versus inflammatory) on the IBDQ and SF-36 physical health component scores. Specifically, study participants with perianal disease indicated significantly higher IBDQ and mastery scores relative to study participants with small bowel disease. Study participants with a fistula or stricture in the bowel reported significantly lower scores on the IBDQ and SF-36 physical health component than individuals with inflammatory disease. There was no significant effect of disease location on 9 of the 11 IBD outcome measures at α = 0.05. Similarly, there was no significant effect of disease behavior on 9 of the 11 IBD outcomes measures at α = 0.05.

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES
  7. APPENDIX:

This longitudinal study evaluated multiple, important variables related to coping with IBD using survey and interview data collected over a 2-year period. Table 7 summarizes the statistically significant results that were identified in the regression analyses.

Table 7. Summary of Regression Analyses for IBD Outcomes
 Explanatory Variables
TimeDisease Activity PatternDisease Type
  • NS, no significant main effect; Sig, significant main effect; statistical significance was assessed at α = 0.05.

  • a

    Statistical significance of main effects is not relevant given that there was a significant time x disease activity pattern interaction.

IBDQN/AaN/AaNS
SF-36 Mental HealthNSSigNS
SF-36 Physical HealthNSNSNS
DistressNSSigNS
StressSigSigNS
Health anxietySigSigNS
Pain anxietySigNSSig
CatastrophizingSigNSSig
Social supportNSNSNS
Well-beingNSSigNS
MasterySigSigNS

The findings of this longitudinal study are consistent with a growing number of studies that have identified disease activity as an important determinant of outcomes in persons with IBD; however, our assessment at multiple occasions strengthens the validity of this association.7, 9, 23 An earlier article by our group10 that focused only on the baseline data from the Manitoba IBD Cohort Study demonstrated substantial differences in disease-specific QOL and positive and negative psychological functioning in groups with self-reported active and inactive disease. However, while that article reported no effect of disease type, the current analyses revealed that for pain anxiety and pain catastrophizing (i.e., characterizations of pain as overwhelming), there were differences between CD and UC participants even after controlling for pattern of disease activity. CD is likely associated with more pain (or less predictable pain) than UC. Patients with CD have higher health care utilization than patients with UC,24 an indicator of greater morbidity, possibly including pain. As individuals progress through the disease course, pain may become a more pronounced factor and maladaptive responses may be more likely to emerge. As well, pain may be more difficult to stabilize among individuals with CD than with UC. Physicians should be encouraged to manage pain early on and not assume that all pain will resolve easily with the prescribing of immunomodulatory therapy. While analgesics are an important aspect of pain management, another may be to explore the emotional and psychological well-being of patients to determine if formal psychological assessment and intervention is indicated. This approach has been taken by our group and others in the management of irritable bowel syndrome,25 and we have adopted this approach for management of chronic pain in IBD as well. Psychological treatment strategies also have good efficacy and minimal iatrogenic effects. If patients develop good coping and relief mechanisms early in the disease course, they may be less likely to experience anxiety and the cognitive reactivity that can further exacerbate pain.

As expected, the disease-specific measure of QOL was sensitive to different patterns of disease activity. The average score for study participants with consistently inactive disease was similar to that reported in the literature for individuals in remission.26 Study participants with fluctuating disease activity showed a significant improvement in disease-specific QOL, while those with active disease showed a relatively flat trajectory. However, there was no significant difference between the consistently active and consistently inactive groups in the magnitude of change over time. This may be partially attributable to the smaller sample size for the latter group, resulting in an increased standard error and reduced sensitivity to detect a difference between disease activity patterns.

Average scores for the SF-36 measures of physical and mental health were below normative values for the Canadian population.27 However, these measures remained largely unchanged over the 2-year observation period, and could not always differentiate among individuals with different patterns of disease activity. While previous research has recommended the use of both disease-specific and general measures of QOL to characterize patients with IBD,28 our study suggests that only the former are sufficiently sensitive to be recommended for use.

Only 1 negative psychological functioning measure, distress, did not show significant improvement or deterioration over time. However, data for this measure were only available for 2 of the 5 measurement occasions. There may not have been sufficient time to detect change in individual's perceptions of distress over the disease course.

The strengths of this study include the sample selection technique, low rate of attrition, and method of analysis. The Manitoba IBD Cohort Study draws participants from a large geographic area and with a diverse demographic profile. The population-based design helps to ensure the sample is representative of a broad range of individuals with IBD. Study attrition was less than 10% over the 2-year observation period and drop-out was not concentrated in any particular cohort subgroup. These findings help to support the conclusion that the observed trajectories of change are representative of those expected in the IBD population. Unlike conventional longitudinal analysis techniques, mixed-effects models account for the inherent variability both among and within individuals.29 Our analyses revealed that there was substantial subject-specific variation on all of the measures, particularly the negative psychological functioning measures.

One potential limitation of this study is that it relies on a self-report measure to assign respondents to disease activity patterns. However, previous research has also used self-reports of disease symptoms and periods of remission to characterize disease activity in IBD populations.5 The feasibility of obtaining clinical measures of inflammation at 6-month intervals in such a large sample of patients is low. Another limitation of the study is that respondents were assigned to 1 of only 3 groups based on their disease activity pattern; for example, in the fluctuating disease activity group there was no distinction between individuals who reported disease activity at all but 1 measurement occasion and those who reported disease inactivity at all but 1 measurement occasion. Subsequent investigations could explore these subgroups to investigate their trajectories of change in IBD outcomes.30

While this study described and tested the longitudinal change in measures of QOL and psychological functioning and their association with disease activity pattern and disease type, it does not test the causal mechanism of these relationships. Increasingly, the multivariate analysis technique of structural equation modeling is being applied in cross-sectional and longitudinal observational studies to test causal theories about health outcomes.31

This study only tested the change in total scores for each outcome measure. Previous cross-sectional studies have examined respondent differences on the components of the IBDQ and SF-36.7, 28, 32 This approach could be extended to longitudinal data, to investigate changes in different aspects of quality of life and their association with disease activity patterns and disease type. Studies of longitudinal change in other outcomes, such as health care use and factors associated with disease flares, could also be pursued.

In summary, having IBD impacts patient QOL, perceived stress, health anxiety, pain anxiety, pain catastrophizing, and mastery over time. Many of the measures used in this study could be adopted in clinical trials of IBD treatment efficacy given their responsiveness to disease activity and change over time. Nevertheless, the findings do raise some questions about using the SF-36 in such studies because of its low discriminative performance. Patients with persistently active disease may have a greater need for psychological care in dealing with their emotional response to illness. Improvements in positive and negative psychological functioning over time were modest, suggesting that disease is having an impact on people's lives even when they are not experiencing active symptoms. Fortunately, those with an active pattern did not show deterioration in quality of life and psychological functioning in spite of ongoing symptoms; rather, they tended to experience poorer functioning that persisted over time.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES
  7. APPENDIX:
  • 1
    Engel GL. The need for a new medical model: a challenge for biomedicine. Science. 1977; 196: 129136.
  • 2
    Drossman D. Gastrointestinal illness and the biopsychosocial model. Psychosom Med. 1998; 60: 258267.
  • 3
    Blondel-Kucharski F, Chircop C, Marquis P, et al. Health-related quality of life in Crohn's disease: a prospective longitudinal study in 231 patients. Am J Gastroenterol. 2001; 96: 29152920.
    Direct Link:
  • 4
    Guthrie E, Jackson J, Shaffer J, et al. Psychological disorder and severity of inflammatory bowel disease predict health-related quality of life in ulcerative colitis and Crohn's disease. Am J Gastroenterol. 2002; 97: 19941999.
    Direct Link:
  • 5
    Petrak F, Hardt J, Clement T, et al. Impaired health-related quality of life in inflammatory bowel diseases: psychosocial impact and coping styles in a national German sample. Scand J Gastroenterol. 2001; 36: 375382.
  • 6
    Porcelli P, Zaka S, Centonze S, et al. Psychological distress and levels of disease activity in inflammatory bowel disease. Ital J Gastroenterol. 1994; 26: 111115.
  • 7
    Casellas F, Arenas JI, Baudet JS, et al. Impairment of health-related quality of life in patients with inflammatory bowel disease: A Spanish multicenter study. Inflamm Bowel Dis. 2005; 11: 488496.
  • 8
    Casellas F, Lopez-Vivancos J, Casado A, et al. Factors affecting health related quality of life of patients with inflammatory bowel disease. Qual Life Res. 2002; 11: 775781.
  • 9
    Gibson PR, Weston AR, Shann A, et al. Relationship between disease severity, quality of life and health-care resources in a cross-section of Australian patients with Crohn's disease. J Gastroenterol Hepatol. 2007; 22: 13061312.
  • 10
    Graff LA, Walker JR, Lix LM, et al. The relationship of inflammatory bowel disease type and activity to psychological functioning and quality of life. Clin Gastroenterol Hepatol. 2006; 4: 14911501.
  • 11
    Han SW, McColl E, Barton JR, et al. Predictors of quality of life in ulcerative colitis: the importance of symptoms and illness representations. Inflamm Bowel Dis. 2005; 11: 2434.
  • 12
    Harvey RF, Bradshaw JM. A simple index of Crohn's disease activity. Lancet. 1980; 1: 514.
  • 13
    Powell-Tuck J, Brown RL, Lennard-Jones JE. A comparison of oral prednisolone given as single or multiple daily doses for active proctocolitis. Scand J Gastroenterol. 1978; 13: 833837.
  • 14
    Guyatt GH, Mitchell A, Irvine EJ, et al. A new measure of health status for clinical trials in inflammatory bowel disease. Gastroenterology. 1989; 96: 804810.
  • 15
    Ware JE, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care. 1992; 30: 473483.
  • 16
    Little RJA, Rubin DB. Statistical analysis with missing data. New York: John Wiley & Sons; 1987.
  • 17
    Cudek R. Mixed-effects models in the study of individual differences with repeated measures data. Multivariate Behav Res. 1996; 31: 371403.
  • 18
    Hedeker D, Gibbons RD. Longitudinal data analysis. Hoboken, NJ: John Wiley & Sons; 2006.
  • 19
    Sullivan S, Dukes KA, Losina E. Tutorial in biostatistics: an introduction to hierarchical linear modeling. Stat Med. 1999; 18: 855888.
  • 20
    Beacon HJ, Thompson SG. Multi-level models for repeated measurement data: application to quality of life data in clinical trials. Stat Med. 1996; 15: 27172732.
  • 21
    Akaike H. A new look at the statistical model identification. IEEE Trans Autom Control. 1974; AC19: 716723.
  • 22
    SAS Institute, SAS/Stat user 's guide, version 9. Cary, NC: SAS Institute; 2004.
  • 23
    Sewitch MJ, Abrahamowicz M, Bitton A, et al. Psychological distress, social support, and disease activity in patients with inflammatory bowel disease. Am J Gastroenterol. 2001; 96: 14701479.
    Direct Link:
  • 24
    Longobardi T, Bernstein CN. Health care resource utilization in inflammatory bowel disease. Clin Gastroenterol Hepatol. 2006; 4: 731743.
  • 25
    Tkachuk G, Graff L, Martin G, et al. Controlled trial of cognitive behavioral group treatment for irritable bowel syndrome in a medical setting. J Clin Psychol Med Settings. 2003; 10: 5769.
  • 26
    Irvine EJ, Feagan B, Rochon J, et al. Quality of life: a valid and reliable measure of therapeutic efficacy in the treatment of inflammatory bowel disease. Gastroenterology. 1994; 106: 287296.
  • 27
    Hopman WM, Towheed T, Anastassiades T, et al. Canadian normative data for the SF-36 health survey. CMAJ. 2000; 163: 265271.
  • 28
    McColl E, Han SW, Barton JR, et al. A comparison of the discriminatory power of the Inflammatory Bowel Disease Questionnaire and the SF-36 in people with ulcerative colitis. Qual Life Res. 2004; 13: 805811.
  • 29
    Fairclough DL, Gagnon D, Papadopoulos G. Planning analyses of quality-of-life studies: a case example with migraine prophylaxis. J Biopharm Stat. 2004; 14: 3151.
  • 30
    Leffondre K, Abrahamowicz M, Regeasse A, et al. Statistical measures were proposed for identifying longitudinal patterns of change in quantitative health indicators. J Clin Epidem. 2004; 57: 10491062.
  • 31
    Raudenbush SW. Comparing personal trajectories and drawing causal inferences from longitudinal data. Annu Rev Psychol. 2001; 52: 501525.
  • 32
    Bernklev T, Jahnsen J, Schulz T, et al. Course of disease, drug treatment and health-related quality of life in patients with inflammatory bowel disease 5 years after initial diagnosis. Eur J Gastroenterol Hepatol. 2005; 17: 10371045.
  • 33
    Sainsbury A, Heatley RV. Review article: psychosocial factors in the quality of life of patients with inflammatory bowel disease. Aliment Pharmachol Ther. 2005; 21: 499508.
  • 34
    Kessler RC, Andrews G, Colpe LJ, et al. Short screening scales to monitor population prevalences and trends in non-specific psychological distress. Psychol Med. 2002; 32: 959976.
  • 35
    Cohen S, Kamarck, Mermelstein R. A global measure of perceived stress. J Health Soc Behav. 1983; 24: 385396.
  • 36
    Cohen S, Williamson GM. Perceived stress in a probability sample of the United States. In: SpacapanSS, OskempS, eds. The social psychology of health. Newbury Park, CA: Sage; 1988: 3167.
  • 37
    Luccock MP, Morley S. The health anxiety questionnaire. Br J Health Psychol. 1996; 1: 137150.
    Direct Link:
  • 38
    McGracken LM, Zayfert C, Gross RT. The pain anxiety symptoms scale: development and validation of a scale to measure the fear of pain. Pain. 1992; 50: 6773.
  • 39
    Rosenstiel AK, Keefee FJ. The use of coping strategies in chronic low back pain patients: relationship to patient characteristics and current adjustment. Pain. 1983; 17: 3344.
  • 40
    Zimet GD, Dahlem NW, Zimet SG, et al. The multidimensional scale of perceived social support. J Pers Assess. 1988; 52: 3041.
  • 41
    Masse R, Poulin C, Dassa C, et al. Elaboration and validation of a tool to measure psychological well-being: WBMMS. Can J Public Health. 1998; 89: 352357.
  • 42
    Pearlin LI, Schooler C. The structure of coping. J Health Soc Behav. 1978; 19: 221.

APPENDIX:

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES
  7. APPENDIX:
Table  . Correlations Among QOL and Psychological Functioning Measures at Baseline Measurement Occasion
 IBDQSF-36MSF-36PDistressStressHAQPASSCatastSoc SupPWBMast
  1. IBDQ, Inflammatory Bowel Disease Questionnaire; SF-36M, SF-36 mental health component; SF-36P, SF-365 physical health component; HAQ, Health Anxiety Questionnaire; PASS, Pain Anxiety Symptom Scale; Catast, pain catastrophizing; Soc Sup, social support; PWB, psychological well-being; Mast, mastery.

IBDQ1.000.320.18−0.64−0.56−0.56−0.42−0.130.220.520.38
SF-36M 1.00−0.48−0.46−0.44−0.34−0.35−0.170.090.250.25
SF-36P  1.00−0.03−0.01−0.01−0.030.170.090.180.12
Distress   1.000.690.570.480.11−0.30−0.71−0.51
Stress    1.000.540.420.03−0.33−0.63−0.56
HAQ     1.000.550.16−0.13−0.44−0.34
PASS      1.000.21−0.09−0.37−0.36
Catast       1.000.040.03−0.01
Soc Sup        1.000.310.26
PWB         1.000.53
Mast          1.00