This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Afzal, M. Articles by Munafò, M. R. Search for Related Content PubMed PubMed Citation Articles by Afzal, M. Articles by Munafò, M. R. Related Collections Gastrointestinal Cognitive Functioning

Selective Processing of Gastrointestinal Symptom-related Stimuli in Irritable Bowel Syndrome

From the Department of Gastroenterology, East Surrey Hospital, Redhill, United Kingdom (M.A.); Academic Unit of Psychiatry, Cotham House, Bristol, United Kingdom (J.P.P.); Clinical Science at South Bristol, Bristol Royal Infirmary, Bristol, United Kingdom (C.S.J.P.); Department of Experimental Psychology, University of Bristol, Bristol, United Kingdom (M.R.M.).

Address correspondence and reprint requests to M. R. Munafò, PhD, Department of Experimental Psychology, University of Bristol, 8 Woodland Road, Bristol OX2 6HE, United Kingdom. E-mail: marcus.munafo{at}bristol.ac.uk

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION NOTES REFERENCES

 
Objectives: We sought to determine whether irritable bowel syndrome (IBS) was associated with attentional bias toward symptom-related cues in IBS patients versus healthy controls, using a modified Stroop task to measure selective processing of gastrointestinal symptom-related cues.

Methods: Fifteen patients with a clinical diagnosis of IBS and 15 healthy controls were recruited into the study. All participants attended a single testing session, during which they completed a modified Stroop task using gastrointestinal symptom-related and neutral control words.

Results: Results indicated a significant main effect of word type (p = .013), with slower color-naming times for IBS-related compared with neutral words, and a significant main effect of exposure (p = .001), with slower color-naming times in the unmasked condition compared with the masked condition. The group x word type x exposure interaction was significant (p = .048). A series of post hoc tests indicated that among patients there was significant interference of symptom-related words in the masked condition but not in the unmasked condition, whereas among controls, the reverse was true.

Conclusions: These results indicate that IBS patients selectively process gastrointestinal symptom-related words compared with neutral words when they are presented subliminally but not when they are presented supraliminally. In contrast, healthy controls demonstrate the opposite pattern. Implications for the cognitive mechanisms in IBS, and future research directions, are discussed.

Key Words: attentional bias • irritable bowel syndrome • modified Stroop

Abbreviations: IBS = irritable bowel syndrome; RT = reaction time; CBT = cognitive behavioral therapy.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION NOTES REFERENCES

 
Although irritable bowel syndrome (IBS) is common, affecting about 15% of the general population (1), its exact cause is unknown. Two consistent (and probably interrelated) characteristics are psychological comorbidity and visceral hypersensitivity. For example, IBS patients have higher scores for anxiety and depression as compared with the general population (2). Enhanced perception of visceral stimuli appears to be an important pathophysiological mechanism in IBS, suggesting a likely role for cognitive and psychosomatic mechanisms in disease and symptom reporting (3).

Selective processing bias refers to the tendency for information processing resources to be allocated disproportionately toward certain categories of stimuli and has been shown to exist in different psychiatric conditions such as generalized anxiety, depression, and addiction (4). Various methods exist for investigating processing bias in experimental contexts, including the modified Stroop task, which requires participants to engage in a color-naming task while lexical stimuli are simultaneously presented. Words may belong to a neutral or relevant category, and evidence for a processing bias toward the relevant category consists of a relative lengthening of color-naming reaction times (RTs) when a relevant stimulus is present relative to when a neutral stimulus is present.

The modified Stroop task has been used extensively to demonstrate selective processing of threat-related cues in anxious individuals from both healthy and psychiatric populations (4). This literature has achieved considerable consensus on the association of these biases with anxiety-related psychiatric case status. In particular, generalized anxiety disorder patients, when compared with healthy volunteers matched for high trait anxiety, appear to demonstrate selective processing of threat-related cues when these are presented both subliminally (i.e., too rapidly to allow conscious processing) and supraliminally (i.e., for sufficient duration to allow conscious processing). In contrast, high-trait-anxiety healthy volunteers appear to only show selective processing of threat-related cues when these are presented subliminally (5). The implication is that generalized anxiety disorder patients are unable to consciously override the selective processing of threat cues, whereas high-trait-anxiety healthy individuals are able to do so (6). Selective processing of physical threat cues has also been reported to predict recovery from day surgery independently of self-reported anxiety (7).

Similar results have been obtained for psychosomatic disorders such as chronic pain. For example, chronic pain patients with a high fear of pain have been shown to demonstrate selective processing of pain-related words compared with patients with a low fear of pain in a visual probe task (8). This bias was not observed for social threat or positive stimuli, suggesting that the observed bias was specific to pain-related material and not generalized to either threatening or highly emotional material. In an extension of this study, students with a low fear of pain oriented attention away from pain-related stimuli, whereas those with a high fear of pain did not (9). This is broadly consistent with the anxiety literature, suggesting that psychopathology is in part related to the inability to override the selective processing of personally relevant threat cues. To date, however, few studies have investigated selective processing of symptom-related words in IBS patients, although there is evidence for selective recall of negative material (10) and gastrointestinal sensation words (11) among IBS patients.

Selective recall of gastrointestinal symptom words among IBS patients (11) might be due to selective attention to IBS-related symptoms, as predicted by cognitive-behavioral theory of IBS (12). We therefore sought to determine whether IBS was associated with attentional bias toward disease-related cues in IBS patients versus healthy controls, using a modified Stroop task to measure selective processing of gastrointestinal symptom-related cues. We used the modified Stroop task as this has been reliably demonstrated to index attentional bias for threat-related stimuli in anxious volunteers and anxiety-spectrum patients.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION NOTES REFERENCES

 
Participants
Fifteen patients with a clinical diagnosis of IBS and 15 age- and sex-matched healthy controls were recruited between June and November 2005. Patients were recruited from the gastroenterology clinic of the Bristol Royal Infirmary, and healthy controls, from the general population. Participants were excluded if they were aged under 18 years or reported color blindness, uncorrected visual impairment, or lack of fluency in English. Participants were reimbursed travel expenses but not paid for participation. The study was approved by the local research ethics committee, and all participants provided fully informed consent.

Of the 15 IBS patients, 6 (40%) reported diarrhea-predominant IBS, 3 (20%) reported constipation-predominant IBS, 3 (20%) reported diarrhea alternating with constipation, and 3 (20%) reported having just pain and bloating. Patients with diarrhea were assessed using Manning criteria. Other patients had a clinical diagnosis of IBS using Rome-like criteria, although without a diary card and full assessment of the duration of their symptoms.

Design
The experimental design included one between-subjects factor of group (patient, control) and two within-subjects factors of word type (symptom-related, neutral) and exposure (masked, unmasked).

Materials
IBS-related stimuli consisted of 12 words chosen in consultation with clinicians to reflect gastrointestinal-related concerns and symptoms (bloating, bowel, constipation, cramp, diarrhea, incontinence, irritable, pain, painful, stomach, watery, wind). Control stimuli consisted of 12 categorized neutral "household" words matched for word length and frequency (antique, bedspread, broom, domestic, handle, linoleum, mattress, pavement, pillow, refrigerator, sofa, wallpaper). Uncategorized neutral words (e.g., cultural, caliper) were used for practice and buffer trials.

Procedure
Participants attended a single testing session, during which they completed the modified Stroop task, followed by a lexical decision task as an awareness check to confirm the masking procedure used in the Stroop task. The modified Stroop and lexical decision tasks were similar to those described elsewhere (13).

The modified Stroop task consisted of 12 practice trials followed by 12 blocks, each consisting of 2 buffer and 12 experimental trials (i.e., 144 experimental trials in total). Each block presented one set of stimulus words (either gastrointestinal symptom-related or control) under either masked (i.e., fixation box for 500 ms, followed by stimulus word and color background for 14 ms, followed by letter mask until response) or unmasked (i.e., fixation box for 500 ms, followed by stimulus word and color background until response) exposure conditions. The order of the 12 blocks was randomized across participants.

The lexical decision task consisted of 12 practice and 24 test trials. On half of the trials, the stimulus presentation was the same as in the masked condition of the color-naming task. The stimulus words were a subset of those in the color-naming task. On the other half of the trials, a nonword was presented. Participants indicated whether a word was present or absent by pressing one of two keys.

The program of events for both the modified Stroop and lexical decision tasks was controlled using MEL2 software (Psychology Software Tools, Pittsburgh, PA).

Statistical Analysis
RT data were excluded from trials with errors or with RT outliers. RTs greater than 1000 ms or 2 SDs above the mean or less than 200 ms or 3 SDs below the mean were considered outliers (14). Mean RTs for each participant were calculated across the 144 experimental trials, grouped by word type and exposure, resulting in four mean RT values (unmasked symptom-related, masked symptom-related, unmasked neutral, masked neutral).

Data were analyzed using a 2 x 2 x 2 mixed-model analysis of variance of mean color-naming times, with group (patient, control) as a between-subjects factor and word type (symptom-related, neutral) and exposure (masked, unmasked) as within-subjects factors.

We calculated that we had greater than 80% power, at an value of 0.05, to detect a between-groups difference in mean interference score of 15 ms with a total sample size of n = 28.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION NOTES REFERENCES

 
Characteristics of Participants
The sample (n = 30) consisted of 15 IBS patients (27% male, M = 30 years 6 months, SD = 8 years 11 months, range 22–53 years) and 15 healthy controls (33% male, M = 30 years 3 months, SD = 9 years 4 months, range 20–52 years). Patients and controls did not differ significantly in age (p = .95) or sex distribution (p = .69).

Modified Stroop Task
Mean color-naming RTs for IBS patients and healthy controls are presented in Table 1.

A 2 x 2 x 2 analysis of variance of mean color-naming RTs to symptom-related and neutral words, with group (patient, control) as a between-subjects factor and word type (symptom-related, neutral) and exposure (masked, unmasked) as within-subjects factors, indicated a significant main effect of word type (F [1, 28] = 7.05, p = .013), with slower color-naming times for IBS-related compared with neutral words, and a significant main effect of exposure (F [1, 28] = 13.94, p = .001), with slower color-naming times in the unmasked condition compared with the masked condition. The main effect of group was nonsignificant (p = .25). The group x word type x exposure interaction was significant (F [1, 28] = 4.29, p = .048). All other interactions were nonsignificant (F values < 1).

In order to clarify the significant three-way interaction, interference scores for symptom-related words compared with control words were calculated by subtracting the mean RTs for neutral words from those for symptom-related words for each participant separately for unmasked and masked conditions, resulting in two interference scores (unmasked, masked). This index reflects the word type effect in the analysis of variance, with positive values indicating a greater interference effect of symptom-related words compared with neutral words.

A series of simple-effects post hoc tests indicated that among patients there was significant interference of symptom-related words in the masked condition (p = .017) but not in the unmasked condition (p = .487), whereas among controls the reverse was true, with significant interference of symptom-related words in the unmasked condition (p = .024) but not in the masked condition (p = .593). These results are presented graphically in Figure 1.

View larger version (11K): [in this window] [in a new window]   Figure 1. Color-naming interference (ms) in unmasked and masked conditions for healthy controls and IBS patients. Mean color-naming interference scores are presented, grouped by clinical status (healthy control or IBS patient) and presentation (supraliminal or subliminal). The interference score index reflects the word type effect in the original analysis of variance, with positive values indicating a greater interference effect of symptom-related words compared with neutral words. Error bars represent 95% confidence intervals.

The percentage of awareness check trials with correct responses was calculated. The overall mean percentage correct was 51%, which was not significantly above chance performance of 50%.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION NOTES REFERENCES

 
These results indicate that IBS patients selectively process gastrointestinal symptom-related words compared with neutral words when these are presented in a masked condition on a modified Stroop task but not when they are presented in an unmasked condition. In contrast, healthy controls demonstrate the opposite pattern, with selective processing of gastrointestinal symptom-related words compared with neutral words when these are presented in an unmasked condition but not when they are presented in a masked condition. Performance on the lexical decision task in both groups indicated that, in the masked condition, words were presented too rapidly to allow conscious processing (i.e., subliminally). We had sufficient power to detect a between-group difference in mean interference score of 15 ms, although in our data this effect was qualified by an interaction with presentation (i.e., supraliminal versus subliminal).

The finding that IBS patients demonstrate a processing bias for subliminally presented gastrointestinal symptom-related words, but not similar words presented supraliminally, suggests that the modified Stroop task may index cognitive processes related to underlying pathology that operate outside of conscious awareness. This is supported by the lack of a similar processing bias for subliminally presented words in the control group and suggests that measures of subliminal processing bias may offer an objective index of features of cognitive processing related to disease state for use in intervention studies. This also suggests that these processing biases operate early in the information processing pathway, possibly influencing initial orienting rather than, for example, a failure to disengage or an inhibition of return, although we could not test this possibility directly.

The lack of a processing bias for supraliminally presented gastrointestinal symptom-related words among IBS patients is difficult to interpret. One possibility might be that IBS patients consciously override the processing bias for symptom-related words when these are presented for a sufficient duration. A similar effect has been reported in the anxiety literature (6) among highly trait-anxious individuals who are not clinically anxious. The presence of a processing bias for supraliminally presented symptom-related words among the control group is more problematic. One explanation for this may be that these words reflect unusual and/or threatening words that capture attention for reasons unrelated to disease. Despite our efforts to match the stimulus words for length and frequency, the symptom-related words composed a distinct category that was perhaps more salient than those in the neutral category. The mechanisms underlying the presence of a subliminal processing bias and lack of supraliminal processing bias among the IBS patients may differ qualitatively from those underlying the opposite pattern observed in controls. Our findings are consistent with central processing models, which predict an increased focus on symptom cues among IBS patients (12), although it is also possible that the effect we observed may be due to an increased familiarity with the stimulus cues used in this study among IBS patients compared with healthy controls.

There are a number of limitations to our study that should be considered when interpreting these results. First, we did not include measures of current or dispositional anxiety, so it was not possible to test whether the patient and control groups were matched on these variables. Given the existing literature on anxiety-related processing biases for threat-related material congruent with current concerns, this is a serious limitation. However, in previous studies conducted by this group at the same clinic, we have shown that IBS patients exhibit anxiety scores in the subsyndromal range, suggesting that increases in anxiety among IBS patients are subtle. Nevertheless, future studies should include measures of anxiety, possibly including measures of visceral-specific anxiety (15) to identify patients who do not demonstrate features of generalized anxiety but do experience gut-specific anxiety. Second, patients were identified on the basis of clinical diagnosis, which was not confirmed with independent reference to, for example, Rome criteria (16). It would have been desirable for clinical diagnosis to have been confirmed with reference to Rome criteria and for measures of symptom severity to have been included (17). This would also have enabled correlational analyses of subliminal processing bias and symptom severity to have been conducted, which would have been informative with respect to the future utility of measures such as these in studies of disease mechanisms and treatment response. Third, controls consisted of age- and sex-matched healthy volunteers. It would have been informative to include a disease control group for whom the symptom-related words would not have been disease-congruent but who were nevertheless currently unwell. The inclusion in future studies of further control groups with gastrointestinal symptoms, such as a group with a similar functional gastrointestinal disorder (e.g., functional dyspepsia) and a group with organic disease (e.g., inflammatory bowel disease), may also be informative. The observed results may also have been due to a greater familiarity among IBS patients with the symptom-related words, so that a control group comprising gastroenterology clinicians might also be desirable in future studies. Fourth, although beyond the explicit scope of this study, it would have been useful to obtain follow-up symptom data to determine the extent to which processing bias among patients predicted future disease and symptom trajectory. Fifth, the modified Stroop task does not allow the component processes of attention, such as initial orienting, disengagement, and so on, to be indexed. Future studies should use other measures of attentional bias, such as the dot probe task (18), to enable the exact attentional processes subserving the observed processing bias to be dissected. Sixth, the symptom-related words selected by clinicians in the current study may not necessarily reflect the IBS symptoms of greatest concern to patients. Future studies should include symptom-related words selected by patients or validation by patients of words selected by clinicians.

Nevertheless, our data suggest that IBS patients preferentially process symptom-related words presented subliminally, which may have important implications for the understanding of cognitive-behavioral mechanisms in IBS. Cognitive behavioral therapy (CBT) has been shown to be effective in some patients with IBS (19), and it would be interesting to see whether the subliminal responses we have seen are modified by CBT and whether this influences treatment response. In addition, these data suggest future, novel interventions for IBS. If the processing biases we observed are shown to be a reliable feature of IBS disease status and predictive of disease trajectory, then interventions to modify these biases may be clinically effective. Similar techniques have been used in the context of anxiety-related disorders, with impressive early success (20). Future studies should use longitudinal study designs in order to assess the predictive value of processing biases in IBS patients, recruit control groups from other disease populations, and use additional measures of attentional bias.

	NOTES

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION NOTES REFERENCES

 

Received for publication December 16, 2005; revision received March 13, 2006.

DOI:10.1097/01.psy.0000232270.78071.28

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION NOTES REFERENCES

 

1. Hungin AP, Whorwell PJ, Tack J, Mearin F. The prevalence, patterns and impact of irritable bowel syndrome: an international survey of 40,000 subjects. Aliment Pharmacol Ther 2003;17:643–50.[CrossRef][Medline]
2. Lydiard RB. Irritable bowel syndrome, anxiety, and depression: what are the links? J Clin Psychiatry 2001;62(suppl 8):38–45.
3. Tillisch K, Mayer EA. Pain perception in irritable bowel syndrome. CNS Spectr 2005;10:877–82.[Medline]
4. Williams JM, Mathews A, MacLeod C. The emotional Stroop task and psychopathology. Psychol Bull 1996;120:3–24.[CrossRef][Medline]
5. Mogg K, Kentish J, Bradley BP. Effects of anxiety and awareness on colour-identification latencies for emotional words. Behav Res Ther 1993;31:559–67.[CrossRef][Medline]
6. Mathews A, MacLeod C. Cognitive approaches to emotion and emotional disorders. Annu Rev Psychol 1994;45:25–50.[CrossRef][Medline]
7. Munafo MR, Stevenson J. Selective processing of threat-related cues in day surgery patients and prediction of post-operative pain. Br J Health Psychol 2003;8:439–49.[CrossRef][Medline]
8. Keogh E, Ellery D, Hunt C, Hannent I. Selective attentional bias for pain-related stimuli amongst pain fearful individuals. Pain 2001;91:91–100.[CrossRef][Medline]
9. Keogh E, Thompson T, Hannent I. Selective attentional bias, conscious awareness and the fear of pain. Pain 2003;104:85–91.[CrossRef][Medline]
10. Gomborone JE, Dewsnap PA, Libby GW, Farthing MJ. Selective affective biasing in recognition memory in the irritable bowel syndrome. Gut 1993;34:1230–3.[Abstract/Free Full Text]
11. Gibbs-Gallagher N, Palsson OS, Levy RL, Meyer K, Drossman DA, Whitehead WE. Selective recall of gastrointestinal-sensation words: evidence for a cognitive-behavioral contribution to irritable bowel syndrome. Am J Gastroenterol 2001;96:1133–8.[CrossRef][Medline]
12. Toner BB. Cognitive-behavioral treatment of irritable bowel syndrome. CNS Spectr 2005;10:883–90.[Medline]
13. Munafo M, Mogg K, Roberts S, Bradley BP, Murphy M. Selective processing of smoking-related cues in current smokers, ex-smokers and never-smokers on the modified Stroop task. J Psychopharmacol 2003;17:310–6.[Abstract/Free Full Text]
14. Ratcliff R. Methods for dealing with reaction time outliers. Psychol Bull 1993;114:510–32.[CrossRef][Medline]
15. Labus JS, Bolus R, Chang L, Wiklund I, Naesdal J, Mayer EA, Naliboff BD. The Visceral Sensitivity Index: development and validation of a gastrointestinal symptom-specific anxiety scale. Aliment Pharmacol Ther 2004;20:89–97.[Medline]
16. Drossman DA. The Rome criteria process: diagnosis and legitimization of irritable bowel syndrome. Am J Gastroenterol 1999;94:2803–7.[Medline]
17. Francis CY, Morris J, Whorwell PJ. The irritable bowel severity scoring system: a simple method of monitoring irritable bowel syndrome and its progress. Aliment Pharmacol Ther 1997;11:395–402.[CrossRef][Medline]
18. Mogg K, Bradley BP. Some methodological issues in assessing attentional biases for threatening faces in anxiety: a replication study using a modified version of the probe detection task. Behav Res Ther 1999;37:595–604.[CrossRef][Medline]
19. Toner BB, Segal ZV, Emmott S, Myran D, Ali A, DiGasbarro I, Stuckless N. Cognitive-behavioral group therapy for patients with irritable bowel syndrome. Int J Group Psychother 1998;48:215–43.[Medline]
20. MacLeod C, Rutherford E, Campbell L, Ebsworthy G, Holker L. Selective attention and emotional vulnerability: assessing the causal basis of their association through the experimental manipulation of attentional bias. J Abnorm Psychol 2002;111:107–23.[CrossRef][Medline]

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Afzal, M. Articles by Munafò, M. R. Search for Related Content PubMed PubMed Citation Articles by Afzal, M. Articles by Munafò, M. R. Related Collections Gastrointestinal Cognitive Functioning