Short-term and Long-term Efficacy of Psychological Therapies for Irritable Bowel Syndrome: A Systematic Review and Meta-analysis

Open AccessPublished:December 22, 2015DOI:https://doi.org/10.1016/j.cgh.2015.11.020

      Background & Aims

      Several meta-analyses have demonstrated the efficacy of psychological therapies for reducing gastrointestinal (GI) symptoms in patients with irritable bowel syndrome (IBS). However, no meta-analysis has investigated the duration of these effects. We performed a meta-analysis to assess the immediate, short-term, and long-term efficacy of psychotherapy for reducing GI symptoms in adults with IBS.

      Methods

      We searched PubMed, PsycINFO, Science Direct, and ProQuest Dissertations and Theses through August 15, 2015 for randomized controlled trials that compared psychological therapy with an active or non-active comparison (control) condition for treatment of GI symptoms in adults with IBS.

      Results

      Forty-one trials were included in our meta-analysis, comprising data from 2290 individuals (1183 assigned to psychotherapy and 1107 assigned to a control condition). Compared with a mixed group of control conditions, psychological therapies had a medium effect on GI symptom severity (d¯ = 0.69) immediately after treatment. On average, individuals who received psychotherapy had a greater reduction in GI symptoms after treatment than 75% of individuals assigned to a control condition. After short-term follow-up periods (1–6 months after treatment) and long-term follow-up periods (6–12 months after treatment), this effect remained significant and medium in magnitude (d¯ = 0.76 and d¯ = 0.73, respectively).

      Conclusions

      Psychological therapies reduce GI symptoms in adults with IBS. These effects remained significant and medium in magnitude after short-term and long-term follow-up periods.

      Keywords

      Abbreviations used in this paper:

      CI (confidence interval), CPSR (Composite Primary Symptom Reduction), GI (gastrointestinal), IBS (irritable bowel syndrome), ITT (intent-to-treat), RCT (randomized controlled trial), SMD (standardized mean difference), TAU (treatment as usual)
      Irritable bowel syndrome (IBS) is a functional gastrointestinal (GI) disorder that affects 7%–16% of the population in the United States.
      • Lovell R.M.
      • Ford A.C.
      Global prevalence of and risk factors for irritable bowel syndrome: a meta-analysis.
      IBS is a serious public health concern, with estimates of annual direct costs in the United States ranging from $950 million to $1.35 billion and productivity costs ranging from $58 to $205 million.
      • Everhart J.E.
      • Ruhl C.E.
      Burden of digestive diseases in the United States part I: overall and upper gastrointestinal diseases.
      • Inadomi J.M.
      • Fennerty M.B.
      • Bjorkman D.
      Systematic review: the economic impact of irritable bowel syndrome.
      Quality of life in individuals with IBS is poor, particularly in the population seeking healthcare, where its influence is comparable to that of ischemic heart disease, heart failure, and diabetes.
      • Lea R.
      • Whorwell P.J.
      Quality of life in irritable bowel syndrome.
      Meta-analyses suggest that psychological therapies are as effective as antidepressants in reducing GI symptoms in IBS immediately after treatment, with a number needed to treat between 2 and 4.
      • Ford A.C.
      • Quigley E.M.
      • Lacy B.E.
      • et al.
      Effect of antidepressants and psychological therapies, including hypnotherapy, in irritable bowel syndrome: systematic review and meta-analysis.
      • Ford A.C.
      • Talley N.J.
      • Schoenfeld P.S.
      • et al.
      Efficacy of antidepressants and psychological therapies in irritable bowel syndrome: systematic review and meta-analysis.
      • Lackner J.M.
      • Mesmer C.
      • Morley S.
      • et al.
      Psychological treatments for irritable bowel syndrome: a systematic review and meta-analysis.
      Whether psychotherapy produces long-lasting changes in GI symptom severity is unknown. The aims of this review were to (1) update previous reviews, (2) expand on previous reviews by examining the longevity of these effects, and (3) investigate whether efficacy is moderated by type of psychological therapy and/or characteristics of the trial.

      Methods

      This meta-analysis was conducted in accordance with the PRISMA statement and followed an a priori established protocol. This study was exempt from ethical approval because analyses involved only de-identified data, and all studies had received local human subject protection approval.

       Search Strategy

      We searched PubMed, PsycINFO, ScienceDirect, and ProQuest Dissertations and Theses through August 15, 2015. Details are available in Supplementary Appendix 1.

       Selection Criteria

      Included studies met the following criteria: (1) randomized controlled trial (RCT), (2) evaluated a psychological intervention, (3) participants were individuals with IBS aged 18 years or older, (4) GI symptoms were assessed before and after treatment by using a continuous measure, and (5) written in English. To be as conservative and consistent in the application of our eligibility criteria as possible, trials that used quasi-randomized methods of allocation (eg, assignment based on order of entry into the trial or social security number) were excluded to the extent that this could be determined.
      • Hunt M.G.
      • Moshier S.
      • Milonova M.
      Brief cognitive-behavioral internet therapy for irritable bowel syndrome.
      • Mahvi-Shirazi M.
      • Fathi-Ashtiani A.
      • Rasoolzade-Tabatabaei S.K.
      • et al.
      Irritable bowel syndrome treatment: cognitive behavioral therapy versus medical treatment.
      • Lowen M.B.
      • Mayer E.A.
      • Sjoberg M.
      • et al.
      Effect of hypnotherapy and educational intervention on brain response to visceral stimulus in the irritable bowel syndrome.
      • van Dulmen A.M.
      • Fennis J.F.M.
      • Bleijenberg G.
      Cognitive-behavioral group therapy for irritable bowel syndrome: effects and long-term follow-up.

      Siegel KM. The effects of emotional disclosure on physical symptoms, healthcare utilization, and psychosocial adjustment in patients with irritable bowel syndrome. PhD dissertation. Alliant International University, San Diego, CA, 2003.

      Cluster randomization was considered acceptable for trials evaluating group therapies. No restrictions were placed on trial size, publication type, or follow-up interval. We included approaches grounded in a mind-body conceptual framework, including relaxation training, biofeedback, and yoga, which can be conceptualized as mindful movement.
      • Schure M.B.
      • Christopher J.
      • Christopher S.
      Mind–body medicine and the art of self-care: teaching mindfulness to counseling students through yoga, meditation, and qigong.
      Studies in which the only psychological intervention was a support group were ineligible because numerous studies used such conditions as controls.
      Control conditions were supportive therapy, education, sham treatments (for biofeedback and hypnosis), online discussion forums, enhanced medical care (medical care not received by the intervention group), treatment as usual (TAU), symptom monitoring, and wait-list controls. Enhanced medical care was only considered a control if no specific antidepressant was administered to all participants. Studies in which the only comparison was to another psychotherapy were excluded.
      Dichotomous outcome measures were ineligible because these were assumed to measure a different underlying construct than continuous measures. This resulted in 1 trial being excluded.
      • Whorwell P.J.
      • Prior A.
      • Faragher E.B.
      Controlled trial of hypnotherapy in the treatment of severe refractory irritable-bowel syndrome.
      Authors were contacted when studies provided insufficient data for effect size computation and when selective reporting was suspected. For studies including both eligible and ineligible participants, data were requested for those participants meeting our inclusion criteria. Details of our screening and coding procedures are provided in Supplementary Appendix 2.

       Outcome Assessment

      Outcomes (in order of preference, based on availability) were (1) GI symptom severity, (2) abdominal pain, and (3) GI dysfunction (constipation and diarrhea). When a trial reported an outcome by using multiple measures, data were extracted for the more frequently used measure. The GI Symptom Diary
      • Neff D.F.
      • Blanchard E.B.
      A multi-component treatment for irritable bowel syndrome.
      was the most commonly used measure. Composite diary data reported by using the Composite Primary Symptom Reduction (CPSR) could not be combined with effect sizes from other studies because this method standardizes change scores by the pretreatment mean as opposed to the change score standard deviation. When no alternative composite score was available, we used abdominal pain as our outcome. Three CPSR studies did not report outcome data in any other format and were therefore excluded.
      • Keefer L.
      • Blanchard E.B.
      The effects of relaxation response meditation on the symptoms of irritable bowel syndrome: results of a controlled treatment study.
      • Payne A.
      • Blanchard E.B.
      A controlled comparison of cognitive therapy and self-help support groups in the treatment of irritable bowel syndrome.
      • Galovski T.E.
      • Blanchard E.B.
      The treatment of irritable bowel syndrome with hypnotherapy.
      Self-reported outcomes were used in all trials except one, which only provided physician-reported data.
      • Svedlund J.
      Psychotherapy in irritable bowel syndrome: a controlled outcome study.

       Data Extraction

      Data were extracted as intent-to-treat (ITT) (ie, analyzed as randomized by using all available follow-up data) when possible. Data were extracted with imputation of missing data when possible, if the method was appropriate (ie, multiple imputation, full information maximum likelihood, expectation maximization, or last observation carried forward).
      The following intervention characteristics were coded: (1) therapy type, (2) delivery method (online, in-person, telephone, or self-help), (3) format (group, individual), and (4) dose (number of in-person sessions, average session duration, therapy duration in weeks, and session frequency). Control groups were classified as either active or non-active. Active controls included supportive therapy, online discussion forums, education, sham treatments, enhanced medical care, and TAU. Non-active controls included wait-list or symptom monitoring. Controls containing both active and non-active components were coded as active.
      • Ljotsson B.
      • Andersson G.
      • Andersson E.
      • et al.
      Acceptability, effectiveness, and cost-effectiveness of internet-based exposure treatment for irritable bowel syndrome in a clinical sample: a randomized controlled trial.
      • Ljotsson B.
      • Falk L.
      • Vesterlund A.
      • et al.
      Internet-delivered exposure and mindfulness based therapy for irritable bowel syndrome-a randomized controlled trial.

       Assessment of Risk of Bias

      The risk of bias assessment tool developed by the Cochrane Collaboration

      Higgins J, Altman D. Assessing risk of bias in included studies. In: Higgins J, Green S, eds. The Cochrane Collaboration: Cochrane handbook for systematic reviews of interventions. Version 5.0.1. 2008.

      for RCTs was used to assess the following possible sources of bias in included trials: (1) adequate generation of allocation sequence, (2) concealment of allocation to conditions, (3) blinding of participants and personnel, (4) handling of incomplete outcome data, and (5) selective outcome reporting. The tool allows for high, low, and unclear risk of bias ratings. Supplementary Appendix 3 describes our criteria for assessing risk of bias.

       Data Synthesis and Statistical Analysis

      Standardized mean differences (SMDs) were computed as a measure of effect size. The SMD is computed from the difference between the 2 groups’ mean change scores divided by their pooled standard deviation. (The variance of this effect size requires information on the pretest-posttest correlation, which was rarely reported by study authors. We therefore calculated the average of all provided pretest-posttest correlations and used this value (0.74) when no pretest-posttest correlation was provided.)
      To calculate pooled mean effect sizes, we used Comprehensive Meta-Analysis (version 3.3.070). Effect sizes were weighted by the inverse of their variances. We followed recommendations to use random effects weights when measures vary across trials.
      • Higgins J.P.T.
      • Thompson S.G.
      • Deeks J.J.
      • et al.
      Measuring inconsistency in meta-analyses.
      In addition to the SMD, 95% confidence intervals (CIs) were calculated; CIs not including 0 were considered statistically significant. Effect size magnitude was interpreted by using the guidelines of Cohen
      • Cohen J.
      Statistical power analysis for the behavioral sciences.
      for SMDs of 0.2–0.3, 0.5, and 0.8 as representing small, medium, and large effect sizes, respectively.
      Eight trials had 2 psychological treatments that were compared with the same control condition. To avoid statistical dependencies in the data, we included outcome data for only the treatment arm hypothesized by the trial authors to be more “potent” in any given analysis. When a trial had more than 1 control arm, we included outcome data for only the less active control (to facilitate comparisons to prior meta-analyses where most trials had non-active control arms
      • Lackner J.M.
      • Mesmer C.
      • Morley S.
      • et al.
      Psychological treatments for irritable bowel syndrome: a systematic review and meta-analysis.
      ).
      We tested whether there were true differences underlying the variability in effect sizes between studies (heterogeneity) or whether variability was due to chance (homogeneity).
      • Higgins J.P.T.
      • Thompson S.G.
      • Deeks J.J.
      • et al.
      Measuring inconsistency in meta-analyses.
      To assess homogeneity, we calculated the Q statistic (a measure of weighted squared deviations). A significant Q rejects the null hypothesis of homogeneity and indicates that the observed variability in effect sizes is unlikely to be due to subject-level sampling error alone. We also calculated the between-studies variance (τ2) and the ratio of true heterogeneity to total observed variation (I2).

      Results

      Of the 1162 records identified through our search, 41 unique trials
      • Svedlund J.
      Psychotherapy in irritable bowel syndrome: a controlled outcome study.
      • Blanchard E.B.
      • Greene B.
      • Scharff L.
      • et al.
      Relaxation training as a treatment for irritable bowel syndrome.
      • Blanchard E.B.
      • Lackner J.M.
      • Sanders K.
      • et al.
      A controlled evaluation of group cognitive therapy in the treatment of irritable bowel syndrome.
      • Blanchard E.B.
      • Schwarz S.P.
      • Suls J.M.
      • et al.
      Two controlled evaluations of multicomponent psychological treatment of irritable bowel syndrome.
      • Boltin D.
      • Sahar N.
      • Gil E.
      • et al.
      Gut-directed guided affective imagery as an adjunct to dietary modification in irritable bowel syndrome.
      • Boyce P.M.
      • Talley N.J.
      • Balaam B.
      • et al.
      A randomized controlled trial of cognitive behavior therapy, relaxation training, and routine clinical care for the irritable bowel syndrome.
      • Corney R.H.
      • Stanton R.
      • Newell R.
      • et al.
      Behavioural psychotherapy in the treatment of irritable bowel syndrome.
      • Craske M.G.
      • Wolitzky-Taylor K.B.
      • Labus J.
      • et al.
      A cognitive-behavioral treatment for irritable bowel syndrome using interoceptive exposure to visceral sensations.
      • Creed F.
      • Fernandes L.
      • Guthrie E.
      • et al.
      The cost-effectiveness of psychotherapy and paroxetine for severe irritable bowel syndrome.
      • Deechakawan W.
      • Cain K.C.
      • Jarrett M.E.
      • et al.
      Effect of self-management intervention on cortisol and daily stress levels in irritable bowel syndrome.
      • Farnam A.
      • Somi M.H.
      • Farhang S.
      • et al.
      The therapeutic effect of adding emotional awareness training to standard medical treatment for irritable bowel syndrome: a randomized clinical trial.
      • Fernández C.
      • Amigo I.
      Efficacy of training in stress and contingency management in cases of irritable bowel syndrome.
      • Fernández C.
      • Perez M.
      • Amigo I.
      • et al.
      Stress and contingency management in the treatment of irritable bowel syndrome.
      • Gaylord S.A.
      • Palsson O.S.
      • Garland E.L.
      • et al.
      Mindfulness training reduces the severity of irritable bowel syndrome in women: results of a randomized controlled trial.
      • Gerson C.D.
      • Gerson M.J.
      A collaborative health care model for the treatment of irritable bowel syndrome.
      • Greene B.
      • Blanchard E.B.
      Cognitive therapy for irritable bowel syndrome.
      • Guthrie E.
      • Creed F.
      • Dawson D.
      • et al.
      A controlled trial of psychological treatment for the irritable bowel syndrome.
      • Heitkemper M.M.
      • Jarrett M.E.
      • Levy R.L.
      • et al.
      Self-management for women with irritable bowel syndrome.
      • Heymann-Monnikes I.
      • Arnold R.
      • Florin I.
      • et al.
      The combination of medical treatment plus multicomponent behavioral therapy is superior to medical treatment alone in the therapy of irritable bowel syndrome.
      • Jang A.L.
      • Hwang S.-K.
      • Kim D.U.
      The effects of cognitive behavioral therapy in female nursing students with irritable bowel syndrome: a randomized trial.
      • Labus J.
      • Gupta A.
      • Gill H.
      • et al.
      Randomised clinical trial: symptoms of the irritable bowel syndrome are improved by a psycho-education group intervention.
      • Lackner J.M.
      • Jaccard J.
      • Krasner S.S.
      • et al.
      Self-administered cognitive behavior therapy for moderate to severe irritable bowel syndrome: clinical efficacy, tolerability, feasibility.
      • Lahmann C.
      • Rohricht F.
      • Sauer N.
      • et al.
      Functional relaxation as complementary therapy in irritable bowel syndrome: a randomized, controlled clinical trial.
      • Lindfors P.
      • Unge P.
      • Arvidsson P.
      • et al.
      Effects of gut-directed hypnotherapy on IBS in different clinical settings: results from two randomized, controlled trials.
      • Ljotsson B.
      • Falk L.
      • Vesterlund A.
      • et al.
      Internet-delivered exposure and mindfulness based therapy for irritable bowel syndrome: a randomized controlled trial.
      • Ljotsson B.
      • Andersson G.
      • Andersson E.
      • et al.
      Acceptability, effectiveness, and cost-effectiveness of internet-based exposure treatment for irritable bowel syndrome in a clinical sample: a randomized controlled trial.
      • Moser G.
      • Trägner S.
      • Gajowniczek E.E.
      • et al.
      Long-term success of GUT-directed group hypnosis for patients with refractory irritable bowel syndrome: a randomized controlled trial.
      • Moss-Morris R.
      • McAlpine L.
      • Didsbury L.
      • et al.
      A randomised controlled trial of a cognitive behavioural therapy-based self-management intervention for irritable bowel syndrome in primary care.
      • Oerlemans S.
      • van Cranenburgh O.
      • Herremans P.J.
      • et al.
      Intervening on cognitions and behavior in irritable bowel syndrome: a feasibility trial using PDAs.
      • Palsson O.S.
      • Turner M.J.
      • Johnson D.A.
      • et al.
      Hypnosis treatment for severe irritable bowel syndrome: investigation of mechanism and effects on symptoms.
      • Roberts L.
      • Wilson S.
      • Singh S.
      • et al.
      Gut-directed hypnotherapy for irritable bowel syndrome: piloting a primary care-based randomised controlled trial.
      • Sanders K.A.
      • Blanchard E.B.
      • Sykes M.A.
      Preliminary study of a self-administered treatment for irritable bowel syndrome: comparison to a wait list control group.
      • Shahabi L.
      • Naliboff B.D.
      • Shapiro D.
      Self-regulation evaluation of therapeutic yoga and walking for patients with irritable bowel syndrome: a pilot study.
      • Shinozaki M.
      • Kanazawa M.
      • Kano M.
      • et al.
      Effect of autogenic training on general improvement in patients with irritable bowel syndrome: a randomized controlled trial.
      • Taneja I.
      • Deepak K.K.
      • Poojary G.
      • et al.
      Yogic versus conventional treatment in diarrhea-predominant irritable bowel syndrome: a randomized control study.
      • Tkachuk G.A.
      • Graff L.A.
      • Martin G.L.
      • et al.
      Randomized controlled trial of cognitive-behavioral group therapy for irritable bowel syndrome in a medical setting.
      • van der Veek P.P.
      • van Rood Y.R.
      • Masclee A.A.
      Clinical trial: short- and long-term benefit of relaxation training for irritable bowel syndrome.
      • Vollmer A.
      • Blanchard E.B.
      Controlled comparison of individual versus group cognitive therapy for irritable bowel syndrome.
      • Zernicke K.A.
      • Campbell T.S.
      • Blustein P.K.
      • et al.
      Mindfulness-based stress reduction for the treatment of irritable bowel syndrome symptoms: a randomized wait-list controlled trial.
      were determined eligible for our meta-analysis. A study flow diagram is presented in Figure 1; a summary of all included trials is presented in Table 1. Most reports included only 1 trial, but 2 reports presented data on 2 studies and are therefore reported in 2 rows. Trials containing multiple eligible intervention groups also span multiple rows. Of the 41 trials, 20 had not been included in the most recent meta-analysis evaluating the efficacy of psychological therapies on GI symptoms.
      Figure thumbnail gr1
      Figure 1Flow diagram summarizing trial identification and selection.
      Table 1Descriptive Statistics for Characteristics of Included Studies
      First authorYearCountryNRaceSexRecruitmentYearsCriteriaTherapyDeliveryFormatSessions
      Blanchard1993US1678Local; clinicPhys+RelaxIn personIndividual10
      Blanchard2007US1648675Local; clinic17RII; phys+CognitiveIn personGroup12
      Blanchard
      Study 1.
      1992US2077Local; clinic16Phys+CognitiveIn personIndividual12
      Blanchard
      Study 2.
      1992US7666Local; clinic13Phys+CognitiveIn personIndividual12
      Boltin2015Israel3576ClinicRIII; phys+RelaxIn personIndividual8
      Boyce2003AU6981Local; clinicRICognitiveIn personIndividual8
      Boyce2003AU6981Local; clinicRIRelaxIn personIndividual8
      Corney1991UK4274ClinicPhys+BehavioralIn personIndividual10.5
      The treatment protocol for this study did not specify a standard number of sessions; by contrast, participants were “mostly seen at weekly intervals for 6-15 one-hourly sessions”.
      Craske2011US697274Local; clinicRII; phys+CognitiveIn personIndividual10
      Creed2003UK1719879clinicRIDynamicIn personIndividual8
      Deechakawan2013US1189186Local; clinicRII; physCognitiveIn personIndividual9
      Deechakawan2013US11886Local; clinicRII; physCognitiveIn person; phoneIndividual9
      Farnam2014Iran7060ClinicRIII; phys+EmotionIn personIndividual2
      Fernández2006Spain2070Clinic9Phys+BehavioralIn personIndividual6
      Fernández2006Spain2070Clinic9Phys+RelaxIn personIndividual6
      Fernández1998Spain4466Clinic8M; phys+RelaxIn personIndividual10
      Gaylord2011US7576100Online; local; clinicRII; physCognitiveIn personGroup9
      Gerson2003US2467Clinic11RICollaborativeIn personIndividual3
      Gerson2003US2467Clinic11RIDynamicIn personIndividual6
      Greene1994US2010075Local; clinic15PhysCognitiveIn personIndividual10
      Guthrie1991UK10277Clinic4Phys+DynamicIn personIndividual7
      Heitkemper2004US9587100Local; clinicRI; physCognitiveIn personIndividual8
      Heitkemper2004US9587100Local; clinicRI; physCognitiveIn personIndividual1
      Heymann-Monnikes2000Germany2688ClinicRICognitiveIn personIndividual10
      Jang2014Korea90100LocalRIIICognitiveIn personGroup8
      Labus2013US698472ClinicRII; phys+CognitiveIn personGroup5
      Lackner2008US509586Local; clinic17RII; phys+CognitiveIn personIndividual10
      Lackner2008US509586Local; clinic17RII; phys+CognitiveIn personIndividual4
      Lahmann2010Germany8066ClinicRII; phys+RelaxIn personGroup10
      Lindfors
      Study 1.
      2012Sweden9079ClinicRII; phys+HypnosisIn personIndividual12
      Lindfors
      Study 2.
      2012Sweden4881ClinicRIIHypnosisIn personIndividual12
      Ljotsson2011Sweden6174Clinic12RIIICognitiveOnlineGroup0
      Ljotsson2010Sweden8685Online; local; clinicRIII; physCognitiveOnlineIndividual0
      Moser2013Austria1279ClinicRIIIHypnosisIn personGroup10
      Moss-Morris2010AU639073ClinicRI or RII; phys+CognitiveIn person; phoneIndividual3
      Oerlemans2011NL7684Local; clinicRIIICognitiveOnlineIndividual0
      Palsson2002US3063ClinicRI; phys+HypnosisIn personIndividual7
      Roberts2006UK7385ClinicPhys+HypnosisIn personIndividual5
      Sanders2007US2810078Local; clinic16RII; physCognitiveSelf-helpIndividual0
      Shahabi2015US3589Online; local; clinicRIIIMindfulnessIn personGroup16
      Shinozaki2010Japan10152Not reportedPhys+RelaxIn personIndividual8
      Svedlund1983Sweden2269Clinic13RIIDynamicIn personIndividual10
      Taneja2004India980Local; clinicRIIMindfulnessIn personIndividual1
      Tkachuk2003US4496Clinic9RII; phys+CognitiveIn personGroup10
      van der Veek2007NL9074ClinicRIII; physRelaxIn personIndividual6
      Vollmer1998US2178Local; clinic13RICognitiveIn personGroup10
      Vollmer1998US2178Local; clinic13RICognitiveIn personIndividual10
      Zernicke2013Canada2590LocalPhysMindfulnessIn personGroup9
      NOTE. Country denotes the country in which the study was conducted: AU, Australia; NL, The Netherlands; UK, United Kingdom; US, United States. N denotes the number of participants who completed baseline study measures. Race denotes the percentage of the post-treatment sample that was white. Sex denotes the percentage of the post-treatment sample that was female (if not available, baseline demographic data were used). Recruitment denotes the method used to recruit participants: local = local advertisement; clinic = in clinic or through physician referral. Years denotes the average number of years since diagnosis or symptom onset. Sessions denotes the number of in-person sessions offered to participants in the intervention group. For online therapies, a group format indicates that participants had access to an online closed discussion forum through which participants could interact. Therapy denotes the active intervention evaluated. Emotion = emotional awareness training. Criteria denotes the diagnostic criteria used to define IBS: RI, RII, and RIII denote Rome I–III criteria, respectively; M = Manning Criteria; phys = physician diagnosed; phys+ = physician diagnosed including a physical exam; self-report = participants indicated that they had been diagnosed by a medical professional.
      a Study 1.
      b Study 2.
      c The treatment protocol for this study did not specify a standard number of sessions; by contrast, participants were “mostly seen at weekly intervals for 6-15 one-hourly sessions”.

       Risk of Bias

      Table 2 presents the risk of bias ratings. The numbers of trials meeting criteria for low risk of bias that was due to (1) allocation sequence, (2) concealment of allocation sequence, (3) blinding, (4) incomplete outcome data, and (5) selective reporting were 21, 16, 1, 22, and 34, respectively.
      Table 2Risk of Bias Ratings for All Studies
      First authorYearSequence generationAllocation concealmentBlinding% DropoutITTAttritionSelective reporting
      Blanchard1993UnclearUnclearHigh30.43NHighLow
      Blanchard2007LowUnclearHigh (Unclear)10.48YLowLow
      Blanchard
      Study 1.
      1992UnclearUnclearHigh (Unclear)0.00YLowLow
      Blanchard
      Study 2.
      1992UnclearUnclearHigh (Low)20.00NUnclearLow
      Boltin2015LowUnclear(High)2.86YLowLow
      Boyce2003LowLow(High)49.52NHighLow
      Corney1991UnclearUnclear(High)2.38YLowLow
      Craske2011LowLow(Unclear)21.81YUnclearLow
      Creed2003LowLow(High)10.53YLowLow
      Deechakawan2013LowLow(High)18.62YLowLow
      Farnam2014UnclearUnclear(High)14.29NUnclearLow
      Fernández2006UnclearUnclearHigh (High)0.00YLowHigh
      Fernández1998UnclearUnclearHigh (High)36.67YUnclearHigh
      Gaylord2011LowLow(Unclear)31.96YUnclearHigh
      Gerson2003LowLow(High)4.87YLowHigh
      Greene1994UnclearUnclearHigh10.00NUnclearLow
      Guthrie1991UnclearUnclearHigh12.75NUnclearLow
      Heitkemper2004LowLow(High)8.33YLowLow
      Heymann-Monnikes2000LowUnclear(Unclear)7.69NUnclearLow
      Jang2014LowUnclear(High)15.56YLowLow
      Labus2013UnclearHighHigh0.00YLowLow
      Lackner2008LowLowHigh20.00YLowLow
      Lahmann2010UnclearUnclear(High)2.50YLowLow
      Lindfors
      Study 1.
      2012LowLow(High)3.33YLowLow
      Lindfors
      Study 2.
      2012LowLowHigh6.25YLowLow
      Ljotsson2011LowLow(High)18.03YLowLow
      Ljotsson2010LowLow(High)5.81NUnclearLow
      Moser2013LowLow(Unclear)4.87YLowLow
      Moss-Morris2010LowLow(High)6.25YLowLow
      Oerlemans2011LowHigh(High)5.26NUnclearLow
      Palsson2002UnclearUnclearHigh20.00NUnclearHigh
      Roberts2006UnclearUnclear(High)18.00YLowLow
      Sanders2007LowLowHigh42.86NHighLow
      Shahabi2015UnclearUnclear(Unclear)22.86NHighLow
      Shinozaki2010UnclearUnclear(High)0.00YLowLow
      Svedlund1983UnclearUnclear(High)1.98NLowLow
      Taneja2004UnclearUnclear(Unclear)4.55YLowHigh
      Tkachuk2003UnclearUnclear(High)34.88NHighLow
      van der Veek2007UnclearLow(High)6.67NUnclearUnclear
      Vollmer1998UnclearUnclearHigh6.25NUnclearLow
      Zernicke2013LowUnclearHigh33.33YUnclearLow
      NOTE. Possible ratings were low, high, or unclear risk of bias. Studies with 2 control groups were rated twice for risk of bias resulting from lack of blinding (ratings for active control groups appear in parentheses). ITT indicates whether the analyses were intent-to-treat (analyzed as randomized).
      a Study 1.
      b Study 2.

       Main Effects

       Immediate post-treatment assessment

      Thirty-nine trials reported continuous data for GI symptom severity immediately after treatment, defined as the first assessment less than 1 month after treatment. Mean effects of psychotherapy on GI symptoms compared with a mixed group of control conditions at first post-treatment follow-up are shown in Figure 2. Positive effect sizes indicate greater average improvement of the treatment group compared with the control group.
      Figure thumbnail gr2
      Figure 2Mean effects of psychological therapies on GI symptoms compared with mixed group of control conditions at the first post-treatment assessment. See discussion section for details on studies with negative effect sizes.
      Compared with a mixed group of (active and non-active) control conditions, psychological therapies were effective at improving GI symptoms (d¯= 0.69, P < .001; 95% CI, 0.52–0.86; Q = 129.07, P < .001, τ2 = 0.19, I2 = 70.56) immediately after treatment. According to the criteria of Cohen,
      • Cohen J.
      Statistical power analysis for the behavioral sciences.
      this effect is medium-sized.

       Impact of Bias

      There was no evidence of significant publication bias at the first post-treatment follow-up, as indicated by the Egger test (b = 0.57, P = .583) and the rank correlation test (τ = 0.09, P = .439). Funnel plots of main effects at all follow-up intervals are provided in Supplementary Figure 1, Supplementary Figure 2, Supplementary Figure 3.
      In a series of pre-specified analyses, we investigated whether the risk of bias ratings were associated with effects of psychotherapy on GI symptoms. There was no statistically significant effect of risk of bias for the domains of sequence generation, allocation concealment, blinding, or selective reporting. There was a significant effect of risk of bias in the domain of attrition (Q = 13.15, P = .001). Specifically, effect sizes were significantly smaller among the 5 studies rated as high risk (d¯= –0.03, P = .875; 95% CI, –0.44 to 0.37) compared with unclear (d¯= 0.83, P < .001; 95% CI, 0.57–1.09; Q = 12.32, P < .001) or low (d¯= 0.73, P < .001; 95% CI, 0.49–0.96; Q = 10.15, P = .001) risk of bias. There was no significant difference in effect sizes between studies rated as having an unclear vs low risk of bias (Q = 0.33, P = .564).

       Short-term Follow-up

      Fourteen trials provided data for GI symptom severity at short-term follow-up, defined as at least 1 month but less than 6 months after treatment. There was a medium effect of psychotherapy on GI symptoms at short-term follow-up compared with mixed control conditions (d¯= 0.76, P < .001; 95% CI, 0.54–0.97; Q = 31.43, P = .003, τ2 = .10, I 2= 58.63). There was no evidence of significant publication bias at this time point, as indicated by the Egger test (b = 0.53, P = .867) and the rank correlation test (τ < 0.001, P = 1.000).

       Long-term Follow-up

      Ten trials provided data for GI symptom severity at long-term follow-up, defined as 6 months to 1 year after treatment. There was a medium-sized effect of psychotherapy on GI symptoms at long-term follow-up (d¯= 0.73, P < .001; 95% CI, 0.43–1.03; Q = 35.80, P < .001, τ2 = 0.17, I2 = 74.86). There was no evidence of significant publication bias at this time point, as indicated by the Egger test (b = 5.09, P = .193) and the rank correlation test (τ < 0.18, P = .474).

       Sensitivity Analyses

      To investigate how the effect of psychotherapy on GI symptoms changes over time within an individual trial, we re-computed effect sizes for the post-treatment time point by using data from only the 13 studies that also provided short-term follow-up data and re-computed effect sizes for short-term follow-up by using data from only the 6 studies that also provided long-term follow-up data.

       Effect size immediately after treatment vs short-term follow-up (13 trials)

      Immediately after treatment, the average effect size of the 13 studies that also included short-term follow-up was d¯= 0.63, P < 0.001; 95% CI, 0.39–0.86; Q = 35.42, P < .001, τ2 = 0.12, I2 = 66.12. At short-term follow-up, the average effect size of these 13 studies was d¯= 0.75, P < .001; 95% CI, 0.52–0.98; Q = 31.22, P = .002, τ2 = 0.11, I2 = 61.56.

       Effect size at short-term vs long-term follow-up (6 trials)

      At short-term follow-up, the average effect size of the 6 studies that also included long-term follow-up was d¯= 0.78, P < .001; 95% CI, 0.58–0.98; Q = 1.35, P = .930, τ2 < 0.001, I2 < 0.001. At long-term follow-up, the average effect size of the 6 studies that also included short-term follow-up was d¯= 0.67, P < .001; 95% CI, 0.44–0.91; Q = 6.48, P = .262, τ2 = 0.02, I2 = 22.84.
      We further investigated the effect of follow-up interval on effect size by limiting our analyses to the 5 studies providing outcome data for all 3 follow-up intervals. The average effect of these studies was d¯= 0.54, P = .001; 95% CI, 0.24–0.84 immediately after treatment, d¯= 0.76, P < .001; 95% CI, 0.54–0.98 at short-term follow-up, and d¯= 0.73, P < .001; 95% CI, 0.50–0.97 at long-term follow-up (Figure 3).
      Figure thumbnail gr3
      Figure 3Mean effects of psychological therapies on GI symptoms compared with control conditions for the 5 studies that included data for all 3 follow-up intervals by follow-up interval.
      In a series of pre-specified analyses, we investigated several intervention and trial characteristics as potential moderators of the effect of psychotherapy on GI symptoms immediately after treatment: modality, delivery method, format, dose, control type, country, and publication date.

       Therapeutic modality

      Cognitive and cognitive-behavioral therapies were evaluated in the largest number of trials (20 trials), followed by relaxation (6 trials) and hypnosis (5 trials). These 3 modalities had similar effect sizes (cognitive: d¯= 0.73, P < .001; 95% CI, 0.48–0.97; hypnosis: d¯= 0.78, P < .001; 95% CI, 0.46–1.11; relaxation: d¯= 0.72, P = .050; 95% CI, 0.00–1.43; Q = 0.08, P = .960).

       Delivery method

      Three studies investigated interventions administered online (with an online therapist). There was no statistically significant difference between the effect sizes of interventions delivered online (d¯= 1.33, P = .006; 95% CI, 0.38–2.27) vs in-person (d¯= 0.64, P < .001; 95% CI, 0.47–0.80; Q = 1.98, P = .160).

       Format

      Within therapies administered in person, there was no significant difference in the effect of interventions administered in a group (d¯= 0.56, P = .004; 95% CI, 0.18–0.94) vs individual format (d¯= 0.66, P < .001; 95% CI, 0.47–0.84; Q = 0.19, P = .666).

       Dose

      We tested whether the “dose” of talk therapy sessions (including in-person or telephone sessions) was associated with larger effect sizes. Self-help and online therapies were excluded from these analyses. Number of sessions (b = –0.02, P = .527), average session duration (b < 0.01, P = .491), therapy duration (b = 0.01, P = .886), and session frequency (b = –0.04, P =.887) were not significantly correlated with effect size.

       Type of control

      Effect sizes were similar for trials that used an active (d¯= 0.66, P < .001; 95% CI, 0.42–0.90) vs non-active control (d¯= 0.68, P < .001; 95% CI, 0.48–0.87) (Q = 0.01, P = .929).

       Country

      Studies were most frequently conducted in the United States (17 studies), Sweden (5 studies), and the United Kingdom (4 studies). Effect sizes were significantly larger in Sweden (d¯= 1.14, P < .001; 95% CI, 0.66–1.60) than in the United States (d¯= 0.54, P < .001; 95% CI, 0.31–0.77; Q = 4.99, P = .026) and the United Kingdom (d¯= 0.32, P = .014; 95% CI, 0.06–0.58; Q = 8.84, P = .003). Two Swedish studies with especially large effect sizes (d¯ = 1.97
      • Ljotsson B.
      • Andersson G.
      • Andersson E.
      • et al.
      Acceptability, effectiveness, and cost-effectiveness of internet-based exposure treatment for irritable bowel syndrome in a clinical sample: a randomized controlled trial.
      ; d¯ = 1.63
      • Ljotsson B.
      • Falk L.
      • Vesterlund A.
      • et al.
      Internet-delivered exposure and mindfulness based therapy for irritable bowel syndrome-a randomized controlled trial.
      ) were conducted by the same lead author.

       Publication date

      Date of publication was not significantly correlated with effect size (b < 0.01, P = 0.720).

      Discussion

      Our work updates previous meta-analyses on the effectiveness of psychological therapies for improving GI symptoms in adults with IBS
      • Ford A.C.
      • Quigley E.M.
      • Lacy B.E.
      • et al.
      Effect of antidepressants and psychological therapies, including hypnotherapy, in irritable bowel syndrome: systematic review and meta-analysis.
      • Lackner J.M.
      • Mesmer C.
      • Morley S.
      • et al.
      Psychological treatments for irritable bowel syndrome: a systematic review and meta-analysis.
      and provides a meta-analysis of the longevity of these effects. Determining the longevity of treatment benefits is especially important in IBS because of the recurrent, intractable nature of this disorder.
      • Canavan C.
      • West J.
      • Card T.
      The epidemiology of irritable bowel syndrome.
      The current study further extends previous work
      • Ford A.C.
      • Quigley E.M.
      • Lacy B.E.
      • et al.
      Effect of antidepressants and psychological therapies, including hypnotherapy, in irritable bowel syndrome: systematic review and meta-analysis.
      by including continuous outcome data and trials with less than a 7-day post-treatment follow-up. We also tested whether treatment characteristics (therapeutic modality, delivery method, format, and dose) and trial characteristics significantly moderated effect sizes.
      As a whole, psychological interventions included in this meta-analysis were found to significantly reduce GI symptoms in adults with IBS. This effect was medium-sized at first post-treatment assessment and was maintained at both short-term and long-term follow-up. This remained true when using data only from studies with more than 1 follow-up, suggesting that the effect of psychotherapy on GI symptoms does not decrease during the course of a 6-month to 12-month follow-up interval. In the 5 studies that provided data at all 3 follow-up intervals, effects were actually greater at short-term (d¯= 0.76) and long-term follow-up (d¯= 0.73) compared with immediately after treatment (d¯= 0.54). The average effect on GI symptoms after treatment (d¯= 0.69) in our analysis was similar to that reported previously for bowel dysfunction (d¯= 0.57) and considerably higher than that reported previously for abdominal pain (d¯= 0.27) in the only previously published meta-analysis on this topic using continuous outcomes.
      • Lackner J.M.
      • Mesmer C.
      • Morley S.
      • et al.
      Psychological treatments for irritable bowel syndrome: a systematic review and meta-analysis.
      Transforming our post-treatment effect size to the U3 Index of Smith and Glass
      • Smith M.L.
      • Glass G.V.
      Meta-analysis of psychotherapy outcome studies.
      indicated that the average individual assigned to psychotherapy experienced a greater decrease in GI symptoms than 75% of individuals assigned to a control condition.
      Cognitive, relaxation, and hypnosis therapies were the most commonly tested treatment modalities within our eligible sample of trials. The results suggest these 3 therapies may be equally effective at improving GI symptoms. However, the majority of studies used interventions that combined several different treatment modalities, making it difficult to sort studies into discrete categories. Thus, this result should be interpreted with caution.
      Meta-regression revealed no significant effect of number of therapy sessions, average duration of sessions, duration of therapy, or frequency of sessions on effect size. These results are comparable to findings from a recent meta-analysis of the efficacy of psychotherapy for depression,
      • Cuijpers P.
      • Huibers M.
      • Ebert D.D.
      • et al.
      How much psychotherapy is needed to treat depression? a metaregression analysis.
      in which number of treatment sessions was not significantly associated with effect size after controlling for other relevant study variables.
      We also observed no significant effect of delivery method (in-person vs online) on effect sizes. This could be good news for efforts to enhance dissemination of treatment; however, this result should be interpreted with caution, because only 3 online therapy trials were included. We found no significant effect of therapy format (group vs individual) on effect size. Thus, group format may be an effective and economical method of delivering psychological therapies to adults with IBS.
      Three trials provided effect sizes indicating greater improvement in the control group.
      • Shahabi L.
      • Naliboff B.D.
      • Shapiro D.
      Self-regulation evaluation of therapeutic yoga and walking for patients with irritable bowel syndrome: a pilot study.
      • Shinozaki M.
      • Kanazawa M.
      • Kano M.
      • et al.
      Effect of autogenic training on general improvement in patients with irritable bowel syndrome: a randomized controlled trial.
      • Tkachuk G.A.
      • Graff L.A.
      • Martin G.L.
      • et al.
      Randomized controlled trial of cognitive-behavioral group therapy for irritable bowel syndrome in a medical setting.
      In 2 of these trials, the treatment group improved more than the control group on a dichotomous outcome measure identified by the authors as the primary outcome.
      • Shinozaki M.
      • Kanazawa M.
      • Kano M.
      • et al.
      Effect of autogenic training on general improvement in patients with irritable bowel syndrome: a randomized controlled trial.
      • Tkachuk G.A.
      • Graff L.A.
      • Martin G.L.
      • et al.
      Randomized controlled trial of cognitive-behavioral group therapy for irritable bowel syndrome in a medical setting.
      The remaining trial
      • Shahabi L.
      • Naliboff B.D.
      • Shapiro D.
      Self-regulation evaluation of therapeutic yoga and walking for patients with irritable bowel syndrome: a pilot study.
      used walking as a comparison condition, which could arguably be conceptualized as an active treatment rather than a control condition.
      Our results were similar to a recently published meta-analysis that used binary outcome data to compare the efficacy of psychotherapy vs antidepressant medications for reducing GI symptoms in IBS. These authors estimated the risk ratio of symptoms not improving with psychotherapy compared with control as 0.68 (roughly equivalent to an odds ratio of 0.34 and d¯ of 0.60).
      • Ford A.C.
      • Quigley E.M.
      • Lacy B.E.
      • et al.
      Effect of antidepressants and psychological therapies, including hypnotherapy, in irritable bowel syndrome: systematic review and meta-analysis.
      Whether the effect of psychotherapy on GI symptoms reduces long-term healthcare costs is unknown. One RCT investigating this question found that mean annual direct healthcare costs were significantly lower after 3 months of psychotherapy, but not 3 months of paroxetine, compared with TAU.
      • Creed F.
      • Fernandes L.
      • Guthrie E.
      • et al.
      The cost-effectiveness of psychotherapy and paroxetine for severe irritable bowel syndrome.
      Furthermore, a higher percentage of individuals who received psychotherapy discontinued their disability benefits during the course of the trial (14.3%) compared with those who received paroxetine (3.6%) or TAU (4.1%). These results are encouraging, but only 3 trials included in our meta-analysis assessed healthcare utilization as an outcome. Future research should investigate the long-term impact of psychotherapy on healthcare utilization and disability.
      This systematic review and meta-analysis used rigorous methodology. Assessment of eligibility and data extraction were each performed independently by 2 authors, and ITT data were used wherever possible. Authors of potentially eligible trials were contacted to obtain unreported data or to exclude ineligible participants. This inclusive approach provided data for 41 RCTs, which is more than any other meta-analysis on this topic to date.
      Several limitations should be considered when interpreting these findings, each of which arise from the nature of the studies available for synthesis. First, there was substantial between-trial variability in many trial characteristics (eg, measures used, inclusion criteria, control conditions). In the future, keeping such variables as consistent as possible across studies will allow for more precise effect size estimates. Effects may also vary according to individual patient characteristics. Unfortunately, this was impossible to assess in the current meta-analysis because of inconsistent reporting and limited diversity (participants were mostly female and mostly white in trials that reported these data). With better reporting, future meta-analyses will be able to investigate this matter.
      Another limitation is that no trial included in the meta-analysis was rated as low risk of bias in every domain. This was partially a result of the difficulty in blinding participants in psychological trials. However, even after excluding this domain, only 9 trials were rated as low risk of bias in all remaining domains. Future studies should follow the CONSORT guidelines for reporting RCTs,
      • Schulz K.F.
      • Altman D.G.
      • Moher D.
      CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials.
      use ITT designs, use active control conditions to control for nonspecific treatment effects, and assess treatment credibility and expectancy. Finally, few studies collected short-term or long-term follow-up data. Future studies should assess treatment outcomes at short-term and long-term follow-up to determine the duration of treatment effects.
      Our finding that psychotherapy improves GI symptoms at short-term and long-term follow-up is particularly noteworthy because of the typically recurrent, persistent nature of IBS symptoms. Future research is needed to compare the longevity of treatment effects for psychotherapy vs pharmacologic therapies such as antidepressants. Although it is beyond the scope of this review, it is also important to consider the mechanisms by which psychotherapies improve GI symptoms as well as to determine the “active ingredients” responsible for this effect. Careful dismantling studies and longitudinal assessment of hypothesized mediators will facilitate the development of even more effective treatments for individuals with IBS.

      Supplementary Appendix 1. Record Search Criteria

      For PubMed, PsycINFO, and ScienceDirect searches, the following terms were required to appear in the title and/or the abstract of the record:
      (IBS OR “irritable bowel syndrome” OR “irritable bowel” OR “irritable colon” OR “spastic colon”) AND (CBT OR “cognitive behavioral” OR “cognitive behavioural” OR “psychological therapies” OR “psychological therapy” OR mindful* OR meditat* OR (psychosocial AND therapy) OR “cognitive therapy” OR “behavior therapy” OR “behaviour therapy” OR psychotherapy OR psychoeducational OR “psychological treatment” OR counseling OR (acceptance AND therapy) OR “psychological intervention” OR “mental health intervention” OR “expressive writing” OR hypnosis OR hypnot* OR “mind-body” OR (intervention AND psychol*) OR psychodynamic OR “applied behavior analysis” OR biofeedback OR “autogenic training” OR “heart rate” OR bioenergetic* OR “support group” OR “group therapy” OR autohypno* OR (emotion* AND express*) OR “T’ai Chi” OR “Tai Chi” OR “Yoga” OR (writ* AND express*) OR “relaxation therapy” OR “self-regulate” OR “self-regulation” OR “regulate emotions” OR “emotion regulation”) AND (random* OR “usual care” OR “treatment as usual” OR control* OR comparison OR placebo OR “wait-list” OR “wait list”).
      For ProQuest Dissertations and Theses, the same search criteria were used except that the term location limit was set to “ALL” (ie, “anywhere except full text”) because a “Title OR Abstract” limit was not available.
      Six hundred eighty-four articles and 47 dissertations were returned by using these criteria on January 14, 2014. Reference sections of 3 previous meta-analyses
      • Lowen M.B.
      • Mayer E.A.
      • Sjoberg M.
      • et al.
      Effect of hypnotherapy and educational intervention on brain response to visceral stimulus in the irritable bowel syndrome.
      • van Dulmen A.M.
      • Fennis J.F.M.
      • Bleijenberg G.
      Cognitive-behavioral group therapy for irritable bowel syndrome: effects and long-term follow-up.

      Siegel KM. The effects of emotional disclosure on physical symptoms, healthcare utilization, and psychosocial adjustment in patients with irritable bowel syndrome. PhD dissertation. Alliant International University, San Diego, CA, 2003.

      were also searched. In addition, the most recently published meta-analysis was forward-searched to locate trials conducted since its publication. To update this search, we conducted the same search again on August 15, 2015; this time we restricted the search to articles published in January 2014 or later. Ninety-nine articles and 1 dissertation were returned by using these criteria.

      Supplementary Appendix 2. Screening and Coding Procedures

      The first author (K.T.L.) screened titles/abstracts for relevance and evaluated all potentially relevant articles in detail to determine eligibility. L.S.W. and S.D.H. were consulted in unclear cases; disagreement was resolved through discussion. Authors were aware of the names of the study authors, results, and other publication information. Effect size data were entered by the first author (K.T.L.) and checked by another (A.C.R.). A single reviewer (K.T.L.) coded all other variables.

      Supplementary Appendix 3. Coding of Risk of Bias

      Ratings of unclear were given when there was insufficient information to permit a judgment of “high” or “low” risk.

       Sequence Generation

      Studies received either an unclear or low risk of bias because of method of allocation sequence generation because studies that used allocation methods that were not purely random were excluded. Studies were judged as having a low risk of bias in this domain if the authors described a random component in the sequence generation process (such as use of a random number generator).

       Allocation Concealment

      Studies were judged as having a low risk of bias because of allocation concealment if participants and investigators enrolling participants could not foresee condition assignment (eg, because of use of sequentially numbered, opaque, sealed envelopes).

       Blinding

      Because of the nature of psychotherapy, it is impossible for patients to be blinded to intervention content. However, it is possible to limit the potential for performance bias, attrition bias, and detection bias because of this knowledge. For this reason, we adopted the practice

      Siegel KM. The effects of emotional disclosure on physical symptoms, healthcare utilization, and psychosocial adjustment in patients with irritable bowel syndrome. PhD dissertation. Alliant International University, San Diego, CA, 2003.

      • Guthrie E.
      • Creed F.
      • Dawson D.
      • et al.
      A controlled trial of psychological treatment for the irritable bowel syndrome.
      of using patient expectancies or credibility as an indicator of risk of bias because of lack of blinding. Studies with active control groups that blinded participants to trial hypotheses were rated as having an unclear risk of bias in this domain (“unclear”, as opposed to “low” because of the possibility that study staff may have inadvertently influenced participants to believe that the psychological intervention was superior). Studies that used an active control group and assessed treatment credibility or expectancies were rated as having an unclear risk of bias because of lack of blinding. (We interpreted an assessment of treatment credibility as evidence that the authors attempted to make their control treatment credible. Incidentally, all of the studies assessing treatment credibility or expectancies reported non-statistically significant differences between their intervention and control group.) Only studies that assessed treatment expectancies or credibility after treatment and in which the treatment group had equal (zero difference) or lower expectancies or credibility than the control group received a rating of low risk in this domain. We did not code risk of bias because of lack of blinding of outcome assessors separately from bias because of blinding of participants because, with one exception, all included data were self-reported.
      • Fernández C.
      • Perez M.
      • Amigo I.
      • et al.
      Stress and contingency management in the treatment of irritable bowel syndrome.

       Attrition

      Studies providing ITT data and that had a participant dropout rate of 20% or less at the first post-treatment follow-up were coded as having a low risk of bias because of handling of incomplete outcome data. Studies that either had greater than 20% dropout or did not provide ITT data (eg, no attempt to collect follow-up data from individuals who dropped out of treatment) were coded as having an unclear risk of bias in this domain. Studies that had greater than 20% dropout and did not provide ITT data were coded as high risk in this domain.

       Selective Reporting

      Studies were rated as having a high risk of bias for selective outcome reporting if they failed to report outcomes that had been collected (as reported in the methods section, on clinicaltrials.gov, or in the Clinical Trial’s registry for the authors’ country). In cases of suspected selective reporting, authors were contacted; when authors provided all requested outcome data or informed us that a particular measure had not been collected at a particular time point, their score on risk of selective reporting was changed from high to low.
      Figure thumbnail fx1
      Supplementary Figure 1Funnel plot of standard error by standard difference in means by immediate post-treatment follow-up.
      Figure thumbnail fx2
      Supplementary Figure 2Funnel plot of standard error by standard difference in means by short-term follow-up.
      Figure thumbnail fx3
      Supplementary Figure 3Funnel plot of standard error by standard difference in means by long-term follow-up.

      References

        • Lovell R.M.
        • Ford A.C.
        Global prevalence of and risk factors for irritable bowel syndrome: a meta-analysis.
        Clin Gastroenterol Hepatol. 2012; 10: 712-721
        • Everhart J.E.
        • Ruhl C.E.
        Burden of digestive diseases in the United States part I: overall and upper gastrointestinal diseases.
        Gastroenterology. 2009; 136: 376-386
        • Inadomi J.M.
        • Fennerty M.B.
        • Bjorkman D.
        Systematic review: the economic impact of irritable bowel syndrome.
        Aliment Pharmacol Ther. 2003; 18: 671-682
        • Lea R.
        • Whorwell P.J.
        Quality of life in irritable bowel syndrome.
        Pharmacoeconomics. 2001; 19: 643-653
        • Ford A.C.
        • Quigley E.M.
        • Lacy B.E.
        • et al.
        Effect of antidepressants and psychological therapies, including hypnotherapy, in irritable bowel syndrome: systematic review and meta-analysis.
        Am J Gastroenterol. 2014; 109: 1350-1366
        • Ford A.C.
        • Talley N.J.
        • Schoenfeld P.S.
        • et al.
        Efficacy of antidepressants and psychological therapies in irritable bowel syndrome: systematic review and meta-analysis.
        Gut. 2009; 58: 367-378
        • Lackner J.M.
        • Mesmer C.
        • Morley S.
        • et al.
        Psychological treatments for irritable bowel syndrome: a systematic review and meta-analysis.
        J Consult Clin Psychol. 2004; 72: 1100
        • Hunt M.G.
        • Moshier S.
        • Milonova M.
        Brief cognitive-behavioral internet therapy for irritable bowel syndrome.
        Behav Res Ther. 2009; 47: 797-802
        • Mahvi-Shirazi M.
        • Fathi-Ashtiani A.
        • Rasoolzade-Tabatabaei S.K.
        • et al.
        Irritable bowel syndrome treatment: cognitive behavioral therapy versus medical treatment.
        Arch Med Sci. 2012; 8: 123-129
        • Lowen M.B.
        • Mayer E.A.
        • Sjoberg M.
        • et al.
        Effect of hypnotherapy and educational intervention on brain response to visceral stimulus in the irritable bowel syndrome.
        Aliment Pharmacol Ther. 2013; 37: 1184-1197
        • van Dulmen A.M.
        • Fennis J.F.M.
        • Bleijenberg G.
        Cognitive-behavioral group therapy for irritable bowel syndrome: effects and long-term follow-up.
        Psychosom Med. 1996; 58: 508-514
      1. Siegel KM. The effects of emotional disclosure on physical symptoms, healthcare utilization, and psychosocial adjustment in patients with irritable bowel syndrome. PhD dissertation. Alliant International University, San Diego, CA, 2003.

        • Schure M.B.
        • Christopher J.
        • Christopher S.
        Mind–body medicine and the art of self-care: teaching mindfulness to counseling students through yoga, meditation, and qigong.
        Journal of Counseling & Development. 2008; 86: 47-56
        • Whorwell P.J.
        • Prior A.
        • Faragher E.B.
        Controlled trial of hypnotherapy in the treatment of severe refractory irritable-bowel syndrome.
        Lancet. 1984; 2: 1232-1234
        • Neff D.F.
        • Blanchard E.B.
        A multi-component treatment for irritable bowel syndrome.
        Behavior Therapy. 1987; 18: 70-83
        • Keefer L.
        • Blanchard E.B.
        The effects of relaxation response meditation on the symptoms of irritable bowel syndrome: results of a controlled treatment study.
        Behav Res Ther. 2001; 39: 801-811
        • Payne A.
        • Blanchard E.B.
        A controlled comparison of cognitive therapy and self-help support groups in the treatment of irritable bowel syndrome.
        J Consult Clin Psychol. 1995; 63: 779-786
        • Galovski T.E.
        • Blanchard E.B.
        The treatment of irritable bowel syndrome with hypnotherapy.
        Appl Psychophysiol Biofeedback. 1998; 23: 219-232
        • Svedlund J.
        Psychotherapy in irritable bowel syndrome: a controlled outcome study.
        Acta Psychiatrica Scandinavica. 1983; 67: 7-86
        • Ljotsson B.
        • Andersson G.
        • Andersson E.
        • et al.
        Acceptability, effectiveness, and cost-effectiveness of internet-based exposure treatment for irritable bowel syndrome in a clinical sample: a randomized controlled trial.
        BMC Gastroenterol. 2011; 11: 110
        • Ljotsson B.
        • Falk L.
        • Vesterlund A.
        • et al.
        Internet-delivered exposure and mindfulness based therapy for irritable bowel syndrome-a randomized controlled trial.
        Behav Res Ther. 2010; 48: 531-539
      2. Higgins J, Altman D. Assessing risk of bias in included studies. In: Higgins J, Green S, eds. The Cochrane Collaboration: Cochrane handbook for systematic reviews of interventions. Version 5.0.1. 2008.

        • Higgins J.P.T.
        • Thompson S.G.
        • Deeks J.J.
        • et al.
        Measuring inconsistency in meta-analyses.
        BMJ. 2003; 327: 557
        • Cohen J.
        Statistical power analysis for the behavioral sciences.
        Lawrence Erlbaum Associates, Inc, Hillsdale, NJ1977
        • Blanchard E.B.
        • Greene B.
        • Scharff L.
        • et al.
        Relaxation training as a treatment for irritable bowel syndrome.
        Biofeedback and Self-regulation. 1993; 18: 125-132
        • Blanchard E.B.
        • Lackner J.M.
        • Sanders K.
        • et al.
        A controlled evaluation of group cognitive therapy in the treatment of irritable bowel syndrome.
        Behav Res Ther. 2007; 45: 633-648
        • Blanchard E.B.
        • Schwarz S.P.
        • Suls J.M.
        • et al.
        Two controlled evaluations of multicomponent psychological treatment of irritable bowel syndrome.
        Behav Res Ther. 1992; 30: 175-189
        • Boltin D.
        • Sahar N.
        • Gil E.
        • et al.
        Gut-directed guided affective imagery as an adjunct to dietary modification in irritable bowel syndrome.
        J Health Psychol. 2015; 20: 712-720
        • Boyce P.M.
        • Talley N.J.
        • Balaam B.
        • et al.
        A randomized controlled trial of cognitive behavior therapy, relaxation training, and routine clinical care for the irritable bowel syndrome.
        Am J Gastroenterol. 2003; 98: 2209-2218
        • Corney R.H.
        • Stanton R.
        • Newell R.
        • et al.
        Behavioural psychotherapy in the treatment of irritable bowel syndrome.
        J Psychosom Res. 1991; 35: 461-469
        • Craske M.G.
        • Wolitzky-Taylor K.B.
        • Labus J.
        • et al.
        A cognitive-behavioral treatment for irritable bowel syndrome using interoceptive exposure to visceral sensations.
        Behav Res Ther. 2011; 49: 413-421
        • Creed F.
        • Fernandes L.
        • Guthrie E.
        • et al.
        The cost-effectiveness of psychotherapy and paroxetine for severe irritable bowel syndrome.
        Gastroenterology. 2003; 124: 303-317
        • Deechakawan W.
        • Cain K.C.
        • Jarrett M.E.
        • et al.
        Effect of self-management intervention on cortisol and daily stress levels in irritable bowel syndrome.
        Biol Res Nurs. 2013; 15: 26-36
        • Farnam A.
        • Somi M.H.
        • Farhang S.
        • et al.
        The therapeutic effect of adding emotional awareness training to standard medical treatment for irritable bowel syndrome: a randomized clinical trial.
        J Psychiatr Pract. 2014; 20: 3-11
        • Fernández C.
        • Amigo I.
        Efficacy of training in stress and contingency management in cases of irritable bowel syndrome.
        Stress Health. 2006; 22: 285-295
        • Fernández C.
        • Perez M.
        • Amigo I.
        • et al.
        Stress and contingency management in the treatment of irritable bowel syndrome.
        Stress Med. 1998; 14: 31-42
        • Gaylord S.A.
        • Palsson O.S.
        • Garland E.L.
        • et al.
        Mindfulness training reduces the severity of irritable bowel syndrome in women: results of a randomized controlled trial.
        Am J Gastroenterol. 2011; 106: 1678-1688
        • Gerson C.D.
        • Gerson M.J.
        A collaborative health care model for the treatment of irritable bowel syndrome.
        Clin Gastroenterol Hepatol. 2003; 1: 446-452
        • Greene B.
        • Blanchard E.B.
        Cognitive therapy for irritable bowel syndrome.
        J Consult Clin Psychol. 1994; 62: 576-582
        • Guthrie E.
        • Creed F.
        • Dawson D.
        • et al.
        A controlled trial of psychological treatment for the irritable bowel syndrome.
        Gastroenterology. 1991; 100: 450-457
        • Heitkemper M.M.
        • Jarrett M.E.
        • Levy R.L.
        • et al.
        Self-management for women with irritable bowel syndrome.
        Clin Gastroenterol Hepatol. 2004; 2: 585-596
        • Heymann-Monnikes I.
        • Arnold R.
        • Florin I.
        • et al.
        The combination of medical treatment plus multicomponent behavioral therapy is superior to medical treatment alone in the therapy of irritable bowel syndrome.
        Am J Gastroenterol. 2000; 95: 981-994
        • Jang A.L.
        • Hwang S.-K.
        • Kim D.U.
        The effects of cognitive behavioral therapy in female nursing students with irritable bowel syndrome: a randomized trial.
        Eur J Gastroenterol Hepatol. 2014; 26: 918-926
        • Labus J.
        • Gupta A.
        • Gill H.
        • et al.
        Randomised clinical trial: symptoms of the irritable bowel syndrome are improved by a psycho-education group intervention.
        Aliment Pharmacol Ther. 2013; 37: 304-315
        • Lackner J.M.
        • Jaccard J.
        • Krasner S.S.
        • et al.
        Self-administered cognitive behavior therapy for moderate to severe irritable bowel syndrome: clinical efficacy, tolerability, feasibility.
        Clin Gastroenterol Hepatol. 2008; 6: 899-906
        • Lahmann C.
        • Rohricht F.
        • Sauer N.
        • et al.
        Functional relaxation as complementary therapy in irritable bowel syndrome: a randomized, controlled clinical trial.
        J Altern Complement Med. 2010; 16: 47-52
        • Lindfors P.
        • Unge P.
        • Arvidsson P.
        • et al.
        Effects of gut-directed hypnotherapy on IBS in different clinical settings: results from two randomized, controlled trials.
        Am J Gastroenterol. 2012; 107: 276-285
        • Ljotsson B.
        • Falk L.
        • Vesterlund A.
        • et al.
        Internet-delivered exposure and mindfulness based therapy for irritable bowel syndrome: a randomized controlled trial.
        Behav Res Ther. 2010; 48: 531-539
        • Ljotsson B.
        • Andersson G.
        • Andersson E.
        • et al.
        Acceptability, effectiveness, and cost-effectiveness of internet-based exposure treatment for irritable bowel syndrome in a clinical sample: a randomized controlled trial.
        BMC Gastroenterol. 2011; 11: 110
        • Moser G.
        • Trägner S.
        • Gajowniczek E.E.
        • et al.
        Long-term success of GUT-directed group hypnosis for patients with refractory irritable bowel syndrome: a randomized controlled trial.
        Am J Gastroenterol. 2013; 108: 602-609
        • Moss-Morris R.
        • McAlpine L.
        • Didsbury L.
        • et al.
        A randomised controlled trial of a cognitive behavioural therapy-based self-management intervention for irritable bowel syndrome in primary care.
        Psychol Med. 2010; 40: 94
        • Oerlemans S.
        • van Cranenburgh O.
        • Herremans P.J.
        • et al.
        Intervening on cognitions and behavior in irritable bowel syndrome: a feasibility trial using PDAs.
        J Psychosom Res. 2011; 70: 267-277
        • Palsson O.S.
        • Turner M.J.
        • Johnson D.A.
        • et al.
        Hypnosis treatment for severe irritable bowel syndrome: investigation of mechanism and effects on symptoms.
        Dig Dis Sci. 2002; 47: 2605-2614
        • Roberts L.
        • Wilson S.
        • Singh S.
        • et al.
        Gut-directed hypnotherapy for irritable bowel syndrome: piloting a primary care-based randomised controlled trial.
        Br J Gen Pract. 2006; 56: 115-121
        • Sanders K.A.
        • Blanchard E.B.
        • Sykes M.A.
        Preliminary study of a self-administered treatment for irritable bowel syndrome: comparison to a wait list control group.
        Appl Psychophysiol Biofeedback. 2007; 32: 111-119
        • Shahabi L.
        • Naliboff B.D.
        • Shapiro D.
        Self-regulation evaluation of therapeutic yoga and walking for patients with irritable bowel syndrome: a pilot study.
        Psychology, Health & Medicine. 2016; 21: 176-188
        • Shinozaki M.
        • Kanazawa M.
        • Kano M.
        • et al.
        Effect of autogenic training on general improvement in patients with irritable bowel syndrome: a randomized controlled trial.
        Appl Psychophysiol Biofeedback. 2010; 35: 189-198
        • Taneja I.
        • Deepak K.K.
        • Poojary G.
        • et al.
        Yogic versus conventional treatment in diarrhea-predominant irritable bowel syndrome: a randomized control study.
        Appl Psychophysiol Biofeedback. 2004; 29: 19-33
        • Tkachuk G.A.
        • Graff L.A.
        • Martin G.L.
        • et al.
        Randomized controlled trial of cognitive-behavioral group therapy for irritable bowel syndrome in a medical setting.
        Journal of Clinical Psychology in Medical Settings. 2003; 10: 57-69
        • van der Veek P.P.
        • van Rood Y.R.
        • Masclee A.A.
        Clinical trial: short- and long-term benefit of relaxation training for irritable bowel syndrome.
        Aliment Pharmacol Ther. 2007; 26: 943-952
        • Vollmer A.
        • Blanchard E.B.
        Controlled comparison of individual versus group cognitive therapy for irritable bowel syndrome.
        Behavior Therapy. 1998; 29: 19-33
        • Zernicke K.A.
        • Campbell T.S.
        • Blustein P.K.
        • et al.
        Mindfulness-based stress reduction for the treatment of irritable bowel syndrome symptoms: a randomized wait-list controlled trial.
        Int J Behav Med. 2013; 20: 385-396
        • Canavan C.
        • West J.
        • Card T.
        The epidemiology of irritable bowel syndrome.
        Clin Epidemiol. 2014; 6: 71
        • Smith M.L.
        • Glass G.V.
        Meta-analysis of psychotherapy outcome studies.
        Am Psychol. 1977; 32: 752
        • Cuijpers P.
        • Huibers M.
        • Ebert D.D.
        • et al.
        How much psychotherapy is needed to treat depression? a metaregression analysis.
        J Affect Disord. 2013; 149: 1-13
        • Schulz K.F.
        • Altman D.G.
        • Moher D.
        CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials.
        BMC Med. 2010; 8: 18

      Linked Article