Systematic Review With Meta-analysis

The Prevalence of Anxiety and Depression in Patients With Irritable Bowel Syndrome

Mohammad Zamani; Shaghayegh Alizadeh-Tabari; Vahid Zamani


Aliment Pharmacol Ther. 2019;50(2):132-143. 

In This Article

Materials and Methods

Information Sources and Search Strategy

A literature search was conducted from the bibliographic databases of Embase, PubMed, Scopus, Web of Science and POPLINE published from the inception to 1 January 2019 with no language restriction. The related terms were searched in the Medical Subject Headings (MeSH) database, and finally, the keywords included "anxiety" OR "anxieties" OR "depression" OR "depressive" OR "depressions" OR "emotional" OR "emotion" OR "mental" OR "psychiatry" OR "psychology" OR "psychological" OR "psychiatric" AND "Irritable Bowel Syndrome" OR "IBS." The search was limited to Title/Abstract. Reference lists of the related reviews and the retrieved papers were manually searched for additional sources.

The present systematic review and meta-analysis was conducted according to the PRISMA (Preferred reporting items for systematic review and meta-analysis) guideline.[13] The preplanned protocol of our study was documented online in the PROSPERO registry (CRD42018108512).[14]

Inclusion and Exclusion Criteria

We included studies reporting the prevalence of anxiety/depressive symptoms/disorders in patients with IBS. To be eligible for inclusion in the analysis, every study had to recruit at least 100 adults (aged ≥15 years old). IBS patients could be diagnosed by questionnaire, Manning criteria, Rome criteria, and/or International Classification of Diseases (ICD) codes. Anxiety/depression should be diagnosed by validated methods and instruments ("symptoms" by questionnaires; "disorders" by a structured psychiatric interview, ICD codes, etc).

The exclusion criteria were as follows:

  1. Reviews, case reports, editorials, letter to the editors and abstracts from conferences.

  2. Duplicate articles or evaluating the same sample.

  3. Studies on IBS patients with concomitant diseases (eg, morbid obesity, surgery, inflammatory bowel disease, etc).

  4. Studies on patients with post-infectious IBS.

  5. Studies with sample size of <100 subjects.

  6. Studies recruited children.

  7. Surveys without clear methodology or results.

  8. Full-texts not being available.

Study Selection and Data Extraction

Two reviewers (MZ, VZ) screened independently the titles and abstracts of all references for potential suitability. Full-texts of the potential articles were obtained for the final assessment of suitability for inclusion. Any discrepancies pertaining to the inclusion of articles were resolved by consensus between the authors. Data were assessed and extracted from the studies by two authors (MZ, SA) and finally included for analysis into a Microsoft Excel spreadsheet (Microsoft Corporation, Redmond, Washington). The following data were extracted: first author's name, study location (country), publication date, diagnostic method of IBS, diagnostic method of anxiety/depression, total sample size, number of subjects by gender (if available), number of different IBS-subgroup patients (constipation-predominant IBS [IBS-C], diarrhoea-predominant IBS [IBS-D], mixed IBS [IBS-M], and undefined IBS [IBS-U], if available), number of healthy control subjects (if available), prevalence rates of anxiety/depressive symptoms/disorders. Healthy controls included the subjects who did not suffer from any functional gastrointestinal disorders, and therefore, we excluded the data on the controls who did not have IBS but had other functional gastrointestinal disorders.

Google Translate was used to translate non-English reports. When a study used different diagnostic instruments for IBS and/or anxiety/depression, we considered each one of the primary analyses as a separate report. If full-texts and/or necessary data of articles were not available, we contacted the corresponding authors by email to send us the information. We excluded duplicates and only selected those with the most comprehensive details. When prevalence rates of anxiety/depressive symptoms were not obviously reported in the papers text (for example, when the symptoms were presented as severity categories based on questionnaire scores), we calculated them using the cut-off for the significantly elevated symptoms reported by the authors based on references. In addition, if the prevalence rates were reported at different time points, we used only the data of the first evaluation. For the present study, we considered anxiety/depressive "symptoms" and anxiety/depressive "disorders" as two separate outcomes. In mental health care, there is a distinction between "symptoms" and "disorders." The symptoms in the "disorders" are more severe and are associated with impairment in life. Besides, "symptoms" are usually measured by a questionnaire, whereas "disorders" are diagnosed using the criteria of the Diagnostic and Statistical Manual of Mental Disorders in a psychiatric interview.[15]

Risk of Bias Assessment

For assessing quality of the included studies, the checklist by Hoy et al[16] was used. It has been designed for the prevalence studies and has nine criteria with two potential responses, including Yes (score 0, low risk) or No (score 1, high risk). The nine criteria were related to target population, sampling frame, selection of sample, response rate by subjects, how to data collection, case definition, study instrument, same mode of data collection, and numerators and dominators for the parameters. The minimum and maximum scores are 0 and 9. Studies with higher scores had higher risk of bias and lower quality.

Study Outcomes and Statistical Analysis

The data extracted from the retrieved articles were combined to give the pooled prevalence rates using the Inverse Variance models (including a double arcsine transformation). The pooled prevalence rates were presented as percent and 95% CI. The heterogeneity between the studies was evaluated using the I2 statistic with a cut-off of 50%, and the chi-squared test with a P < 0.10, used to define a statistically significant degree of heterogeneity.[17] Publication bias was assessed using a funnel plot and Egger's test. Data were pooled by using a random-effects model[18] to give a more conservative estimate of the prevalence rate of anxiety/depressive symptoms/disorders and the odds ratio (OR) of the symptoms/disorders in IBS patients and healthy subjects. Meta-regression was used to explore potential influence of gender (% male) and publication year on the outcomes, and P < 0.05 was considered statistically significant. Subgroup analyses were performed according to IBS subgroup (IBS-C, IBS-D, IBS-M), gender (male and female) and IBS diagnostic instrument (Rome I, II and III). A subgroup difference of P < 0.05 was considered to be indicative of significant difference between subgroups. All statistical analyses were done using STATA (StataCorp, College Station, TX).