Hypotheses:

1. Primary null hypothesis: The rate of clinical response, assessed as patient-reported global symptom rating and "adequate relief of IBS symptoms," does not differ between non-depressed IBS patients treated with the SSRI citalopram and patients treated with placebo.

2. Secondary null hypotheses:

   1. Changes in disease-related quality of life, assessed with the IBS-QOL instrument, do not differ between patients treated with the SSRI citalopram and patients treated with placebo.
   2. Changes in rectosigmoid visceral sensitivity, assessed by barostat balloon distention, do not differ between patients treated with the SSRI citalopram and patients treated with placebo.

Irritable bowel syndrome (IBS) affects an estimated 15 million Americans at a cost of $1.7 billion per year. Visceral hypersensitivity is present in many patients with IBS, but its contribution to clinical symptoms is unclear. Tricyclic antidepressants may be beneficial in IBS, but their side effects can be unacceptable. Because they are better tolerated, selective serotonin reuptake inhibitors (SSRIs) are often used to treat IBS, but their efficacy in IBS has not been examined in controlled studies. We propose a randomized, placebo-controlled trial of SSRI treatment in IBS. Non-depressed patients will be studied in order to assess SSRI effects on IBS independent of depression. Our specific aims are: 1) To determine whether the SSRI citalopram is superior to placebo in improving clinical symptoms, disease-related quality of life, and tolerance to rectal balloon distension; 2) To assess whether symptomatic improvement is correlated with improvement in quality of life and/or visceral sensitivity. Subjects will fulfill Rome II IBS criteria, will have normal screening studies, and will not be depressed or on antidepressants. Global and specific symptoms, and satisfaction will be rated daily during a 1-week baseline. Subjects will then be randomized in concealed, double-blind fashion to citalopram or placebo, will complete the validated IBS-QOL instrument, and will undergo rectal compliance and sensory testing with a barostat. Subjects will be treated and will rate symptoms and satisfaction weekly for a total of 8 weeks, and also daily during the final week for comparison with the baseline. At study end, subjects will again complete the IBS-QOL and undergo a barostat study. For the primary outcome, we estimate that to detect a standardized effect size of 0.9 in global symptom rating with 2-sided α=0.05 and β=0.1, 54 subjects are needed. We plan to enroll 60 subjects, which will allow detection of an odds ratio for response (adequate relief) of 4.5 with 2-sided α=0.05 and β=0.2. If the odds ratio for this dichotomous outcome is smaller, this study will provide pilot data for a larger trial. Clinical symptoms are expected to fluctuate. Even if citalopram is not superior to placebo, prospectively collected data will illuminate the relationship between symptoms and visceral sensitivity, and the placebo effect.

Inclusion Criteria:

Inclusion criteria are:

1. Fulfilling Rome II IBS definition;
2. Age ≥18 yrs and able to give informed consent;
3. Normal sigmoidoscopy, colonoscopy or barium enema within 5 years, normal complete blood count and thyroid studies, and negative stool ova and parasite exam for patients with diarrhea.

Exclusion Criteria:

Exclusion criteria are:

1. Current psychiatric diagnosis or active treatment with antidepressants;
2. Pregnancy;
3. Major systemic illness, or illness that could explain IBS-like symptoms;
4. Active IBS therapy other than fiber or loperamide.

8 weeks

During the study, subjects were asked to call a research coordinator weekly to report completion of questionnaires. If the telephone call was not received, a research coordinator contacted the subject. Subjects were asked to call if they experienced side effects.

Results have been entered as available in the published manuscript; the PI is no longer at UCSF. This trial has the limitation of relatively small sample size. A potential limitation is that patients could have been unblinded based on side effects.