Original Research

Randomized, Double-Blind Placebo-Controlled Trial to Assess the Effect of Probiotics on Irritable Bowel Syndrome in Veterans With Gulf War Illness

Author and Disclosure Information



Randomization was computer generated and maintained by the study pharmacist so that study personnel and patients were blinded to the trial groups. All investigators were blinded and allocation was concealed. The medication was supplied in a numbered container by the pharmacist after patient enrollment. After a 2-week run-in period, veterans were randomized (1:1) to receive either 1 sachet of probiotic (De Simone Formulation; formally known as VSL#3 and Visbiome) or placebo once daily for 8 weeks.

Each probiotic packet contains 900 billion probiotic bacteria per sachet.11 This formulation contained 8 viable strains of bacteria: 4 strains of Lactobacillus (L acidophilus, L plantarum, L paracasei, L delbrueckii subsp. bulgaricus); 3 strains of Bifidobacteria (Bifidobacterium breve, B lactis, B infantis); and 1 strain of Streptococcus thermophilus. This formulation had been commercialized and studied as VSL#3 and is currently available in the United States under the Visbiome trade name. While branding changed during the study, the formulation did not. The investigational medicine (VSL#3, Visbiome, and placebo) were shipped from the manufacturer Dupont/Danisco in Madison, Wisconsin. The subjects received placebo or probiotic (VSL#3/Visbiome) and both were identical in appearance. The medication was supplied in a numbered container by the pharmacist after patient enrollment.


Veterans completed the bowel disease questionnaire to record baseline bowel habits.22 All veterans recorded daily bowel symptoms to confirm the presence of IBS during the 2-week pretreatment period, at baseline, and at the end of the 8-week treatment. The symptoms assessed included severity of abdominal pain (0, none to 100, severe); severity of bloating (0, none to 100, severe); stool frequency; Bristol stool scale (1, very hard to 7, watery); severity of diarrhea (0, none to 100, severe); severity of constipation (0, none to 100, severe); satisfaction with bowel habits (0, none to 100, severe); and IBS affecting or interfering with life (0, none to 100, severe). The bowel symptom score is the sum of the 5 symptom scores.23,24

IBS-specific QOL (IBS-QOL) was recorded at baseline and at the end of treatment.25 The IBS-QOL consists of a 34-item validated disease-specific questionnaire that measures 8 domains relevant to subjects with IBS: dysphoria, interference with activity, body image, health worry, food avoidance, social reaction, sexual life, and relationships. We used the Somatic Symptom Checklist to detect the following extra-intestinal symptoms that are common among veterans with GWI: headache, backache, wheeziness, insomnia, bad breath, fatigue, general stiffness, dizziness, weakness, sensitivity to hot and cold, palpitation, and tightness in chest. Subjects rated symptoms on a scale of 1 to 5: how often (1, none; 2, monthly; 3, once weekly; 4, several times weekly; 5, daily), and how bothersome (1, not at all to 5, extremely).26

Subjects completed the Posttraumatic Stress Disorder (PTSD) Checklist–Military, which is specific to military experience with 17 items on a 1 to 5 scale (1, not at all to 5, extremely). Scores were summed to produce a total symptom severity score (range, 17-85).27 Subjects also completed the Brief Symptom Inventory 18 (BSI-18) during the baseline evaluation.28 BSI-18 measures subjects’ reported overall psychological distress. It assesses 3 symptoms dimensions (somatization, depression, and anxiety) and a global severity index. The raw scores were transferred to normative T scores based on samples of nonpatient normal men and women.

Trial Flowchart

Symptom data were compared after 8 weeks of treatment. The primary study endpoint was change in bowel symptom score. The secondary endpoints were mean change in symptoms, QOL, extra-intestinal symptoms, and PTSD score. The study was approved by the Salt Lake City Veterans Affairs Medical Center and the University of Utah Institutional Review Board and registered in ClinicalTrials.gov (NCT03078530).

Statistical Methods

Comparisons of the probiotic vs placebo groups for demographic variable were analyzed using a 2-sample t test for continuous variables, and with a χ2 test or Fisher exact test for categorical variables. The primary and secondary outcome variables were recorded daily for 2 weeks as pretreatment baseline and for 2 weeks at the end of treatment. These symptoms were recorded as ordered categorical variables, which were then averaged across the week to produce a continuous measurement for statistical analysis. For the primary outcome of GI symptoms, posttreatment comparisons were made between the study groups using a 2-sample t test of the baseline vs posttreatment values. All P values were calculated for 2-sided comparisons. The planned sample size in our study protocol was to recruit 40 individuals per group in order to achieve 80% power to detect a 30% improvement between baseline and end of treatment in the primary bowel symptom score. This study recruited 53 subjects. With this sample size, the study had 80% power to detect a 0.8 SD in any of the outcomes.


Recommended Reading