Sixty-four IBS patients were recruited from an outpatient clinic, which is only devoted to gastrointestinal functional disorders. The study protocol had previously been approved by the ethical committee of Federico II University of Naples and therefore had been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki. All participants signed a consent form prior to being enrolled in the study.
The diagnosis of IBS was made on the basis of the Rome III criteria  together with the exclusion of any other organic disease. The severity of IBS was scored using the validated Functional Bowel Disorders Severity Index (FBDSI), developed by Drossman et al., which provides an easy-to-use scale to appraise illness severity in this complex group of patients. It is comprised of three variables: current pain (by 0–100 visual analog scale (VAS)), diagnosis of chronic abdominal pain and the number of physician visits within the past 6 months. Severity was rated as mild (1–36 points), moderate (37–110 points) and severe (≥111 points). Details of the index calculation are presented in the original article .
The eligibility criteria included females and males between the ages of 18 and 75 having the ability to understand and the willingness to comply to the study procedures; females with childbearing potential were required to have a negative pregnancy blood test within 48 h of the two colonic transit studies. They were also required to have an average daily abdominal bloating score or flatulence ≥24 on a 100-mm VAS during the 2-week run-in period. Patients were excluded if they had serious, unstable medical condition, insulin-dependent diabetes mellitus, major psychiatric diagnosis, previous history of drug or alcohol abuse 6 months prior to screening, a diagnosis of lactase deficiency or if symptoms resolved with lactose-free diet and if they had undergone previous abdominal surgery except appendectomy, caesarean section, tubal ligation, laparoscopic cholecystectomy, hysterectomy or abdominal wall hernia repair. They were also excluded if they currently used medications that may alter gastrointestinal motility, or long-term antibiotics or proton pump inhibitors. Patients were also carefully interrogated on use of over-the-counter medication within 3 days of transit studies and antibiotic use 1 month prior to screening.
The following medications were permitted during the course of the study, as long as they had been used at a constant dosage and were commenced at least 1 month prior to the start of the 2-week run-in period: birth control pill or depot intramuscular contraceptive preparation, oestrogen–progesterone replacement therapy, l-thyroxine, low-dose antidepressants (up to 25 mg day−1 of amitriptyline, nortriptyline or selective serotonin reuptake inhibitor) or antihypertensives in the diuretic, angiotensin-converting enzyme inhibitor or angiotensin II inhibitor classes.
A parallel-group, double-blinded, placebo-controlled study with a total of 64 patients being randomised to either placebo (n = 32) or symbiotic (n = 32) was planned. Randomisation was carried out according to a computer-generated, blocked randomisation list independent of the research group (Statistical Department of Istituto Di Sanità, Roma) and with a concealed block size of 4. The patients, the investigators, the doctors and the study nurse were blinded using randomisation codes, which were kept confidential until the end of data analysis. Eligible patients were required to pursue a baseline 2-week run-in period recording daily symptoms by means of a diary and, followed by a 4-week treatment period (Fig. 1). All patients underwent colonic transit measurement and filled in the quality of life questionnaire at the end of run-in and of the treatment phase.
The symbiotic mixture (Probinul, CaDi Group, Rome, Italy) contains lyophilised bacteria (5 × 109
Lactobacillus plantarum, 2 × 109
Lactobacillus casei subp. rhamnosus and 2 × 109
Lactobacillus gasseri, 1 × 109
Bifidobacterium infantis and 1 × 109
Bifidobacterium longum, 1 × 109
Lactobacillus acidophilus, 1 × 109
Lactobacillus salivarus and 1 × 109
Lactobacillus sporogenes and 5 × 109
Streptococcus termophilus), prebiotic inulin (2.2 g; VB Beneo Synergy 1) and 1.3 g of tapioca-resistant starch. The study preparation was provided in a powder form (5-g sachets) containing the symbiotic mixture or the matching placebo (CaDi Group, Rome, Italy) comparable in colour, texture and taste to the symbiotic mixture. The patients were instructed to ingest the investigational powder preparation twice daily, preferably in the morning and in the evening far from meals, dissolving it in water.
Assessment of symptoms and bowel functions
Patients filled in a daily diary to evaluate stool frequency, bowel function and symptoms. Bowel function was assessed on validated adjectival scales: stool consistency by Bristol stool form scale [18–21], and the sensation of incomplete evacuation was evaluated by a binomial answer (yes or no). Bowel symptoms were bloating, sensation of flatulence, pain and faecal urgency which were assessed on 100-mm VAS. Moreover, they tracked the intake of all medications during the trial. If they received antibiotics for acute infections (for example bacterial throat or urinary infections), the symptom data for the time of antibiotic administration and 1 week after completion were excluded from analysis and the treatment period was prolonged to ensure availability of the requisite number of weeks of study treatment. At the end of each week, patients were required to record a weekly response to a question on the global satisfactory relief of abdominal bloating and sensation of flatulence by a single statement (yes or no).
The patients recorded all the food they had eaten during two weekdays in the diary during the run-in and the treatment period. Nutrients were manually calculated.
The patients reported the daily consumption of the study product in the diary. The compliance was calculated as percent of planned ingestion of the study product, and compliance above 80 % was set as a minimum requirement in order to be regarded as acceptable.
Tolerability and safety
Tolerability and safety were assessed by recording all reported adverse events.
Assessment of colonic transit time
Total and segmental colonic transit time (CTT) was assessed according to Metcalf et al. . The abdominal X-rays were recorded at 9:00 AM on the day after 3 days ingestion of 20 differently shaped radioopaque markers at 9:00 AM each day (Marquat Genie Biomedical, Boissy-St Léger, France). The X-rays were taken using a rapid high-kilovoltage technique to reduce radiation exposure to less than 0.5 mSv. The markers located to the right of the vertebral spinous process and above an imaginary line running from the fifth lumbar vertebra to the pelvic outlet were assigned to the right colon; those left of the vertebral spinous process and above an imaginary line running from the fifth lumbar vertebra to the anterior superior iliac crest were assigned to the left colon; and those below an imaginary line running from the pelvic brim on the right to the anterosuperior iliac crest on the left were judged to be in the rectosigmoid. Total and segmental CTT (left side, right side and rectosigmoid) were calculated according to the formula: Colonic transit (hours) = 1.2 × number of markers. A Spearman correlation explored the association between Bristol stool form scale and colonic transit time and any significant change in symptoms.
Assessment of quality of life
Before the start of the randomisation period and at the end of it, the patients answered a self-administrated quality of life questionnaire, SF-36 (validated Italian version) . This is a generic measure of perceived health status, widely used in medical and health service research, that incorporates behavioural functioning, subjective well-being and perception of health by assessing eight health concepts: physical function (how patients perceive their ability to perform physical tasks), role-physical (how patients perceive their ability to fulfil their life role physically), bodily pain (how patients perceive their level of pain), general health (how patients perceive their overall health and well-being), vitality (how patients perceive their level of ‘energy’), social function (how patients perceive their ability to participate in social activities), role-emotional (how patients perceive their ability to fulfil their life role emotionally) and mental health (how patients perceive their emotional and psychological well-being). The scores on all scales range from 0 to 100, with higher scores reflecting better health.
The primary endpoint for analysis in this study was the global satisfactory relief of bloating and flatulence. The primary analysis was a comparison by Fisher's exact test of the proportion of responders treated with symbiotic mixture vs placebo. Responders were defined as individuals whose response to the weekly global satisfactory relief questions was yes on at least 50 % of weekly assessment of the 4-week randomisation .
The secondary endpoints included severity of VAS scores for symptoms and bowel function scores. The daily diary was used to appraise the bowel function scores (frequency, consistency and the proportion of days with the sense of incomplete evacuation) and the severity of individual symptoms (abdominal bloating, sensation of flatulence, pain and urgency). The VAS symptom scores are rounded to the nearest integer in millimetres. The adjectival scales for stool consistency (Bristol stool form scale) were recorded as numerical values. The data on bowel function scores and VAS scores of individual symptoms were summarised weekly as the means and standard errors in each treatment group.
In the secondary analysis, the comparison of the two groups focused on contrasting week-by-week symptoms and bowel function scores during the treatment period using a repeated-measures analysis of covariance (ANCOVA) model with the corresponding value at baseline as the covariate. Age was also included as a covariate for analyses of the symptoms. ANCOVA was used to account for the potential effects of differences in the mean baseline measurements and scores between the two treatment groups .
The remaining total and segmental CTT and eight-item SF-36 scores were compared within groups with paired t test or Wilcoxon signed-rank test for paired comparison samples, as warranted. Post-treatment between group values were computed based on two-sample t test or Mann–Whitney U (M-W) for unpaired comparisons, as warranted.
Data are presented as mean (±standard error (SE)), unless otherwise indicated; Significance was expressed at p < 0.05 level. The SPSS software package for Windows (release 18.0; SPSS Inc, Chicago, IL, USA) was used for statistical analysis.
From previous clinical trials [24, 25] that considered the intention to be able to detect a 30 % difference in the proportion of responders between the two groups with 80 % power at a = 0.05, a minimum of 60 patients would be required. Our study was designed to include 64 patients, 32 per treatment group. Knowing that this was a very optimistic calculation based on previous clinical trials, we also planned to perform sample size calculations for future studies based on the proportion of responders in our treatment group.