Subjects and Methods
Study Subjects
A randomised double-blind, crossover, placebo-controlled study was conducted over a period of 2 years from October 2010, in the Tayside area of Scotland. Healthy volunteers aged 65–90 years, with BMI 18.5–30.0 kg/m, were recruited by flyers and posters from retirement homes and the local community. Exclusion criteria included (i) asplenia and other acquired or congenital immunodeficiencies; (ii) autoimmune diseases; (iii) connective tissue disease; (iv) self-reported symptoms of acute or recent infection, including antibiotic use within the previous 3 months; (v) taking probiotics or prebiotics, including lactulose, for constipation; (vi), chronic gastrointestinal problems (inflammatory bowel disease, irritable bowel syndrome, cancer); and (vii) use of immunosuppressive or anti-inflammatory drugs. Each volunteer gave written informed consent, and the trial protocol was assessed and approved by the Tayside Committee for Research Ethics. The study was registered at clinicaltrials.gov as NCT01226212.
Random Assignment and Blinding
Random assignment was carried out by an independent clinician before volunteer enrolment. Study numbers were assigned and randomised to start on either the synbiotic or placebo, using a table of random digits. Participants, retirement home staff, research nurses and investigators were blinded to the allocation of synbiotic or placebo. To ensure this, the appearance and packaging of the synbiotic and placebo were identical, and were distributed to the research nurse by an independent administrator, who was not a study participant. The research nurse delivered these to the volunteers at home. Assignment of the synbiotic, or placebo, was consecutive to the first free treatment (not yet allocated) on the randomisation sheet.
Study Design
All of the study visits and enrolment were carried out by the research nurse. At the initial visits, the research nurse assigned study numbers, measured heights and weights, and explained details of the trial, while the volunteers completed a Medical and Lifestyle Questionnaire (Gastroenterology Department, Ninewells Hospital, Dundee, UK). They were then instructed on how to complete the daily bowel habit diary, which commenced 1 week before trial start. This involved the volunteers' assessment of abdominal pain, stool consistency (hard, soft or loose) and stool frequency. At the second visit, the prestudy bowel habit diary was collected, and new bowel habit and compliance diaries were supplied to the volunteers. Fasting venous blood samples were taken, and they were asked to send a faecal sample to the laboratory before commencement of the actual trial period. The study lasted for 12 weeks. Volunteers commenced the synbiotic or placebo period for 4 weeks, followed by a 4-week washout, before switching to the opposite feeding regimen for a further 4 weeks.
The synbiotic comprised ca. 2 × 10 freeze-dried viable B. longum in a gelatin capsule, and 6 g of the prebiotic Synergy I, twice daily. Synergy 1™ is composed of a mixture of inulin and oligofructose (DP2–60). Placebos were provided in identical capsules, and contained potato starch and six grams of powdered maltodextrose, which are completely absorbed in the upper gut. The synbiotic/placebo was taken after breakfast and following the evening meal, to minimise killing of the probiotic by gastric acid. Bowel habit and compliance diaries were completed throughout the study, and weight and general health of the subjects were assessed at every visit by the research nurse. Peripheral fasting bloods were taken at home by the research nurse, at the start, mid-point and end of each trial period for analysis of immune modulators. Results were retained in participant case notes, and recorded in an anonymised database. Stool samples were obtained at the same time points for microbiological analysis by fluorescent in situ hybridisation (FISH), and measurements of bacterial fermentation products. Stools were placed in a sterile plastic bag in a plastic container and collected by taxi from the volunteers' residences. Samples were delivered to the laboratory within 1 h.
Health assessments included full blood counts, blood lipids, blood glucose, blood insulin, immunoglobulins, C-reactive protein (CRP) and the Medical and Lifestyle Questionnaires. At the end of each study period, the participants were asked to complete questionnaires on bowel habit and general wellbeing. Bowel habit scores were on a 1-point scale from −3, indicating an extreme decrease in frequency of bowel movements, 0, indicating normal frequency and +3, a substantial increase in bowel movements. The general well-being scores were on a 1-point scale from −2, indicating feeling poorly, 0, indicating no change in energy levels or well-being to +3, indicating a substantial increase in well-being and energy levels. Compliance in consumption of the synbiotic or placebo was verified at each visit by the study nurse, as well as on the basis of daily records and returns of the remaining products.
Primary Objective
The principal outcome measurement of the study was an increase in faecal bifidobacteria in volunteers consuming the synbiotic compared with the placebo. Secondary outcomes were improvements in bacterial composition of the elderly colonic microbiota, inflammatory markers linked to ageing, and bowel habit and health status. Patients were withdrawn if they required antibiotics for any reason, or for noncompliance. Any adverse events were reported to the ethics committee that approved the study. Patients were also allowed to withdraw without giving a reason, in accordance with the ethics committee guidelines.
Manufacture of the Probiotic
The probiotic preparation was prepared and packaged in the University Microbiology Laboratory in Ninewells Hospital, as described previously. All microbial standardisation, quantitation and purity tests were checked by two independent microbiologists.
Sample Size Calculation
The sample size was determined on the basis of published data on bacterial populations in elderly people. The standard deviation for bifidobacteria was log10 1.2. Thus, to achieve 90% power at a two-sided 5% significance level, and to detect a minimum difference of 1.0 log in bifidobacterial numbers between the synbiotic and placebo periods, it was calculated that a sample size of 33 subjects would be required (MGH Biostatistics Software, Massachusetts General Hospital, Boston, MA, USA) in the crossover study. Therefore, to allow for dropouts of people that could not complete the investigation, it was planned to recruit 40 volunteers.
Microbiological Analysis
Measurements of faecal bacterial communities were carried out by FISH, using group-, genus- and species-specific probe sets, together with a eubacterial probe to assess numbers of total bacteria, as detailed in Child et al. Phyla were determined by combinations of probes; Firmicutes (Clostridium cluster XIVa, F. prausnitzii group, ruminococci, Roseburia intestinalis, lactic acid bacteria), Bacteroidetes (bacteroides/prevotella), Actinobacteria (Atopobium group, bifidobacteria), Proteobacteria (Enterobacteriaceae, desulphovibrio). Bifidobacterial-specific probes used in the study were as described previously. Probes were purchased from Thermohybaid, Interactiva Division (Ulm, Germany), and 5'-labelled with the fluorochrome Cy3.
Clinical Measurements
Fasting venous bloods were used for measurements of CRP, full blood counts, blood lipids, blood glucose, blood insulin and immunoglobulins (Department of Biochemical Medicine, Ninewells Hospital, Dundee, UK).
Immune Parameters
For cytokine and chemokine analysis, blood samples were collected and whole blood cultures were carried out in flat-bottom microtiter plates, using methods described previously. Blood was diluted 1/10 in RPMI 1640-glutamine medium (Gibco, Life Technologies, Paisley, UK) supplemented with 1% penicillin/streptomycin (Gibco) distributed in 24-well plates (Corning Costar, High Wycombe, UK), and stimulated by LPS, 10 μg/mL for IL-1β, IL-6, TNF-α, or with phytohaemagglutinin (10.0 μg/mL) for measurements of IL-10, IFN-γ, IL-4, IL-8 and MCP-1. Plates were incubated for 24 h. One set of unstimulated wells served as controls. The plates were centrifuged (900g, 5 min), and supernatants harvested and frozen at −20 °C until use. Production of cytokines was measured with specific immunoassays, according to the manufacturer's instructions (R & D Systems Inc., Minneapolis, MN, USA). Limits of detection for these assays were as follows: IL-6, 0.7 pg/mL; IL-10, 3.9 pg/mL; IL-1β, 1.0 pg/mL; TNF-α, 1.6 pg/mL; IFN-γ, 8.0 pg/mL; IL-4, 10 pg/mL; IL-8, 3.5 pg/mL; and MCP-1, 1.7 pg/mL. All supernatants were tested simultaneously in duplicate.
Bacterial Metabolites
Faecal samples were diluted 1 in 10 (wt:vol) in sodium phosphate buffer (0.1 mol/L, pH 6.8) and homogenised in a Stomacher 400 (Seward, Norfolk, UK) for 2 min at standard speed. Aliquots were frozen and stored at −20°C. SCFA, lactate and succinate concentrations were measured using gas chromatography, as described previously.
Statistical Analyses
The statistical package GraphPad Prism Version 4 (La Jolla, CA, USA) was used for data analysis. Logarithms of bacterial counts were used to achieve normal distribution, and mean values ± standard deviation. As the groups were not independent, comparison of baseline, 2- and 4-week values between the synbiotic and placebo group, and within each group from baseline, were compared by a two-tailed paired t-test for normally distributed data, and a two-tailed Wilcoxon signed-rank test for data that were not normally distributed. The length of the washout period was assessed by comparing baseline values at the start of each period in the same subjects (synbiotic/placebo or placebo/synbiotic) by a two-tailed paired t-test. To determine whether the order of consumption had an effect on outcomes, values for the synbiotic and placebo periods from both arms of the study at baseline and 4 weeks were compared by a t-test for normally distributed data, or Mann–Whitney U test for nonparametric data, as the groups were independent. All tests were two-tailed. P < 0.05 was taken to indicate statistical significance.
Chemicals
Unless stated otherwise, all chemicals were obtained from Sigma (Poole, Dorset, UK).