The Effect of Probiotic Supplementation

Intervention

The intervention product (Yakult) (supplied as fermented milk) and placebo will be delivered in sealed pots of 65 mL with date stamped expiry. Yakult contains Lactobacillus casei Shirota (a minimum of 6.5 × 10 9 live cells of Lactobacillus casei Shirota are contained in each pot). The pots will be stored at 4-7 °C (domestic refrigerator) on premises at University Hospital of Leicester and the University of Leicester until provided to participants. These products have a shelf life of four weeks but fresh deliveries will be sent every two weeks. They will be stored at approximately seven degrees Celsius (refrigerated at the University and subsequently, after delivery to participants, in domestic refrigerators) in a restricted area where only members of the research team will have access to them. Participants will have supplies provided to them in person every two-weeks and will be required to ingest two pots of Yakult, every day for six months. Participants will be instructed to ingest one 65ml pot in the morning (prior to breakfast) and one bottle in the evening (prior to an evening meal). They will also be instructed to avoid any other dietary supplement aimed at modulating the gut microbiota during the six month intervention period. Researchers will keep a log of the amount of pots supplied to participants and will visit the participants at the haemodialysis unit to supply more pots.

Placebo

The placebo will be indistinguishable (identical in taste and colour but will not contain Lactobacillus casei Shirota) to both participants and trial investigators. It will be stored and provided in exactly the same manner as the intervention product.

Compliance

A record of compliance for supplement ingestion will be completed by all participants (including days where they may have missed taking the supplement). Following feedback from a research patient group, all participants will be offered any or all of the following steps in any combination to aid adherence to the product:

• Phone call reminders (daily or weekly)
• Text or email alerts at any preferred schedule
• Regular visits / reminders in person on the dialysis units

Patient Numbers - Feasibility and Statistical Power Patients will be recruited from within the Leicester Renal Network, which includes ten dialysis units treating over 800 haemodialysis patients. The number of participants required is therefore readily attainable. Based on a previously reported (Wang et al., 2015), post-intervention change (compared to pre-intervention) in serum endotoxin following probiotic supplementation in peritoneal dialysis patients (-1.11 ± 1.5 EU/mL for probiotic and 0.86 ± 2.3 endotoxin units/mL for placebo), it was calculated (Stata IC version 15.1, StataCorp, Texa, USA) that n=44 (n=50 accounting for 10% dropout for death and transplantation over 6 months) was required to detect a significant difference between probiotic and placebo groups with 90% power and alpha 0.05.

Dropout rate A drop-out rate of 10% in line with previous studies is expected. This is also entirely in-keeping with previous experience of interventional studies in the haemodialysis population which show a 10-20% drop out rate due to death, transplantation and non-adherence (Graham-Brown et al., 2016).

Randomisation The trial design will be a randomised-controlled trial (RCT). Participants will be individually randomised to either the Lactobacillus casei Shirota or a well-matched placebo. After recruitment, participants will be randomised to one of two groups using the REDCap system by the bioinformatics team within the National Institute of Health Research Biomedical Research Centre at the University of Leicester.

Blinding This trial will be conducted in a double-blind manner. Both participants and researchers will be blind to the treatment allocation. Both the Lactobacillus casei Shirota (Yakult) and the placebo will be supplied, and simply marked as, 'A' or 'B', by Yakult Honsha. Neither the researchers nor the participants will know which is Yakult and which is placebo and will therefore both be unaware which product they are taking. Yakult Honsha will have no knowledge of which patients are randomised to which group. Once all patients have completed the study, the database has been locked and statistical analyses performed, the nature of the two product groups will be revealed by un-blinding.

Main assessments (at baseline and six months) will take place over one or more sessions dependent on patient preference. Blood and saliva samples will be taken at the start of dialysis, eliminating the need for additional venepuncture; relevant demographic data will be extracted from medical records throughout the trial and also collected from participants prior to their usual haemodialysis appointment. Participants will have the choice of completing the questionnaires during their usual haemodialysis appointment (with the assistance of a researcher) or taking them home to complete at their convenience and return them to the researcher at their next haemodialysis appointment. The questionnaires will take no longer than 30 minutes to complete.

Participants will have the choice of giving their faecal sample either at their normal dialysis appointment or will be given a kit to collect their sample at home. All data can be collected from participants around their usual haemodialysis treatment so as to reduce the additional time participants need to attend over and above their out-patient appointment. Although this may require them to arrive earlier (e.g. 30 minutes) than their usual appointment time.