Professional Documents
Culture Documents
Clinical Microbiology and Infection: M. Ko Łodziej, H. Szajewska
Clinical Microbiology and Infection: M. Ko Łodziej, H. Szajewska
Original article
a r t i c l e i n f o a b s t r a c t
Article history: Objectives: To assess the effectiveness of Lactobacillus reuteri DSM 17938 for the prevention of diarrhoea
Received 26 July 2018 and antibiotic-associated diarrhoea (AAD) in children.
Received in revised form Methods: Hospitalized children who received antibiotics were assigned by a computer-generated list to
14 August 2018
receive L. reuteri (at 2 108 CFU) or placebo, twice daily, for the duration of antibiotic treatment. Follow
Accepted 15 August 2018
Available online 25 August 2018
up was for 1 week after antibiotic cessation. The primary outcome measures were diarrhoea and AAD.
Both were defined according to one of three definitions (i) three or more loose or watery stools per day
Editor: L. Leibovici for 48 h; (ii) three or more loose or watery stools per day for 24 h; or (iii) two or more loose or watery
stools per day for 24 h. For AAD, it had to be diarrhoea caused by Clostridium difficile or otherwise
Keywords: unexplained diarrhoea.
Antibiotic Results: A total of 250 children were randomized and 247 were analysed (L. reuteri n ¼ 123, placebo
Lactobacillus reuteri n ¼ 124; median age 4 months). The occurrences of diarrhoea and AAD were similar in both groups,
Randomized clinical trial regardless of the definition used. Using the strictest definition (i.e. definition (i)), the occurrence of
Diarrhea
diarrhoea in the L. reuteri group was 25 (20%) compared with 16 (13%) in the placebo group (absolute risk
Children
reduction e0.07 (e0.17 to 0.02). The occurrence of AAD was 14 (11.4%) in the L. reuteri group compared
with 8 (6.5%) in the placebo group (absolute risk reduction e0.05 (e0.13 to 0.02)). The groups were
similar with respect to all secondary outcome measures, including adverse events.
Conclusions: Lactobacillus reuteri was not effective in the prevention of diarrhoea or AAD in children.
M. Kołodziej, Clin Microbiol Infect 2019;25:699
© 2018 European Society of Clinical Microbiology and Infectious Diseases. Published by Elsevier Ltd. All
rights reserved.
https://doi.org/10.1016/j.cmi.2018.08.017
1198-743X/© 2018 European Society of Clinical Microbiology and Infectious Diseases. Published by Elsevier Ltd. All rights reserved.
700 M. Kołodziej, H. Szajewska / Clinical Microbiology and Infection 25 (2019) 699e704
Children younger than 18 years who received oral or intrave- Study procedure
nous antibiotic therapy, which was started within 24 h of enrol-
ment, were included in this study. We excluded participants who Throughout the study period, healthcare providers and/or
had pre-existing acute or chronic diarrhoea, a history of chronic caregivers recorded daily the number and consistency of stools in a
gastrointestinal disease (e.g. inflammatory bowel disease, cystic standard stool diary. To record stool consistency in children
fibrosis, coeliac disease, food allergy) or other severe chronic dis- younger than 1 year, we used the Amsterdam Infant Stool Scale, and
ease (e.g. neoplastic diseases), immunodeficiency, use of probiotics loose or watery stools corresponded to A-consistency [19]. In
within the 2 weeks before enrolment, use of antibiotics within the children older than 1 year, we used the Bristol Stool Form scale, and
4 weeks before enrolment, prematurity, or exclusive breastfeeding. loose or watery stools corresponded to scores of 5 to 7 [20]. In the
case of missing or incomplete data, we used data from the hospital
Randomization and masking chart.
In the event of loose or watery stools, we investigated the stool
A computer-generated randomization list prepared by a person sample for the presence of viral or bacterial pathogens. For this
unrelated to the trial was used to allocate participants with a block analysis, we used a standard rapid, qualitative, chromatographic
of eight. All investigators were blinded to the block size until all the immunoassay that simultaneously detects rotaviruses, adenovi-
data were analysed. Children were stratified according to partici- ruses and noroviruses. Standard microbiological techniques were
pating hospital. Sequentially numbered, sealed, opaque envelopes used to isolate and identify bacterial pathogens (Salmonella spp.,
containing the treatment assignment were prepared from Shigella spp., Campylobacter spp., Yersinia spp.). Clostridium difficile
computer-generated randomization schedules by a person not toxins A and B were identified by standard enzyme immunoassay.
involved directly in the trial and were concealed from the enrolling
physicians. When all eligibility criteria were met by the participant, Sample size
an envelope containing the treatment assignment was opened by
the enrolling physician. The patient was assigned to the group ac- The primary outcome of the study was the frequency of diar-
cording to the code. The active product and placebo were packaged rhoea. The frequency of AAD in earlier trials varied, depending on
in identical bottles. The contents of the bottles looked and tasted the definition of AAD used in the study [21e23]. Based on data from
the same. All the investigators, caregivers, outcome assessors, and studies previously conducted at The Medical University of Warsaw
the person responsible for the statistical analysis remained blinded [17], we assumed the baseline rate of AAD to be 23% in the placebo
to the intervention until the completion of the study and the group. To detect an absolute reduction of 15% difference between
analysis of the data. groups, with a power of 80% and a significance level of 5% and
M. Kołodziej, H. Szajewska / Clinical Microbiology and Infection 25 (2019) 699e704 701
taking into account that 20% of the patients would be lost to follow occurrence of diarrhoea appeared to be less common with placebo
up, we calculated that 250 children would be needed. compared with L. reuteri DSM 17938 (16 of 124 (13%) versus 25 of
123 (20%), respectively); however, this difference between groups
Statistical analysis was not significant (RR 1.58, 95% CI 0.9e2.79, ARR e0.07, 95% CI
e0.17 to 0.02) The occurrence of diarrhoea and AAD using other
Data were analysed on an intention-to-treat basis, including all definitions were similar in both groups.
participants in the groups to which they were randomized for There also was no statistically significant difference between the
whom outcomes were available. The computer software STATSDIRECT L. reuteri DSM 17938 and placebo groups for any of the secondary
ver 3.1.20 (version 3.1.20) was used to calculate the relative risk outcomes. Both study products were well tolerated. The incidence
(RR) and absolute risk reduction (ARR) with a 95% CI. The difference of adverse effects (i.e. abdominal pain, spitting) was similar in both
between study groups was considered significant when the 95% CI groups.
for RR did not include 1.0. All statistical tests were two-tailed and There were no differences between groups in the frequencies of
performed at the 5% level of significance. diarrhoea and AAD according to the type of antibiotic used or the
number of antibiotic classes used in the same patient (see
Results Supplementary material, Table S1).
Table 2 presents primary and secondary outcomes. Using the Summary of findings
strictest definition (three or more loose or watery stools per day for
a minimum of 48 h), the occurrence of AAD appeared to be less In this randomized, double-blind, placebo-controlled trial, the
common with placebo compared with L. reuteri DSM 17938 (8 of occurrences of diarrhoea and AAD were similar in the L. reuteri DSM
124 (6.5%) versus 14 of 123 (11.4%), respectively). However, this 17938 and placebo groups, regardless of the definition used. The
difference between groups was not significant (RR 1.76, 95% CI groups were similar with respect to all secondary outcome mea-
0.79e3.98, ARR e0.05, 95% CI e0.13 to 0.02). Similarly, the sures, including adverse events. The findings from our study
Table 1
Baseline characteristics and data on antibiotic therapy by indications and class
Characteristics at entry
n 125 125
Age (months), mean (SD) 25.8 (33.8) 25.7 (35.2)
Age distribution
<6 months, n (%) 43 (34.4%) 49 (39.2%)
6 to 24 months, n (%) 36 (28.8%) 31 (24.8%)
24 to 48 months, n (%) 26 (20.8%) 21 (16.8%)
48 months to 5 years, n (%) 7 (5.6%) 11 (8.8%)
5 years, n (%) 13 (10.4%) 13 (10.4%)
Sex, male/female, n 71/54 72/53
Fever 38 C, n (%) 93 (74.4%) 73 (58.4%)
Body weight (kg), mean (SD) 12.1 (10.3) 12.8 (11.8)
Antibiotic therapy by indications and class
Reasons for antibiotic therapy
Respiratory tract infection, n (%) 73 (58.4%) 68 (54.4%)
Urinary tract infection, n (%) 14 (11.2%) 15 (12.0%)
Otolaryngology infection, n (%) 22 (17.6%) 23 (18.4%)
Fever of unknown source, n (%) 12 (9.6%) 15 (12.0%)
Skin infection, n (%) 0 2 (1.6%)
Other infections, n (%) 4 (3.2%) 2 (1.6%)
Antibiotic classes
Aminopenicillins, n (%) 26 (20.8%) 37 (29.6%)
Cephalosporins second-generation, n (%) 79 (63.2%) 70 (56%)
Cephalosporins third-generation, n (%) 13 (10.4%) 15 (12%)
Macrolides, n (%) 6 (4.8%) 3 (2.4%)
Lincosamides, n (%) 1 (0.8%) 0
Combination antibiotic therapy
One class, n (%) 121 (96.8%) 123 (98.4%)
Two classes, n (%) 4 (3.2%) 2 (1.6%)
Length of antibiotic treatment, days, mean (SD) 9.1 (2.2) 8.9 (2.3)
Table 2
Primary and secondary outcomes
Primary outcomes
Diarrhoea
three or more loose or watery stools per day for a minimum of 48 h, n (%) 16 (12.9%) 25 (20.3%) 1.58 (0.9e2.79) e0.07 (e0.17 to 0.02)
three or more loose or watery stools per day for a minimum of 24 h, n (%) 26 (21%) 27 (21.7%) 1.05 (0.65e1.68) e0.01 (e0.11 to 0.09)
two or more loose or watery stools per day for a minimum of 24 h, n (%) 26 (21%) 27 (21.7%) 1.05 (0.65e1.68) e0.01 (e0.11 to 0.09)
Antibiotic-associated diarrhoea (caused by Clostridium difficile or for otherwise unexplained diarrhoea, i.e. negative laboratory stool tests for
infectious agents)
three or more loose or watery stools per day for a minimum of 48 h, n (%) 8 (6.5%) 14 (11.4%) 1.76 (0.79e3.98) e0.05 (e0.13 to 0.02)
three or more loose or watery stools per day for a minimum of 24 h, n (%) 17 (13.7%) 16 (13%) 0.95 (0.51e1.77) 0.01 (e0.08 to 0.09)
two or more loose or watery stools per day for a minimum of 24 h, n (%) 17 (13.7%) 16 (13%) 0.95 (0.51e1.77) 0.01 (e0.08 to 0.09)
Secondary outcomes
Infectious diarrhoea
Rotavirus, n (%) 5 (4%) 10 (8.1%) 2.02 (0.74e5.51) e0.04 (e0.11 to 0.02)
Adenovirus, n (%) 0 2 (1.6%) e0.02 (e0.06 to 0.01)
Norovirus, n (%) 0 0
Salmonella spp., n (%) 4 (3.2%) 1 (0.8%) 0.25 (0.04e1.65) 0.02 (e0.02 to 0.07)
Shigella spp., n (%) 0 0
Campylobacter spp., n (%) 0 0
Yersinia spp., n (%) 0 0
Clostridium difficile, n (%) 1 (0.8%) 0 0.01 (e0.02 to 0.04)
Unknown infectious aetiology, n (%) 17 (13.7%) 16 (13%) 0.95 (0.51e1.77) 0.01 e 0.08 to 0.09)
Need for discontinuation of the antibiotic treatment, n (%) 5 (4%) 3 (2.4%) 0.60 (0.16e2.24) 0.02 (e0.03 to 0.07)
Need for intravenous rehydration, n (%) 83 (66.9%) 76 (62%) 0.92 (0.76e1.11) 0.05 e 0.07 to 0.17)
Need for hospitalization to manage diarrhoea (in outpatients) 0 0
Adverse effects, n (%) 7 (5.6%) 3 (2.4%) 0.43 (0.12e1.5) 0.03 (e0.02 to 0.09)
Abdominal pain 3 (2.4%) 1 (0.8%) 0.34 (0.05e2.31) 0.02 (e0.02 to 0.06)
Spitting up 2 (1.6%) 1 (0.8%) 0.50 (0.07e3.80) 0.01 (e0.03 to 0.05)
Flexing 2 (1.6%) 1 (0.8%) 0.50 (0.07e3.80) 0.01 (e0.03 to 0.05)
complement previous studies that examined the effects of probiotic effective dosing remains unclear. Two earlier studies carried out by
administration on the risk of AAD. our group that evaluated the efficacy of using L. reuteri DSM 17938
There are several possible reasons for the lack of an effect of at different doses (108 and 109 CFU/day, respectively) showed no
L. reuteri DSM 17938 in our study. First, the dose of the studied effect in preventing nosocomial diarrhoea in children [24,25].
probiotic (2 108 CFU). With regard to diarrhoeal diseases, the However, other studies have documented the effectiveness of
M. Kołodziej, H. Szajewska / Clinical Microbiology and Infection 25 (2019) 699e704 703
L. reuteri DSM 17938 at a dose of 108 CFU/day both in the man- Conclusions
agement of diarrhoeal diseases and other indications [26e28]. Still,
the optimal dose of L. reuteri DSM 17938 has not been clearly In this randomized, double-blind, placebo-controlled trial, we
established. In our study, the timing of administration of L. reuteri found no evidence that the administration of L. reuteri DSM 17938
DSM 17938 in relation to the administration of antibiotics was not at a daily dose of 2 108 CFU was effective in preventing diarrhoea
predefined. Although it is likely that the study products were or AAD, regardless of the definition used, in children <18 years
administered between antibiotic doses, simultaneous administra- (median age 4 months) treated with antibiotics administered orally
tion by parents or care-providers cannot be excluded. As lactoba- or intravenously.
cillus species, including L. reuteri DSM 17938, are sensitive to Until more data are available, according to current guidelines, if
antibiotics [29], the lack of effect may be due to the insufficient the use of probiotics for preventing AAD is considered, the use of
colonization rate of L. reuteri, and hence its inability to exert its Lactobacillus GG or Saccharomyces boulardii is recommended [12].
probiotic effects. However, the prudent use of antibiotics, only in situations of clear
clinical benefit, remains the best method of preventing AAD.
Strengths and limitations of the study
Transparency declaration
The overall occurrence of diarrhoea, (not including AAD), ranged
from 13% to 21.7% and it ranged from 6.5% to 21.7% if both diarrhoea MK declares no competing interests. HS has participated as a
and AAD are considered; thus, it was generally in line with our clinical investigator, and/or advisory board member, and/or
initial assumptions, and the study was not underpowered. As there consultant, and/or speaker for Arla, Biogaia, Biocodex, Danone,
is no one standard definition of AAD, and the rate depends on the Dicofarm, Hipp, Nestle, Nestle Nutrition Institute, Nutricia and
definition used, we used three different definitions. This approach Mead Johnson.
allows a more precise comparison with other studies. The majority
of included children (82%) were younger than 4 years of age and Acknowledgements
received the most commonly used antibiotics in the paediatric age
group such as second-generation cephalosporins and amino- We thank Dr Anna Kołodziej from the General Hospital in
penicillins, so the findings are generalizable. w for her help in the recruitment.
Łuko
Nevertheless, the trial has several limitations. First, patients
were only followed up for 1 week after their antibiotic treatment.
Authors' contributions
As diarrhoea may occur up to 2 months after the end of antibiotic
treatment [2], some cases of AAD may have been missed. Second,
HS initially conceptualized this study. All authors contributed to
only relatively minor overall percentages of the patients in the
the study protocol. MK conducted the study. MK analysed the data
placebo (7.2%) and L. reuteri group (12.9%) had infectious diarrhoea
under the supervision of HS. HS and MK wrote the manuscript. Both
of known aetiology. However, in the case of diarrhoea, the stool
authors contributed to (and agreed upon) the final version. Both
samples were tested for most common enteropathogens only; so,
authors have full access to all the data in the study. HS is the
one cannot exclude the possibility that some of the cases of un-
guarantor for the data.
explained diarrhoea were caused by unidentified infectious agents.
Moreover, this study also did not allow one to reach conclusions
Funding
about the efficacy of administering L. reuteri DSM 17938 in the
prevention of C. difficile-associated diarrhoea, a clinically important
The study products were manufactured and supplied by BioGaia
condition, because only one patient developed C. difficile. Finally,
(Lund, Sweden), free of charge. This trial was fully funded by The
the study was also not powered to detect significant differences in
Medical University of Warsaw (Young Investigators research grant
any of the secondary outcomes.
number: 1W44/PM1/17). The funder of the study had no role in
study design, data collection, data analysis, data interpretation or
Comparison with other studies
writing of the report.
A 2016 systematic review investigated the efficacy of L. reuteri
DSM 17938 in the management of various types of diarrhoeal dis- Appendix A. Supplementary data
eases in children. Overall, eight RCTs, involving 1229 participants,
were identified. In preventive trials carried out in hospitalized Supplementary data related to this article can be found at
children, based on the findings from two RCTs (n ¼ 290), there was https://doi.org/10.1016/j.cmi.2018.08.017.
no significant reduction in the risk of nosocomial diarrhoea, rota-
virus diarrhoea or diarrhoea of any origin with L. reuteri adminis- References
tration. Based on one RCT (n ¼ 97), there was no effect of L. reuteri
[1] Bartlett JG. Clinical practice. Antibiotic associated diarrhea. N Engl J Med
use on the risk of AAD. However, as stated earlier, the evidence is 2002;346:334e9.
limited because the overall frequency of diarrhoea was surprisingly [2] McFarland LV, Ozen M, Dinleyici EC, Goh S. Comparison of pediatric and adult
low [25]. antibiotic-associated diarrhea and Clostridium difficile infections. World J
Gastroenterol 2016;22:3078e104.
A more recent open RCT involved 1013 children aged 3e14 years [3] Cote GA, Buchman AL. Antibiotic-associated diarrhoea. Expert Opin Drug Saf
in primary care settings who received amoxicillin ± clavulanic acid 2006;5:361e72.
or cephalosporins. In this trial, children who received L. reuteri DSM [4] McFarland LV. Epidemiology, risk factors and treatments for antibiotic-
associated diarrhea. Dig Dis 1998;16:292e307.
17938 at a dose of 2 108 CFU (sachets) compared with any pre-
[5] Sammons JS, Toltzis P, Zaoutis TE. Clostridium difficile infection in children.
ventive therapy had a reduced risk of antibiotic-related gastroin- JAMA Pediatr 2013;167:567e73.
testinal symptoms. However, the findings are available in abstract [6] Ho€genauer C, Hammer HF, Krejs GJ, Reisinger EC. Mechanisms and manage-
form only [30]. Until a full publication is available, the methodo- ment of antibiotic-associated diarrhea. Clin Infect Dis 1998;27:702e10.
[7] McFarland LV. Meta-analysis of probiotics for the prevention of antibiotic
logical details, including the definition of outcome measures, associated diarrhea and the treatment of Clostridium difficile disease. Am J
remain unclear. Gastroenterol 2006;101:812e22.
704 M. Kołodziej, H. Szajewska / Clinical Microbiology and Infection 25 (2019) 699e704
[8] Hill C, Guarner F, Reid G, Gibson GR, Merenstein DJ, Pot B, et al. Expert plantarum PL02 in the prevention of antibiotic-associated diarrhea in chil-
consensus document. The International Scientific Association for Probiotics dren: a randomized controlled pilot trial. Digestion 2008;78:13e7.
and Prebiotics consensus statement on the scope and appropriate use of the [19] Ghanma A, Puttemans K, Deneyer M, Benninga MA, Vandenplas Y. Amsterdam
term probiotic. Nat Rev Gastroenterol Hepatol 2014;11:506e14. infant stool scale is more useful for assessing children who have not been
[9] Surawicz CM. Probiotics, antibiotic associated diarrhoea and Clostridium toilet trained than Bristol stool scale. Acta Paediatr 2014;103:91e2.
difficile diarrhoea in humans. Best Pract Res Clin Gastroenterol 2003;17: [20] Lewis SJ, Heaton KW. Stool form scale as a useful guide to intestinal transit
775e83. time. Scand J Gastroenterol 1997;32:920e4.
[10] Szajewska H, Kolodziej M. Systematic review and meta-analysis: Saccharo- [21] Vanderhoof JA, Whitney DB, Antonson DL, Hanner TL, Lupo JV, Young RJ.
myces boulardii in the prevention of antibiotic-associated diarrhoea. Aliment Lactobacillus GG in the prevention of antibiotic-associated diarrhea in chil-
Pharmacol Ther 2015;42:793e801. dren. J Pediatr 1999;135:564e8.
[11] Szajewska H, Kołodziej M. Systematic review with meta-analysis: Lactoba- [22] Szyman ski H, Arman ska M, Kowalska-Duplaga K, Szajewska H. Bifidobacte-
cillus rhamnosus GG in the prevention of antibiotic-associated diarrhoea in rium longum PL03, Lactobacillus rhamnosus KL53A, and Lactobacillus plantarum
children and adults. Aliment Pharmacol Ther 2015;42:1149e57. PL02 in the prevention of antibiotic-associated diarrhea in children: a ran-
[12] Szajewska H, Canani RB, Guarino A, Hojsak I, Indrio F, Kolacek S, et al., domized controlled pilot trial. Digestion 2008;78:13e7.
ESPGHAN Working Group for Probiotics Prebiotics. Probiotics for the pre- [23] Bin Z, Ya-Zheng X, Zhao-Hui D, Bo C, Li-Rong J, Vandenplas Y. The efficacy of
vention of antibiotic-associated diarrhea in children. J Pediatr Gastroenterol Saccharomyces boulardii CNCM I-745 in addition to standard Helicobacter
Nutr 2016;62:495e506. pylori eradication treatment in children. Pediatr Gastroenterol Hepatol Nutr
[13] Savino F, Fornasero S, Ceratto S, De Marco A, Mandras N, Roana J, et al. Pro- 2015;18:17e22.
biotics and gut health in infants: a preliminary caseecontrol observational [24] Wanke M, Szajewska H. Lack of an effect of Lactobacillus reuteri DSM 17938 in
study about early treatment with Lactobacillus reuteri DSM 17938. Clin Chim preventing nosocomial diarrhea in children: a randomized, doubleblind,
Acta 2015;451:82e7. placebo-controlled trial. J Pediatr 2012;161:40e43.e1.
[14] Georgieva M, Pancheva R, Rasheva N, Usheva N, Ivanova L, Koleva K, et al. Use [25] Urban ska M, Gieruszczak-Białek D, Szajewska H. Systematic review with
of the probiotic Lactobacillus reuteri DSM 17938 in the prevention of antibiotic meta-analysis: Lactobacillus reuteri DSM 17938 for diarrhoeal diseases in
associated infections in hospitalized Bulgarian children: a randomized, children. Aliment Pharmacol Ther 2016;43:1025e34.
controlled trial. IMABdAnnu Proc (Sci Pap) 2015;21:895e900. [26] Gutierrez-Castrello n P, Indrio F, Bolio-Galvis A, Jime
nez-Gutie rrez C, Jimenez-
[15] Kołodziej M, Szajewska H. Lactobacillus reuteri DSM 17938 in the prevention Escobar I, Lo pez-Vel
azquez G. Efficacy of Lactobacillus reuteri DSM 17938 for
of antibiotic-associated diarrhoea in children: protocol of a randomised infantile colic: systematic review with network meta-analysis. Medicine (Balt)
controlled trial. BMJ Open 2017;7:e013928. 2017;96:e9375.
[16] Moher D, Hopewell S, Schulz KF, Montori V, Gøtzsche PC, Devereaux PJ, et al. [27] Savino F, Cordisco L, Tarasco V, Palumeri E, Calabrese R, Oggero R, et al.
CONSORT 2010 explanation and elaboration: updated guidelines for reporting Lactobacillus reuteri DSM 17938 in infantile colic: a randomized, double-blind,
parallel group randomised trials. BMJ 2010;340:c869. placebo controlled trial. Pediatrics 2010;126:e526e33.
[17] Kotowska M, Albrecht P, Szajewska H. Saccharomyces boulardii in the pre- [28] Coccorullo P, Strisciuglio C, Martinelli M, Miele E, Greco L, Staiano A. Lactobacillus
vention of antibiotic-associated diarrhoea in children: a randomized reuteri (DSM 17938) in infants with functional chronic constipation: a double-
double-blind placebo-controlled trial. Aliment Pharmacol Ther 2005;21: blind, randomized, placebo-controlled study. J Pediatr 2010;157:598e602.
583e90. [29] Neut C, Mahieux S, Dubreuil LJ. Antibiotic susceptibility of probiotic strains: is
[18] Szyman ski H, Arman
ska M, Kowalska-Duplaga K, Szajewska H. Bifidobacte- it reasonable to combine probiotics with antibiotics? Med Mal Infect 2017;47:
rium longum PL03, Lactobacillus rhamnosus KL53A, and Lactobacillus 477e83.