Effects of Probiotic and Synbiotic Supplementation on Inflammatory Markers: A Systematic Review and Meta-Analysis

(1)

Meta-analyses

Effect of probiotic and synbiotic supplementation on in ﬂ ammatory markers in health and disease status: A systematic review and meta-analysis of clinical trials

Asma Kazemi

^a

, Sepideh Soltani

^b

, Sima Ghorabi

^c

, Abbas Keshtkar

^d

, Elnaz Daneshzad

^c

, Fatemeh Nasri

^e

, Seyed Mohammad Mazloomi

^a^,^*

aNutrition Research Center, Shiraz University of Medical Sciences, Shiraz, Iran

bDepartment of Nutrition, School of Public Health, Iran University of Medical Sciences, Tehran, Iran

cDepartment of Clinical Nutrition, School of Nutritional Sciences and Dietetic, Tehran University of Medical Sciences, Tehran, Iran

dDepartment of Health Science Educational Development, School of Public Health, Tehran University of Medical Sciences, Tehran, Iran

eDepartment Immunology, School of Medicine, Shiraz University of Medical Sciences, Shiraz, Iran

a r t i c l e i n f o

Article history:

Received 1 March 2019 Accepted 1 April 2019

Keywords:

Probiotic Synbiotic Inﬂammation Interleukin Cytokine Meta-analysis

s u m m a r y

The current systematic review and meta-analysis investigated the effect of probiotic/synbiotic on a wide range of inﬂammatory and anti-inﬂammatory markers in healthy and various disease conditions.

PubMed, SCOPUS and Web of Science databases were searched. All clinical trials which investigated the effect of oral administration of probiotic or synbiotic on inﬂammatory markers (C-reactive protein (CRP), interleukin (IL) 1

b

, IL-4, IL-6, IL-8, IL-10, IL-12, tumor necrosis factor (TNF)

a

, interferon (IFN)

g

and transforming growth factor (TGF)

b

) for more than one week with concurrent control groups were included. One-hundred sixty seven publications was analysed. Results were as follows: CRP decreased in healthy, metabolic disorders, inﬂammatory bowel disease (IBD), arthritis and critically ill condition but not in renal failure. IL-1B: no change in healthy subjects and arthritis. TNF-

a

: decreased in healthy, fatty liver, IBD and hepatic cirrhosis, no change in diabetes, metabolic syndrome (MS)þPCOS (polycystic ovary syndrome) and arthritis. IL-6: no change in healthy, metabolic disorders and arthritis, increased in cirrhosis and renal failure, decreased in PCOSþMS. IL-10: no change in healthy, IBD and metabolic disorders, increased in arthritis. IL-4, IL-8, IL-12, IFN-g and TGF-b: no change in healthy subjects. In conclusion, probiotic/synbiotic decreased some of the inﬂammatory markers. The intervention was most effective in CRP and TNF-

a

reduction in healthy or disease state. Moreover, the intervention decreased inﬂammation most effectively in the following disease conditions, respectively: IBD, arthritis, fatty liver.

PROSPERO registration number: CRD42018088688.

1. Introduction

Probiotics are

“

live microorganisms that, when administered in adequate amounts, confer a health bene

ﬁ

t to the host

” [1]. The

health-bene

ﬁ

cial effects of probiotics are increasingly recognised.

Oral delivery of probiotics that are then integrated into the gut ecosystem has the potential to healthfully modulate the gut microbiome

[2]. A great deal of research focus has been attracted on

probiotics in recent years and the effects of probiotic on many disease conditions such as gastrointestinal disease

[3,4], allergic

and viral disease

[5],Helicobacter pylori

infection

[6], nonalcoholic

fatty liver disease

[7], diabetes[8e10], hypertension[11], hyper-

lipidemia

[12], autoimmune disease[13], cancer[14], postoperative Abbreviations used:Coef, coefﬁcient; CRF, chronic renal failure; CRP, C-reactive

protein; CVD, cardiovascular disease; IBD, inﬂammatory bowel disease; IBS, in- ﬂammatory bowel syndrome; IFN, interferon; IL, interleukin; LPS, lipopolysaccharides; MS, metabolic syndrome; NAFLD, nonalcoholic fatty liver disease; PCOS, polycystic ovary syndrome; PBMC, peripheral blood mononuclear cell; RA, rheumatoid arthritis; SMD, Standardized mean difference; TGF, transforming growth factor; TNF, tumor necrosis factor.

*Corresponding author. Department of Food Hygiene and Quality Control, School of Nutrition and Food Sciences, Shiraz University of Medical Sciences, PO Box 71645-111, Shiraz, Iran.

E-mail addresses: [email protected], [email protected]

(S.M. Mazloomi).

Contents lists available atScienceDirect

Clinical Nutrition

j o u r n a l h o m e p a g e : h t t p : / / w w w . e l s e v i e r . c o m / l o c a t e / c l n u

https://doi.org/10.1016/j.clnu.2019.04.004

Clinical Nutrition xxx (xxxx) xxx

Please cite this article as: Kazemi A et al., Effect of probiotic and synbiotic supplementation on in

ﬂ

ammatory markers in health and disease

status: A systematic review and meta-analysis of clinical trials, Clinical Nutrition, https://doi.org/10.1016/j.clnu.2019.04.004

(2)

complications

[15,16], obesity [17], critically ill patients[18]

and osteopenia

[19]

have been studied. Probiotics induce their effects principally due to their role in immune system modulation and the anti-in

ﬂ

ammatory response

[20]. Disturbance in the equilibrium of

the gut microbiota leads to change in gut barrier function and in- creases in permeability of intestinal mucosa and as a consequence, translocation of gram negative bacteria into lymph node and blood stream

[21]. Lipopolysaccharides (LPS) of gram-negative bacteria

activate the innate immune system resulting in the production of pro-in

ﬂ

ammatory cytokines

[22,23]. Modulation of intestinal

microbiota via probiotic bacteria has been claimed to induce epithelial healing and prevent bacterial translocation across the epithelium

[24].

“

Prebiotic is a non-digestible compound that, through its metabolism by microorganisms in the gut, modulates composition and/or activity of the gut microbiota, thus conferring a bene

ﬁ

cial physiologic effect on the host

” [25]. The combination of one or

more probiotics with prebiotics is termed as synbiotic.

Several meta-analyses have investigated the effects of probiotics on in

ﬂ

ammatory markers in various disease conditions such as diabetes

[8,9,26,27], nonalcoholic fatty liver disease (NAFLD) [7]

and rheumatoid arthritis (RA)

[13,28,29]

and to the best of our knowledge, no study have examined the effect of probiotic on in

ﬂ

ammation in healthy individuals. The aim of the current study was to provide a strong combination of evidence that demonstrates the direct and biological effects of probiotics on in

ﬂ

ammatory markers. Moreover, cytokines and in

ﬂ

ammatory mediators operate in network

[30], and while several meta-analyses have examined

some standard in

ﬂ

ammatory markers such as CRP, TNF- a and interleukin (IL)-6, no study has simultaneously examined a wide range of in

ﬂ

ammatory and anti-in

ﬂ

ammatory markers re

ﬂ

ecting the cytokine network. The current systematic review and meta- analysis investigates the effect of probiotic and synbiotic supplementation on a wide range of in

ﬂ

ammatory and anti-in

ﬂ

ammatory markers in healthy and various disease conditions.

2. Method

The protocol of this systematic review has been registered on PROSPERO website (www.crd.york.ac.uk/PROSPERO) (PROSPERO registration number

¼

CRD42018088688). This systematic review protocol has been developed based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) state- ment guidelines

[31].

2.1. Study selection criteria 2.1.1. Type of studies

The systematic review included all clinical trials with either parallel or cross-over design and at least two arms that investigated the effect of oral administration of probiotic or synbiotic on in-

ﬂ

ammatory markers (CRP, IL-1B, IL-4, IL-6, IL-8, IL-10, IL-12, TNF- a and IFN- g ) for more than one week with concurrent control groups.

Cytokines were measured in serum, peripheral blood mononuclear cell (PBMC), or secreted from mitogen (LPS, phytohemagglutinin, phorbol myristate acetate-ionomycin, concanavalin A) stimulated PBMC. The studies were open label (single, double or triple blind) and either randomised or non-randomised.

2.1.2. Type of participants

Both male and female adults with the age

>

18 year were included while studies with adolescents participants (under 18 years of age), pregnant and lactating women were excluded.

Four reviewers (AK, SS, SG, ED) assessed titles and abstracts of all primary articles that met the search strategy in order to

determine studies eligible for inclusion. Then, two reviewers independently evaluated the full text of potentially relevant non- duplicated articles (AK, SS). Any disagreements were resolved by discussion to reach consensus. When consensus was not reached, a third reviewer (SMM) acted as an arbitrator.

2.2. Search strategy

We searched the following electronic database from 1990 until 31 May 2018 with no restriction in English language: PubMed, Scopus, Web of Science, National Institute of Health Clinical Trials Register (https://clinicaltrials.gov/) to identify unpublished studies and the Cochrane Central Register of Controlled Trials (Clinical Trials).

Relevant search terms in accordance with an intervention and outcome component of current systematic review were extracted from Mesh and key word of the studies in the primary search. Full search strategy for PubMed, Scopus, and Web of Science are pre- sented in supplementary (sup) Table 1. Moreover, bibliographies of all relevant prior reviews and primary studies identi

ﬁ

ed by search strategy were scanned for additional relevant paper. Scopus and Web of Science and annual meetings were searched for conference papers.

2.3. Data extraction

Data extraction from primary articles was performed independently by two reviewers (AK, SS) using a quantitative data extraction form. The data extraction form has been piloted previously.

Any discrepancies were resolved by consensus between the two reviewers and, when this was not possible, a third reviewer (SMM) acted as an arbitrator and made a decision on the data entered. Data collected from the studies included participants' characteristics such as gender, age, disease, BMI and country; study design, duration of the study, sample size, the sample in which the outcomes were measured (serum, PBMC, stimulated PBMC), publication year, composition and dose of supplements and methodological quality. If data was presented as graph only, it was extracted using Plot Digitiser software (http://plotdigitizer.

sourceforge.net/). Corresponding authors of included studies, in

which required data were not provided, were contacted to request the data needed for the purpose of meta-analysis.

2.4. Risk of bias (quality) assessment

Quality assessments were performed with the Jadad scale

[32],

based on the following parameters: randomisation, random description, blinding of participants and personnel, incomplete outcome data and selective reporting. Two investigators independently rated each study and settled any differences by consensus or referring to a third investigator.

2.5. Statistical analysis

Statistical analysis of data was performed using Stata software version 13 (StataCorp LP, College Station, TX, USA). Data from clinical trials was analysed using mean difference with standard deviation. Mean change and its corresponding standard deviation (SD) of in

ﬂ

ammatory markers within the intervention and placebo groups were used to calculate the unstandardised difference in means (MD) to be used as effect size for meta-analysis. In case the studies had reported the baseline and after intervention values, based on the studies that had reported standard deviation for pre- and post-intervention and changes in outcomes, we estimated the correlation coef

ﬁ

cient for cytokines levels and used it to calculate

A. Kazemi et al. / Clinical Nutrition xxx (xxxx) xxx

2

ﬂ

ammatory markers in health and disease

status: A systematic review and meta-analysis of clinical trials, Clinical Nutrition, https://doi.org/10.1016/j.clnu.2019.04.004

(3)

the standard deviation for change. Study heterogeneity was measured using the Q test and

I²

test. Sources of heterogeneity were explored in meta-regression and subgroup analysis. Subgroup analysis was conducted by the following variables: sample in which cytokines were measured (serum, PBMC, secretion from stimulated PBMC), strains of bacteria (where it was probable), sex, age (

<

49 vs.

49), single vs. multiple strains of probiotics, BMI ( 25 (kg/m

²

) vs.

>

25 (kg/m

²

), dose and study duration (lower than 8 weeks/8 weeks and more).

The potential for publication bias was assessed using funnel plot, the Begg rank correlation method, and the Egger weighted regression method. A p-value of 0.05 was considered to be sta- tistically signi

ﬁ

cant. When publication bias was suspected based on visual inspection, Duval and Tweedie's Trim and Fill method was applied to estimate the impact of publication bias on the observed summary effect size.

A sensitivity analysis was performed to test the small study ef- fect. We also implemented sensitivity analyses to explore the ef- fects of methodological quality and sample size on the robustness of review conclusions.

3. Results

Figure 1

shows the process for the inclusion of studies. A total of 183 publications were considered to have met the eligibility criteria and were included. We could not gain access to enough data to analyse six of the studies

[33e39]. Seven studies were

published two or three times while they were related to the same studies, so eight studies were excluded. Finally, 169 were

analysed.

Table 1

shows a description of the included studies that examined the effect of probiotic/symbiotic on in

ﬂ

ammatory markers in healthy participants,

Table 2

shows patients with metabolic disorders and

Table 3

shows patients with disease other than metabolic disorders. Quality assessment of studies are presented in Sup Table 2. The results have been summarised in

Table 4. Sensitivity analysis for the small study effect was signif-

icant for none of the outcomes. Since non-randomized and non- blind studies were also included in the meta-analysis, we con- ducted the analysis either in the presence or in the absence of non-randomized and non-blind studies to assess the effect of methodological quality. No signi

ﬁ

cant difference was seen except for TNF- a in healthy patients. Such that after excluding non- randomized and non-blind studies, the reduction in TNF- a was no more signi

ﬁ

cant (Sup Table 3). Publication bias was signi

ﬁ

cant for none of outcomes except IL-12; however, no studies was imputed with the trim-and-

ﬁ

ll method.

3.1. CRP

A total of 136 clinical trials measured CRP (healthy (30), diabetes (16), NAFLD (5), PCOS (4), metabolic syndrome (MS) (4), arthritis (8), in

ﬂ

ammatory bowel disease (IBD) (7), in

ﬂ

ammatory bowel syndrome (IBS) (1), (chronic renal failure) CRF(4), surgery (10), HIV (4), critically ill (8), cirrhosis (3), atopic dermatitis (1), multiple sclerosis (1), allergy (2), Alzheimer's disease (1), schizophrenia (1), patients with pulmonary symptoms due to mustard (1). The studies with three arms were considered as two and the studies with four arms were considered as three studies.

Records identified through PubMed, Scopus, WOS

(n =43235)

ScreeningIncludedEligibilityIdentification

Records after duplicates removed (n =30682)

Records screened (n =30682)

Records excluded (n =28225)

Full-text articles assessed for eligibility

(n = 2457)

Full-text articles excluded, (n =2282)

Studies included in qualitative synthesis

(n =175)

Studies included in quantitative synthesis

(meta-analysis) (n =168) Fig. 1.Flow diagram.

A. Kazemi et al. / Clinical Nutrition xxx (xxxx) xxx 3

Please cite this article as: Kazemi A et al., Effect of probiotic and synbiotic supplementation on in

ﬂ

(4)

Table 1

Characteristic of randomized controlled trials that evaluated the effect of the probiotic/synbiotic on the serum level of inﬂammatory biomarkers in apparently healthy participants.

Author,Year Country Sex Age Design Duration (weeks) Participants

(Intervention, Control)

Intervention (Probiotic strains) Control status Outcome

Ahn, 2015[46] South Korea Both 53.4 Parallel 12 46, 46 L. curvatus HY7601 and L. plantarum

KY1032

Placebo hs-CRP

Arunachalam, 2000[47] New Zealand Both 69 Parallel 6 13, 12 Milk supplemented with B. lactis Milk IFN-g

Brahe, 2015[48] Denmark Female 59.9 Parallel 6 18, 16 L. paracasei F19 Placebo hs-CRP, IL-6, TNF

Burton, 2017[49] Switzerland Male 24 Cross-over 2 40, 40 L. rhamnosus GG, L. delbrueckii

Bulgaricus, S. thermophilus

Placebo IL-6, TNF

Cavallini, 2016[50] Brazil Male 45.75 Parallel 6 17, 15 Synbiotic (Soy product fermented with

Enterococcus faecium andL. helveticus)

Unfermented soy product

hs-CRP

Childs, 2014[51] UK Both 43 Cross-over 3 11, 11, 11, 11 1. Probiotic (B. animalis subsp. lactis Bi-

07)þ

2. xylo-oligosaccharide 3.synbiotic (B. animalis subsp.

lactisþxylo-oligosaccharide)

Placebo IL-4, IL-6, IL-10

Christensen, 2006[52] Denmark Both 27 Parallel 3 42, 42 B. animalis ssp. lactis (BB-12) and L.

paracasei ssp. paracasei (CRL-431)

Placebo IL-10, INF

Costabile, 2017[53] UK Both 70 Cross-over 1.2 55, 56 L. rhamnosus GG combined with SCF Placebo hs-CRP, IL-6, IL-8

Cox, 2010[54] Australia Male 27.3 Cross-over 4 20, 20 L. fermentum VRI-003 Placebo IL-4, IL-12, INF

Cox, 2014[55] Australia Both 40.55 20 39, 45 B. animalis subsp. lactis Bl-04 hs-CRP

Dong, 2013[56] UK Both 55e74 Cross-over 4 30, 30 probiotic drink containing 1.31010 L.

casei Shirota

Skimmed milk IL1B, IL-6, TNF, IL-8, IL-10, IL-12, IL-17, INF

Gleeson, 2011[57] Spain Both Parallel 16 25, 20 L. casei Shirota Milk IL1B, IL-4, IL-6, TNF, IL-8, IL-10,

INF

Gomes, 2017[58] Brazil Female Parallel 8 21, 22 Symbiotic (prebioticþL. acidophilus, L.

casei, Lactococcus lactis, B. biﬁdum, and B. lactis BL-4)þDietary intervention

Dietary intervention

IL-6, TNF, IL-10

Hirose, 2013[59] Japan Both 50.6 Parallel 12 20, 15 L. plantarum L-137 Placebo INF

Hor, 2018[60] Malaysia Both 44.6 Parallel 52 43, 37 L. casei Zhang Placebo IL1B, IL-4, IL-10

Irwin, 2017[61] Australia Both 27.75 Parallel 8 15, 14 L. acidophilus and B. lactis Bi-07 Placebo hs-CRP

Iwasa, 2013[41] Japan Male 1 day 12, 12 L. helveticusplus exercise Exercise hs-CRP, TNF

Jung, 2015[62] North Korea Both 39 Parallel 12 49, 46 L. curvatus HY7601 and L. plantarum

KY1032

Placebo hs-CRP

Kazemi, 2019[63] Iran Both Parallel 8 27, 25 L. Helvetius, B. longum Placebo IL1B, IL-10, IL-6, TNF

Kekkonen, 2008[64] Finland Both 44 Parallel 3 13, 16 Milk based drink containing L.

Rhamnsus GG

Placebo hs-CRP, IL1B, IL-6, TNF

kim, 2018[40] Korea Both 38.7 Parallel 12 30, 30 L. gasseri BNR17 Placebo hs-CRP, INF

Lamprecht, 2012[65] Austria Male 37.9 Parallel 14 11, 12 6 probiotic strains: B. biﬁdum, B. lactis,

Enterococcus faecium, L. acidophilus, L.

brevis, and Lactococcus lactis

Placebo IL-6, TNF

Lee, 2017[66] USA Both 28 Cross-over 4 25, 25 Yogurt smoothie with B. animalis subsp.

lactis BB-12

Yogurt hs-CRP

Lee, 2017[67] Korea Both 65.7 Parallel 12 73, 79 Yogurt containing L. casei, B. lactis and

heat-treated L. plantarum

Yogurt hs-CRP, TNF, IL-12, INF

Lefevre, 2015[68] France Both 63.1 Parallel 50, 50 Biﬁdobacterium Placebo INF

Macfarlane,2013 Scotland Both 71.9 Cross-over 4 42, 42 Synbiotic (B. longum and an insulin-

based prebiotic)

Placebo hs-CRP, IL1B, IL-4, IL-6, TNF, IL- 8, IL-10, INF

Maneerat, 2013[69] UK Both 66.75 Cross-over 3 10, 9 Synbiotic (galacto-oligosaccharides

plus B. lactis Bi-07)

Placebo IL1B, IL-6, IL-8, IL-10, INF

Mangalat, 2012[70] USA Both 33.75 Parallel 8 23, 10 L. reuteri DSM 17938 Placebo IL1B, IL-6, TNF, IL-8, IL-12, INF

Manzoni, 2017[71] Brazil Both >65 Parallel 6 14, 15 Synbiotic (Soy and yacon extracts

containing B. animalis ssp. lactis)

Placebo IL-6, TNF, IL-10

Marcial, 2017[72] USA Both 23.2 Parallel 8 21, 20 L. johnsonii N6.2 Placebo TNF, INF

Marcos, 200[73]4 Spain Both Parallel 6 69, 53 Milk fermented with yogurt cultures

plus Lacto- bacillus casei

Placebo IL-4, IL-5, TNF, INF

Marinkovic, 2016[74] Serbia Both 23.2 Parallel 14 20, 19 L. helveticusLafti L10 Placebo IL-10, INF

Meyer, 2007[75] Austria Female 24.4 Parallel 4 17, 16 Yogurt IL1B, IL-6, TNF, IL-10, INF

A.Kazemietal./ClinicalNutritionxxx(xxxx)xxx4

Please cit e this article as: Kazemi A e t al., Effect of prob io tic and synbio tic supplem entation on in

ﬂ

ammat ory mark ers in health and disease status: A sys tema tic rev iew and meta-ana ly sis of clinical trials, Clinical N utrition, https://doi. org/1 0. 1 0 1 6/j.clnu. 20 1 9.04 .0 0 4

(5)

Yogurt containing L. bulgaricus and S.thermophilusþL. casei

Minami, 2015[76] Japan Both 60.5 Parallel 12 19, 25 B. breve B-3 Placebo hs-CRP

Mohammadi,[77]2015 Iran Male 32.65 Parallel 6 12, 13, 10 1.Probiotic yoghurt containing L.

acidophilus and B. lactis2.Multispecies probiotic capsule contains L. casei, L.

acidophilus, L. bulgaricus, B. longum, L.

rhamnosus, B. breve, S. thermophilus

Conventional yogurt

hs-CRP, INF

MoroGarcia, 2013[78] Spain Both 70 Parallel 24 26, 21 L. delbrueckii subsp. Bulgaricus Placebo IL-8

Naruszewicz, 2002[79] Poland Both 42.3 Parallel 6 s 18, 18 L. plantarum Placebo IL-6

Neto, 2013[80] Brazil Both 67.9 Parallel 12 9, 8 Synbiotic (FOSþL. paracasei, L.

rhamnosus, L. acidophilus and B. lactis

Placebo IL-6, TNF

Nova, 2011[81] Spain Both Parallel 6 18, 18 Synbiotic (L. acidophilus, B. animalis

ssp. lactis, L. delbrueckii ssp. bulgaricus, S.thermophilus, and L. paracasei ssp.

paracaseiþFOS)

Placebo hs-CRP

Nyangale, 2014[82] UK Both 65e80 Cross-over 4 18, 18 Bacillus coagulans GBI-30 Placebo hs-CRP, IL1B, IL-6, TNF, IL-10

Olivares, 2006[83] Spain Both 23e43 Parallel 4 15, 15 Dairy product contained S.

thermophilus, L. coryniformis, L. gasseri

Dairy product IL-4, TNF, IL-10, IL-12

Osterberg, 2015[84] USA Male 22.65 Parallel 4 9, 11 S.thermophilus, L. acidophilus, L.

delbrueckii ssp. Bulgaricus, L. paracasei, L. plantarum, B. longum, B. infantis, and B. breveþdietary intervention

Placebo and hypercaloric diet

hs-CRP, IL-6, TNF

Ouwehand, 2008[85] Finland Both Parrallel 24 19, 18 1.Oat-based drink supplemented with

B. longum2.fermented oat drink containing B. animalis ssp. lactis

Oat drink TNF, IL-10

Rajkumar, 2015[86] India Parallel 6 15, 15, 15 1- L. salivariusþFOS 2. L. salivarius Placebo hs-CRP, IL1B, IL-6, TNF

Ramijn, 2017 Australia Both 35.4 Parallel 8 38, 39 L. helveticusþB. longum Placebo hs-CRP, IL1B, IL-6, TNF

Sanchez, 2014[87] Canada Both 35 Parallel 24 52, 53 L. rhamnosusþweight loss Placebo plus

weight loss

hs-CRP

Seifert, 2011[88] Germany Male 31.5 Parallel 4 34, 34 probiotic drink contain L. casei Shirota Placebo hs-CRP, TNF

Simons, 2006[89] Australia Both 51.5 Parallel 10 23, 21 L. fermentum Placebo hs-CRP

Spanhaak, 1998[90] Netherlands Male 55.8 4 10, 10 Fermented milk containing L.casei

shirota

Unfermented milk hs-CRP

Stenman, 2016[91] Finland Both 48.7 Parallel 24 25, 36 B. animalis ssp. lactis 420 Placebo hs-CRP, IL-6

Tomohiko Ogawa, 2006[92] Japan Both 40.3 Parallel 4 6, 6 Symbiotic (L. casei subsp. casei together

and dextran)

Placebo INF

Valentini, 2015[93] France, Germany, Italy Both 70.1 Parallel 8 31, 31 B. infantis, B. longum, B. breve, L.

acidophilus, L. delbrückii ssp.

bulgaricus, L. paracasei, L. plantarum, and S.thermophilus

Placebo hs-CRP, IL-6, TNF, IL-10

West, 2014[94] Australia Both 49.5 Parallel 20 39, 39 B. animalis subsp. lactis Placebo IL-4, IL-6, IL-8, IL-12

Wilms, 2016[95] Netherlands Both 20.7 Parallel 2 10, 10 Synbiotic (B. biﬁdum, B. lactis, B. lactis,

L. acidophilus, L. casei, L. paracasei, L.

plantarum, L. salivarius, Lactococcus lactisþFOS

Placebo IL1B, IL-6, TNF, IL-8

Zarrati, 2013[96] Iran Both 36 Parallel 8 25, 25 L. acidophilus La5, B. BB12, and L. casei

plus weight loss diet

Yogurtþ weight loss

IL-4, TNF, IL-10, IL-17, INF

Zhang, 2018[97] china Both 33.4 Parallel 12 67, 67 L. paracasei, L. casei 431, L. fermentium

PCC

Yogurt hs-CRP

B., biﬁdobacterium; hsCRP high sensitive C-reactive protein; FOS, fructooligossacaride; IFN, interferon; IL, interleukin; L., lactobacillus; S, streptococcus; TNF, tumor necrosis factor.

Please cit e this article as: Kazemi A e t al., Effect of prob io tic and synbio tic supplemen tation on in

ﬂ

ammat ory mar kers in health and disease status: A syst ematic revi ew and meta-anal y sis of clinical trials, Clinical N utrition, https://doi. org/1 0. 1 0 1 6/j.clnu. 20 1 9.04 .0 04

(6)

Table 2

Characteristic of randomized controlled trials that evaluated the effect of the probiotic/synbiotic on the serum level of inﬂammatory biomarkers in patients with metabolic disorders.

Author,Year Country Sex Age Design Duration Participants

(Intervention, Control)

Intervention (Probiotic strains) Control status Participants disease Outcome

Abbaszadeh, 2016[98] Iran Both 44.7 Parallel 8 21, 21 L. casei, L. acidophilus, L. rhamnosus, L. bulgaricus, B. breve, B. longum, and S.thermophilus

Placebo NAFLD IL-6, TNF

Ahmadian, 2017[99] Iran Both 60 Parallel 6 30, 29 L. casei, L. acidophilus, L. bulgaricus, L. rhamnsus, L.

longum,B. breve, S. thermophilus,þFOS

Placebo T2DM IL-6, TNF

Aller, 2011[100] Spain Both 0 Parallel 12 14, 14 L. bulgaricus and S.thermophilus Placebo NAFLD IL-6, TNF

Andreasen, 2010[33] Denmark Male 57.5 Parallel 4 21, 24 L. acidophilus NCFM Placebo T2DM hs-CRP, IL-6, TNF

Asemi, 2013[101] Iran Both 51.5 Parallel 8 27, 27 L. acidophilus, L. casei, L. rhamnosus, L. bulgaricus, B. breve, B. longum, S.thermophiles,þFOS

Placebo T2DM hs-CRP

Asemi, 2014[102] Iran Both 53 Cross-over 6 62, 62 Synbiotic (L. sporogenes and inulin) Placebo T2DM hs-CRP

Asgharian, 2016[103] Iran Both 47 Parallel 8 38, 36 L. acidophilus, L. casei, L. rhamnosus, L. bulgaricus, B. breve, B. longum, S.thermophilesþFOS

Placebo NAFLD hs-CRP

Barreto, 2013[104] Brazil Female 62.5 Parrallel 12 12, 12 L. plantarum Fermented milk Metabolic syndrome hs-CRP, IL-6, TNF

Bayat, 2016[105] Iran Both 50.5 Parallel 8 20, 20 Not mentioned Dietary advisement NAFLD hs-CRP

Ekhlasi, 2017[106] Iran Both 44 Parallel 8 15, 15 Syniotic (L. casei, L. rhamnosus, S.thermophilus, B.

breve,L. acidophilus, B. longum, L. bulgaricusþFOS)

Placebo NAFLD TNF

Eslamparast, 2014[107] Iran Both 46 Parallel 28 26, 26 Synbiotic (L. casei, L. rhamnosus, S.thermophilus, B.

breve, L. acidophilus, B. longum, L. bulgaricus)þFOS

Placebo NAFLD hs-CRP, TNF

Farrokhian, 2017[108] Iran Both 64 Parallel 12 30, 30 L. acidophilus, L. casei, B. biﬁdum, plus 800 mg inulin

Placebo T2DMþCHD hs-CRP

Feizollahzadeh, 2017[109] Iran Both 55.25 Parallel 8 20, 20 L. planetarum Placebo T2DM hs-CRP, IL1B

Firuzi, 2017[110] Iran Both 53.5 Parallel 12 48, 53 L. acidophilus, L. casei, L. lactis and B. biﬁdum, B.

longum and B. infantis

Placebo T2DM hs-CRP

Ghanei, 2018[111] Iran Female 29.5 Parallel 12 30, 30 L.acidophilus, L. Plantarum, L. Fermentum, L.Gasseri Placebo PCOShttps://www.

mayoclinic.org/

diseases-conditions/

pcos/symptoms- causes/syc-20353439

hs-CRP, IL-6, TNF, IL-10

Hove, 2015[112] Denmark Male 59.5 Parallel 12 23, 18 Milk fermented withL. helveticus Artiﬁcially acidiﬁed

milk

T2DM hs-CRP, TNF

Karamali, 2018[113] Iran Female 27 Parallel 12 30, 30 L. acidophilus, L. casei and B. biﬁdum Placebo PCOShttps://www.

mayoclinic.org/

diseases-conditions/

pcos/symptoms- causes/syc-20353439

hs-CRP

Karimi, 2018[114] Iran Female 28.5 Parallel 12 50, 49 synbiotic (L. acidophilus, L. casei, L. bulgaricus, L.rhamnosus, B. longum, B. breve,

S.thermophilusþFOS)

Placebo PCOS hs-CRP

Kobyliak, 2018[115] Ukraine Both 54.5 Parallel 8 31, 22 Synbiotic (14 probiotic strains: Lactobacillus, Lactococcus, Biﬁdobacterium, Propionibacterium, Acetobacter)

Placebo T2DM IL1B,IL-6,TNF,IL-8,INF

Kooshki, 2015[116] Iran Both 54 prallel 8 22, 22 Synbiotic Placebo T2DM hs-CRP, IL-6, TNF

Mazloom, 2013[117] Iran Both 53 Parallel 6 16, 18 L. acidophilus, L. bulgaricus, L. biﬁdum, and L. casei Placebo T2DM hs-CRP, IL-6

Mobini, 2017[118] Sweden Both 64.5 parallel 12 14, 15 L. reuteri DSM 17938 Placebo T2DM hs-CRP

Moﬁdi, 2017[119] Iran Both 45.35 Parallel 28 25, 25 Symbiotic (seven strains (L. casei, L. rhamnosus,

S.thermophilus, B. breve, L. acidophilus, B. longum and L. bulgaricus) and fructooligosaccharide)

Placebo NAFLD hs-CRP, TNF

Mohamadshahi, 2014[120] Iran Both 51 Parallel 8 21, 21 probiotic yogurt: L. delbrueckii subsp. bulgaricus and S. thermophilusþB. animalis subsp. lactisþL.

acidophilus

Conventional yogurt

T2DM hs-CRP, IL-6, TNF

Mohseni, 2018[121] Iran Both 60.55 Parallel 12 30, 30 L. acidophilus, L. casei, L. Fermentum and B. biﬁdum with antibiotic

Placebo T2DM hs-CRP

Nasri, 2018[122] Iran Female 25.8 Parallel 12 30, 30 L. acidophilus, L. casei and B. biﬁdum Placebo PCOS hs-CRP

Rabiei, 2018[123] Iran Both 59 Parallel 12 20, 20 Placebo Metabolic Syndrome hs-CRP, IL-6

Please cit e this article as: Kazemi A e t al., Effect of prob io tic and synbio tic supplem entation on in

ﬂ

(7)

3.1.1. Healthy

From 30 clinical trials, two were excluded since data were reported as geometric mean

[33]

and least square mean

[40], so 28

studies in healthy subjects with normal weight or obese with 2395 participants were included. Twenty studies had two arms,

ﬁ

ve had three arms and three had four arms. Participants of three studies were subjects with elevated serum cholesterol, hyper- triglyceridemia and migraine that were analysed in healthy class.

In the pooled analysis of studies, a signi

ﬁ

cant effect of probiotic on serum CRP reduction (SMD

¼

0.20 mg/l, 95%CI -0.33 to 0.06, p

¼

0.005) was observed with a heterogeneity of 59% (p

¼<

0.001) (Fig. 2 and

Table 5). Effect of probiotic on CRP remained signiﬁ

cant with a larger effect on subjects with BMI 25 (kg/m

²

) (SMD

¼

0.29, p

¼<

0.001,

I²¼

68.8%, p

¼<

0.001), whereas for BMI

>

25 (kg/m

²

), it was not signi

ﬁ

cant (SMD

¼

0.008, p

¼

0.24,

I²¼

0.0 p

¼

0.46) (Sup Fig. 1). Moreover, the supplement was most effective at dose 5 10

¹⁰

compared to the

>

10

⁹

-

<

5 10

¹⁰

and

<

10

⁹

.

Meta-regression for age, duration of studies, and BMI did not show a signi

ﬁ

cant trend. Meta-regression for age after adjusting for BMI, indicated a signi

ﬁ

cant inverse association (Coef

¼

0.01, p

¼

0.03,

I²

_residual

¼

51.39%, Adjusted R

²¼

17.93%).

We also conducted subgroup-analysis by the strains of bacteria (we pooled healthy individuals with patients with metabolic disorders).

L. casei

increased CRP, while B.breve,

L. rhamnosus, L. sporogenesþ

inulin and two multiple strains formula containing (Bi

ﬁdobacterium longum, B. infantis, B. breve, L.acidophilus, L. paracasei,L. bulgaricus,Lactobacillus plantarum) and (Lactobacillus acidophilus,L. casei,B. biﬁdum) decreased CRP.Lactobacillus helveticus,L. curvatusþL. plantarum,B. lactis,L. reuteri and L. plantarum

had no effect. The results are presented in

Table 6

and

Sup Fig. 2.

3.1.2. Metabolic disorders

From 29 clinical trials in participants with metabolic disorders (diabetes (16), polycystic ovary syndrome (PCOS) (4), fatty liver (5) and metabolic syndrome (MS) (4)), 1815 participants were included. Twenty-six studies had two arms and three had three arms. In the pooled analysis of studies, a signi

ﬁ

cant effect of probiotic on serum CRP reduction (SMD

¼

0.32 mg/l, 95% CI -0.57 to 0.08, p

¼

0.009) was observed with a heterogeneity of 84.2%

(p

¼<

0.001) (Fig. 3 and

Table 5). Meta-regression for age, duration

of studies and BMI was not signi

ﬁ

cant.

The subgroup analysis by disease revealed a signi

ﬁ

cant decrease of CRP in diabetes (SMD

¼

0.41, p

¼<

0.001, heterogeneity (79.5%, p

¼<

0.001)) and fatty liver (SMD

¼

0.38, p

¼<

0.0011, heterogeneity (82.4%, p

¼<

0.001)) while no change was observed in PCOS (SMD

¼

0.20, p

¼

0.08, heterogeneity (92.4%, p

¼<

0.001)) and metabolic syndrome (SMD

¼

0.13, p

¼

0.11, heterogeneity (28%, p

¼

0.25) (Sup Fig. 3,

Table 5).

3.1.3. IBD

From seven clinical trials in patients with IBD, 1868 participants were included. Four studies had two arms and one had four arms.

In the pooled analysis of studies, a signi

ﬁ

¼

1.37, 95% CI -1.81 to 0.47, p

¼

0.002) was observed with a heterogeneity of 84.1% (p

¼<

0.001) (Fig. 4,

Table 5). Meta-regression for duration of studies was not

signi

ﬁ

cant.

3.1.4. Arthritis

From eight clinical trials in patients with arthritis (seven rheumatoid arthritis, one spondyloarthritis), 750 participants were included. In the pooled analysis of studies, a signi

ﬁ

¼

0.58, 95% CI -1.15 to 0.01, p

¼

0.04) was observed with a heterogeneity of 90.3%

Symbiotic(L.casei,L.rhamnosus,S.thermophilus,B. breve,L.acidophilus,B.longum,L.bulgaricusþFOS Raygan,2018[124]IranBoth61.25Parallel1230,30Bifdobacteriumbifdum,L.casei,L.acidophilusPlaceboT2DMþCHDhs-CRP Rezaei,2017[125]IranBoth50Parallel445,45L.acidophilusLa5andB.lactisBb12PlaceboT2DMhs-CRP Sato,2017[126]JapanBoth64.5Parallel1634,34L.caseistrainShirota-fermentedmilkPlaceboT2DMhs-CRP,IL-6,TNF Sherf-Dagan,2018[127]IsraelBoth43Parallel2440,40L.acidophilus,B.biﬁdum,L.rhamnosus,Lactococcus lactis,L.casei,B.breve,S.thermophiles,B.longum,L. paracasei,L.plantarum,B.infatis

PlaceboNAFLDhs-CRP,IL-6,TNF,IL-10 Shoaei,2017[128]IranFemale26.1Parallel832,33Familact(L.casei,L.acidophilus,L.rhamnosus,L. bulgaricus,B.breve,B.longum,S.thermophiles)PlaceboPCOShs-CRP Soleimani,2017[129]IranBoth56.5Parallel1230,30L.acidophilus,L.caseiandB.biﬁdumPlaceboDiabetichemodialysishs-CRP Tajadadi,2014[130]IranBoth52.3Parallel827,27Symbioticbread(L.sporogenesandinulin)BreadT2DMhs-CRP Tonucci,2015[131]BrazilBoth51.39Parallel623,22FermentedmilkcontainingL.acidophilusLa-5and B.animalissubsplactisBB-12Conventional fermentedmilkT2DMIL-6,TNF,IL-10 Xavier-Santos,2018[132]BrazilBoth48.25Parallel823,22Synbiotic(moussecontainingL.acidophilusLa-5)PlaceboMetabolicSyndromeIL1B,IL-6,TNF,IL-8,IL- 10,IL-12 B.,biﬁdobacterium;CHD,coronaryheartdisease;FOS,fructooligossacaride;hsCRPhighsensitiveC-reactiveprotein;IFN,interferon;IL,interleukin;L.,lactobacillus;NAFLD,non-alcoholicfattyliverdisease;PCOS,polycystic ovarysyndrome;S,streptococcus;TNF,tumornecrosisfactorT2DM,type2diabetesmellitus.

ﬂ

ammatory markers in health and disease

(8)

Table 3

Characteristic of randomized controlled trials that evaluated the effect of the probiotic/synbiotic on the serum level of inﬂammatory biomarkers in unhealthy participants.

Author,Year Country Sex Age (year) Design Duration

(weeks)

Participants (Intervention, Control)

Abbas, 2014[98] Pakistan Both 35.3 Parallel 6 37, 35 Saccharomyces boulardii Placebo IBS TNF, IL-8, IL-10, IL-12

Akbari, 2016[133] Iran Both 80 Parallel 12 30, 30 L. acidophilus, L. casei, B. biﬁdum, and L.

fermentum

pacebo Alzheimer's Disease hs-CRP

Akkasheh, 2016[134] Iran Both 37.25 Parallel 8 20, 20 L. acidophilus, L. casei, B. biﬁdum Placebo Major depression hs-CRP

Alberda, 2007[135] Canada Both 62 Parallel 1 10, 9 L. casei, L. plantarum, L. acidophilus, and

L. Bulgaricus), 3 strains of Biﬁdobacterium (B. longum, B.

breveand B. infantis) and S.salivarius subsp. Thermophilus

Placebo Multiple organ dysfunction

syndrome

hs-CRP

Alipour, 2014[136] Iran Female 42.5 Parallel 8 22, 24 L. casei Placebo RA hs-CRP, IL1B, IL-6, TNF, IL-

10, IL-12

Anderson, 2004[137] UK Both 71 Parallel 2 67, 55 Synbiotic (L.acidophilus, B. lactis,

S.thermophilus, L. bulgaricusþFOS

Placebo Elective abdominal surgery hs-CRP, IL-6

Bajaj, 2008[138] USA 53 Parallel 8 14, 6 Probiotic yogurt No placebo Nonalcoholic minimal

hepatic encephalopathy cirrhotics

IL-6,TNF

Bajaj, 2014[139] USA Both 58.5 8 14, 16 L. GG Placebo Patients with cirrhosis IL1B,IL-6,TNF, IL-10,IL-17

Borges, 2018[140] Brazil Both 51.95 Parallel 12 16, 17 S. thermophilus, L. acidophilus, B.

longum

Placebo Hemodialysis hs-CRP, IL-6

Costanza, 2015[141] Brazil Parallel 12 8, 8 Saccharomyces boulardii Placebo Heart failure hs-CRP

Cui, 2013[142] China 2 23, 25 BiﬁdobacteriumþEnteral feeding Enteral feeding Acute pancreatitis hs-CRP, TNF, IL-8

Pineda, 2011[143] Canada Female 61.45 Parallel 12 15, 14 L. rhamnosus GR-1 and L. reuteri RC-14 Placebo Rheumatoid arthritis IL1B, IL-6, TNF, IL-8, IL-10,

IL-12, IL-17 De Roos, 2017[144] Netherlands Female 40 Parallel 12 31, 32 Probiotic (B. biﬁdum, B. lactis, L.

acidophilus, B. brevis, B. casei, B.salivarius, Lactococcus lactis)

Placebo Migraine hs-CRP, TNF

Dhiman, 2014[145] India Both 49 Parallel 24 16, 13 4 Lactobacillus species (paracasei,

plantarum, acidophilus, and delbrueckii subspecies bulgaricus), 3

Biﬁdobacterium species (longum, infantis and breve), and S.thermophilus.

Placebo Cirrhosis IL1B, IL-6, TNF

Drago[146],2012 Italy Both 30.46 Parallel 16 19, 19 L. salivarius Placebo Atopic Dermatitis IL-4, INF

Ebrahimi-Mameghani, 2013[147]

Iran Both 34.6 Parallel 1 20, 20 VSL#3 (4 strains of Lactobacillus (casei,

plantarum, acidophilus, Bulgaricus), 3 strains of Biﬁobacterium (longum, breve, infantis) and S.Thermophilus)

Placebo Surgical ICU hs-CRP, IL-6

Fang, 2018[148] China Both 46 Parallel 1 34, 34 Probioticþantibiotics Antibiotics Severe acute pancreatitis hs-CRP, IL-6

Federico, 2009[149] Italy Both 47 Parallel 8 9, 9 Synbiotic (L. paracaseiþGlutamine and

Zinc and XOSþInulin and Vitamin B6)

Placebo Ulcerative colitis IL1B, IL-6, TNF, IL-8, IL-10

Fernandes, 2016[150] Brazil 37 Parallel 2 3, 3 Synbiotic (FOSþL.paracasei, L.

rhamnosus, L.acidophilus, and B.lactis)

Placebo Roux-en-Y Gastric Bypass hs-CRP, IL1B, IL-6, TNF

Fujimori, 2009[151] Japan Both 36 Parallel 4 10,10,12 Synbiotic (B. longumþpsyllium) Psyllium Ulcerative colitis hs-CRP

Giamarellos-Bourboulis, 2009[152]

Greece Both 54.4 Parallel 2 31, 23 Synbiotic 2000FORTE, (Pediococcus

pentoseceus, Leuconostoc mesenteroides, L. paracasei ssp paracasei and L. plantarum,þinulin, oat bran, pectin and resistant starch)

Placebo Multiple Injuries hs-CRP

Gilbey, 2015[153] Israel Both 31.6 Parallel 1.4 27, 26 Streptococu Salivariusþantibiotics Antibiotics Acute tonsillitis hs-CRP

Groeger, 2013[154] Ireland Both Parallel 8 22, 26, 48 Biﬁobacterium infantis 35264 Placebo UC, Psoriasis, chronic

fatigue syndrome

hs-CRP, IL-6, TNF

Gupta, 2013[155] India Both 44 Parallel 8 25, 26 Probiotic (four strains of lactobacillus,

three strains of biﬁdobacterium and S.

Thermophilus)þPropranolol

Placeboþ Propranolol

Cirrhosis IL-6, TNF

Han, 2015[156] South Korea Both 52.7 Parallel 1 60, 57 L. subtilis/S.faecium Placebo Alcoholic hepatitis IL1B,TNF

Please cit e this article as: Kazemi A e t al., Effect of prob io tic and synbio tic supplem entation on in

ﬂ

(9)

Hatakka, 2003[157] Finland Both 51 Parallel 52 8, 13 L. rhamnosus GG Placebo RA hs-CRP, IL1B, IL-6, TNF, IL- 10, IL-12

Hod, 2017[158] Israeal Female 29.5 Parallel 8 54, 53 L. rhamnosus; L. casei; L. paracasei; L.

plantarum; L. acidophilus; B. bifdum; B.

longum; B. breve; B. infantis;

S.thermophilus; L. bulgaricus; and Lactococcus lacti

Placebo IBS hs-CRP

Horvath, 2016[159] Austria Both 58 Parallel 24 44, 36 B. biﬁdum, B. lactis, L. acidophilus, L.

brevis, L. casei W56, L. salivariu, Lactococcus lactis and Lactococcus lactis

Placebo Cirrhosis hs-CRP

Hummelen, 2011[160] Netherlands Both Parallel 10 19, 25 L. rhamnosus GR-1 and L. reuteri RC-14 Placebo HIV IL-10, INF

Inoue, 2014[161] Japan Both 29.6 Parallel 8 24, 25 L. acidophilus Placebo Atopic Dermatitis IL-4, IL-5, IL-6, TNF, IL-10,

IL-12, IL-17, INF

Ivory, 2008[162] UK Both 18e45 Parallel 20 10, 10 Milk containing L. casei Shirota Milk Seasonal allergic rhinitis IL1B, IL-5, IL-6, TNF, INF

Jafarnejad, 2017[19] Iran Female 58 Parallel 24 20, 21 (L. casei, B. longum, L. acidophilus, L.

rhamnosus, L. bulgaricus, Biﬁobacteriumbreve, and S.thermophilus)þCa VitD

Calcium and Vitamin D

Osteopenic

Postmenopausal Women

IL1B, TNF

Jain, 2004[163] UK Both 72.5 Parallel 1 45, 45 Synbiotic (L. acidophilus, B. lactis, S

thermophilus and L. bulgaricus with oligofructose)

Placebo Critically ill patients hs-CRP

Jenks, 2010[164] New Zealand Both 43.3 Parallel 12 32, 31 S.salivarius, B. lactis, and L. acidophilus Placebo Spondylo arthritis hs-CRP

Kanazawa, 2005[165] Japan Both 63.75 Parallel 2 21, 23 Synbiotic (B. breve and L. casei strain

ShirotaþGOS)

Entral feeding Biliary cancer undergoing high-risk hepatectomy

hs-CRP

Karbaschian, 2018[166] Iran Both 34.65 Parallel 16 23, 23 Synbiotic (L. casei, L. rhamnosus,

S.thermophilus, B. breve, L. acidophilus, B. longum, and L.

bulgaricusþFOS)þmultivitamin and mineral

Multivitamin and mineral

Mini Gastric Bypass hs-CRP, IL-6, TNF

Karlsson, 2010[167] Sweden Male 68.5 Parallel 4 9, 7 Synbiotic (oat drink drink with L.

plantarum)

Oat drink Cardiovascular disease hs-CRP,,,,IL-6,TNF,,,,,

Kawase, 2009[168] Japan Both 36.8 Parallel 6 20, 18 L. GG, L. gasseri TMC0356 Placebo Japanese cedar pollinosis hs-CRP

Koga, 2013[169] Japan Both 53.2 Parallel 2 18, 19 L. casei Shirota YIT 9029 Placebo Alcoholic liver cirrhosis hs-CRP, IL-6

kouchaki, 2016[170] Iran Both 33.5 Parallel 12 30,30 Probiotic (L. acidophilus, L. casei, B.

biﬁdum and L. fermentum)

Placebo Multiple sclerosis hs-CRP

Krebs, 2013[171] Slovenia Both Parallel 20, 16 Symbiotic (Pediacoccus pentosaceus,

Leuconostoc mesenteroides, L.

paracasei, L. plantarum and bioactive plantﬁbers:beta- glycan, inulin, pectin, resistant starch)

No placebo Patients with preceding large bowel operation for colorectal cancer

hs-CRP, IL-6

Krebs, 2016[172] Slovenia Both 64.5 Parallel 0.5 18, 16 Prebiotic (2.5 g of each of the four

fermentableﬁbres: betaglucan, inulin, pectin and resistant starch.)

IL-6

Krebs, 2016 Slovenia Both 62 1 18, 16 Synbiotic: (Pediacoccus pentosaceus,

paracasei subsp paracasei, and L.

plantarum)þprebiotic (2.5 g of each of the four fermentableﬁbres: betaglucan, inulin, pectin and resistant starch.)

IL-6,

Lei, 2017[173] China Both 66.8 Parallel 24 215, 218 Skimmed milk containg L. casei Shirota Skimmed milk Osteoarthritis hs-CRP

Liu, 2014[174] China Both 41.5 Parallel 8 21, 21 Biﬁdobacterium triple viable bacterial

preparationþSalofalk

Salofalk Ulcerative colitis TNF, IL-8, IL-10

Liu, 2015[175] China Both 39.88 Parallel 8 30, 30 Probiotic and diet therapy Diet therapy Ulcerative colitis hs-CRP

Mandel, 2010[176] USA Both 62.5 Parallel 8 22, 22 Bacillus coagulans Placebo RA

Malaguarnera, 2012[177] Italy Both 46.8 Parallel 24 34, 32 Symbiotic (B. longum and

fructooligosaccharide) plus lifestyle modiﬁcation and vitamin B

Lifestyle modiﬁcation

Non Alcoholic Steatohepatitis

hs-CRP, TNF

McNaught, 2002[178] UK Both 68.5 Parallel 1.2 64, 65 Without placebo Elective surgical patients hs-CRP

(continued on next page)

Please cit e this article as: Kazemi A e t al., Effect of prob io tic and synbio tic supplemen tation on in

ﬂ

ammat ory mar kers in health and disease status: A syst ematic revi ew and meta-anal y sis of clinical trials, Clinical N utrition, https://doi. org/1 0. 1 0 1 6/j.clnu. 20 1 9.04 .0 04

(10)

Table 3(continued)

Author,Year Country Sex Age (year) Design Duration

(weeks)

Participants (Intervention, Control)

Synbiotic (oatmeal based drink per ml of L. plantarum 299v)

McNaught, 2005[178] UK Both 71 Parallel 1.2 52, 51 Synbiotic (oatmeal based drink per ml

of L. plantarum 299v)

Placebo Critically ill patient IL-6

Mimura, 2017[179] Brazil Both 37.45 Parallel 16 22, 23 Synbiotic (FOS and L. paracasei, L.

rhamnosus, L. acidophilus, and of B.

lactis)

Placebo Recurrent aphthous

stomatitis

IL-4, TNF, IL-10, IL-12, IL- 17, INF

Minami, 2018[180] Japan Both 45.5 Parallel 12 40, 40 B. breve Placebo pre-obese adults hs-CRP

Mizuta, 2016[181] Japan Both 70 Parallel 3 31, 29 B. longum BB536 Without placebo Patients undergoing

colorectal resection

hs-CRP, IL-6

Nagata, 2010[182] Japan Female 22 Parallel 6 16, 17 L. plantarum No.14 Placebo Seasonal allergic disease hs-CRP

Natarajan, 2014[183] USA Both 54 Cross-over 8 21, 21 S. thermophilus KB 19, L. acidophilus KB

27, and B. longum KB 31

Placebo Dialysis hs-CRP

Nilsson, 2018[184] Sweden Female Parallel 48 45, 45 L. reuteri Placebo old women with low bone

mineral density

hs-CRP, TNF

Panahi, 2017[185] Iran Male 41.8 Parallel 6 40, 20 Probiotic (L. acidophilus, L. bulgaricus, L.

rhamnosus, L. casei, B. breve, B. longum and S.thermophilusthe)

Placebo Pulmonary diseaes due to

sulfur mustard exposure

hs-CRP

Rayes, 2002[186] Germany Both 61 Parallel 1.2 30, 30,30 1. Enteral feeding contains synbiotic

(ﬁber and lactobacillus) 2. Enteral feeding contains heat killed lactobacillus

Enteral feeding (no placebo)

Major abdominal surgery hs-CRP

Roller, 2007[187] Ireland Both Parallel 12 19, 15 Synbiotic (L. rhamnosus GG,B. lactis

Bb12 and inulin enriched with oligofructose)

Placebo Cancer patients TNF, IL-10, IL-12, INF

Rossi, 2016[188] Australia Both 69 Cross-over 6 31, 31 Synbiotic (nine different strains

contained Lactobacillus, Biﬁdobacteria, and StreptococcusþFOS and GOS

Placebo Renal Failure IL1B, IL-6, TNF, IL-10

Schunter, 2012[189] USA Female 47.6 Parallel 4 14, 13 Symbiotic (Pediococcus pentosaceus,

paracasei subsp paracasei,L. plantarum and betaglucan, inulin, pectin, and resistant starch)

Betaglucan, inulin, pectin, and resistant starch

HIV hs-CRP

Shadnoush, 2013[190] Iran 37.69 Parallel 8 86, 90 B. BB-12 and L. acidophilus Placebo IBD hs-CRP, IL1B, IL-6, TNF, IL-

10

Shariaty, 2017[191] Iran Both 57.8 Parallel 4 18, 18 L. casei, L. acidophilus, L. rhamnosus, L.

bulgaricus, B. breve, B. longum, and S.thermophilus

Placebo CRF under hemodialysis hs-CRP

Sharma, 2011[192] India Both 40.5 Parallel 1 22, 18 L. acidophilus, B. longum, B. biﬁdum,

and B. infantis

Placebo Acute Pancreatitis hs-CRP

Singh, 2013[193] Switzerland Both 30 Parallel 8 s 10, 10 B. lactis NCC2818 Placebo Seasonal allergic rhinitis IL1B, IL-5, TNF

Smecuol, 2013[194] Argentina Both 43 (median) Parallel 3 10, 7 B. infantis Placebo Celiac Disease IL-4, IL-5, IL-6, TNF, IL-10,

IL-12, IL-17, INF

Stadlbauer, 2008[195] UK Both 54.2 Parallel 4 10, 8 L. casei Shirota Without placebo Cirrhosis hs-CRP

Tan, 2011[196] China Both 40 Parallel 3 22, 21 B. longum, L. bulgaricus, S.

thermophilus

No placebo Traumatic brain injury hs-CRP, IL-4, IL-6, IL-10, IL-12, INF

Tomasik, 2015[197] USA Both 46.45 Parallel 14 30, 27 L. rhamnosus strain GG and B. animalis

subsp. lactis

Placebo Schizophrenia hs-CRP

Usami, 2011[198] Japan Both 65.6 Parallel 4 32, 29 Synbiotic (B. breve, L casei strain

ShirotaþGOS)

Without placebo Hepatic surgery with or without cirrhosis

hs-CRP, IL-6 Vaghef-Mehrabany, 2014

[199]

Iran Female 42.7 Parallel 8 22, 24 L. casei Placebo Rheumatoid arthritis IL1B, IL-6, TNF, IL-10, IL-

12,

Villar-García, 2015[200] Spain Both 47.5 Parallel 12 22, 22 Saccharomyces boulardii Placebo HIV hs-CRP, IL-6

Viramontes-Horner, 2014 [201]

Mexico Both 39.8 Parallel 8 20, 15 Synbiotic gel (L. acidophilus, B. lactis

and Inulin) plus nutritional counseling

Placebo Hemodialysis hs-CRP

Please cit e this article as: Kazemi A e t al., Effect of prob io tic and synbio tic supplem entation on in

ﬂ

(11)

(p

¼<

0.001) (Fig. 5,

Table 5). Meta-regression for age and duration

of studies was not signi

ﬁ

cant.

3.1.5. Critically ill patients

From eight clinical trials in critically ill patients, 428 participants were included (Multiple organ dysfunction syndrome (1), acute pancreatitis (2), critically ill patients (1), surgical ICU (1), patients with multiple injuries (1), traumatic brain injury (1), acute tonsil- litis (1)). Seven studies had two arms and one had three arms.

In the pooled analysis of studies, a signi

ﬁ

¼

0.66, 95% CI 1.03 to 0.29, p

¼<

¼

0.001) (Fig. 6,

Table 5).

Meta-regression for duration was not signi

ﬁ

cant. Egger test showed no publication bias (p

¼

0.11). In

ﬂ

uence analysis showed that none of the trials had signi

ﬁ

cant effect on pooled effect size.

3.1.6. Surgery

Effect of probiotic on CRP was examined in 10 studies where participants underwent surgery. In these studies, serum level of CRP was assessed immediately before surgery and maximum up to two weeks after the surgery except one that measure CRP four weeks before and 12 weeks after the surgery so this study was excluded.

In the pooled analysis of nine studies in subjects who underwent surgery (519 participants), no signi

ﬁ

cant change in serum CRP (SMD

¼

0.06, 95% CI -0.35 to 0.22, p

¼

0.66) was observed with a heterogeneity of 59.9% (p

¼

0.008) (Fig. 7,

Table 5).

3.1.6.1. CRF patients under hemodialysis.

In the pooled analysis of four studies in CRF patients under hemodialysis (134 participants), no signi

ﬁ

cant effect of probiotic on serum CRP (SMD

¼

0.12, 95%

CI -0.47 to 0.22, p

¼

0.19) (Table 5).

3.1.7. HIV

In the pooled analysis of four studies in patients with HIV (118 participants), no signi

ﬁ

cant effect of probiotic on serum CRP (SMD

¼

0.23, 95% CI -0.60 to 0.13, p

¼

0.12) (Table 5).

3.1.8. Cirrhosis

In the pooled analysis of three studies in cirrhotic patients (135 participants), no signi

ﬁ

cant effect of probiotic on serum CRP (SMD

¼

0.80, 95% CI -2.80 to 1.14, p

¼

¼<

0.001) (Table 5).

3.2. IL-10

Forty-two clinical trials measured IL-10 (healthy (22), diabetes (1), NAFLD (1), PCOS (1), MS (2), RA (4), IBD (4), IBS (1), CRF (2), cancer (1), HIV (1), aphthous stomatitis (1), trauma (1), atopic dermatitis (1).

3.2.1. Healthy

From 22 clinical trials in healthy subjects with normal weight or obese, 1555 participants were included. Seventeen studies had two arms, three had three arms, three had four arms and one had

ﬁ

ve arms. Nine studies measured IL-10 in serum, four in PBMC and seven measured IL-10 secretion from mitogen stimulated PBMC.

In the pooled analysis of studies, no signi

ﬁ

cant effect of probiotic on IL-10 (SMD

¼

0.14 pg/ml, 95% CI -0.35 to 0.08, p

¼

¼<

0.001) (Fig. 8,

Table 5).

Subgroup analysis by the sample in which IL-10 was measured revealed a signi

ﬁ

cant increase in IL-10 measured in PBMC

Wang,2015[202]ChinaBoth52.5Parallel2421,18B.bifidum.catenulatum,B.longum,and L.plantarumPlaceboPeritonealdialysisIL-5,IL-6,TNF,IL-10,IL- 17,INF Wassenberg,2011[203]SwitzerlandBoth26.8Cross-over430,30L.helveticus,L.paracaseiPlaceboAllergicrhinitisIL-4,IL-5,IL-8,IL-10,INF Xiao,2006[204]JapanBoth36.6Parallel1420,20YogurtcontainingB.bifidum536YogurtJCPIL-10,INF Yamamoto,2016[205]JapanBoth26.25Parallel2424,26L.acidophilusL-92PlaceboAtopicdermatitisTNF,IL-8,IL-12 Yang,2014[206]USAMale50Parallel1210,7BacilluscoagulansGBI-30PlaceboHIVhs-CRP,TNF Yokoyama,2014[207]JapanBoth65.5Parallel321,21Synbiotic(L.caseistrainShirota,B. breveandGOS)WithoutplaceboPatientswithoesophageal cancerhs-CRP Zamani,2016[208]IranBoth51.4Parallel830,30L.acidophilus,L.casei,B.bifidumPlaceboRheumatoidarthritishs-CRP Zamani,2017[209]IranBoth49.4Parallel827,27Symbiotic(L.acidophilus,L.caseiandB. bifidumplusinulin)PlaceboRheumatoidarthritishs-CRP Zhang,2018[97]ChinaBoth36Parallel838,38Symbiotic(Bifidobacteriumplus mesalazine)PlaceboModerateulcerativecolitishs-CRP,IL-4,IL-8 Zhou,2017[210]ChinaBothParallel45,45BifidobacteriumtripleviableDiarrheasecondaryto chemotherapyIL-6,TNF Zhou,2017[211]ChinaBoth39.8Parallel633,32Symbiotic(Bifidobacteriumplus mesalazine)PlaceboActiveulcerativecolitisIL-6,TNF,INF B.,bifidobacterium;FOS,fructooligossacaride;GOS,galactooligosaccharide;hsCRPhighsensitiveC-reactiveprotein;IFN,interferon;IL,interleukin;IBS,inflammatorybowelsyndrome;JCP,Japanesecedarpollinizes;L., lactobacillus;RA,rheumatoidarthritis;S,streptococcus;TNF,tumornecrosisfactor.

Please cite this article as: Kazemi A et al., Effect of probiotic and synbiotic supplementation on in

ﬂ

(12)

NOTE: Weights are from random effects analysis Overall (I-squared = 59.2%, p = 0.000) Lee A (2017)

Cavallini (2016)

Irwin (synbiotic) (2017) Iwasa (2013)

Costabile (2017)

Simons (2006) ID

Lee (2017)

Minami (2018)

Rajkumar (2015) Kekkonen (2008)

Seifert (2011)

Stenman (2016) Valentini (2015) Brahe (2015)

Nilsson (2018) Kekkonen (2008)

Stenman (2016) Macfarlane (2013)

Stenman (2016) De Roos (2017)

Osterberg (2015)

cox (2014) Lee (2017) Akkasheh (2016)

Costabile (2017)

Kekkonen (2008) Ahn (2015)

Jung (2015)

Rajkumar (2015) Rajkumar (2014) Nyangale (2014) Mohammadi (2015)

Sanchez (2014) Lee (2017)

Irwin (probiotic) (2017)

Zarrati (2014) Minami (2015)

cox (2014) Mohammadi (2015) Study

-0.20 (-0.33, -0.06) -0.12 (-0.44, 0.20) -0.32 (-1.02, 0.38)

-0.74 (-1.49, 0.02) -0.31 (-1.12, 0.49)

-0.70 (-1.08, -0.32)

-0.37 (-0.97, 0.22) SMD (95% CI)

0.13 (-0.42, 0.68)

-0.31 (-0.75, 0.13)

-0.00 (-0.72, 0.71) -0.43 (-1.12, 0.26)

0.27 (-0.21, 0.75)

-0.13 (-0.59, 0.33) -1.39 (-1.94, -0.83) 1.14 (0.41, 1.87)

0.03 (-0.38, 0.44) 0.03 (-0.67, 0.72)

-0.50 (-1.02, 0.02) -0.51 (-0.95, -0.08)

-0.20 (-0.66, 0.26) 0.19 (-0.31, 0.68)

-0.56 (-1.46, 0.34)

0.00 (-0.43, 0.43) -0.15 (-0.71, 0.40) -0.78 (-1.43, -0.14)

-1.01 (-1.49, -0.52)

-0.62 (-1.37, 0.13) 0.04 (-0.37, 0.45)

-0.19 (-0.59, 0.22)

-0.00 (-0.72, 0.71) -0.69 (-1.42, 0.05) 0.59 (-0.07, 1.26) 0.07 (-0.77, 0.91)

0.01 (-0.40, 0.42) 0.63 (0.07, 1.19) -0.18 (-0.91, 0.55)

0.37 (-0.19, 0.93) -0.51 (-1.12, 0.10)

-0.28 (-0.70, 0.13) -0.22 (-1.05, 0.61)

100.00 3.69 2.10

1.93 1.78

3.39

2.46 Weight

2.65

3.12

2.05 2.13

2.96

3.03 2.62 2.00

3.25 2.12

2.78 3.15

3.04 2.88

1.55

3.18 2.63 2.29

2.93

1.94 3.27

3.30

2.05 1.98 2.20 1.69

3.28 2.60 2.00

2.61 2.43

3.24 1.72

%

-0.20 (-0.33, -0.06) -0.12 (-0.44, 0.20) -0.32 (-1.02, 0.38)

-0.74 (-1.49, 0.02) -0.31 (-1.12, 0.49)

-0.70 (-1.08, -0.32)

-0.37 (-0.97, 0.22) SMD (95% CI)

0.13 (-0.42, 0.68)

-0.31 (-0.75, 0.13)

-0.00 (-0.72, 0.71) -0.43 (-1.12, 0.26)

0.27 (-0.21, 0.75)

-0.13 (-0.59, 0.33) -1.39 (-1.94, -0.83) 1.14 (0.41, 1.87)

0.03 (-0.38, 0.44) 0.03 (-0.67, 0.72)

-0.50 (-1.02, 0.02) -0.51 (-0.95, -0.08)

-0.20 (-0.66, 0.26) 0.19 (-0.31, 0.68)

-0.56 (-1.46, 0.34)

0.00 (-0.43, 0.43) -0.15 (-0.71, 0.40) -0.78 (-1.43, -0.14)

-1.01 (-1.49, -0.52)

-0.62 (-1.37, 0.13) 0.04 (-0.37, 0.45)

-0.19 (-0.59, 0.22)

-0.00 (-0.72, 0.71) -0.69 (-1.42, 0.05) 0.59 (-0.07, 1.26) 0.07 (-0.77, 0.91)

0.01 (-0.40, 0.42) 0.63 (0.07, 1.19) -0.18 (-0.91, 0.55)

0.37 (-0.19, 0.93) -0.51 (-1.12, 0.10)

-0.28 (-0.70, 0.13) -0.22 (-1.05, 0.61)

100.00 3.69 2.10

1.93 1.78

3.39

2.46 Weight

2.65

3.12

2.05 2.13

2.96

3.03 2.62 2.00

3.25 2.12

2.78 3.15

3.04 2.88

1.55

3.18 2.63 2.29

2.93

1.94 3.27

3.30

2.05 1.98 2.20 1.69

3.28 2.60 2.00

2.61 2.43

3.24 1.72

%

0

-1.94 0 1.94

Fig. 2.Forest plot displaying standard mean difference and 95% conﬁdence intervals for the impact of probiotic administration on C-reactive protein (CRP) levels in healthy subjects.

(The studies with more than two arms (different strains of probiotic was investigated in each arm), were considered as more than one studies.).

Table 4

Summary of the results.

CRP TNF-a IL-1B IL-6 IL-10 IFN-g IL-4 IL-8 IL-12 TGF-B

Healthy (In PBMC) No change (In PBMC) (IFN secreted from PBMC) No change No change No change No change

Metabolic disorders No change No change No change

Diabetes No change No change

NAFLD No change

MSþPCOS No change No change

MS No change No change

IBD No change

Arthritis No change No change No change

Cihrrosis Critically ill

Surgery No change

CRF No change

B., biﬁdobacterium; CHD, coronary heart disease; FOS, fructooligossacaride; hsCRP high sensitive C-reactive protein; IFN, interferon; IL, interleukin; L., lactobacillus; NAFLD, non-alcoholic fatty liver disease; PCOS, polycystic ovary syndrome; S, streptococcus; TNF, tumor necrosis factor T2DM, type 2 diabetes mellitus.

or : Small effect Or : Moderate effect Or : Large effect.

Please cite this article as: Kazemi A et al., Effect of probiotic and synbiotic supplementation on in

ﬂ

(13)

(SMD

¼

0.47, p

¼

0.001), while serum IL-10 decreased signi

ﬁ

cantly (SMD

¼

0.29, p

¼<

0.001) and secreted IL-10 from mitogen stimulated PBMC (SMD

¼

0.13, p

¼

0.11) did not change signi

ﬁ

cantly (Sup Fig. 4). Subgroup analysis by age indicated a signi

ﬁ

cant reduction of IL-10 in subjects with age

>

49 (SMD

¼

0.44, p

¼

0.0.0,

I²¼

87.0%, p

¼<

0.001) while for age 49, no signi

ﬁ

cant change as observed (SMD

¼

0.02, p

¼

0.74,

I² ¼

60.0%, p

¼<

0.001). The greatest decrease was observed at dose 5 10

¹⁰

compared to the

>

10

⁹

-

<

5 10

¹⁰

and

<

10

⁹

(Sup Table 4). Results of subgroup- analysis by the strains of bacteria (in healthy individuals and patients with metabolic disorders) are presented in

Table 6

and

Sup Fig. 5. IL-10 was decreased only in one of the subgroups(B. lactis).

Meta-regression for age and duration of studies was not sig- ni

ﬁ

cant while meta-regression for BMI (Coef

ﬁ

cient (Coef)

¼

0.13, p

¼

0.02, I

²

_residual

¼

79.21%, Adjusted R

²¼

31.7%) revealed a signi

ﬁ

cant association. Meta-regression for age after adjusting for BMI (Coef

¼

0.03, p

¼

0.009, I

²

_residual

¼

66.44%, Adjusted R

²¼

63.34%) was signi

ﬁ

cant.

3.2.2. Metabolic disorders

From

ﬁ

ve clinical trials in subjects with metabolic disorders (diabetes (1) fatty liver (1), PCOS (1), metabolic syndrome (2)), 358 participants were included. In the pooled analysis of studies, no

signi

ﬁ

cant effect of probiotic on IL-10 (SMD

¼

0.38, 95% CI -0.39 to 1.14, p

¼

(p

¼<

0.001) (Fig. 9,

Table 5).

3.2.3. IBD

In the pooled analysis of four studies in IBD patients (253 participants), no signi

ﬁ

cant effect of probiotic on IL-10 was seen (SMD

¼

3.1, 95%CI -2.06 to 8.27, p

¼

0.24) with a heterogeneity of 99% (p

¼<

0.001).

3.2.4. Arthritis

In the pooled analysis of four studies in patients with arthritis (142 participants), a signi

ﬁ

cant increase was observed in IL-10 (SMD

¼

0.51, 95%CI 0.17 to 0.86, p

¼

0.003) with a heterogeneity of 0.00% (p

¼

0.57).

3.3. IL-1B 3.3.1. Healthy

Thirty-three clinical trials measured IL-1B (healthy (17), diabetes (1), MS (1), alcoholic hepatitis (1), arthritis (4), IBD (2), allergy (2), CRF (1), cirrhosis (2), gastric bypass (1), celiac disease (1).

Table 5

Meta-analysis of the change in inﬂammatory markers in health and disease conditions.

Cytokine Participants health condition Num study (sample size) SMD 95% CI p Heterogeneity p

CRP Healthy 28 (2395) 0.20 0.33 to0.06 0.005 58.5% <0.001

Metabolic dis. 29 (1815) 0.37 0.61 to0.13 0.002 84.1% <0.001

Diabetes 16 (1065) 0.41 0.53 to0.28 <0.001 79.5% <0.001

NAFLD 5 (322) 0.38 0.60 to0.15 0.001 82.4% <0.001

MS 3 (84) 0.33 0.10 to 0.77 0.13 28% 0.25

PCOS 5 (344) 0.2 0.02 to 0.42 0.08 92.4% <0.001

IBD 7 (1868) 1.37 1.81 to0.47 0.002 84.1% <0.001

Arthritis 8 (750) 0.67 1.25 to0.09 0.02 89.2% <0.001

CRF 4 (134) 0.12 0.47 to 0.22 0.48 36.5% 0.19

HIV 3 (88) 0.16 0.59 to 0.26 0.45 63.8% 0.06

Surgery 9 (519) 0.065 0.35 to 0.22 0.66 59.9% 0.008

Cirrhosis 3 (135) 0.8 2.80 to 1.14 0.41 94.9% <0.001

Critically ill 8 (428) 0.66 1.03 to0.29 <0.001 69.4% 0.001

IL-1B Healthy 15 (838) 0.32 0.64 to0.005 0.053 79.5% <0.001

Arthritis 4 (142) 0.16 0.51 to 0.18 0.35 71.4% 0.02

IL-10 Healthy 21 (1477) 0.14 0.35 to 0.08 0.21 75.7% <0.001

Metabolic dis. 5 (358) 0.38 0.39 to 1.14 0.34 88.4% <0.001

IBD 4 (253) 3.1 2.06 to 8.27 0.24 99.0% <0.001

Arthritis 4 (142) 0.51 0.17 to 0.86 0.003 0.0% 0.57

IFN-g Healthy 24 (1515) 0.02 0.25 to 0.21 0.86 77.7% <0.001

IL-4 Healthy 11 (869) 0.07 0.21 to 0.06 0.29 0.0 0.57

IL-8 Healthy 11 (1055) 0.06 0.18 to 0.07 0.35 22.1% 0.19

IL-12 Healthy 7 (567) 0.01 0.33 to 0.30 0.94 68.3% 0.001

TGF-B Healthy 6 (411) 0.13 0.18 to 0.21 0.9 0.0% 0.65

IL-6 Healthy 23 (1456) 0.09 0.24 to 0.06 0.24 52.2% <0.001

Metabolic dis. 16 (707) 0.025 0.25 to 0.20 0.83 54.1% 0.005

DM 6 (344) 0.14 0.08 to 0.35 0.21 10.7% 0.35

Fatty liver 3 (150) 0.04 0.28 to 0.36 0.81 59.5% 0.08

MSþPCOS 6 (213) 0.31 0.58 to 0.03 0.03 66% 0.01

Cirrhosis 5 (167) 0.36 0.05 to 0.68 0.02 41.8% 0.14

Surgery 5 (319) 0.9 1.6 to0.18 0.003 85.5% <0.001

Arthritis 4 (142) 0.32 0.71 to 0.08 0.11 21.3% 0.28

CRF 4 (167) 0.35 0.029 to 0.67 0.03 0.0% 0.94

IBD 5 (4 0.11 1.35 to 1.57 0.88 97.7% <0.001

TNF-a Healthy 32 (2087) 0.30 0.49 to0.10 0.006 73.9% <0.001

Metabolic dis. 20 (955) 0.30 0.62 to 0.03 0.07 84.8% <0.001

IBD 6 (450) 0.87 1.23 to0.50 <0.001 67.5% 0.003

Cirrhosis 4 (130) 0.43 0.78 to0.08 0.02 0.0% 0.78

Fatty liver 7 (378) 0.75 0.97 to0.54 <0.001 83.9% <0.001

PCOSþMS 5 (178) 0.25 0.05 to 0.56 0.1 46% 0.12

Diabetes 7 (357) 0.02 0.23 to 0.19 0.86 84.2% <0.001

Arthritis 4 (142) 0.51 0.17 to 0.86 0.003 0.0% 0.57

CRF: chronic renal failure; dis., disorders; DM, diabetes mellitus; IBD: inﬂammatory bowel disease; MS: metabolic syndrome; NAFLD: non-alcoholic fatty liver disease; PCOS, Polycystic ovary syndrome.

Please cite this article as: Kazemi A et al., Effect of probiotic and synbiotic supplementation on in

ﬂ

(14)

From 17 clinical trials in healthy subjects, one was excluded as it measured IL-1. Therefore, 16 studies (916 participants) were included. Eleven studies had two arms, two studies had three and one had four arms. Five studies measured IL-1B in serum, two in PBMC and seven measured IL-10 secretion from mitogen stimulated PBMC.

In the pooled analysis of studies, effect of probiotic on IL-1B (SMD

¼

0.32 pg/ml, 95% CI -0.64 to 0.005, p

¼

0.053) was marginally signi

ﬁ

cant with a heterogeneity of 79.5% (p

¼<

0.001) (Fig. 10,

Table 5).

A subgroup analysis for the sample revealed a signi

ﬁ

cant decrease in IL-1B measured in PBMC (SMD

¼

0.53, p

¼

0.021),

Table 6

Sub group analysis by the strains of bacteria.

Strains SMD P I2 P

CRP

L. helveticus 0.30 0.23 0.0 0.98

B.breve 0.38 0.04 0.0 0.60

L. sporogenesþinulin 0.36 0.02 0.0 0.46

B.(longum, infantis, breve)þL.(acidophilus, paracasei, bulgaricus, plantarum) 1.02 <0.001 0.42 0.17

L. acidophilus, casei, B,biﬁdum 0.56 <0.001 62.3 0.004

L. casei 0.65 <0.001 89.8 <0.001

L. curvatusþL. plantarum 0.08 0.60 0.0 0.44

B.lactis 0.04 0.71 43.2 0.12

L. reuteri 0.02 0.90 0.0 0.91

L. plantarum 0.16 0.46 26.3 0.26

L.rhamnosus 0.53 <0.001 82.3 0.003

L.aidophilusþB.lactis 0.30 0.03 21.7 0.28

L. (casei, rhamnosus, acidophilus, bulgaricus)þB.(breve, longum)þS. thermophilesþFOS 0.17 0.08 90.9 <0.001

Other strains 0.13 0.045 79.9 <0.001

Overall 0.22 79.3 79.3 <0.001

TNF-a

L.rhamnosus 0.01 0.97 0.0 0.89

B.lactis 0.19 0.22 58.1 0.07

L. casei 0.01 0.89 56.2 0.02

L.reuteri 0.13 0.48 0.0 0.62

B.longumþprebiotic 0.09 0.55 95.4 <0.001

L.plantarum 0.32 0.32 0.0 0.80

FOSþL. paracasei,L. rhamnosus,L. acidophilus, andB. lactis 0.26 0.30 0.0 0.91

B.(infantis, longum, breve)þL.(acidophilus, bulgaricus, paracasei, plantarum)þS.thermophilus 0.04 0.86 0.0 0.47 L.(casei, rhamnosus, acidophilus, bulgaricus)þB.(breve, longum)S.thermophilusþFOS 0.26 0.02 80.9 <0.001

Other strains 0.2 0.002 65.6 <0.001

Overall 0.03 0.44 70.8 <0.001

IL-6

L.rhamnosus 0.21 0.10 0.0 0.73

B.lactis 0.18 0.13 54.6 0.05

L. casei 0.05 0.67 0.0 0.47

L.plantarum 0.71 0.005 87.1 <0.001

L.(casei, rhamnosus, acidophilus, bulgaricus)þB.(breve, longum)S.thermophilusþFOS 0.51 0.02 0.0 0.88

Other strains 0.19 0.02 57.1 0.002

IL-1B

L.casei 0.20 0.14 0.0 0.99

B.lactis 0.12 0.50 8.2 0.34

Other strains 0.05 0.54 86.1 <0.001

Overall 0.05 0.26 58.3 <0.001

IL-10

L.casei 0.04 0.69 54.1 0.04

B.lactis 0.47 0.003 65.9 0.03

B. longum 0.09 0.55 0.0 0.50

L.helveticus 0.36 0.08 30.8 0.23

B.lactisþL.casei 0.31 0.08 0.0 0.94

B.lactisþL.acidophilus 0.22 0.25 96.8 <0.001

Other strains 0.13 0.14 88.8 <0.001

Overall 0.07 0.17 79.7 <0.001

IFN-g

L.casei 0.04 0.68 3.5 0.40

B.lactis 0.21 0.24 0.0 0.59

B.lactisþL.acidophilus 0.30 0.14 0.0 0.40

B.longum 0.20 0.27 51.2 0.15

Other strains 0.26 0.001 86.3 <0.001

Overall 0.14 0.009 75.8 <0.001

IL-4

L.casei 0.15 0.24 9.1 0.64

B.lactis 0.14 0.43 0.0 0.33

Other strains 0.17 0.09 0.0 0.77

Overall 0.07 0.36 0.0 0.57

B, biﬁdobacterium; hsCRP high sensitive C-reactive protein; IFN, interferon; IL, interleukin; L, lactobacillus; S, streptococcus; TNF, tumor necrosis factor.

Please cite this article as: Kazemi A et al., Effect of probiotic and synbiotic supplementation on in

ﬂ

(15)

while serum IL-1B (SMD

¼

0.059, p

¼

0.59) and secreted IL-1B from mitogen stimulated PBMC (SMD

¼

0.188, p

¼

0.09) did not change signi

ﬁ

cantly (Sup Fig. 6). Results of subgroup-analysis by the strains of bacteria (in healthy individuals) are presented in

Table 6

and

Sup Fig. 7. IL-1

b was decreased only in one of the subgroups (B. lactis). Meta-regression for age and duration of studies was not signi

ﬁ

cant while metareg for BMI (Coef

¼

0.23, p

¼

0.13, I

²

_residual

¼

80.65%, Adjusted R

² ¼

15.14%) revealed a signi

ﬁ

cant association.

3.3.2. Arthritis

In the pooled analysis of four studies in patients with arthritis (142 participants), no signi

ﬁ

cant effect of probiotic on serum IL-1B (SMD

¼

0.16, 95% CI -0.51 to 0.18, p

¼

0.02).

3.4. IFN-

g

Thirty four studies measured IFN- g (healthy (23), diabetes (1), atopic dermatitis (2), trauma (1), HIV(1) cancer (1), celiac (1), allergy (2), ulcerative colitis (1), CRF (1).

From 23 clinical trials in healthy subjects, two studies were excluded as they measured IFN- a and IFN- b . Seventeen studies had two arms, three had three arms and one had four arms. Eight

studies measured IFN- g in serum, three in PBMC and ten measured IFN- g secretion from mitogen stimulated PBMC.