Maternal probiotic supplementation for prevention of morbidity and mortality in preterm infants

Jacquelyn Grev; Marie Berg; Roger Soll

doi:10.1002/14651858.CD012519.pub2

. 2018 Dec 12;2018(12):CD012519. doi: 10.1002/14651858.CD012519.pub2

Maternal probiotic supplementation for prevention of morbidity and mortality in preterm infants

Jacquelyn Grev ^1,^✉, Marie Berg ², Roger Soll ³

Editor: Cochrane Neonatal Group

PMCID: PMC6516999 PMID: 30548483

Abstract

Background

Inflammation may contribute to preterm birth and to morbidity of preterm infants. Preterm infants are at risk for alterations in the normal protective microbiome. Oral probiotics administered directly to preterm infants have been shown to decrease the risk for severe necrotizing enterocolitis (NEC) as well as the risk of death, but there are safety concerns about administration of probiotics directly to preterm infants. Through decreasing maternal inflammation, probiotics may play a role in preventing preterm birth and/or decreasing the inflammatory milieu surrounding delivery of preterm infants, and may alter the microbiome of the preterm infant when given to mothers during pregnancy. Probiotics given to mothers after birth of preterm infants may effect infant bacterial colonization, which could potentially reduce the incidence of NEC.

Objectives

1. To compare the efficacy of maternal probiotic administration versus placebo or no intervention in mothers during pregnancy for the prevention of preterm birth and the prevention of morbidity and mortality of infants born preterm.

2. To compare the efficacy of maternal probiotic administration versus placebo, no intervention, or neonatal probiotic administration in mothers of preterm infants after birth on the prevention of mortality and preterm infant morbidities such as NEC.

Search methods

We used the standard search strategy of Cochrane Neonatal to search the Cochrane Central Register of Controlled Trials (CENTRAL 2017, Issue 2), MEDLINE via PubMed (1966 to 21 March 2017), Embase (1980 to 21 March 2017), and CINAHL (1982 to 21 March 2017). We also searched clinical trials databases, conference proceedings, and the reference lists of retrieved articles for randomized controlled trials and quasi‐randomized trials.

Selection criteria

We included randomized controlled trials in the review if they administered oral probiotics to pregnant mothers at risk for preterm birth, or to mothers of preterm infants after birth. Quasi‐randomized trials were eligible for inclusion, but none were identified. Studies enrolling pregnant women needed to administer probiotics at < 36 weeks' gestation until the trimester of birth. Probiotics considered were of the genera Lactobacillus, Bifidobacterium or Saccharomyces.

Data collection and analysis

We used the standard methods of the Cochrane Collaboration and Cochrane Neonatal to determine the methodologic quality of studies, and for data collection and analysis.

Main results

We included 12 eligible trials with a total of 1450 mothers and 1204 known infants. Eleven trials administered probiotics to mothers during pregnancy and one trial administered probiotics to mothers after birth of their preterm infants. No studies compared maternal probiotic administration directly with neonatal administration. Included prenatal trials were highly variable in the indication for the trial, the gestational age and duration of administration of probiotics, as well as the dose and formulation of the probiotics. The pregnant women included in these trials were overall at low risk for preterm birth. In a meta‐analysis of trial data, oral probiotic administration to pregnant women did not reduce the incidence of preterm birth < 37 weeks (typical risk ratio (RR) 0.92, 95% confidence interval (CI) 0.32 to 2.67; 4 studies, 518 mothers and 506 infants), < 34 weeks (typical risk difference (RD) 0.00, 95% CI ‐0.02 to 0.02; 2 studies, 287 mothers and infants), the incidence of infant mortality (typical RD 0.00, 95% CI ‐0.02 to 0.02; 2 studies, 309 mothers and 298 infants), or the gestational age at birth (mean difference (MD) 0.15, 95% CI ‐0.33 to 0.63; 2 studies, 209 mothers with 207 infants).

One trial studied administration of probiotics to mothers after preterm birth and included 49 mothers and 58 infants. There were no significant differences in the risk of any NEC (RR 0.44, 95% CI 0.13 to 1.46; 1 study, 58 infants), surgery for NEC (RR 0.15, 95% CI 0.01 to 2.58; 1 study, 58 infants), death (RR 0.66, 95% CI 0.06 to 6.88; 1 study, 58 infants), and death or NEC (RR 0.53, 95% CI 0.19 to 1.49; 1 study, 58 infants). There was an improvement in time to reach 50% enteral feeds in infants whose mothers received probiotics, but the estimate is imprecise (MD ‐9.60 days, 95% CI ‐19.04 to ‐0.16 days; 58 infants). No other improvement in any neonatal outcomes were reported. The estimates were imprecise and do not exclude the possibility of meaningful harms or benefits from maternal probiotic administration. There were no cases of culture‐proven sepsis with the probiotic organism. The GRADE quality of evidence was judged to be low to very low due to inconsistency and imprecision.

Authors' conclusions

There is insufficient evidence to conclude whether there is appreciable benefit or harm to neonates of either oral supplementation of probiotics administered to pregnant women at low risk for preterm birth or oral supplementation of probiotics to mothers of preterm infants after birth. Oral supplementation of probiotics to mothers of preterm infants after birth may decrease time to 50% enteral feeds, however, this estimate is extremely imprecise. More research is needed for post‐natal administration of probiotics to mothers of preterm infants, as well as to pregnant mothers at high risk for preterm birth.

Plain language summary

Maternal probiotic supplementation for prevention of morbidity and mortality in preterm infants

Review question  Do probiotics given to pregnant women prior to delivery and/or probiotics given to mothers of premature babies after birth compared with administration of placebo, no intervention or postnatal administration of probiotics to premature babies themselves reduce the risk of morbidity and mortality in premature babies?

Background  Inflammation may contribute to premature birth and to morbidity of premature babies. Premature babies are at risk for alterations in the growth of normal "good" bacteria in their intestinal tract. Probiotics are supplements of "good bacteria." Giving these bacteria directly to premature babies may decrease inflammation and decrease the risk for severe gastrointestinal disease (necrotizing enterocolitis) as well as the risk of death. However, there are safety concerns about administration of probiotics directly to premature babies.

Study characteristics  Twelve eligible trials with a total of 1450 mothers and 1204 known infants were included from searches, up to date as of March 2017. Eleven trials gave probiotics to mothers during pregnancy and one trial gave probiotics to mothers after the birth of their preterm infants. No studies compared maternal probiotic administration directly versus neonatal administration.The studies of pregnant women focused on various aspects of the studies with different probiotics given at different times. The pregnant women included in these trials were overall at low risk for preterm birth. The one trial with mothers given probiotics after birth was to mothers of infants who weighed less than 1500 g at birth.

Key results  There is insufficient evidence to conclude whether there is appreciable benefit or harm to neonates of either oral supplementation of probiotics administered to pregnant women at low risk for preterm birth or oral supplementation of probiotics to mothers of preterm infants after birth. There were no trials that gave mothers at high risk for having a premature infant probiotics, so the effects of probiotics given to those mothers is unknown. More studies are needed to know if probiotics given to mothers of preterm infants decreases death, necrotizing enterocolitis, or other problems related to prematurity.

Quality of evidence  In general, the evidence was of low to very low quality due to imprecision (small sample sizes) and indirectness (enrolled mothers were not necessarily at high risk for delivering their babies early).

Summary of findings

Background

Description of the condition

The microbiome is defined as the genomes and gene products of microbes that live within and on humans (Johnson 2012). Under normal conditions the microbiome of an infant is established through exposure with bacteria both prenatally and postnatally. The placenta, amniotic fluid and meconium have their own microbiota that influences the microbiome of the infant. Postnatal influences of the microbiome come from exposures mainly from the maternal microbiome which includes exposure to vaginal, oral, fecal, skin and breast milk bacteria, with a typical predominance of protective Lactobacillus and Bifidobacterium in the neonatal period.

Preterm infants are at risk for alterations in the normal protective microbiome due their increased cesarean delivery rate, exposure to antibiotics, nosocomial exposures to pathogens, lack of typical skin contact with maternal flora, as well as alterations in typical exposure to breast milk (Cilieborg 2012). Cesarean delivery alters colonization of neonates so their stool has lower bifidobacteria and more skin flora such as Staphylococcus, Corynebacterium and Propionibacterium (Faa 2013). Maternal intrapartum antibiotics and antibiotics given to mothers with prolonged rupture of membranes are associated with lower transfer of Lactobacillus to the infant, and even antibiotics given days before birth have been shown to alter the premature infant microbiome with less bacterial diversity of the first stool samples (Faa 2013; Keski‐Nisula 2013). Antibiotics given in the neonatal period have also been associated with decreased stool bifidobacteria (Faa 2013). This may lead to delayed colonization with low diversity of bacteria as well as colonization with pathogens and bacterial overgrowth which put infants at risk for necrotizing enterocolitis (NEC), sepsis and death in the neonatal period (Cilieborg 2012).

The development of NEC is typically associated with a dysregulation of inflammation in the gut that results in gas‐producing bacteria translocating through the bowel mucosa and is not usually caused by one single organism. It is a disorder that affects approximately 7% of infants with a birth weight of less than 1500 g, with a death rate of approximately 20% to 30% (Neu 2011). Strategies to prevent NEC have been employed including exclusive human milk diet, standardized feeding protocols, human milk fortifiers and postnatal probiotics, but prematurity and alterations of gut flora still put premature infants at risk.

Modifications of the neonatal microbiome may have long‐term effects on health that may last until adulthood. Altered microbiomes have been associated with the development of atopy (hereditary tendency to be hypersensitive to certain allergens), inflammatory bowel disease, obesity and impaired glucose regulation. It is unknown to what extent early preterm infant alterations in the microbiome may influence their lifelong risk for these diseases.

Description of the intervention

Probiotics are defined as living microorganisms that confer a health benefit to the host (Othman 2007); they are often given enterally to colonize the gastrointestinal tract. Typical probiotic bacteria administered are bifidobacteria, Lactobacillus or Saccharomyces (Dugoua 2009; Thomas 2010). They have been given directly to preterm infants with the intention to decrease the rate of NEC (AlFaleh 2014; Costeloe 2016; Deshpande 2010; Neu 2011), and there is emerging interest in promoting gastrointestinal colonization in preterm infants through administration of probiotics to their mothers instead of directly to infants. Probiotics have already been studied during pregnancy for other indications with the intention to treat genitourinary infections, prevent infant atopy, enhance metabolism and prevent preterm labor (Dugoua 2009; Gomez Arango 2015; Lahtinen 2009; Lindsay 2013; Reid 2003; VandeVusse 2013).

The focus of this investigation is the maternal oral administration of probiotics. Maternal probiotics can be administered to pregnant women at risk for preterm birth or to mothers of preterm infants after birth. Pregnant women administered probiotics in pregnancy will be compared to pregnant women receiving placebo or no intervention. Mothers of preterm infants administered probiotics after birth will be compared to either mothers of preterm infants administered placebo or no intervention, or to preterm neonates administered probiotics directly.

How the intervention might work

The microbiome of the pregnant and postpartum woman is influenced by the established microbiome, diet, probiotic exposures, and pregnancy itself as there are alterations of the microbiome that occur during pregnancy that are likely hormonally mediated. During pregnancy probiotics have been generally regarded as safe (Elias 2011; Gomez Arango 2015; Lindsay 2013), and beneficial for mothers in preventing and treating bacterial vaginitis, and reducing the risk of gestational diabetes, with hypotheses but no definitive evidence of their efficacy in preventing preterm birth (Gomez Arango 2015; Othman 2007). While studies have shown that probiotics taken by adults only change their microbiome while they are being taken, these short‐term alterations in maternal microbiome may have long‐term effects on establishment of the fetal and neonatal microbiome (Matamaros 2013).

Normal microbiome colonization occurs prenatally with bacteria found in the placenta, amniotic fluid and fetal meconium (Matamaros 2013), and postnatal colonization from the mother and environment, including vaginal Lactobacillus obtained at birth, maternal skin flora and bacteria in breast milk (Cilieborg 2012; Faa 2013). Especially predominant and beneficial to the healthy neonatal microbiome are lactobacilli and Bifidobacterium. Lactobacilli are facultative anaerobes that create an environment in which bifidobacteria, anaerobic gram‐positive bacilli, can colonize and predominate in the colon (Bergmann 2014; Dugoua 2009). These bacteria help maintain the mucosa of the intestines, prevent pathogens from colonizing the colon, modulate inflammation along the mucosa and activate the immune system (Hickey 2012), which may help to decrease the risk of NEC and subsequent development of sepsis.

Breast milk consumption greatly contributes to colonization of newborn infants. Studies have shown that there are more than 200 species of bacteria in human milk with a predominance of Lactobacillus and bifidobacteria in addition to Staphylococcus, Streptococcus and corynebacteria (Bergmann 2014). There is evidence that the neonatal gut microbiome reflects the bacteria found in breast milk and the mother’s stool (Bergmann 2014; Cilieborg 2012; Jost 2014). According to the theory of the entero‐mammary pathway, bacteria are deposited into the mammary milk ducts from the maternal gastrointestinal tract via active transport through the blood (Bergmann 2014; Jost 2014). Dendritic cells in the maternal gut lumen are thought to trap bacteria, and with the help of mononuclear cells, transport them through the blood to the breast (Bergmann 2014). This is thought to be hormonally mediated, occurring late in the third trimester, with the most bacteria present in the breast in the peripartum period (Bergmann 2014).

Studies have shown that lactobacilli taken orally by the mother are present in breast milk when taken prenatally and postnatally (Bergmann 2014). Supplementation with Lactobacillus reuteri (L. reuteri) during the third trimester prior to delivery resulted in a statistically significant increase in L reuteri in maternal colostrum (Abrahamsson 2009), and Lactobacillus GG taken by mothers for one month prior to delivery increased the diversity of fecal Bifidobacterium in neonates (Gueimonde 2006).

Why it is important to do this review

Probiotic administration in high‐risk very low birth weight (VLBW) infants has been shown to reduce the incidence of NEC as well as mortality (AlFaleh 2014; Deshpande 2010). Postnatal probiotic administration decreased the risk of developing stage II‐III NEC by more than half (AlFaleh 2014). In their analysis, Deshpande and coworkers suggest that routine use be implemented without the need for more placebo‐controlled trials (Deshpande 2010).

However, the studies included in these meta‐analyses are quite heterogeneous in regards to the specie(s), dose, timing and duration of administration, and there is concern about safety and quality assurance of probiotics which lie outside the purview of the Food and Drug Administration (FDA) regulation. A recent systematic review cautions universal use of probiotics in VLBW infants citing the lack of evidence of efficacy for specific strains (Mihatsch 2012), while others in the neonatal community cite both methodologic and safety concerns as cautions against universal adoption (Garland 2010; Soll 2010). The American Academy of Pediatrics 2010 Clinical Report written by the Committee on Nutrition states that there is insufficient data to recommend probiotics in infants weighing less than 1000 g, and that since not all probiotics have been studied, they cannot all be generally recommended (Thomas 2010). The European Society for Paediatric Gastroenterology, Hepatology and Nutrition Committee on Nutrition counsels caution in introducing potentially infectious agents to VLBW infants due to their immunologic immaturity, and states that there are not enough data to conclude that probiotic administration to preterm neonates is safe (Agostoni 2010). While there are many studies that suggest the safety of probiotics (AlFaleh 2014; Costeloe 2016), there have been cases of sepsis with the strains of probiotics given to preterm infants (Bertelli 2015), as well as a death due to gastrointestinal mucormycosis from Rhizopus oryzae‐contaminated probiotic directly administered to a preterm infant (CDC HAN 2014). With these emerging safety concerns, the FDA has issued a warning against the use of probiotics in immunocompromised patients (CDC 2015; FDA 2014).

These same safety concerns may not be universal where there are regulatory organizations responsible for overseeing the safety of probiotics, such as the Natural Health Products Directorate in Canada (Janvier 2013).

Currently, probiotic use in the VLBW population is not routine in many neonatal intensive care units (NICUs) in the USA. A recent survey of US NICUs that are members of Vermont Oxford Network showed that 14% of these NICUs give VLBW infants probiotics (Viswanathan 2016). Of these 14%, 8.8% of NICUs routinely give all VLBWs probiotics, while another 5.2% of NICUs give probiotics to select VLBW infants (Viswanathan 2016).

With these safety concerns regarding the direct administration of probiotics to preterm infants and evidence showing that probiotics given to preterm infants decreases the rate of NEC and mortality (AlFaleh 2014; Deshpande 2010), there is interest in alternate methods for inducing a normal microbiome in a safe manner. Since maternal flora and breast milk bacteria are known to be transmitted naturally to fetuses and neonates, an alternative method of exposure to probiotics that avoids direct administration to preterm infants, and therefore direct administration of possible contaminants, is maternal probiotic administration to mothers of preterm infants or pregnant women at risk for preterm birth. While not all species of probiotics or timing of administration within pregnancy have been studied thoroughly enough to make definitive statements about safety for all probiotics (Dugoua 2009), probiotics in pregnancy have been generally regarded as safe (Elias 2011; Gomez Arango 2015; Lindsay 2013).

Maternal probiotic administration in pregnancy for prevention of preterm labor and preterm birth has been addressed in one systematic Cochrane review focusing on probiotics specifically as a treatment or prevention of urogenital infections (Othman 2007). Our review will also address probiotic administration to pregnant women at risk of preterm birth, but its focus is on the preterm infant and includes pregnant women at risk for preterm birth for any reason. To our knowledge, this will be the first systematic review of maternal probiotic administration with a focus on mothers of preterm infants given probiotics after birth, the first to address the comparison of probiotics administered to mothers of preterm infants versus administration to the preterm infants themselves, and the first to address prenatal and postnatal probiotic administration to mothers at risk for preterm birth.

Objectives

To determine whether maternal probiotic administration to pregnant women and/or probiotic administration to mothers after preterm birth compared to administration of placebo, no intervention or postnatal administration to preterm infants reduces the risk of morbidity and mortality in preterm infants.

Comparisons

Probiotics administered to pregnant women (< 36 weeks' gestation) versus placebo or no intervention:
1. In pregnant women (< 36 weeks' gestation), maternal probiotics only prior to birth versus maternal placebo or no intervention prior to birth;
2. In pregnant women (< 36 weeks' gestation), maternal probiotics both prior to and after birth versus maternal placebo or no intervention prior to and after birth.
Probiotics administered exclusively after birth in mothers of preterm infants (< 37 weeks' gestation) versus maternal placebo or no intervention.
Probiotics administered exclusively after birth in mothers of preterm infants (< 37 weeks' gestation) versus neonatal probiotic administration.

Methods

Criteria for considering studies for this review

Types of studies

We included randomized controlled trials. Quasi‐randomized controlled trials, and cluster trials were eligible for inclusion but none were identified. Cross‐over trials were not eligible for inclusion.

Types of participants

Comparison 1: probiotics administered to pregnant women versus placebo or no intervention

Pregnant women at risk for preterm birth (birth at < 36 weeks' gestation).

Comparison 2: probiotics administered exclusively after birth in mothers of preterm infants versus maternal placebo or no intervention

Postpartum mothers who have given birth to a preterm infant born < 37 weeks' gestation.

Comparison 3: probiotics administered exclusively after birth in mothers of preterm infants versus neonatal probiotic administration

Postpartum mothers who have given birth to a preterm infant born < 37 weeks' gestation.
Newborn infants born < 37 weeks' gestation.

Types of interventions

Included probiotics included Lactobacillus species, Bifidobacterium species, or Saccharomyces, and mixed preparations of these probiotics at any dose with the intention to treat for a minimum of seven days.

For antenatal probiotic administration, probiotics were taken at some point during the trimester in which the mother gives birth, or within one week of birth if born early in the third trimester.

Types of outcome measures

Data on the mothers and the preterm infants of these mothers was collected.

Primary outcomes

Preterm birth (1. any preterm birth < 37 weeks' gestation and 2. preterm birth < 34 weeks' gestation) (limited to infants in Comparison 1).
Any NEC (Bell stage I to III) (Bell 1978).*
Severe NEC (Bell stage II or more) (Bell 1978).
Death (before discharge).
Death or severe NEC.
Death or any NEC.*

Secondary outcomes

Gestational age at birth (limited to infants in Comparison 1).*
Early onset bacterial sepsis, defined by positive blood culture on 'day of life 3' or earlier (limited to infants in Comparison 1).
Late‐onset sepsis, defined by positive blood culture on 'day of life 3' or later.
Culture‐proven sepsis with supplemented probiotic(s) before hospital discharge.
Neonatal sepsis (at any time)*
Surgical NEC (Bell Stage III) (Bell 1978).*

For infants born < 37 weeks

Duration of hospital stay (days).
Duration of parenteral nutrition (days).
Days to full enteral feeds.
Days to 50% enteral feeds*
Growth (g/kg/day) during hospitalization (prior to discharge).
Growth Z score at 36 to 40 weeks' postmenstrual age (PMA).
Retinopathy of prematurity (ROP) (any stage, severe stage 3 or greater) (ICCROP 2005).
Intraventricular hemorrhage (IVH) (any grade and severe (Grade III‐IV)) (Papile 1978).
Cystic periventricular leukomalacia (PVL) (diagnosed by cranial imaging).
Patent ductus arteriosus (PDA) (treated either medically or surgically).
Bronchopulmonary dysplasia (BPD) assessed at 28 days and at 36 weeks' PMA.
Long‐term major neurodevelopmental disability assessed at 18 to 24 months in survivors (cerebral palsy (CP)), developmental delay (Bayley or Griffith assessment more than two standard deviations (SD) below the mean) or intellectual impairment (intelligence quotient (IQ) more than two SD below mean), blindness (vision < 6/60 in both eyes), sensorineural deafness requiring amplification) (Jacobs 2013).

Maternal secondary outcomes

Maternal chorioamnionitis: suspected or confirmed intrauterine inflammation or infection or both ("Triple I") based upon the criteria from the 2016 chorioamnionitis workshop Higgins 2016). Suspected Triple I is defined by a fever without a clear source plus baseline fetal tachycardia, maternal white blood count greater than 15,000/mm³ in the absence of corticosteroids or definite purulent fluid from the cervical os. Confirmed Triple I is defined as meeting criteria for suspected Triple I plus amniocentesis‐proven infection through a positive Gram stain, low glucose or positive amniotic fluid culture or placental pathology revealing diagnostic features of infection. For comparison 1 only.
Maternal endometritis: clinical diagnosis by obstetric provider consisting of an oral temperature of ≥ 38.0 °C on any two of the first 10 postpartum days or a temperature of ≥ 38.7 °C during the first 24 hours postpartum based upon the American Committee of Maternal Welfare's standard definition) (Mackeen 2015).
Maternal mastitis diagnosed clinically by obstetrics provider at any time during the study.
Maternal sepsis as defined by positive maternal blood culture and assessed at any time during the study.
Maternal Group B Streptococcus colonization (based upon most recent vaginal/rectal swab results prior to birth).

* outcome added post‐hoc after review of available study outcomes

Search methods for identification of studies

We used the criteria and standard methods of the Cochrane Neonatal (see the Cochrane Neonatal Group search strategy for specialized register).

Electronic searches

We conducted a comprehensive search including: Cochrane Central Register of Controlled Trials (CENTRAL 2017, Issue 2) in the Cochrane Library; MEDLINE via PubMed (1966 to 21 March 2017); Embase (1980 to 21 March 2017); and CINAHL (1982 to 21 March 2017) using the following search terms: (probiotic OR lactobacillus OR bifidobacter* OR saccharomyces), plus database‐specific limiters for RCTs and neonates (see Appendix 1 for the full search strategies for each database). We did not apply language restrictions.    We searched clinical trials registries for ongoing or recently completed trials (clinicaltrials.gov; the World Health Organization’s International Trials Registry and Platform www.whoint/ictrp/search/en/, and the ISRCTN Registry).

We searched the Cochrane Pregnancy and Childbirth Group’s Trials Register by contacting the Information Scientist (see Appendix 2).

Searching other resources

We searched the abstracts of the Society for Pediatric Research (US) (published in Pediatric Research) for the years 1985 to 1999 using the following key words: (probiotic OR lactobacillus OR bifidobacteria OR saccharomyces) AND (neonates OR infants).

We also searched for conference abstracts from Pediatric Academic Societies (PAS) and European Society for Paediatric Research (ESPR). We carried out searches in Abstracts to View (2000 to present) (www.abstracts2view.com/pasall/) and Pediatric Research as well as the Cochrane Pregnancy and Childbirth Group's Trials Register.

Data collection and analysis

We collected information regarding the method of randomization, blinding, drug intervention, stratification, and whether the trial was single or multicenter for each included study. We noted the information regarding trial participants including gestational age criteria, birth weight criteria, and other inclusion or exclusion criteria. We analyzed the information on clinical outcomes including death, severe NEC (Bell stage II or more), early onset sepsis (if probiotics were administered before delivery), late onset sepsis, prematurity (gestational age) (if probiotics were administered prior to delivery), culture‐proven sepsis with supplemented probiotic(s), duration of hospital stay (days), duration of parenteral nutrition (days), days to full enteral feeds, growth (g/kg/day), ROP (any, severe), IVH (Grade III‐IV), cystic PVL and PDA (treated either medically or surgically), chronic lung disease (CLD), and long‐term major neurodevelopmental disability. Maternal data collected included maternal chorioamnionitis, endometritis, mastitis and sepsis. We assessed the quality of evidence at the outcome level using the Grading of Recommendations Assessment Development and Evaluation (GRADE) approach, as outlined in the GRADE Handbook (Schünemann 2013).

Selection of studies

We assessed the titles and abstracts resulting from the electronic searches. The full copy of all relevant or potentially relevant trials was obtained and assessed according to the 'Criteria for considering studies for this review'.

We included randomized controlled trials fulfilling the selection criteria described in the previous section. We planned to include cluster‐randomized and quasi‐randomized controlled trials but none were identified. Cross‐over studies were not eligible for inclusion. Both superiority trials and non‐inferiority trials were eligible for inclusion. All review authors reviewed the results of the search and separately selected the studies for inclusion. Disagreements about whether a trial should be included were resolved by discussion and consensus was reached. In cases where additional information was needed before a decision could be made as to whether to include a trial, we planned to obtain this information from the study investigator.

Data extraction and management

Two review authors extracted, assessed, and coded all data for each study, using a form designed specifically for this review. Any standard error (SE) of the mean was replaced by the corresponding standard deviation (SD). We resolved any disagreement by discussion. For each study, final data were entered into Review Manager 5 (RevMan) by one review author (JG) and then checked by the other review author (RS) (Review Manager 2014). If needed, we planned to contact authors of trials to obtain missing data.

Assessment of risk of bias in included studies

Two review authors (JG, RS) independently assessed the risk of bias (low, high, or unclear) of all included trials using the Cochrane ‘Risk of bias’ tool (Higgins 2011) for the following domains.

Sequence generation (selection bias)
Allocation concealment (selection bias)
Blinding of participants and personnel (performance bias)
Blinding of outcome assessment (detection bias)
Incomplete outcome data (attrition bias)
Selective reporting (reporting bias)
Any other bias

Any disagreements were resolved by discussion or by a third assessor. See Appendix 3 for a more detailed description of risk of bias for each domain.

Measures of treatment effect

We performed the statistical analyses using Review Manager 5 software (Review Manager 2014). We analyzed categorical data using risk ratio (RR), and risk difference (RD). For statistically significant outcomes, we calculated the number needed to treat for an additional participant with a beneficial outcome (NNTB) or number needed to treat for an additional participant with a harmful outcome (NNTH). We analyzed continuous data using mean difference (MD) and the standardized mean difference (SMD). We reported the 95% confidence interval (CI) on all estimates.

Unit of analysis issues

Had we included any cluster‐randomized trials, we planned to use the intracluster correlation coefficient (ICC) to calculate effective sample sizes in an approximate analysis (Cochrane Handbook for Systematic Reviews of Interventions, 15.3.4). Once trials had been changed to their effective sample size, we planned to enter the data into RevMan in a manner similar to how data from randomized control trials are entered.

Dealing with missing data

The review was analyzed using an intention‐to‐treat paradigm. For studies with missing data, we attempted to contact the authors to obtain this data.

Assessment of heterogeneity

We estimated the treatment effects of individual trials and examined heterogeneity among trials by inspecting the forest plots and quantifying the impact of heterogeneity using the I² statistic. We graded the degree of heterogeneity as: less than 25% — no heterogeneity; 25% to 49% — low heterogeneity; 50% to 75% — moderate heterogeneity; more than 75% — substantial heterogeneity. If statistical heterogeneity (an I² > 50%) was noted, the possible causes were explored (for example, differences in study quality, participants, intervention regimens, or outcome assessments).

Assessment of reporting biases

We planned to create a funnel plot if there were at least 10 studies meeting our inclusion criteria.

Data synthesis

If multiple studies were identified and thought to be sufficiently similar in participant population, intervention and outcomes, meta‐analysis was carried out using Review Manager 2014. For categorical outcomes, we calculated the typical estimates of RR and RD, each with its 95% CI; and for continuous outcomes the MD or a summary estimate for the MD, each with its 95% CI, was calculated. We used a fixed‐effect model for meta‐analysis. If meta‐analysis was judged to be inappropriate, we analyzed individual trials and interpreted them separately.

Quality of the evidence

We used the GRADE approach, as outlined in the GRADE Handbook (Schünemann 2013), to assess the quality of evidence for the following (clinically relevant) outcomes: preterm birth (any preterm birth (< 37 weeks) and preterm birth < 34 weeks) (limited to infants in Comparison 1); severe NEC (Bell stage II or more) (Bell 1978); death (mortality before discharge); death or severe NEC; culture‐proven sepsis with supplemented probiotic(s) before hospital discharge; late‐onset sepsis, defined by positive blood culture on 'day of life 3' or later.

We added several clinically important outcomes to the 'Summary of findings' tables based on review of available studies post hoc: for Comparison 1: gestational age at birth (weeks); miscarriage/stillbirth; for Comparison 2: any NEC, surgery for NEC, feeding with more than 50% breast milk (days).

Two review authors independently assessed the quality of the evidence for each of the outcomes listed above. We considered evidence from randomized controlled trials as high quality but downgraded the evidence one level for serious (or two levels for very serious) limitations based upon the following: design (risk of bias), consistency across studies, directness of the evidence, precision of estimates and presence of publication bias. We used the GRADEpro GDT GDT Guideline Development Tool to create ‘Summary of findings’ tables to report the quality of the evidence.

The GRADE approach results in an assessment of the quality of a body of evidence in one of four grades.

High: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Subgroup analysis and investigation of heterogeneity

We examined the effect of differences in the definitions and measurement of outcomes, such as diagnosis of NEC and infections and differences in length of follow‐up using methods for addressing clinical heterogeneity: subgroup analysis and meta‐regression. If any other significant heterogeneity was detected between studies, we planned to carry out subgroup analysis and avoid pooling of results.

Subgroups for analysis

The following pre‐planned subgroup analyses were performed.

Probiotic type: Lactobacillus sp ecies, Bifidobacterium species, Saccharomyces, or mixed preparations of these probiotics
Economic country setting (low‐income, lower‐middle income, upper‐middle income and high‐income economies as defined by the World Bank (The World Bank 2016)

In addition, a post hoc subgroup analysis was performed based on maternal gestational age at enrollment.

There were insufficient data to perform subgroup analyses based upon birth weight, race, ethnicity, sex, provision of breast milk, and maternal risk for preterm birth.

Sensitivity analysis

Excluding studies with high risk of bias (performed).
Excluding unpublished studies (planned but not performed due to lack of data).

Results

Description of studies

See the Characteristics of included studies.

See the Figure 1 study flow diagram.

Results of the search

Twelve studies were found that met the inclusion criteria; 11 for the comparison of probiotics administered to pregnant women (< 36 weeks' gestation) versus placebo or no intervention, (Comparison 1) (Badehnoosh 2018; Dolatkhah 2015; Fernández 2016; Jacobsson 2016; Kopp 2008; Laitinen 2009; Lindsay 2014; Lindsay 2015; Mantaring 2016; Ou 2012; Rautava 2012), and one for the comparison of probiotics administered exclusively after birth in mothers of preterm infants < 37 weeks' gestation versus maternal placebo or no intervention, (Comparison 2) (Benor 2014). The studies in Comparison 1 and 2 combined included 1450 mothers and 1204 known infants (as not all studies reported the number of infants born). No studies were found meeting the criteria for the comparison of probiotics administered exclusively after birth in mothers of preterm infants < 37 weeks' gestation versus neonatal probiotic administration (Comparison 3).

Nine studies reported on at least one outcome of interest and included 1199 mothers and 1048 known infants (as not all studies reported the number of infants born) (Badehnoosh 2018; Benor 2014; Fernández 2016; Kopp 2008; Laitinen 2009; Lindsay 2014; Lindsay 2015; Ou 2012; Rautava 2012), while three studies (Dolatkhah 2015; Jacobsson 2016; Mantaring 2016) with 260 mothers with 156 known infants did not report on outcomes of interest. Three studies conducted in pregnant women with atopy or with a fetus at risk for atopy that included 524 mothers and 484 infants (Kopp 2008; Ou 2012; Rautava 2012), only had gestational age at birth as an outcome of interest, and did not report the data in a format suitable for meta‐analysis. The remaining six studies reported on at least one outcome of interest for a total of 675 mothers and 564 infants. Five of these studies were in Comparison 1 and included 626 mothers and 506 infants; four administered the intervention only to mothers during pregnancy (Badehnoosh 2018; Fernández 2016; Lindsay 2014; Lindsay 2015), and one administered the intervention during pregnancy and after birth (Laitinen 2009). One study met the criteria for Comparison 2, administering the intervention to mothers of preterm infants after birth, for a total of 49 mothers with 58 infants (Benor 2014). See the 'Rrisk of bias' tables for data on risk of bias (Figure 2; Figure 3).

'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.

Included studies

Comparison 1: Probiotics administered to pregnant women (< 36 weeks' gestation) versus placebo or no intervention

Before delivery (probiotics given only prior to delivery)

Summary: six studies were identified (Badehnoosh 2018; Dolatkhah 2015; Fernández 2016; Jacobsson 2016; Lindsay 2014; Lindsay 2015).

Participants: there were six studies that administered probiotics to women only during pregnancy: three to pregnant women with gestational diabetes (Badehnoosh 2018;Dolatkhah 2015; Lindsay 2015), one to obese women at risk for metabolic derangements (Lindsay 2014), one to low‐risk women to assess immunological or inflammatory responses (Jacobsson 2016), and one to pregnant women with a history of mastitis during lactation (Fernández 2016).

Interventions: two studies used combination probiotics with various species of lLctobacillus, Bifidobacterium and/or Streptococcus (Badehnoosh 2018; Dolatkhah 2015), while four studies used single probiotic agents, three administering Lactobacillus salivarius (L. salivarius) (Fernández 2016; Lindsay 2014; Lindsay 2015) and one using Lactobacillus rhamnosus (L. rhamnosus) (Jacobsson 2016). There was a degree of heterogeneity in terms of the timing and duration of the intervention, ranging from eight to 10 weeks until delivery (Jacobsson 2016), starting at 24 to 28 weeks and lasting four to eight weeks (Badehnoosh 2018; Dolatkhah 2015; Lindsay 2014), or starting at 30 weeks or < 34 weeks until delivery (Fernández 2016; Lindsay 2015). All control group participants received a placebo in these studies.

Outcomes of interest reported upon were preterm birth < 37 weeks (Badehnoosh 2018; Lindsay 2014; Lindsay 2015), preterm birth < 34 weeks (Lindsay 2014), gestational age at birth (Badehnoosh 2018; Lindsay 2014; Lindsay 2015), neonatal mortality (Lindsay 2014), miscarriage/stillbirth (Lindsay 2015), and mastitis (Fernández 2016).

Review of individual studies (based on study aims)

Treatment of maternal gestational diabetes

Badehnoosh 2018 conducted a randomized, double‐blind, placebo‐controlled clinical trial at the Akbarabadi Clinic in Tehran, Iran (April 2016 to September 2016), evaluating the effects of probiotic supplementation on biomarkers of inflammation, oxidative stress and pregnancy outcomes among participants with gestational diabetes (GDM). Sixty pregnant women diagnosed between 24 to 28 weeks with GDM who were not on oral hypoglycemic agents were randomly allocated to intake either probiotic capsule containing Lactobacillus acidophilus (L.acidophilus), Lactobacillus casei (L. casei) and Bifidobacterium bifidum (B. bifidum) ( 2 x 10^9 colony‐forming units (CFU)/g each) (n = 30) or placebo (n = 30) for six weeks. Primary outcomes were inflammatory markers, while secondary outcomes were biomarkers of oxidative stress and pregnancy outcomes, including preterm birth, gestational age at birth, Apgar scores, hospitalization, hypoglycemia, macrosomia, and cesarean delivery rates.

Dolatkhah 2015 conducted a double‐blind placebo‐controlled randomized clinical trial at Alzahra University Hospital in Tabriz, Iran (Spring and Summer 2014), to assess the effect of a probiotic supplementation on glucose metabolism and weight gain in pregnant women with newly diagnosed GDM. Sixty‐four pregnant women diagnosed between 24 to 28 weeks with GDM were randomly assigned to receive either aL.acidophilus LA‐5, Bifidobacterium BB‐12, Streptococcus thermophilus (S. thermophilus) STY‐31 and Lactobacillus delbrueckii bulgaricus (L. delbrueckii bulgaricus) LBY‐27 (4 biocap > 4 × 10^9 CFU) probiotic (n = 32) or placebo (n = 32) capsule along with dietary advice for eight consecutive weeks. Primary outcomes assessed were trend in weight gain and glucose metabolism indices in the mother. No outcomes of interest for this systematic review were reported.

Lindsay 2015 conducted a double‐blind placebo‐controlled randomized trial at the National Maternity Hospital in Dublin, Ireland (March 2012 to May 2014) to investigate the effects of a probiotic capsule intervention on maternal metabolic parameters and pregnancy outcome among pregnant women with gestational diabetes. One‐hundred forty‐nine singleton pregnant women with a new diagnosis of GDM in this pregnancy and who were < 34 weeks' gestation were randomly assigned to receive a capsule of probiotic (n = 74 ) with 100 mg of L. salivarius UCC118 at a target dose of 10^9 CFUs, or placebo (n = 75) until delivery. The primary outcome was the maternal fasting glucose and secondary outcomes were need for pharmacologic treatment of GDM, neonatal birth weight, stillbirths or miscarriages, gestational age at birth, preterm birth, cord blood metabolic parameters, Apgar score, small for gestational age (SGA) or large for gestational age (LGA) birth weight percentiles, and need for admission to the neonatal intensive care unit (NICU).

Treatment of maternal obesity

Lindsay 2014 (Probiotics in Pregnancy Study (ProP Study) conducted a placebo‐controlled, double‐blind, randomized trial at the National Maternity Hospital, Dublin, Ireland (March 2012 to March 2013) that investigated the effect of a probiotic capsule on maternal fasting glucose in obese pregnant women. One‐hundred seventy‐five pregnant obese women were assigned to probiotic administration with 100 mg L. salivarius UCC118 at 10^9 CFU (n = 83) or placebo capsule (n = 82) from 24 to 28 weeks' gestation. The primary outcome was the change in maternal fasting glucose and secondary outcomes included differences in the incidence of impaired glycemic control, neonatal anthropometric measures, delivery complications, cord blood metabolic variables, 5‐minute Apgar score, preterm birth rate < 37 weeks or < 34 weeks, gestational age at birth, perinatal mortality and admission to the NICU.

Inflammation/lipids

Jacobsson 2016 conducted a randomized, placebo‐controlled trial to assess the effects of supplementation with L. rhamnosus (LGG) in low‐risk pregnant women on maternal immunological or inflammatory responses. Forty low‐risk pregnant women at eight to 10 weeks of pregnancy were randomized to receive L. rhamnosus (10^8 CFU) (n = 20) or placebo (n = 20) until delivery. Main outcomes assessed were inflammatory and immunologic assays at recruitment, 25 weeks and 35 weeks of gestation. No outcomes of interest for this systematic review were reported.

Prevention of mastitis

Fernández 2016 conducted a double‐blind, randomized controlled trial at the Hospital Clínico San Carlos in Madrid, Spain that investigated the effect of an oral probiotic given to pregnant women to prevent infectious mastitis. One‐hundred‐eight pregnant women with a history of infectious mastitis after a prior pregnancy were assigned to probiotic capsule administration with L. salivarius PS2 at 9 x 10^10 CFU (n = 55) or placebo capsule (n = 53) from approximately week 30 of pregnancy until delivery. The primary outcome was the diagnosis of mastitis in the first three months after giving birth. Secondary outcomes included acute mastitis, subacute mastitis, breast pain scores, and breast milk bacterial counts.

Before and after delivery (probiotics given both prior to and after delivery)

Summary: five studies were identified that administered probiotics to women both prior to and after delivery (Kopp 2008, Laitinen 2009; Mantaring 2016;Ou 2012, Rautava 2012).

Participants: of the five studies, three studies enrolled pregnant women with a fetus at risk for atopy (Kopp 2008; Ou 2012; Rautava 2012), while two studies focused on healthy women with a focus on nutrition (Laitinen 2009; Mantaring 2016).

Interventions: two studies used a single agent probiotic of Lactobacillus GG (Kopp 2008; Ou 2012), while three used combination probiotics, one using L. rhamnosus and Bifidobacterium longum (B. longum) versus Lactobacillus paracasei (L. paracasei) and B. longum (Rautava 2012), and the other two using L. rhamnosus and Bifidobacterium lactis (B. lactis) (Laitinen 2009; Mantaring 2016). The beginning gestational age at first administration and the duration varied between studies from 14 weeks until end of exclusive breastfeeding (Laitinen 2009), four to six weeks before delivery until six months (Kopp 2008), 24 weeks until six months (Ou 2012), two months before delivery until two months after delivery (Rautava 2012), and the beginning of the third trimester until two months postpartum (Mantaring 2016). The studies that focused on atopic risk all had placebo capsules as the control (Kopp 2008, Ou 2012, Rautava 2012), while the other two studies had either nutritional counseling (Laitinen 2009), or nutritional beverage supplement (Mantaring 2016) that were administered in both the probiotic and control group.

Outcomes of interest reported upon were preterm birth < 37 weeks (Laitinen 2009), gestational age at birth (Kopp 2008; Ou 2012; Rautava 2012), although these studies did not report average and standard deviation (SD) and thus cannot be included quantitative analysis, neonatal mortality (Laitinen 2009), miscarriage/stillbirth (Laitinen 2009), and neonatal and maternal sepsis (Laitinen 2009).

Review of individual studies (based on study aims)

Multiple aims

Laitinen 2009 conducted a randomized controlled trial in maternal welfare clinics in Turku and South‐West Finland (April 2002 to November 2005) that investigated the impact of dietary counseling and/or probiotics on pregnant women with an aim to optimize maternal dietary intake and metabolism, to advance maternal health and reduce the risk of disease in the child. After final recruitment, 256 healthy pregnant women were recruited, 85 were in the control + placebo group, and 171 were in the nutritional intervention group. These 171 pregnant women were further randomized in a double‐blind fashion to receive probiotics (L. rhamnosus GG, and B. lactis Bb12, 10^9 CFU/day) and the nutritional intervention (n = 85), or placebo and the nutritional intervention (n = 86) from enrollment to the study around 14 weeks' gestation until the end of exclusive breastfeeding. Fourteen articles have been published from the original trial (Characteristics of included studies), with one only including a subset of enrolled patients as it was published before final recruitment. Primary and secondary outcomes vary between the articles, and include outcomes focused on maternal dietary compliance, adiposity, pregnancy weight gain, anthropometric measurements during and after pregnancy, maternal diagnosis of gestational diabetes, maternal laboratory indices for lipids, leptin, and glucose metabolism, as well as placental fatty acids, breast milk fatty acids and inflammatory markers, maternal sepsis and infant sensitization. Outcomes pertinent to the neonate include birth weight, birth length, birth head circumference, gestational age at birth, miscarriage rate, preterm birth < 37 weeks, preterm birth < 32 weeks, preterm birth 32 to 36 weeks, cesarean delivery rate, perinatal mortality, infant sepsis, and 5‐minute Apgar score. There are also reports of illness in mother and illness in child without further description. For this review, the two arms that received the nutritional intervention and either probiotic or placebo were included and compared.

Atopy

Kopp 2008 conducted a double‐blind, placebo‐controlled prospective trial at outpatient gynecology offices co‐ordinated with the University of Freiburg in Freiburg, Germany (July 2002 to June 2004), to investigate the immunologic proliferative response and cytokine release in cultures of isolated mononuclear cells from pregnant women and their neonates supplemented withLactobacillus GG (LGG) or placebo. Ninety‐four pregnant women with at least one first‐degree relative or a partner with an atopic disease were randomly assigned to receive either the probiotic LGG (ATCC 53103; 5 x 10^9 CFUs LGG twice daily) or placebo four to six weeks before expected delivery. This was followed by a post‐natal period of six months in which breastfeeding mothers took the randomized interventions for three months, but formula‐fed infants (n = 2 in the LGG group, n = 3 in the placebo group) were given the agents directly. After three months, the capsules were given only directly to infants until six months. Primary outcomes were proliferative response of cord blood mononuclear cells and peripheral blood mononuclear cells from the corresponding mother. Secondary outcomes included gestational age at birth. Only outcomes pertaining to the prenatal period of probiotic administration were included since formula‐fed infants were directly given the probiotic or placebo.

Ou 2012 conducted a prospective, double‐blind, placebo‐controlled clinical trial at Kaohsiung Chang Gung Memorial Hospital in Taiwan (August 2002 and January 2006), to evaluate the effectiveness of prenatal and postnatal probiotics in the prevention of early childhood and maternal allergic diseases. One‐hundred ninety‐one pregnant women with atopic diseases were randomly assigned to receive either probiotics (Lactobacillus GG; 1x10^ 10 CFUs daily) (n = 95) or placebo (n= 96) from 24 weeks of pregnancy until delivery. After delivery, the interventions were given exclusively to breastfeeding mothers or to non‐breastfeeding neonates until six months of age. Outcomes assessed were the point and cumulative prevalence of sensitization and development of allergic diseases, and improvement of maternal allergic symptom score and plasma immune parameters before and after intervention. Mother‐infant pairs were followed until 36 months. Secondary outcomes include gestational age at birth. Only outcomes pertaining to the prenatal period of probiotic administration were included in this review since formula‐fed infants were directly given probiotics or placebo.

Rautava 2012 conducted a parallel, double‐blind placebo‐controlled trial at a single tertiary center in Turku, Finland (August 2005 and April 2009), to investigate whether maternal probiotic supplementation during pregnancy and breast‐feeding reduces the risk of developing eczema in high‐risk infants. Two‐hundred forty‐one mothers with allergic disease and atopic sensitization were randomized to receive (1) L. rhamnosus 1 x 10^9 CFIU and B. longum 1 x 10^9 CFUs (n = 81), (2) L paracasei 1 x1 0^9 CFUs and B longum 1 x1 0^9 CFUs (n = 82), or (3) placebo (n = 78), beginning two months before delivery and during the first two months of breast‐feeding. The primary outcome was cumulative incidence of eczema in the infant up to two years old. Secondary outcomes were atopic sensitization, gestational age at birth, birth weight, and duration of exclusive breast‐feeding.

Nutrition

Mantaring 2016 conducted a double‐blind, randomized controlled study at the Community Hospital of Muntinlupa City, Philippines to assess the effects of maternal nutritional supplement beverages formulated both with and without a probiotic mixture on maternal health, fetal/infant growth and health when given to pregnant women in the third trimester and during lactation. Two‐hundred thirty‐three healthy women in the beginning of the third trimester of pregnancy who were willing to exclusively breast feed for at least two months were assigned to receive a daily nutritional supplement (S, n = 78); the same supplement with probiotics L. rhamnosus (CGMCC 1.3724) 7 x 10^8 CFU and B. lactis (CNCC I‐3446) 7 x 10^8 CFU per serving (S pro, n = 78), or no supplement (no‐S, n = 77) for the duration of their pregnancy and eight weeks' postpartum. Outcomes included maternal health outcomes at gestational months six, seven, eight, and delivery, as well as neonatal health outcomes, including fetal ultrasound at 24 to 28 weeks, birth weight, Apgar scores, and infant growth over the first year of life. No outcomes of interest for this systematic review were reported.

Comparison 2: probiotics administered exclusively after birth in mothers of preterm infants < 37 weeks' gestation versus maternal placebo or no intervention

Summary: there was one study that administered probiotics to postpartum mothers who have given birth to a preterm infant born < 37 weeks' gestation identified (Benor 2014).

Participants: mothers of infants born with birth weight of 1500 g or less who intended to breast feed and were enrolled within 48 hours of birth, and their infants were included in Benor 2014.

Interventions:Benor 2014 administered L. acidophilus and B. lactis. Placebo capsules were used in the control group.

Outcomes of interest:Benor 2014 reported on the following outcomes of interest: any necrotizing enterocolitis (NEC), death prior to hospital discharge, death or NEC, surgery for NEC, bacterial sepsis, culture‐proven sepsis with probiotic organism, bronchopulmonary dysplasia (BPD), patent ductus arteriosus (PDA), intraventricular hemorrhage (IVH), periventricular leukomalacia (PVL), retinopathy of prematurity (ROP), days to feeding with more than 50% breast milk.

Benor 2014 conducted a prospective, randomized, double‐blind, placebo‐controlled trial at the Tel Aviv Medical Center in Israel (June 2007 to November 2009) which examined the effects of maternal oral probiotic supplementation on the incidence of NEC, death, and sepsis in very low birth weight (VLBW) infants fed with maternal breast milk. Mothers were assigned to supplementation with L. acidophilus and B. lactis 2 × 10(E) CFU/day or to placebo starting from one to three days postpartum. The primary outcome measures were NEC, sepsis, and death. A total 49 mothers of 58 VLBW infants were recruited (25 infants were in the probiotic group and 33 in the placebo group).

Comparison 3: probiotics administered exclusively after birth in mothers of preterm infants < 37 weeks' gestation versus neonatal probiotic administration

Postpartum mothers who have given birth to a preterm infant born < 37 weeks' gestation; and
Newborn infants born < 37 weeks' gestation.

Summary: no studies identified.

Excluded studies

We excluded 17 studies from this review. Four studies were excluded due to timing of administration of probiotics that did not align with the criteria that the intervention be given in the trimester of birth, or intention to administer within one week of birth if stopped in the second trimester. Three studies were excluded due to probiotics administered only until the second trimester (Jain 2017; Jamilian 2016; Krauss‐Silva 2011), and one was excluded due to post‐natal administration of probiotics to mothers of term infants (Ortiz‐Andrellucchi 2008). Two studies administered the intervention vaginally, not orally, and were excluded on the grounds of incorrect route of administration (Daskalakis 2014; Karampelas 2013). As a placebo or no intervention was a requirement for the control group, four studies were excluded due to inappropriate comparator, with three studies using either yogurt or fermented product which presumably contain probiotics in both arms (Asemi 2011; Asemi 2012; Nishijima 2005), and one study was excluded due to a comparison of probiotic treatment versus antibiotic treatment (Hantoushzadeh 2012). Bisanz 2014 was excluded due to a co‐intervention only given to the probiotic group and not the control group. Gonai 2014 was excluded due to unclear gestational age at study entry, and Gronlund 2011; Grzeskowiak 2012; Hanson 2014; and Vitali 2012 were excluded due to issues with study design, including non‐random assignment and non‐random assessment of a larger randomized trial. Wickens 2008 was excluded due to multiple factors including recruitment at 35 weeks, mothers who gave birth prior to 37 weeks or had infants with a NICU admission were explicitly excluded, and probiotics were administered directly to the infant after birth starting on day two. See the table of Characteristics of excluded studies for more details about individual studies.

Post hoc the decision was made to exclude studies that administered probiotics starting at 36 weeks' gestation to pregnant women, as the majority of infants of these mothers would be born at term, especially with the inclusion criteria that the probiotics were intended to be administered for at least one week. The population of interest in this review is preterm infants, and the outcomes of interest relate to prematurity. Including these studies would not fit with the overall goal of this review. There is a large body of literature that fits this criteria and listing those studies is beyond the scope of this review.

Risk of bias in included studies

Allocation

Sequence generation: four studies did not describe how participants were randomized into groups (Dolatkhah 2015; Jacobsson 2016; Mantaring 2016; Ou 2012) and therefore have an unclear risk of selection bias in this domain. Eight studies used computer‐generated random numbers (Badehnoosh 2018; Benor 2014; Fernández 2016; Kopp 2008; Laitinen 2009; Lindsay 2014; Lindsay 2015; Rautava 2012,) with three of these studies having this sequence generation done by independent researchers or statisticians removed from the rest of the study (Laitinen 2009; Lindsay 2014; Lindsay 2015). Overall, there is a low risk of allocation bias due to issues with sequence generation, as those studies with unclear risk of bias did not contribute data to the outcomes included.

Allocation concealment: five studies did not describe their allocation concealment and were deemed to have an unclear risk (Dolatkhah 2015; Jacobsson 2016; Mantaring 2016; Ou 2012; Rautava 2012), while two studies did not specifically describe their allocation concealment but were deemed to be low risk due to the use of placebo capsules that resembled the probiotic capsules (Fernández 2016; Kopp 2008). In one study, allocation was said to be concealed from the researcher and participants until final analyses but not fully described (Badehnoosh 2018), and in another study sealed envelopes were opened in the room with the patient (Laitinen 2009). Sequentially‐numbered, sealed, opaque envelopes were used in two studies (Lindsay 2014; Lindsay 2015), and opaque medication boxes with identical capsules was used in one study (Benor 2014) as methods for allocation concealment. Overall, the risk of bias due to allocation concealment is low.

Blinding

Blinding of participants and personnel: nine studies were deemed to be low risk for performance bias with eight studies that were double blinded (Benor 2014 ;Fernández 2016; Kopp 2008; Laitinen 2009; Lindsay 2014; Lindsay 2015; Ou 2012; Rautava 2012), and one study (Badehnoosh 2018) in which it was unclear if the research nurse who randomized participants was blinded, but other personnel and participants were blinded. Three studies did not describe blinding of participants or personnel and are of unclear risk of performance bias (Dolatkhah 2015; Jacobsson 2016; Mantaring 2016).

Blinding of outcome assessors: four studies described that groups were concealed from the researchers at least until after outcomes were assessed (Badehnoosh 2018; Benor 2014; Ou 2012; Rautava 2012), three studies had no clear description of blinding but the outcomes assessed were thought to be straightforward enough to have low risk of detection bias (Fernández 2016; Kopp 2008; Laitinen 2009), while two studies had adequate blinding of assessors for most outcomes (Lindsay 2014; Lindsay 2015), totaling nine studies that are deemed to be low risk for detection bias. Three studies had no description of blinding of outcome assessors and are thus of unclear risk (Dolatkhah 2015; Jacobsson 2016; Mantaring 2016).

Incomplete outcome data

Six studies were deemed to have low attrition bias due to < 10% of participants lost to follow up in five studies (Badehnoosh 2018; Fernández 2016; Kopp 2008; Laitinen 2009; Lindsay 2015), and in one study that had 89% full outcome data, with the other 11% only being excluded due to a lack of ability to provide > 50% breast milk by one week postpartum (Benor 2014). Four studies were at high risk for attrition bias due to > 10% participants lost to follow‐up (Lindsay 2014; Mantaring 2016; Ou 2012; Rautava 2012), while two studies had unclear risk due to no description of their completeness of data (Dolatkhah 2015; Jacobsson 2016). Of note, the number of infants was lower than the number of mothers due to miscarriage and dropouts from the studies.

Selective reporting

Nine studies were deemed to be low risk, five since outcomes that were described in the study protocols were reported (Badehnoosh 2018; Benor 2014; Laitinen 2009; Lindsay 2014; Lindsay 2015), and four that did not have published study protocols but reported data for outcomes that were stated (Fernández 2016; Kopp 2008; Ou 2012; Rautava 2012). One study was deemed to have unclear risk (Jacobsson 2016) due to lack of information in the abstract, and two studies that did not report data for all planned outcomes were considered high risk of reporting bias (Dolatkhah 2015; Mantaring 2016).

Other potential sources of bias

Ten of the studies have low risk of other sources of bias (Badehnoosh 2018; Benor 2014; Dolatkhah 2015; Fernández 2016; Kopp 2008; Laitinen 2009; Lindsay 2014; Lindsay 2015; Ou 2012; Rautava 2012), while two have an unclear risk due to their lack of description in their abstracts (Jacobsson 2016; Mantaring 2016).

Effects of interventions

See: Table 1; Table 2

Summary of findings for the main comparison. Probiotics administered to pregnant women at risk of preterm birth compared to placebo or no intervention for prevention of morbidity and mortality in preterm infants.

Probiotics administered to pregnant women compared to placebo or no intervention for prevention of morbidity and mortality in preterm infants
Patient or population: prevention of morbidity and mortality in preterm infants  Setting:   Intervention: probiotics administered to pregnant women < 36 weeks' gestation  Comparison: placebo or no intervention.
Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Certainty of the evidence  (GRADE)	Comments
	Risk with placebo or no intervention.	Risk with Probiotics administered to pregnant women < 37 weeks' gestation
Preterm birth < 37 weeks' gestation	Study population		Typical RR 0.92  (0.32 to 2.67)	518  (4 RCTs)	⊕⊕⊝⊝  LOW ^{1, 2}
	26 per 1,000	24 per 1,000  (8 to 70)
Preterm birth < 34 weeks' gestation	Study population		‐	287  (2 RCTs)	⊕⊝⊝⊝  VERY LOW ^{1, 3}	No preterm birth < 34 weeks' gestation in either group.
	see comment	see comment
Gestational age at birth (weeks)	The mean gestational age at birth (weeks) was 0	MD 0.15 higher  (0.33 lower to 0.63 higher)	‐	207  (2 RCTs)	⊕⊕⊝⊝  LOW ¹	Outcome added post hoc
Death	Study population		‐	298  (2 RCTs)	⊕⊕⊝⊝  LOW ³	No death reported in either group.
	see comment	see comment
Miscarriage/stillbirth	Study population		Typical RR 0.79  (0.20 to 3.14)	320  (2 RCTs)	⊕⊕⊝⊝  LOW ²	Outcome added post hoc
	25 per 1,000	20 per 1,000  (5 to 78)
Neonatal sepsis ≤ 3 days	Study population		not estimable	162  (1 RCT)	⊕⊝⊝⊝  VERY LOW ^{1, 3, 4}
	0 per 1,000	0 per 1,000  (0 to 0)
Severe necrotizing enterocolitis (NEC) (Bell stage II or more)						Not reported
Death or severe NEC						Not reported
Late‐onset sepsis						Not reported
Culture‐proven sepsis with supplemented probiotic						Not reported
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).    CI: Confidence interval; RR: Risk ratio
GRADE Working Group grades of evidence  High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.  Moderate certainty: we are moderately confident in the effect estimate. The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.  Low certainty: our confidence in the effect estimate is limited. The true effect may be substantially different from the estimate of the effect.  Very low certainty: we have very little confidence in the effect estimate, The true effect is likely to be substantially different from the estimate of effect.

Probiotics administered exclusively after birth in mothers of preterm infants compared to maternal placebo or no intervention for prevention of morbidity and mortality in preterm infants
Patient or population: prevention of morbidity and mortality in preterm infants  Setting:   Intervention: probiotics administered exclusively after birth in mothers of preterm infants < 37 weeks' gestation  Comparison: maternal placebo or no intervention
Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Certainty of the evidence  (GRADE)	Comments
	Risk with maternal placebo or no intervention	Risk with Probiotics administered exclusively after birth in mothers of preterm infants < 37 weeks' gestation

Any necrotizing Enterocolitis (NEC)	Study population		RR 0.44  (0.13 to 1.46)	58  (1 RCT)	⊕⊕⊝⊝  LOW ¹	Outcome added post hoc
	273 per 1,000	120 per 1,000  (35 to 398)
Death prior to hospital discharge	Study population		RR 0.66  (0.06 to 6.88)	58  (1 RCT)	⊕⊕⊝⊝  LOW ¹
	61 per 1,000	40 per 1,000  (4 to 417)
Death or NEC	Study population		RR 0.53  (0.19 to 1.49)	58  (1 RCT)	⊕⊕⊝⊝  LOW ¹
	303 per 1,000	161 per 1,000  (58 to 452)
Surgery for NEC	Study population		RR 0.15  (0.01 to 2.58)	58  (1 RCT)	⊕⊕⊝⊝  LOW ¹	Outcome added post hoc
	121 per 1,000	18 per 1,000  (1 to 313)
Neonatal sepsis	Study population		RR 1.32  (0.48 to 3.61)	58  (1 RCT)	⊕⊕⊝⊝  LOW ¹
	182 per 1,000	240 per 1,000  (87 to 656)
Culture‐proven sepsis with probiotic organism	Study population		not estimable	58  (1 RCT)	⊕⊕⊝⊝  LOW ¹
	0 per 1,000	0 per 1,000  (0 to 0)
Feeding with more than 50% breast milk (days)	The mean feeding with more than 50% breast milk (days) was 0	MD 9.6 lower  (19.04 lower to 0.16 lower)	‐	58  (1 RCT)	⊕⊕⊝⊝  LOW ¹	Outcome added post hoc Although statistically significant, the result is imprecise and only derived from the one included study
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).    CI: Confidence interval; RR: Risk ratio
GRADE Working Group grades of evidence  High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.  Moderate certainty: we are moderately confident in the effect estimate. The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.  Low certainty: our confidence in the effect estimate is limited. The true effect may be substantially different from the estimate of the effect.  Very low certainty: we have very little confidence in the effect estimate. The true effect is likely to be substantially different from the estimate of effect.

Date	Event	Description
25 January 2018	New search has been performed	Performed the search, included results, discussion and conclusion.
6 February 2017	Amended	Added external source of support

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Preterm birth < 37 weeks' gestation (age at enrollment)	4	518	Risk Ratio (M‐H, Fixed, 95% CI)	0.92 [0.32, 2.67]
1.1 Enrollment at < 20 weeks	2	309	Risk Ratio (M‐H, Fixed, 95% CI)	0.74 [0.22, 2.51]
1.2 Enrollment at > 20 weeks	2	209	Risk Ratio (M‐H, Fixed, 95% CI)	2.0 [0.19, 20.90]
2 Preterm birth < 37 weeks' gestation (probiotic preparation)	4	518	Risk Ratio (M‐H, Fixed, 95% CI)	0.92 [0.32, 2.67]
2.1 Lactobaccillus preparation	2	287	Risk Ratio (M‐H, Fixed, 95% CI)	1.79 [0.31, 10.35]
2.2 Mixed probiotic preparation	2	231	Risk Ratio (M‐H, Fixed, 95% CI)	0.60 [0.15, 2.48]
3 Preterm birth < 34 weeks' gestation	2	287	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.02, 0.02]
4 Gestational age at birth (weeks)	2	207	Mean Difference (IV, Fixed, 95% CI)	0.15 [‐0.33, 0.63]
5 Death prior to hospital discharge	2	298	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.02, 0.02]
6 Miscarriage/stillbirth	2	320	Risk Ratio (M‐H, Fixed, 95% CI)	0.79 [0.20, 3.14]
7 Maternal sepsis	1	171	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.02, 0.02]
8 Maternal mastitis	1	108	Risk Ratio (M‐H, Fixed, 95% CI)	0.45 [0.27, 0.75]
9 Neonatal sepsis ≤ 3 days	1	159	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.02, 0.02]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Any necrotizing enterocolitis (NEC)	1	58	Risk Ratio (M‐H, Fixed, 95% CI)	0.44 [0.13, 1.46]
2 Death prior to hospital discharge	1	58	Risk Ratio (M‐H, Fixed, 95% CI)	0.66 [0.06, 6.88]
3 Death or NEC	1	58	Risk Ratio (M‐H, Fixed, 95% CI)	0.53 [0.19, 1.49]
4 Surgery for NEC	1	58	Risk Ratio (M‐H, Fixed, 95% CI)	0.15 [0.01, 2.58]
5 Neonatal sepsis	1	58	Risk Ratio (M‐H, Fixed, 95% CI)	1.32 [0.48, 3.61]
6 Culture‐proven sepsis with probiotic organism	1	58	Risk Difference (M‐H, Fixed, 95% CI)	0.0 [‐0.07, 0.07]
7 Bronchopulmonary dysplasia (BPD)	1	58	Risk Ratio (M‐H, Fixed, 95% CI)	0.73 [0.28, 1.92]
8 Patent ductus arteriosus	1	58	Risk Ratio (M‐H, Fixed, 95% CI)	1.21 [0.64, 2.28]
9 Intraventricular hemorrhage (any grade)	1	58	Risk Ratio (M‐H, Fixed, 95% CI)	1.58 [0.54, 4.60]
10 Periventricular leukomalacia (PVL)	1	58	Risk Ratio (M‐H, Fixed, 95% CI)	5.28 [0.63, 44.38]
11 Retinopathy of prematurity	1	58	Risk Ratio (M‐H, Fixed, 95% CI)	1.98 [0.36, 10.97]
12 Days to achieve 50% enteral feeds (days)	1	58	Mean Difference (IV, Fixed, 95% CI)	‐9.60 [‐19.04, ‐0.16]

Methods	The aim of the study was to evaluate the effects of probiotic supplementation on biomarkers of inflammation, oxidative stress and pregnancy outcomes among participants with GDM. Randomized, double‐blind, placebo‐controlled, parallel clinical trial Method of generating randomization sequence: computer‐generated random numbers Allocation concealment: allocation was concealed from the researchers and participants until the final analyses were completed Blinding of intervention: patients and researchers were blinded to the intervention. It is unclear if the trained midwife at the Gynecology Clinic who created the randomized allocation sequence by computer, enrolled patients and allocated patients was blinded. Blinding of outcome measurement: patients and researchers were blinded to the intervention until final analyses were completed Complete follow‐up: yes, 100%
Participants	60 pregnant women diagnosed between 24 to 28 weeks with GDM who were not on oral hypoglycemic agents, and their 60 infants Exclusion criteria: placental abruption, pre‐eclampsia, eclampsia, hypo‐ and hyperthyroidism, urinary tract infection, smokers, those with kidney or liver diseases and required commencing insulin therapy during intervention, taking any probiotic products including probiotic yogurt and kefir during the trial. Demographic data (maternal) ‐ Maternal age: probiotics (N = 30): 28.8 ± 5.4 years; placebo (N = 30): 27.8 ± 3.7 years ‐ Gestational age: probiotics (N = 30): gestational age (prior to intervention): 25.7 ± 1.0 weeks; placebo (N = 30): gestational age (prior to intervention): 25.6 ± 1.2 weeks
Interventions	Probiotics group (N = 30) received a daily probiotic capsule with Lactobacillus acidophilus, Lactobacillus casei and Bifidobacterium bifidum ( 2 x 10^9 CFU/g each) Placebo group (N = 30) received a daily placebo capsule The intervention or placebo were administered for 6 weeks during the second and/or third trimester, starting at 24 to 28 weeks' gestation.
Outcomes	Primary outcomes were inflammatory markers, while secondary outcomes were biomarkers of oxidative stress and pregnancy outcomes, including preterm birth, newborn weight, length, and head circumference, Apgar scores, hyperbilirubinemia, newborn hospitalization, newborn hypoglycemia, macrosomia, cesarean delivery rates, maternal pre‐eclampsia, polyhydramnios, maternal hospitalization, and need for maternal insulin therapy. Outcomes of interest were preterm birth < 37 weeks and gestational age at birth.
Notes	Period of study: April 2016 to September 2016 Published: 2017 Source of funding: grant from the Vice chancellor for Research, IUMS, Tehran, Iran Sponsorship Source: Iran University of Medical Sciences (IUMS) Country: Iran Setting: single‐center, Akbarabadi Clinic in Tehran, Iran Author's contact details: Zatollah Asemi asemi_r@yahoo.com, jamilian.mehri@gmail.com TRIAL REGISTRATION: IRCT201611115623N91
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated random numbers
Allocation concealment (selection bias)	Low risk	Allocation was concealed from the researchers and subjects until the final analyses were completed
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Group was concealed from the researchers and participants until the final analyses were completed. It is unclear if the trained midwife at the gynecology clinic who created the randomized allocation sequence by computer, enrolled patients and allocated patients was blinded.
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Group was concealed from the researchers and participants until the final analyses were completed
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No participants were lost to follow‐up
Selective reporting (reporting bias)	Low risk	Outcomes described in the study protocol were reported in the study.
Other bias	Low risk

Methods	The aim of the study was to determine the effects of maternal oral probiotic supplementation on the incidence of NEC, death, and sepsis in very low birth weight (VLBW) infants fed with maternal breast milk. Prospective, randomized, double‐blind, placebo‐controlled trial at the Tel Aviv Medical Center (June 2007‐November 2009). Method of generating randomization sequence: computer‐generated random numbers Allocation concealment: opaque medication boxes Blinding of intervention: nursing staff, physicians, researchers and participants were masked to group assignment. Blinding of outcome measurement: NICU physicians who determined clinical outcomes were masked to study assignment. Complete follow‐up: yes
Participants	Infants with a birth weight of 1500 g or less. Entry into the study during the first 48 hours after delivery. Mothers were recruited at 0 to 72 hours after birth. Exclusion criteria for the mothers: Quote: "Chronic disease of the heart, lung, gastrointestinal tract or kidneys, chronic use of antibiotics, chronic use of probiotic products, or inability to breast feed or express breast milk". Exclusion criteria for infants: Quote: "Too ill to receive enteral nutrition, the presence of congenital chromosomal, neurological, gastrointestinal or major cardiac anomalies, and/or hypoxic brain injury". Demographic data (infant) In total, 49 mothers with 58 VLBW infants were recruited. A total of 25 infants were in the probiotic group and 33 in the placebo group. Gestational age of enrolled infants: probiotics (N = 25) 29.49 ± 2.69 weeks; placebo (N = 33) 29.47 ± 2.57 weeks Birth weight of enrolled infants: probiotics (N = 25) 1105 ± 267 g; placebo (N = 33) 1080 ± 237 g Gestational age, birth weight and exposure to antenatal corticosteroids were similar between groups. There were more infants in the control group that were delivered by cesarean section than in the probiotic group (94% versus 72%, P = 0.03).
Interventions	Intervention was started as soon as the mother was recruited at 0 to 72 hours of life of their infant, and continued until the infant was discharged. Probiotic capsules that contained a mixture of Lactobacillus acidophilus (L. acidophilus NCFM) and Bifidobacteria lactis (Bi‐07) at a concentration of 2 x 10^10 CFU manufactured by Danisco (Copenhagen, Denmark)., Placebo capsule containing cellulose manufactured by Danisco (Copenhagen, Denmark).
Outcomes	Neonatal outcomes of interest: severe NEC, death, death or NEC, retinopathy of prematurity, Intraventricular hemorrhage, any grade, culture‐proven sepsis with probiotic, mean duration to reach full feeds
Notes	Sponsorship Source: Department of Pediatrics, Dana Children’s Hospital, Tel Aviv Sourasky Medical Center Department of Neonatology, Lis Maternity Hospital, Tel Aviv Sourasky Medical Center Country: Israel Setting: Tel Aviv Medical Center NICU Author's contact details; Shira Benor, MD, Department of Pediatrics, Tel Aviv Sourasky Medical Center Address: 6 Weizman Street, Tel Aviv 64239, Israel Email: shirabenor@yahoo.com TRIAL REGISTRATION: ClinicalTrials.gov NCT00835874. (Dolberg 2007).
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Participating mothers were randomized by computer‐generated random numbers with opaque medication boxes without any marking on the box.
Allocation concealment (selection bias)	Low risk	Opaque medication boxes containing identical capsules (without any marking on the box)
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Nursing staff, physicians, researchers and participants were masked to group assignment. Only the study pharmacist was aware of treatment assignment
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	NICU physicians who determined clinical outcomes were masked to study assignment.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	49 of 55 mothers (89%), and 58 of 65 infants (89%) were included in analysis. The exclusion of 6 mothers to 7 infants was based upon the inability to provide at least 50% of infant milk at 1 week postpartum.
Selective reporting (reporting bias)	Low risk	Reported on outcomes described in the protocol
Other bias	Low risk

Methods	The aim of the study was to assess the effect of a probiotic supplementation on glucose metabolism and weight gain in pregnant women with newly diagnosed GDM. Double‐blind placebo‐controlled randomized clinical trial Method of generating randomization sequence: not described Allocation concealment: not described Blinding of intervention: not described Blinding of outcome measurement: not described Complete follow‐up: 100%
Participants	64 nulliparous pregnant women diagnosed between 24 to 28 weeks of gestation with GDM and their infants Exclusion criteria for mothers: age < 18 years or > 45 years, BMI < 18.5, history of type 2 diabetes mellitus, history of chronic diseases, smoking or alcohol consumption, probiotic food consumption 2 weeks before the intervention, antibiotics 1 month before intervention, acute gastrointestinal issues one month before the trial, use of glucocorticoids or immunosuppressive drugs, need for insulin or other diabetic drugs during study period Demographic data (maternal) Maternal age: probiotics (N = 29) 28.14 ± 6.24 years, placebo (N = 27) 26.48 ± 5.23 years; originally 32 mothers enrolled in each group.
Interventions	Probiotics group (N = 32) received a daily probiotic capsule containing Lactobacillus acidophilus LA‐5, Bifidobacterium BB‐12, Streptococcus thermophilus STY‐31 and Lactobacillus delbrueckii bulgaricus LBY‐27 (4 biocap > 4 × 109 CFU) Placebo group (N = 32) received a daily placebo capsule Participants were allocated to receive either probiotic supplement or placebo capsules once daily for 8 weeks at or after 24 to 28 weeks' gestation.
Outcomes	Primary outcomes assessed were trend in weight gain and glucose metabolism indices in the mother. No infant outcomes of interest were reported.
Notes	Period of study: spring and summer 2014 Published: 2015 Source of funding: Tehran Darou Pharmaceuticals provided the probiotic supplement Sponsorship Source: Department of Nutrition Sciences, Shahid Beheshti University of Medical  Sciences (SBUMS), International Branch, Tehran, Iran Country: Iran Setting: Alzahra University Hospital in Tabriz, Northwest of Iran Author's contact details: Majid Hajifaraji; m.hajifaraji@nnftri.ac.ir TRIAL REGISTRATION: IRCT201405181597N3.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not described
Allocation concealment (selection bias)	Unclear risk	Not described
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Not described
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Not described
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	Not described
Selective reporting (reporting bias)	High risk	Did not report the following planned laboratory outcomes from the study protocol: hs‐CRP, TNF‐α, Interleukine‐6, total antioxidant capacity/TAC, malondialdehyde/MDA, superoxide dismutase/SOD, Glutathione Reductase/GSH‐Red, Glutathione Peroxidase/GSH‐Px, Uric Acid
Other bias	Low risk

Methods	The aim of the trial was to evaluate the potential of probiotics to prevent mastitis when orally administered during late pregnancy to women who had experienced infectious mastitis after previous pregnancies. Double‐blind randomized controlled trial Method of generating randomization sequence: computer‐generated allocation sequence Allocation concealment: yes Blinding of intervention: control group received capsules with powdered milk Blinding of outcome measurement: yes Complete follow‐up: two women dropped out of study for Quote: "reasons not related to the study"
Participants	A total of 108 pregnant women, chronological age 24 to 35 and gestational age 27 to 32 weeks participated in this study. All met the following criteria: normal pregnancy, healthy status, and a history of lactational mastitis after at least one previous pregnancy. Demographic data Maternal age: Probiotics group N = 55: 31 years (95% CI 31 to 32) Placebo group N = 53: 31 years (95% CI 30 to 31) Gestational age at study enrollment (mean weeks (95% CI)) Probiotics group N = 55: 29.82 weeks (29 to 30) Placebo group N = 53: 30.06 weeks (30 to 30)
Interventions	Probiotic group (n = 55) ingested daily 9 log10 CFU of L. salivarius PS2, whereas those in the placebo group (n = 53) received a placebo. The intervention was given from approximately week 30 of pregnancy until delivery.
Outcomes	Mastitis, breast pain score, bacterial counts in milk sample, probiotic in milk sample. Outcome of interest reported: maternal mastitis.
Notes	Period of study: unknown Published: 2016 Source of funding: supported by AGL2013‐41980‐P project from the Ministerio de Economía y Competitividad (Spain). Sponsorship Source: approved by the Ethical Committee of Clinical Research of Hospital Clínico San Carlos Madrid. Country: Spain Setting: Hospital Clínico San Carlos Madrid, Spain Author's contact details: Department of Nutrition, Food Science, andFood Technology, Universidad Complutense de Madrid, Spain TRIAL REGISTRATION: Clinical Trials Registration. NCT01505361.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated allocation sequence
Allocation concealment (selection bias)	Low risk
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Control group received placebo capsules with powdered milk
Blinding of outcome assessment (detection bias)   All outcomes	Low risk
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Two women (out of 108 enrolled) dropped out of study for Quote: "reasons not related to the study"
Selective reporting (reporting bias)	Low risk
Other bias	Low risk

Methods	Jacobsson 2016 studied the immunological and inflammatory response after antenatal supplementation with Lactobacillus rhamnosus in low‐risk pregnant women. The objective was to evaluate whether supplementation with Lactobacillus rhamnosus (LGG) in low‐risk pregnant women interfere on maternal immunological or inflammatory response. Randomized controlled trial Method of generating randomization sequence: unknown Allocation concealment: unknown Blinding of intervention: unknown Blinding of outcome measurement: unknown Complete follow‐up: unknown
Participants	Low‐risk pregnant women at 8 to 10 weeks' gestation (risk factors undefined) Exclusion criteria: unknown Demographic data Probiotics group N = 20 Placebo group N = 20
Interventions	Low‐risk women randomized at week 8 to 10 weeks' gestation to supplementation with LGG (108 CFU) or placebo until delivery.
Outcomes	Maternal blood was sampled at recruitment, at weeks 25 and 35. Fluorescence activated cell‐scanning analysis; FOXP3 coloring and Toll‐like receptor stimulation with Lactobacillus plantarum, Pseudomonas aeruginosa and lipopolysaccharides (LPS) was used to evaluate the response of 26 parameters. Variables were log transformed and data presented with geometric mean and geometric standard deviation. Baseline values and changes from baseline were compared with two sample t‐tests. No outcomes of interest reported.
Notes	Period of study: unknown Published: 2016 (abstract) Source of funding: unknown Sponsorship Source: unknown Country: Sweden and Norway Setting: conducted by investigators at Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden; Chalmers University, Gothenburg, Sweden; Institute of Public Health, Oslo, Norway. Author's contact details: not stated TRIAL REGISTRATION: not stated
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Published as abstract only. Little detail available on which to judge risk of bias.
Allocation concealment (selection bias)	Unclear risk	Published as abstract only. Little detail available on which to judge risk of bias.
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Published as abstract only. Little detail available on which to judge risk of bias.
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Published as abstract only. Little detail available on which to judge risk of bias.
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	Published as abstract only. Little detail available on which to judge risk of bias.
Selective reporting (reporting bias)	Unclear risk	Published as abstract only. Little detail available on which to judge risk of bias.
Other bias	Unclear risk	Published as abstract only. Little detail available on which to judge risk of bias.

Methods	The aim of the study was to investigate the immunologic proliferative response and cytokine release in cultures of isolated mononuclear cells in pregnant women with a first degree relative or partner with atopic disease and their neonates supplemented with probiotics during pregnancy. Double‐blind, randomized placebo‐controlled trial Method of generating randomization sequence: random computer generation in blocks of 4 Allocation concealment: placebo controlled Blinding of intervention: identical appearance, taste, smell and packaging of capsules Blinding of outcome measurement: yes Complete follow‐up: total enrolled 94 (cord blood obtained on 68 infants)
Participants	94 pregnant women who had uneventful pregnancies and a family history of atopic disease (eczema, allergic rhinitis or asthma) and their 94 Infants probiotics N = 50, Placebo N = 44). Exclusion criteria: underlying chronic disease (i.e. diabetes mellitus, rheumatoid arthritis, chronic infectious disease) Maternal Demographics Maternal age at study enrollment: unknown
Interventions	Two capsules of placebo (microcrystalline cellulose) or L. rhamnosus GG (ATCC 53103) containing 5x10^9 CFU of LGG (Infectopharm, Heppenheim, Germany) daily for 4 to 6 weeks before expected delivery. After birth, breastfeeding mothers took the randomized interventions for 3 months, but formula‐fed infants (n = 2 in the LGG group, n = 3 in the placebo group) were given the agents directly. After 3 months, the capsules were given only directly to infants until 6 months .
Outcomes	Primary outcomes were proliferative response of cord blood mononuclear cells and peripheral blood mononuclear cells from the corresponding mother. Secondary outcomes related to the neonate were gestational age and birth weight. Outcome of interest: gestational age at birth.
Notes	Five formula‐fed infants were given the interventions directly after birth. Thus, only data pertaining to the time period of prenatal supplementation was included in this review. Period of study: July 2002 to June 2004 Published: 2007 Source of funding: University of Freiburg and by Infectopharm, Heppenheim, Germany Sponsorship Source: University of Freiburg Country: Germany Setting: outpatient Gynecology offices coordinated with the University of Freiburg Author's contact details: Dr Matthias Kopp, University Children’s Hospital, Mathildenstraße 1, D‐79106 Freiburg, Germany. E‐mail: Matthias.kopp@ uniklinik‐freiburg.de TRIAL REGISTRATION: None
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated random number assignment
Allocation concealment (selection bias)	Low risk	No description of how investigators received and distributed the sequence. Placebo‐controlled (capsules of placebo (microcrystalline cellulose) or L. rhamnosus GG (ATCC 53103) containing 5 x 10^9 CFU of LGG daily for 4 to 6 weeks before expected delivery).
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	No clear description of blinding, but numbered identical appearing interventions were given.
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	No specific description of blinding of assessors.
Incomplete outcome data (attrition bias)   All outcomes	Low risk	All infants are accounted for in an ITT analysis. Cord blood values only reported on 68/94 enrolled infants.
Selective reporting (reporting bias)	Low risk
Other bias	Low risk

Methods	Laitinen 2009 investigated the impact of dietary counseling and/or probiotics on pregnant women with an aim to optimize maternal dietary intake and metabolism, to advance maternal health,and reduce the risk of disease in the child. Randomized, prospective, parallel‐group, combined dietary counseling and probiotics intervention study at a single center. Mothers were randomized to either 1. probiotics and dietary intervention, 2. placebo and dietary intervention, or 3. placebo and no dietary intervention. The data from the placebo group and no dietary intervention group were not included in this review as there is no appropriate comparator. Multiple reports were generated by this trial and are briefly noted below in the "Notes" section. Method of generating randomization sequence: computer‐generated block randomization of six women done by a statistician not involved in recruitment or study visits. Allocation concealment: sealed envelopes with participant group assignments Blinding of intervention: probiotic versus placebo intervention was double‐blinded and adequate. Nutritional counseling component was unblinded. For the included comparisons both groups received identical nutritional counseling. Blinding of outcome measurement: yes. Complete follow‐up: unclear (depending on outcome assessed). Greater than 10% loss to follow‐up for reasons varying from "moved" to "illness in child".
Participants	256 pregnant women who were less than 17 weeks' gestation at enrollment. Exclusion criteria: metabolic or chronic diseases such as diabetes, rheumatoid arthritis, inflammatory bowel disease, thyroid diseases, malignancies, > 17 weeks' gestation pregnant, Demographic data Probiotics and dietary intervention group N = 85 pregnant women, N = 82 infants; maternal age 29·7 ± 4·1 years Placebo and dietary intervention group N = 86, N = 80 infants; maternal age 30·1 ± 5·2 years Placebo and control N = 85 (data were not included in review due to no probiotic and control group for comparison)
Interventions	Probiotics and dietary intervention group (N = 85) received Lactobacillus rhamnosus GG, and Bifidobacterium lactis 10^10 CFUs/day each. Placebo and dietary intervention group (N = 86) received identical‐looking capsules that contained cellulose and dextrose anhydrate. The intervention was given from the first study visit around 14 weeks until the end of exclusive breastfeeding. Dietary counseling intervention occurred at each study visit and aimed to modify dietary intake to conform to current dietary recommendations with a specific focus on dietary fat intake.
Outcomes	Primary and secondary outcomes vary between the articles, and include outcomes focused on maternal dietary compliance, adiposity, pregnancy weight gain, anthropometric measurements during and after pregnancy, maternal diagnosis of gestational diabetes, maternal laboratory indices for lipids, leptin, and glucose metabolism, as well as placental fatty acids, breast milk fatty acids and inflammatory markers, and infant sensitization. Outcomes pertinent to the neonate include birth weight, birth length, birth head circumference, gestational age at birth, miscarriage rate, preterm birth < 32 weeks, preterm birth 32 to 36 weeks, cesarean delivery rate, 5‐minute Apgar score. Outcomes of interest include preterm birth < 37 weeks, maternal miscarriage, maternal sepsis, neonatal mortality and neonatal early onset sepsis.
Notes	Period of study: April 2002 to November 2005 Published: Main study 2009; various publications from 2006 to 2014 Source of funding:grants from the Social Insurance Institution of Finland, the Sigrid Juselius Foundation and the Academy of Finland. Provision of food products was by Raisio plc (Raisio, Finland), B. lactis Bb12 by Chr. Hansen (Hoersholm, Denmark) and L. rhamnosus GG by Valio Ltd (Helsinki). Sponsorship Source: University of Turku Country: Finland Setting: study clinic in Turku University Central Hospital Author's contact details: Dr Kirsi Laitinen, fax þ358 2 333 6862, email kirsi.laitinen@utu.fi TRIAL REGISTRATION: ClinicalTrials.gov, NCT00167700 Study reports from the Laitinen 2009 cohort with descriptions of primary outcomes of each study: Piirainen 2006: focused on the outcomes of maternal dietary intake, compliance with diet, maternal weight gain. This study was published with only a subset of participants from the final study (n = 231; control n = 69, intervention n = 140 without description of probiotic versus placebo). Aaltonen 2008: primary outcome of interest: Infant blood pressure at 6 months of age. Secondary outcomes: maternal weight gain and blood pressure, dietary intake, 6‐month old weight, head circumference, length. Huurre 2008: primary outcome of interest: Infant sensitization at 6 and 12 months (Only including the mother‐infant pairs assigned to nutritional counseling); other outcomes included rates of exclusive breastfeeding, cesarean section rate, allergy testing of mothers, inflammatory markers in colostrum and breast milk. Hoppu 2012: breast milk fatty acid and cytokine composition (n= 125 due to breast milk sample availability). Other outcomes reported were maternal dietary intake. Hoppu 2014: blood lipid concentrations during and after pregnancy. Ilmonen 2011: maternal anthropometric measurements during and after pregnancy including BMI, adiposity, proportion of body fat and waist circumference. Secondary outcomes were dietary intakes of foods and nutrients and a healthy eating index during the postpartum period. Kaplas 2007: placental fatty acids (only looked at placenta and umbilical cord serum samples of 30 healthy women from the study) Laitinen 2009: maternal glucose metabolism characterized by plasma glucose concentration, blood glycated Hb A1C, serum insulin and HOMA and QUICKI indices. Other outcomes measured included dietary energy‐yielding nutrients assessed from food diaries, which were analyzed to explain changes in glucose metabolism. Laitinen 2011: maternal central adiposity (abstract only for European nutrition conference) Luoto 2009: gestational diabetes diagnosis (PAS abstract only). Secondary outcomes included adverse effects, prenatal growth, proportion of large for gestational age infants. Luoto 2010: gestational diabetes diagnosis. Secondary outcomes included miscarriage rate, preterm birth < 32 weeks, preterm birth 32 to 36 weeks, cesarean delivery rate, 5‐minute Apgar score, birth weight, birth length, birth head circumference. Luoto 2012: colostrum adiponectin concentration. Other outcomes reported include diagnosis of gestational diabetes, weight gain during pregnancy Vahamiko 2013: serum leptin concentrations in women, cord blood and infants at 1 month of age. Other outcomes include weight and rate of exclusive breastfeeding at 1 and 6 months of age.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated block randomization done by a statistician not involved in recruitment or study visits.
Allocation concealment (selection bias)	Low risk	Sealed envelopes with participant group assignments that were opened in the room with the patient.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Nutritional counseling component was single‐blinded but both groups being compared in this review received the counseling, and the probiotic versus placebo intervention was double‐blinded
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	No specific description of blinding of assessors
Incomplete outcome data (attrition bias)   All outcomes	Low risk	92% of patients have birth data included
Selective reporting (reporting bias)	Low risk	Registered clinical trial outcomes are reported
Other bias	Low risk

Methods	The aim of the Probiotics in Pregnancy Study was to investigate the effect of a probiotic capsule on maternal fasting glucose in obese pregnant women. Randomized controlled double‐blind placebo‐controlled trial Method of generating randomization sequence: computer‐generated randomization process by independent researcher Allocation concealment: allocation concealed from the research dietician in sealed sequentially‐numbered opaque envelopes Blinding of intervention: the identity of the capsules unknown to either the participants, researchers or the primary investigators. Blinding of outcome measurement: adequate for most outcomes Complete follow‐up: 138/175 completed and reported (mostly due to assessment of poor compliance)
Participants	175 pregnant women less than 20 weeks' gestation with an early pregnancy body mass index (BMI; in kg/m2) from 30.0 to 39.9 were recruited from antenatal clinics at the National Maternity Hospital, Dublin, Ireland. Exclusion criteria were BMI < 30.0 or > 39.9, prepregnancy or gestational diabetes, age < 18 years, multiple pregnancy, and fetal anomaly. 138 women who completed the study (63 women in the probiotic group; 75 women in the placebo group) Demographic data Maternal age: Probiotics group: 31.4 ± 5.0 years (N = 63) Placebo group: 31.0 ± 5.2 (N = 75)
Interventions	Women were randomly assigned to receive either a daily probiotic (Lactobacillus salivarius at 10⁹ CFUs) or a placebo capsule from 24 to 28 weeks of gestation in addition to routine antenatal care.
Outcomes	The primary outcome was the change in fasting glucose between groups from pre‐intervention to post‐intervention. Secondary outcomes were the incidence of gestational diabetes and neonatal anthropometric measures. Outcomes of interest: preterm birth < 37 weeks, preterm birth < 34 weeks, neonatal mortality.
Notes	Period of study: recruitment from March 2012 to March 2013 Published: 2014 Source of funding: funded by the National Maternity Hospital Medical Fund with support from Ivo Drury Award Sponsorship Source: Alimentary Health Ltd. provided probiotic and placebo capsules free of charge Country: Ireland Setting: recruited from antenatal clinics at the National Maternity Hospital, Dublin, Ireland. Author's contact details: F. M. McAuliffe, University College Dublin, Obstetrics and Gynaecology, School of Medicine and Medical Science, National Maternity Hospital, Dublin Ireland. Fionnuala.mcauliffe@ucd.ie TRIAL REGISTRATION: This trial was registered at Current Controlled Trials as ISRCTN97241163 (part A). Am J Clin Nutr 2014;99:1432–9.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomized controlled double‐blind, placebo‐controlled trial Method of generating randomization sequence: computer‐generated randomization process by independent researcher.
Allocation concealment (selection bias)	Low risk	Allocation concealment: Allocation concealed from the research dietician in sealed sequentially‐numbered opaque envelopes.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Blinding of intervention: the identity of the capsules unknown to either the participants, researchers or the primary investigators.
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Blinding of outcome measurement: adequate for most outcomes (unclear which outcomes determined by the research nurse).
Incomplete outcome data (attrition bias)   All outcomes	High risk	Complete follow‐up: 138/175 completed and reported (mostly due to assessment of poor compliance)
Selective reporting (reporting bias)	Low risk
Other bias	Low risk

Methods	The aim of this study was to investigate the effects of a probiotic capsule intervention on maternal metabolic parameters and pregnancy outcome among women with gestational diabetes. Single center, double‐blind, placebo‐controlled randomized trial Method of generating randomization sequence: computer‐generated randomization process by independent researcher Allocation concealment: allocation concealed from the research dietician in sealed sequentially‐numbered opaque envelopes Blinding of intervention: the identity of the capsules unknown to either the participants, researchers or the primary investigators. Blinding of outcome measurement: adequate for most outcomes Complete follow‐up: 115/149 (77%) completed and reported (mostly due to assessment of poor compliance)
Participants	Recruited pregnant women < 34 weeks' gestation with a new diagnosis of gestational diabetes or impaired glucose tolerance following a 3‐hour 100 g glucose tolerance test. Exclusion criteria: prepregnancy diabetes, age < 18 years, gestational age ≥ 34 weeks' gestation, multiple pregnancy, and fetal anomaly. Excluded if planned medical treatment with insulin or oral hypoglycemic agents immediately after diagnosis. Demographic data Probiotics group N = 74 (48 in secondary per protocol analysis), maternal age 33.5 ± 5.0 years Placebo group N = 75 (52 in secondary per protocol analysis) maternal age 32.6 ±4.5 years
Interventions	Women were randomized to a daily probiotic (Lactobacillus salivarius UCC118 10^9 CFUs) or placebo capsule from diagnosis until delivery.
Outcomes	Fasting blood samples were collected at baseline and 4 to 6 weeks after capsule commencement for analysis of glucose, insulin, C‐peptide, and lipids. The primary outcome was difference in fasting glucose postintervention, first analyzed on an ITT basis and followed by per‐protocol analysis that excluded women commenced on pharmacological therapy (insulin or metformin). Secondary outcomes were changes in insulin, C‐peptide, homeostasis model assessment and lipids, requirement for pharmacological therapy, and neonatal anthropometry. Outcomes of interest: preterm birth < 37 weeks, preterm birth < 34 weeks, gestational age at birth and maternal miscarriage.
Notes	Period of study: recruitment from March 2012 to March 2014 Published: 2015 Source of funding: funded by the National Maternity Hospital Medical Fund with support from Ivo Drury Award and the European Unions Seventh Framework Program, Project EarlyNutrition Sponsorship Source: Alimentary Health Ltd. provided probiotic and placebo capsules free of charge Country: Ireland Setting: recruited from antenatal clinics at the National Maternity Hospital, Dublin, Ireland. Author's contact details: F. M. McAuliffe, University College Dublin, Obstetrics and Gynaecology, School of Medicine and Medical Science, National Maternity Hospital, Dublin Ireland. Fionnuala.mcauliffe@ucd.ie TRIAL REGISTRATION: This trial was registered at Current Controlled Trials as ISRCTN97241163 (part B). Am J Clin Nutr 2014;99:1432–9.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomized controlled, double‐blind, placebo‐controlled trial Method of generating randomization sequence: computer‐generated randomization process by independent researcher.
Allocation concealment (selection bias)	Low risk	Allocation concealment: allocation concealed from the research dietician in sealed sequentially‐numbered opaque envelopes.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Blinding of intervention: the identity of the capsules unknown to either the participants, researchers or the primary investigators.
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Blinding of outcome measurement: adequate for most outcomes (unclear which outcomes determined by th research nurse).
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Complete follow‐up Probiotics group N = 74 (48 in secondary per protocol analysis) Placebo group N = 75 (52 in secondary per protocol analysis)
Selective reporting (reporting bias)	Low risk
Other bias	Low risk

Methods	This study evaluated the effectiveness of prenatal and postnatal probiotics in the prevention of early childhood and maternal allergic diseases. Prospective, double‐blind, randomized, placebo‐controlled trial Method of generating randomization sequence: unclear Allocation concealment: unclear Blinding of intervention: treatment codes were kept by the supplier until all data had been collected and analyzed Blinding of outcome measurement: treatment codes were kept by the supplier until all data had been collected and analyzed Complete follow‐up: 78% of participants randomized were analyzed in the newborn period
Participants	191 pregnant women with any allergic diseases (asthma, eczema, food allergy or allergic rhinitis) AND elevated IgE (elevated total IgE and/or positive allergen specific IgE meeting specific thresholds) Exclusion criteria: maternal diabetes mellitus, rheumatoid arthritis, or infectious disease; no family history of allergic diseases Demographic data: Probiotics group N = 95; maternal age 31 years, range 23–40 years Placebo group N = 96; maternal age 30 years, range 25–42 years
Interventions	Probiotics group (N = 95) received Lactobacillus rhamnosus GG; 1 x 10^10 CFU daily Placebo group (N = 96) received placebo daily During pregnancy, the interventions began at 24 weeks' gestation and continued until delivery. After delivery, the interventions were given to exclusively to breastfeeding mothers or directly administered to non‐breastfeeding neonates until 6 months of age.
Outcomes	Outcome of interest reported: gestational age at birth Other outcomes reported: point and cumulative prevalence of sensitization and development of allergic diseases, and improvement of maternal allergic symptom score and plasma immune parameters before and after intervention at 6, 18 and 36 months.
Notes	Only outcomes at birth regarding the probiotics being given during pregnancy were included, as postnatally, some infants received the probiotics directly. Period of study: August 2002 and January 2006 Published: 2012 Source of funding: unclear Sponsorship Source: Chang Gung Memorial Hospital Country: Taiwan Setting: Single center, Kaohsiung Chang Gung Memorial Hospital Author's contact details: Kuender D. Yang, Department of Medical Research, Show Chwan Memorial Hospital in Chang Bing, No. 6‐1 Lu‐Kung Road, Chang Bing Industrial Center, Lu‐Kang, Changhua 505, Taiwan. E‐mail: yangkd.yeh@hotmail.com TRIAL REGISTRATION: ClinicalTrials.gov NCT00325273.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No description of sequence generation
Allocation concealment (selection bias)	Unclear risk	No description of allocation concealment
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Treatment codes were kept by the supplier until all data had been collected and analyzed
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Treatment codes were kept by the supplier until all data had been collected and analyzed
Incomplete outcome data (attrition bias)   All outcomes	High risk	Only 78% of randomized participants were analyzed at the newborn period.
Selective reporting (reporting bias)	Low risk	Changed the timing of post‐natal follow up laboratory tests of allergic sensitization from protocol to study
Other bias	Low risk

Methods	This study investigated whether maternal probiotic supplementation during pregnancy and breast‐feeding reduces the risk of developing eczema in high‐risk infants. Parallel, randomized, double‐blind, placebo‐controlled trial Method of generating randomization sequence: computer‐generated by the manufacturer of the study products, independent of the researchers Allocation concealment: study group code only known by manufacturer Blinding of intervention: double‐blind with group only known to researchers after all follow‐up and outcomes assessed Blinding of outcome measurement: double‐blind with group only known to researchers after all follow‐up and outcomes assessed Complete follow‐up: 12/241 infants were lost to follow‐up completely (5%), while 205/241 (85%) had outcomes assessed.
Participants	241 pregnant women mothers with allergic disease and atopic sensitization, and their 241 infants Exclusion criteria: immune‐mediated disease other than atopic or allergic disease, and infants born of multiple pregnancies Demographic data Probiotics group #1, N = 81, maternal age 31 years, range 22‐43 years Probiotics group #2, N = 82, maternal age 30 years, range 22‐42 years Placebo group N = 78, maternal age 30 years, range 22‐40 years
Interventions	Probiotics group #1 receiving Lactobacillus rhamnosus 1 x 10^9 CFU and Bifidobacterium longum 1 x 10^9 CFU with a dietary food supplement of vitamins and minerals, N = 81 Probiotics group #2 receiving L paracasei 1 x 10^9 CFU and B longum 1 x10^9 CFU with a dietary food supplement of vitamins and minerals, N = 82 Placebo consisted of the same dietary food supplement of vitamins and minerals without added probiotics N = 78 Intervention or placebo began 2 months before delivery and continued until 2 months of breast feeding.
Outcomes	The primary outcome was cumulative incidence of eczema in the infant up to 2 years old. Secondary outcomes were atopic sensitization, gestational age at birth, birth weight, and duration of exclusive breast‐feeding. Outcome of interest: gestational age at birth.
Notes	Period of study: August 2005 to April 2009 Published: 2012 Source of funding: Sigrid Juselius Foundation and Turku University Hospital EVO research funding. Sponsorship Source: Turku University Hospital Country: Finland Setting: single tertiary center in Turku, Finland Author's contact details: Samuli Rautava, MD, PhD, Department of Paediatrics, Turku University Central Hospital, Kiinamyllynkatu 4‐8, 20520, Turku, Finland. E‐mail: samrau@utu.fi. TRIAL REGISTRATION: ClinicalTrials.gov NCT00167700.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated random sequence
Allocation concealment (selection bias)	Unclear risk	Method of concealment not described, but since only the drug manufacturer knew the group until after the study was complete, it is likely to be low.
Blinding of participants and personnel (performance bias)   All outcomes	Low risk	Researchers and participants blinded until after all follow‐up and outcomes assessed.
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Researchers and participants blinded until after all follow‐ up and outcomes assessed.
Incomplete outcome data (attrition bias)   All outcomes	High risk	15% with incomplete outcome
Selective reporting (reporting bias)	Low risk
Other bias	Low risk

Study	Reason for exclusion
Asemi 2011	Asemi 2011 conducted a single‐blinded randomized clinical trial in Iran to assess the effects of daily consumption of probiotic yogurt versus conventional yogurt on inflammatory factors in pregnant women. This was excluded since yogurt, which inherently contains probiotic bacteria, was given to both the control and intervention group.
Asemi 2012	Asemi 2012 conducted a randomized, single‐blind controlled trial in Iran to determine the effects of daily consumption of probiotic yogurt versus conventional yogurt on lipid profiles of pregnant women. This study was excluded due to both the intervention and the control receiving probiotics since the conventional yogurt contained S. thermophilus and L. bulgaricus.
Bisanz 2014	Bisanz 2014 conducted a randomized, clinical trial in the Mwanza Region of Tanzania evaluating the effects of probiotic and Moringa plant matter supplementation on maternal on maternal blood metal levels. Excluded due to co‐intervention (Moringa, a micronutrient‐rich plant) given only to probiotic group. Trial registration: ClinicalTrials.gov NCT02021799
Daskalakis 2014	Daskalakis 2014 reported preliminary results from their ongoing double‐blind clinical trial of pregnant women between 24 to 34 weeks of gestation with premature rupture of membranes designed to evaluate the effects of vaginal probiotic suppositories on time to delivery and neonatal outcomes. This study was excluded from this review since the route of administration of the probiotics was vaginal, not oral.
Gonai 2014	Gonai 2014 conducted a randomized control trial in Japan of pregnant women with gestational diabetes to examine the effects of probiotic yogurt versus regular diet on glucose metabolism indices. This study was excluded due to the unclear gestational age at study entry.
Gronlund 2011	Gronlund 2011 analyzed a subset of mother‐infant dyads from a larger randomized controlled trial in high risk allergic families in Finland comparing two different probiotic groups to each other and a placebo arm, with probiotics given 2 months before and after birth. Only a subset of dyads were included in this study, based upon the duration of breast‐feeding, thus, these comparison groups were not allocated via randomization.
Grzeskowiak 2012	Grzeskowiak 2012 evaluated the impact of probiotic administration to mother or infant on gut microbiota composition in 6‐month‐old Finnish and German infants. This study included patients from the Gronlund 2011 trial, a trial of post‐natal administration of probiotic formula to infants, and a few non‐randomized infants who were exclusively breast fed. Reasons for exclusion include probiotics directly administered to infants post‐natally, and non‐randomization in the exclusively breast fed infants in the third arm of this study.
Hanson 2014	Hanson 2014 conducted a non‐randomized pilot feasibility trial in the USA in mid‐gestation pregnant women to determine the effects of probiotic administration on Group B strep colonization. This study was excluded because the study groups were not randomized.
Hantoushzadeh 2012	Hantoushzadeh 2012 conducted a randomized, placebo‐controlled clinical trial in Iran investigating the efficacy of oral probiotic versus clindamycin for one week to treat bacterial vaginosis. Excluded due to an antibiotic as the comparator.
Jain 2017	Jain 2017 conducted a randomized, placebo‐controlled, double‐blind study in the USA in pregnant women ≤ 18 weeks to assess the safety and efficacy of probiotics versus placebo for 60 days. This study was excluded since the intervention not continued into the third trimester when presumably the majority, if not all, of infants were born.
Jamilian 2016	Jamilian 2016 conducted a randomized clinical trial in Iran in pregnant women to determine the effects of probiotics on metabolic profiles by administering probiotic vs placebo from 9 to 21 weeks' gestation. This study was excluded since the intervention was not continue into the third trimester.
Karampelas 2013	Karampelas 2013 conducted a double‐blind study in Greece in pregnant patients with confirmed PPROM between 24th and 34th week of gestation to determine the effects of probiotics on latency period. This study was excluded due to vaginal, not oral, probiotic administration.
Krauss‐Silva 2011	Krauss‐Silva 2011 conducted a randomized, placebo‐controlled trial in Brazil to evaluate the efficacy of the early administration of probiotic lactobacilli to pregnant women with asymptomatic bacterial vaginosis to prevent spontaneous premature delivery and associated neonatal morbidity. While the intervention began at less than 20 weeks, this study was excluded since the intervention was stopped during the second trimester at 24 to 26 weeks.
Nishijima 2005	Nishijima 2005 conducted a randomized, controlled study in Japan to evaluate the effect of fermented milk with probiotics on vaginal colonization with probiotics and potentially pathogenic bacteria in pregnant women enrolled 24 healthy Japanese women ≥ 35 weeks' gestation. Excluded due to the potential for probiotics in fermented milk products given to the control group.
Ortiz‐Andrellucchi 2008	Ortiz‐Andrellucchi 2008 conducted a randomized controlled trial in postpartum, lactating women to determine if six weeks of probiotic intake would modulate the immune system. This study was excluded since probiotics were only administered after birth to mothers of infants born at term.
Vitali 2012	Vitali 2012 conducted a pilot, non‐randomized trial in Italy to investigate the impact of probiotic supplementation on the vaginal microbiota and immunological profiles of healthy women during late pregnancy. This study was excluded since study groups were based upon ability to consume probiotic product and were not randomized.
Wickens 2008	Wickens 2008 conducted a randomized, controlled trial in New Zealand to determine the efficacy of two different probiotics given to pregnant and postpartum women on the development of allergic sensitization and eczema in their offspring. Excluded since probiotics begun late in the third trimester at 35 weeks, mothers who gave birth prior to 37 weeks or had infants with a NICU admission were explicitly excluded, and probiotics were administered directly to the infant after birth starting on day 2.

Trial name or title	SPRING: an RCT study of probiotics in the prevention of gestational diabetes mellitus in overweight and obese women
Methods	Prospective, double‐blind, randomized controlled trial
Participants	Inclusion criteria: overweight and obese pregnant women less than 16 weeks' gestation at recruitment, singleton pregnancy, BMI > 25, >18 years old, able to read and understand English and provide informed consent. Exclusion criteria: pregnancy at > 16 weeks' gestation at recruitment; multiple pregnancy; known pre‐existing diabetes, impaired fasting glucose or impaired glucose tolerance; GDM prior to recruitment as diagnosed by early pregnancy glucose testing, taking medications likely to influence glucose metabolism, medical conditions associated with altered glucose metabolism, known major fetal abnormality noted on 12‐week ultrasound examination; and known ingestion of probiotics via capsules or sachets.
Interventions	Once daily oral probiotic combination of Lactobacillus rhamnosus GG and Bifidobacterium animalis subsp. lactis BB‐12 (Chr. Hansen A/S, Horsholm, Denmark) at a dose of >1 x 10^9 CFUs each per day or placebo (microcrystalline cellulose and dextrose anhydrate capsules –Chr. Hansen). Intervention or placebo will occur from enrollment prior to 16 weeks' gestation until birth.
Outcomes	Primary outcome: frequency of diagnosis of GDM at 28 weeks' gestation by a 75 g oral glucose tolerance test (OGTT). Secondary outcomes Maternal: gestational weight gain, pre‐eclampsia, induction of labor, cesarean delivery, change in gut microbiome, change in lipids and inflammatory profile, change in dietary indices and physical activity levels between baseline and 28 weeks Neonatal: body composition, anthropometry, preterm delivery, shoulder dystocia, hypoglycemia, treatment with supplemental fluids or feeds, nerve palsy, admission to the NICU, jaundice requiring phototherapy, bone fracture, neonatal death or stillbirth.
Starting date	November 2012
Contact information	Prof Leonie Callaway Level 9  Health Sciences Building  Royal Brisbane and Women's Hospital  Buttereld Street  HERSTON QLD 4029  Country Australia  Phone +61 7 3346 5273  Email l.callaway@uq.edu.au
Notes	ACTRN12611001208998 DOI:10.1186/1471‐2393‐13‐50

Trial name or title	The Probiotics in Pregnancy Study (PiP Study): rationale and design of a double blind randomized controlled trial to improve maternal health during pregnancy and prevent infant eczema and allergy
Methods	Two‐center, randomized, double‐blind, placebo‐controlled trial
Participants	Inclusion criteria: English‐speaking pregnant women who intend to breast feed their infant are eligible to be enrolled in the study between 14 and 16 weeks' gestation if either they or the unborn child’s biological father have a history of asthma or eczema treated by a doctor, or allergic rhinitis treated by a doctor or pharmacist. Exclusion criteria: under age 16, does not intend to stay in either of the study centers for the 18 months following enrollment, has a serious immunological disorder that suppresses immune function or is taking immune suppressant drugs, has known cardiac valve disease for which antibiotic prophylaxis is required when undergoing dental procedures, has a history of a transplant or human immunodeficiency virus, were on long‐term of continuous antibiotic therapy, required in vitro fertilisation to establish the current pregnancy, has a pre‐enrollment scan showing major fetal abnormalities, at the time of enrollment is using or intended to use probiotic drinks or supplements themselves or in their child, is participating in another randomized controlled trial, has a severe allergy to cow’s milk (as probiotic capsules may contain traces of milk), has previously participated in the study with an older child or is deemed unsuitable for study inclusion for any other medical reason. Any pregnant woman with pre‐existing type 1 or 2 diabetes is eligible for the study but is excluded  from the OGTT and gestational diabetes outcomes. Infant inclusion/exclusion: all infants born in the study are eligible for inclusion in study outcomes. In the case of multiple births, only the first‐born infant will be included in the study.
Interventions	Probiotic supplementation by Lactobacillus rhamnosus HN001 (6 × 10^9 CFU/day) administered to mothers from 14 to 16 weeks' gestation until delivery and continuing until 6 months postpartum, if breastfeeding, versus placebo capsules with con‐derived maltodextrin with an identical appearance.
Outcomes	Primary Outcomes Infant eczema Infant eczema severity Infant atopic sensitization Secondary Outcomes Maternal outcomes: maternal gestational diabetes mellitus, maternal BV, maternal vaginal GBS colonization, maternal postpartum depression and anxiety, gestation at delivery, maternal anthropometric measures at birth, and 6 and 12 months postpartum, caesarean section, fasting lipids and bile acids from fasting maternal blood taken at time of OGTT, gut microbiota and faecal SCFA from maternal faecal samples taken in second trimester, maternal postpartum depression and anxiety, maternal bacteremia/septicemia, hospitalizations, gastrointestinal symptoms. Infant related outcomes: anthropometric measures at birth, and 6 and 12 months, NICU admission, PROM, preterm birth (< 37 weeks), infant GBS, bacteremia/septicemia, hospitalizations (excluding birth admissions), 5‐minute Apgar scores, infant symptoms
Starting date	Not stated
Contact information	Christine Barthow
Notes	Wellington and Auckland, New Zealand DOI 10.1186/s12884‐016‐0923‐y Australian New Zealand Clinical Trials Registration: ACTRN12612000196842. Date Registered: 15/02/12.

Trial name or title	A randomized controlled demonstration trial of multifaceted nutritional intervention and or probiotics: the healthy mums and babies (HUMBA) trial
Methods	Two by two factorial randomized controlled demonstration trial
Participants	Inclusion criteria: women with a singleton pregnancy, BMI ≥ 30 kg/m², between 12 and 17 6/7 weeks of gestation and able to provide informed written consent. Exclusion criteria: maternal pre‐existing diabetes or HbA1c at booking ≥ 50 mmol/mol, taking probiotic supplements, known congenital abnormality, medications or medical conditions which alter glucose metabolism, multiple pregnancy, bariatric surgery, and severe hyperemesis.
Interventions	Oral probiotic capsule consisting of Lactobacillus rhamnosus GG and Bifidobacterium lactis BB12 at a dose of 7 × 10^9 CFUs per day each, or placebo from enrollment between 12 to 17 6/7 weeks until birth. Dietary intervention versus routine dietary advice There are four groups: dietary intervention with probiotic capsule, dietary intervention with placebo capsules, routine dietary advice with probiotic capsule and routine dietary advice with placebo capsules.
Outcomes	Primary outcomes Maternal: excessive gestational weight gain between enrollment and 36 weeks' gestation Infant: infant birth weight Secondary outcomes Maternal: maternal pregnancy glucose metabolism as assessed by OGTT parameters at 26 to 28 weeks and HbA1c at 28 and 36 weeks and 5 months postpartum, changes in diet quality and dietary patterns between recruitment, 28 to 30 weeks and 5 months postpartum, functional health and well‐being at 36 weeks and 5 months postpartum, depression and anxiety scores at 36 weeks’ and 5 months postpartum, maternal adiposity at 5 months postpartum; GDM, pregnancy‐induced hypertension (pre‐eclampsia and gestational hypertension), mode of birth, blood lipid concentrations at 28 to 30 and 36 weeks’ gestation and 5 months postpartum, maternal feedback about participation in the study Infant: neonatal anthropometry: head circumference and length, and associated Z‐scores, birth weight adjusted for length, girths (chest, arm and abdominal) adjusted for length, subscapular, triceps and supra iliac skin fold thickness adjusted for length, and arm muscle area adjusted for arm length, neonatal body composition (via PEA POD®) including fat mass and lean mass adjusted for length and fat mass adjusted for lean mass, gestational age at birth, LGA, small for gestational age (SGA, < 10th centile), admission to neonatal care unit (and reason), neonatal composite morbidity, including birth trauma (fracture, brachial plexus injury, cephalohematoma, subgaleal hematoma), hypoxic ischemic encephalopathy, sepsis, respiratory distress requiring continuous positive airway pressure support, hypoglycemia requiring dextrose treatment (buccal or intravenous), Initiation and establishment of breast feeding, including feeding in first two postnatal weeks (collected by phone call at 6 weeks and questionnaires), infant anthropometry and body composition at 5 months of age, infant feeding over first 5 months (breast feeding, formula use, complementary feeding with solids), feeding behavior as assessed by BEBQ scores, infant nutritional intake at 5 months Other secondary outcomes include: attendance at study visits, adherence to probiotic/placebo regimen, cost‐effectiveness of the intervention
Starting date	Unknown
Contact information	Lesley M. E. McCowan l.mccowan@auckland.ac.nz Department of Obstetrics and Gynaecology, Faculty of Medical and Health  Science, University of Auckland, Private Bag 92019, Auckland, New Zealand
Notes	DOI 10.1186/s12884‐016‐1149‐8 ACTRN12615000400561

Trial name or title	Effect of oral probiotic supplementation during pregnancy and lactation in modulating immune response in breast milk, fetus and neonate
Methods	Randomized, parallel group, placebo‐controlled trial
Participants	Pregnant and lactating women starting at 34 weeks' gestation who will exclusively breast feed for 14 weeks.
Interventions	Probiotic capsule contains a mix of 8 different strains of freeze‐dried lactic acid bacteria and bifidobacteria; each capsule containing not less than 112.5 billion CFU. Study Drug will be given orally 3 times a day starting from 34 weeks of gestation till 14 weeks postpartum. Placebo identical looking in size, color and appearance. The dose will be 3 times a day orally starting from 34 weeks of gestation till 14 weeks' postpartum.
Outcomes	Primary outcomes: proportion of mothers who will have detectable levels of immune markers in mature breast milk including cytokines‐IFN‐γ, IL‐6, IL‐10 antimicrobial agents‐ Total IgA, Lactoferrin and growth factors –IGF‐1. Secondary outcomes: proportion of fetus and neonates who will have detectable levels of immune markers in cord blood, including cytokines ‐IFN‐γ, IL‐6, IL‐10, and growth factors ‐ IGF‐1. Proportion of neonates with protective levels of antibody against Hepatitis B surface antigen (anti‐HbsAg) 6 weeks after vaccination with hepatitis B vaccination.
Starting date	12/18/2014
Contact information	Dr Neena Malhotra Department of Obstetrics & Gynaecology All India Institute of Medical Sciences, New Delhi‐110029  South 011‐26593361  malhotraneena@yahoo.com
Notes

PERMALINK

Maternal probiotic supplementation for prevention of morbidity and mortality in preterm infants

Jacquelyn Grev

Marie Berg

Roger Soll

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Summary of findings

Background

Description of the condition

Description of the intervention

How the intervention might work

Why it is important to do this review

Objectives

Comparisons

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Comparison 1: probiotics administered to pregnant women versus placebo or no intervention

Comparison 2: probiotics administered exclusively after birth in mothers of preterm infants versus maternal placebo or no intervention

Comparison 3: probiotics administered exclusively after birth in mothers of preterm infants versus neonatal probiotic administration

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

For infants born < 37 weeks

Maternal secondary outcomes

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

Data extraction and management

Assessment of risk of bias in included studies

Measures of treatment effect

Unit of analysis issues

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Quality of the evidence

Subgroup analysis and investigation of heterogeneity

Subgroups for analysis

Sensitivity analysis

Results

Description of studies

1.

Results of the search

2.

3.

Included studies

Comparison 1: Probiotics administered to pregnant women (< 36 weeks' gestation) versus placebo or no intervention

Before delivery (probiotics given only prior to delivery)

Review of individual studies (based on study aims)

Treatment of maternal gestational diabetes

Treatment of maternal obesity

Inflammation/lipids

Prevention of mastitis

Before and after delivery (probiotics given both prior to and after delivery)

Review of individual studies (based on study aims)

Multiple aims

Atopy

Nutrition

Comparison 2: probiotics administered exclusively after birth in mothers of preterm infants < 37 weeks' gestation versus maternal placebo or no intervention

Comparison 3: probiotics administered exclusively after birth in mothers of preterm infants < 37 weeks' gestation versus neonatal probiotic administration

Excluded studies

Risk of bias in included studies

Allocation

Blinding

Incomplete outcome data

Selective reporting

Other potential sources of bias

Trial name or title	To study the transit of bifidobacteria from mother to infant: a randomized controlled trial
Methods	Single‐centre, double‐blind, interventional, randomized controlled trial
Participants	Pregnant and postpartum women
Interventions	Bifidobacteria probiotic capsule or daily placebo capsule from 16 weeks of pregnancy until the baby is 3 months old
Outcomes	Primary outcome measures: presence of the supplemented bifidobacterial strain in the infant’s stool in the 1st week postpartum, at 1 month postpartum and at 3 months postpartum. To measure this, the infant's stool will be analyzed for the presence of the supplemented bifidobacterial strain. Secondary outcome measures Differences in the following between the probiotic and placebo groups: Maternal lipids Maternal insulin resistance as measured by fasting glucose and HOMA‐IR Maternal C‐reactive Peptide Maternal bloods will be taken in early (approximately 16 weeks' gestation) and late (approx. 34 weeks' gestation) pregnancy and will be analyzed for these metabolic parameters.
Starting date	December 2015
Contact information	Prof. Fionnuala McAuliffe The National Maternity Hospital (Ireland) UCD Department of Obstetrics & Gynaecology  65‐66 Lower Mount Street  Dublin 2
Notes

Trial name or title	Randomized, double‐blind, placebo‐controlled trial of probiotics in pregnancy (PIP) and its effect on Group‐B streptococcal colonization‐ study protocol.
Methods	Single‐centre, double‐blinded, randomized controlled trial
Participants	Inclusion criteria: pregnant women who are 24 weeks' pregnant will be enrolled
Interventions	Probiotics or placebo starting at 24 weeks for 12 weeks.
Outcomes	Primary outcome: GBS colonization at 35 to 37 weeks' gestation
Starting date	Unknown
Contact information	Mehta S  Fiona Stanley Hospital, Perth, WA, Australia Email: shailender.mehta@health.wa.gov.au
Notes	Sample size of 460 is expected Funded by Telethon

Trial name or title	Effect of probiotic Lactobacilli on vaginal flora of pregnant women at high risk for preterm delivery
Methods	Double‐blind, randomized, placebo‐controlled trial
Participants	One hundred and sixty (160) women at high risk for PTL, based on a prior history of preterm birth, will be approached at their first antenatal visit to participate. Recruitment of 54 patients with symptomatic or asymptomatic BV (based on Nugent Scoring). Inclusion criteria: pregnant women with previous incidence of preterm labour, otherwise healthy, over 18 years of age, able to provide informed consent, less than or equal to 16 weeks' gestation, singleton pregnancy, and normal uterine cavity. Exclusion criteria: significant medical complications (pre‐eclampsia, thrombophilia, hypertension), multiple pregnancy, less than 18 years of age, patients receiving antibiotics or other antimicrobial therapies at time of recruitment, fetal complications such as intrauterine growth restriction or other abnormalities, diabetes, documented need for cervical cerclage, patient enrolled in other clinical trials
Interventions	Women with documented BV will be randomized to either treatment with lactobacilli preparation (n = 27) or placebo (n = 27). Women with symptomatic BV will be treated with oral metronidazole prior to starting the lactobacilli or placebo.
Outcomes	Vaginal flora and cytokine profile and the incidence of preterm labour.
Starting date	February 2005
Contact information	Alan Bocking, MD, FRCSC Chief, Department of Obstetrics and Gynaecology, Mount Sinai Hospital Mount Sinai Hospital  Toronto, Ontario, Canada, M5G 1X5
Notes	NCT00217308 This study has suspended participant recruitment. (Tightly defined inclusion criteria were making recruitment very slow.)

Trial name or title	Probiotics (Lactobacillus rhamnosus) in reducing glucose intolerance during and after pregnancy (GRIP)
Methods	Double‐blind, placebo‐controlled, randomized trial
Participants	Inclusion criteria: pregnant women who are 35 years old or older, have a family history of diabetes, or are overweight (BMI > 23) and have a singleton pregnancy with delivery planned at the study hospital at 12 to 14 weeks of gestation Exclusion criteria: history of gestational diabetes, known diabetes mellitus, known chronic diseases (hypothyroidism ,cardiac, renal, rheumatoid arthritis, carcinoma), women maintained on medications such as: corticosteroids, azathioprin, antiepileptic epileptic drugs, known Polycystic Ovarian Syndrome, and non‐residents of Karachi
Interventions	Lactobacillus rhamnosus 1 x 10^10 CFUs orally once daily or placebo from 12 to 14 weeks of gestation until delivery
Outcomes	Glucose intolerance at 24 to 28 weeks of pregnancy and at 6 to 8 weeks' postpartum. Feasibility‐ process of recruitment rate, recruitment rate and reasons for non‐participation Compliance at 36 weeks including maternal side effects Maternal safety: maternal mortality, weight gain, pre‐eclampsia, mode of delivery, induction of labor Fetal/neonatal safety from time of birth until 6 to 8 weeks of life: death, still births, preterm birth, birth trauma, macrosomia, SGA, polyhydramnios, hypoglycemia, large for gestational age, shoulder dystocia, 5‐minute Apgar < 7, hyperbilirubinemia, respiratory distress
Starting date	October 2011
Contact information	Bilal Ahmed, Senior Instructor Research, Aga Khan University Karachi‐Pakistan
Notes	NCT01436448 The recruitment status of this study is unknown. The completion date has passed and the status has not been verified in more than two years.

Trial name or title	Probiotics and early microbial contact in preterm neonates (ProPre)
Methods	Randomized, placebo‐controlled, double‐blind, controlled trial
Participants	Inclusion criteria Premature infant born at < 35 weeks' gestational age Exclusion Criteria Severe asphyxia Significant anomalies
Interventions	Probiotics Lactobacillus rhamnosus GG 10E9 CFU / day (LGG) Lactobacillus rhamnosus GG and Bifidobacterium lactis Bb‐12 administered 10E9 CFU/ day each (LGG+Bb‐12) Placebo: Microcrystalline cellulose Group 1: lactating mother gets placebo, infant gets LGG Group 2: lactating mother gets placebo, infant gets placebo Group 3: lactating mother gets LGG, infant gets placebo Group 4: lactating mother gets LGG+Bb‐12, infant gets placebo Group 5: lactating mother gets placebo, infant gets LGG+Bb‐12
Outcomes	Primary outcomes: assessment of indigenous intestinal microbiota composition in premature neonates during the first month of life, neonatal intestinal immune gene expression from fecal samples during the first month of life, breast milk composition at one month. Secondary outcomes: intestinal immunity, breast milk composition
Starting date	April 2014
Contact information	Samuli Rautava, Turku University Hospital Collaborators: University of Turku Massachusetts General Hospital
Notes	The recruitment status of this study is unknown. The completion date has passed and the status has not been verified in more than two years.

Trial name or title	Effects of oral probiotic supplementation on Group B Strep (GBS) rectovaginal colonization in pregnancy
Methods	Double‐blind, randomized, placebo‐controlled trial
Participants	Inclusion criteria: pregnant women between 20 to 28 weeks' gestation who are 18 years of age or older and have a singleton gestation. Exclusion criteria Pre‐existing morbidity: Immunocompromised status (HIV +; malignancy; history of organ transplant; chronic steroid therapy; autoimmune disease requiring treatment during pregnancy, and other immunocompromised states); type 1 diabetes and type 2 diabetes;congenital cardiac disease and cardiac valvular disease requiring antibiotic prophylaxis during procedure/labor; pulmonary disease (except mild asthma); renal disease; chronic hepatic disease (Hepatitis B, C); inflammatory bowel disease (Crohn's disease or ulcerative colitis); stomach or duodenal ulcer; bowel resection, gastric bypass, and chronic indwelling venous, bladder, or gastric catheter. Multifetal gestation. Use of probiotics preparations in the 3 months prior to beginning of the study treatment or use of any additional probiotics preparations (other than study treatment) at any time during the study period (including over the counter food supplements such as Activia, BioK, other oral or vaginal probiotics products (BUT not including other common forms of yogurt). Chronic (daily) use of broad spectrum antibiotics. History of infant with GBS sepsis. IUGR, fetal anomalies, major, diagnosed at time of second trimester anatomy ultrasound Anticipated delivery < 35 weeks for maternal/fetal indication. Placenta previa or accreta (with anticipated delivery prior to 35 weeks).
Interventions	Beginning at 20 to 28 weeks' gestation until delivery
Outcomes	Primary outcome measure Group B Streptococcus rectovaginal colonization at 35 to 37 weeks' gestational age. Secondary outcome measures (assessed at 6 weeks' postpartum) Maternal antepartum, intrapartum, and postpartum outcomes: urinary tract infections, chorioamnionitis, endometritis, cellulitis, bacteremia, sepsis, and other infectious morbidity. Neonatal outcomes: gestational age at delivery, Apgar scores, bilirubin levels, C‐reactive protein, rule out sepsis evaluation, sepsis, pneumonia, meningitis, NICU admission, and length of hospital stay
Starting date	November 2011
Contact information	Natali Aziz, MD; 6507248222; naziz@stanford.edu Aptos Women's Health Center  Aptos, California, USA, 95003 Contact: Karen Erlich khemidwife@earthlink.net Dominican Hospital  Santa Cruz, California, USA, 95065 Contact: Karen Erlich khemidwife@earthlink.net Stanford University School of Medicine/Lucile Packard Children's Hospital  Stanford, California, USA, 94305 Contact: Anna Girsen, PhD; 6507255720; agirsen@stanford.edu
Notes	This study is currently recruiting participants NCT01479478 Estimated Enrollment: 372 Estimated Study Completion Date: November 2019

Trial name or title	Oral probiotics Lactobacillus rhamnosus GR‐1 and Lactobacillus reuteri RC‐14 reduce Group B Streptococci colonization in pregnant women
Methods	Prospective, double‐blind, randomized clinical trial
Participants	Pregnant women with positive GBS screening culture at 35 to 37 weeks of gestation with a singleton gestation.
Interventions	U‐relax oral capsule, Lactobacillus GR‐1 and RC‐14 or Placebo daily for 14 days
Outcomes	Primary outcome measures: number of GBS‐positive pregnant women who became GBS‐negative at childbirth (time frame: 2 weeks after taking probiotic)
Starting date	April 2011
Contact information	Ming Ho, MD China Medical University Hospital Taichung, Taiwan, 403
Notes

Trial name or title	Probiotic therapy for the reversal of bacterial vaginosis in pregnancy (ProVIP)
Methods	Randomized controlled trial
Participants	Pregnant women who test positive for bacterial vaginosis prior to 17 weeks' gestation
Interventions	Probiotic lactobacilli vs. placebo for 12 weeks
Outcomes	Primary outcome measures: microbial DNA profile at 28 and at 35 weeks' gestation; a change in the vaginal microbial DNA profile.    Secondary outcome measures: microbial function at 28 and at 35 weeks' gestation; a change in microbial function as measured by RNA transcriptomics.
Starting date	May 2012
Contact information	Alan Bocking, Mount Sinai Hospital Mount Sinai Hospital Toronto, Ontario, Canada
Notes

Trial name or title	Probiotics capsule as supplemental therapy for Group B Streptococci infection and vaginitis during pregnancy
Methods	Double‐blind, randomized, placebo‐controlled trial
Participants	Pregnant women at 35 weeks' gestation
Interventions	Lactobacillus + cranberry vs. cranberry vs. placebo
Outcomes	Quote: "all cases vaginal group B streptococcus colonies"
Starting date	June 2012
Contact information	Meng‐Hsing Wu, M.D., Ph.D. Department of Obstetrics and Gynecology, College of Medicine, National Cheng Kung University, Tainan, Taiwan +88662353535 ext 5222 mhwu68@mail.ncku.edu.tw
Notes	The recruitment status of this study is unknown. The completion date has passed and the status has not been verified in more than two years.

Trial name or title	Nutrition and pregnancy intervention study
Methods	Randomized double blind placebo‐controlled trial with 2 interventions and 4 arms
Participants	Inclusion Criteria: pregnant, less than 16 weeks' gestation, overweight, healthy Exclusion Criteria: diabetes (type 1 or 2), celiac disease, Increased bleeding tendency
Interventions	The intervention will involve consumption of fish oil and/or probiotic capsules from early pregnancy until 6 months after delivery. Lactobacillus and Bifidobacterium  Fish oil  Placebo: microcrystalline cellulose  Placebo: medium chain triglycerides
Outcomes	Primary outcome measures Prevalence of GDM, GDM at gestational weeks 24 to 28 Fasting glucose levels at the third trimester of pregnancy Prevalence of allergy in child at 12 and 24 months of age Secondary outcome measures Need for medication for management of GDM (insulin or metformin) during pregnancy Body composition of mother during and after pregnancy Immunologic and metabolic markers during and after pregnancy Fecal microbiota before, during and after intervention Other outcome measures: Body composition, growth, development and metabolic markers of the child 0 to 24 months of age
Starting date	September 2013
Contact information	Kirsi Laitinen Adjunct professor + 358 02 333 6063 kirsi.laitinen@utu.fi Turku University Hospital Turku, Finland, 20521 University of Turku Turku, Finland, 20521
Notes	Estimated Enrollment: 440  Estimated Study Completion Date: December 2019  Estimated Primary Completion Date: March 2017 ClinicalTrials.gov Identifier: NCT01922791