Numerous articles have questioned the effectiveness of probiotics for many of the conditions for which they have been promoted. A recent systematic review (T Newlover-Delgado, et al.
The authors discuss some of the limitations such as variations in definitions, choice of probiotic and dosage, and short-term duration. There is not a discussion of selection bias. It is quite possible that some negative studies were completed which were not published which could further lower or eliminate the potential benefit.
My take: Probiotics may be helpful for children with recurrent abdominal pain; it is certainly not a panacea.
Related blog posts:
Moon over Zabriskie Point, Death Valley -just before sunrise
Almost sunrise at Zabriskie Point, Death Valley
Two recent papers in Cell provide additional questions about the effectiveness of probiotics.
The coverage of these studies in the media has created some controversy; enough so that the International Scientific Association of Probiotics and Prebiotics disseminated a very critical review:
CLINICAL EVIDENCE AND NOT MICROBIOTA OUTCOMES DRIVE VALUE OF PROBIOTICS
Here is a small excerpt:
Two recent papers have generated much adverse publicity for the probiotic field. Headlines driven by sensationalism, not data, claim “Probiotics labelled ‘quite useless’” (BBC) and “Probiotics ‘not as beneficial for gut health as previously thought’” (The Guardian). The quotes are from author Eran Elinav, who generalizes the study findings to all ‘probiotics’ as a class – a generalization that ignores that specific probiotic are meant for specific purposes…
The scope of these papers is limited to microbiome data; no clinical endpoints are assessed. Without clinical evidence, it is not possible to conclude about the tested probiotic’s usefulness, and it is certainly not possible to conclude about probiotic usefulness in general…. The authors discount the existing body of evidence for probiotic health benefits, including Level 1 placebo-controlled, randomized trials. Cochrane reviews (the gold standard used by physicians and public health policy makers) of the totality of evidence show that specific probiotics can prevent antibiotic associated diarrhea (AAD) and C. difficile diarrhea. This evidence has been translated into evidence-based recommendations for probiotics issued by medical groups. Regardless of an effect on the microbiota, these are established, evidence-based benefits of probiotics.
My take: This controversy points to the problem that probiotics are often considered more effective than the science merits. While there are some conditions that may respond to probiotics, it should be understood that each probiotic needs to be looked at for each specific clinical scenario.
Related blog posts:
A recent retrospective study (AF Kane et al. J Pediatr 2018; 195: 73-9) with 640 VLBW infants found that the probiobiotic, Lactobacillus rhamnosus GG (LGG), was associated with an increased risk of necrotizing enterocolitis (NEC).
LGG supplementation was started at a median age of 6 days at a dose of 2.5 to 5 x 10 to the 9th CFU/day.
- LGG group had an aOR of 2.10 for developing NEC. LGG group NEC incidence was 16.8% whereas NEC incidence was 10.2% prior to institution of LGG.
The authors note their findings are in contrast to findings from 38 randomized trials (10,520) which have found that probiotics lowered the risk of NEC.
My take: This study reinforces the need for further studies to identify which factors and probiotic strains are likely to lead to reduced rates of NEC.
Related blog posts:
A recent review (I Hojsak et al. JPGN 2018; 66: 3-9) examined published trials regarding the role of probiotics in the prevention of nosocomial diarrhea. The review was conducted by a working group on behalf of ESPGHAN.
- “Recommendation: If probiotics for preventing nosocomial diarrhea in children are considered, the WG [working group] recommends using L rhamnosus GG (at least 10 to the 9th CFU/day, for the duration of hospital stay).
- Quality of evidence: Moderate
- Strength of recommendation: Strong
- Number needed to treat (in order for beneficial effect in one): 12 patients
The authors do not recommend L reuteri DSM17938 due to lack of efficacy; other probiotics did not receive a recommendation either due to lack of data or lack of efficacy.
It is possible that there have been unpublished negative probiotic studies which would alter the calculation of a beneficial effect.
My take: While the working group recommends L rhamnosus GG if probiotics are used to prevent diarrhea, the absolute benefit is low.
Related blog posts:
Bright Angel Trail, Grand Canyon
There is some debate about whether colic is truly a GI disorder. A recent commentary (V Sung, MD Cabana. J Pediatr 2017; 191: 6-8) provides some insight.
- “‘Colic’ is a term coined by the ancient Greeks…derived from ‘kolikos,’ meaning crampy pain, sharing its root with the the word colon.”
- “Since 1994, there have been at least a dozen case-control studies that have indicated differences in the gut microbiota between infants with and without colic.”
- Studies have had conflicting results with whether calprotectin levels are increased in infants with colic compared with controls.
- Among probiotics, L reuteri DSM17938 “is the best studied strain.” Despite several studies suggesting efficacy, “the largest and only double-blind randomized trial that included both breastfed and formula-fed infants with colic (n=167) in Australia was ineffective.
- The commentary reviews a recent study (Fatheree NY et al. J Pediatr 2017; 191: 170-8) “although very small in comparison, adds to this literature, being the second double-blind randomized, placebo-controlled trial of L reuteri DSM17938 shown to be ineffective in breastfed infants with colic.” Sample size =20. “It is the first to document increased fecal calprotectin levels that decrease with reduced crying” …though this “may be reflections of normal levels in healthy young infants, which change over time.” In addition, this study did not find evidence of systemic inflammation. The authors speculate that the frequent use of antireflux medications could dampen the effects of probiotics.
My take: We still do not know whether efforts at changing an infant’s microbiome improve clinical outcomes in colic.
Related blog posts:
Near Bright Angel Trail, Grand Canyon
The enthusiasm for probiotics is generally greater than expected based on the data available to support their use for many indications. This has been discussed several times on this blog (see below). The reasons why probiotics are sometimes not effective can be related to being poorly regulated/lack rigorous production standards; even in conditions in which there is some effectiveness (eg. antibiotic-associated diarrhea [AAD]), the number of persons needed to treat for one person to benefit is fairly high. Furthermore, as a recent study (A Olek et al. J Pediatr 2017; 186: 82-6) shows, even in conditions like AAD in which probiotics have proven efficacy, the effects may be strain-specific and/or dose-related.
Olek et al showed that Lactobacillus planatarum DSM9843 (LP299V) was NOT beneficial compared to placebo in reducing the incidence of loose/watery stools or mean number of stools among 438 children receiving outpatient antibiotic therapy.
Specifics: This was a prospective, double-blind, randomized, placebo-controlled parallel-group study. The treatment group received LP299V during antibiotic therapy and for 1 week afterwards. In addition to monitoring the number of stools, the authors determined the frequency of AAD which they defined according to WHO guidelines (>3 loose/watery stools/24 hours after initiation of antibiotics). In this study, AAD was confined to study duration rather than over 2 months.
- Overall, 44.5% of children developed loose/watery stools among placebo group and 39% among probiotic group
- 4.1% developed AAD among placebo group and 2.8% among LP299V
- LP299V showed no significant beneficial effects in reducing AAD or loose/watery stools
The authors note that LP299V has been effective in studies involving adult hospitalized patients. They question whether healthy children, therefore, may be less likely to benefit from probiotics and whether a higher dose could have been more effective.
My take: “Data from clinical studies on probiotics are conflicting” for many conditions, including antibiotic-associated diarrhea.
Lovers Leap, near Ashville (Appalachian Trail)
Related blog posts:
In a recent study (O Jadresin et al. JPGN 2017; 64: 925-9), 55 children with functional abdominal pain or irritable bowel syndrome were randomized (prospective, double-blind, placebo-controlled study) to either L reuteri DSM or placebo.
- The intervention group had more days without pain: median 89.5 days vs. 51 days (P=.029)
- Abdominal pain was less severe in the intervention group at some time points (second month, and fourth month)
- The two groups did not differ with regard to duration of abdominal pain, stool type, or absence from school
Limitation: Small number of patients -the estimated samples size was not reached
My take: This study suggests that probiotics may help some pediatric patients with irritable bowel syndrome. Trying to identify which patients should receive a probiotic and which probiotic should be selected remains unclear.
Related blog posts:
Pierre Renoir’s portrait of Claude Monet, Musee d’Orsay
Families are often surprised to learn my opinion about probiotics. The “truth” about probiotics is that they are poorly regulated/lack rigorous production standards and are mostly ineffective for many of the conditions for which they have been promoted. Even in conditions in which there is some effectiveness (eg. antibiotic-associated diarrhea), the number of persons needed to treat for one person to benefit is fairly high.
In addition, when someone says that they are taking a probiotic, many families do not understand the idea of “strain” specific effects. I tell families that if they see a “dog in yard” sign that they do not know if that is a poodle of a pit bull. With probiotics, similarly you often do not know if you are getting a pit bull or a poodle.
As a consequence, I think negative studies like a recent report (K Wojtyniak et al. J Pediatr 2017; 184: 101-05) are helpful. In this study, the authors examined the effectiveness of Lactobacillus casei rhamnosus Lcr35 (Lcr35) in the management of constipation.
This randomized, double-blind, placebo-controlled trial was conducted in 94 children <5 years of age. Dose: 8 x10 to the 8th CFU twice daily x 4 weeks.
- “Lcr35 as a sole treatment was not more effective than placebo in the management of functional constipation.” In fact, the placebo group had a greater increase in bowel movement frequency than the treatment group.
- Both groups had improvement -more than half in each group (total 52 of 81 who completed study) had reached endpoint of 3 or more BMs/week without soiling.
My take: Probiotics often are ineffective. This study showed that Lcr35 was NOT helpful for pediatric constipation.
Related blog posts:
Claude Monet, La Rue Montorgueil
Probiotics and alterations in the microbiome are being examined for a range of ailments. However, as noted in previous blog posts, the current evidence shows only a limited number of disorders where probiotics have been proven effective. There is more evidence, now, that probiotics may be beneficial for nonalcoholic fatty liver disease (NAFLD).
- Am J Clin Nutr 2014; 99: 425-6. editorial
- Am J Clin Nutr 2014; 99: 535-42.
The referenced article examined 52 nondiabetic patients with fatty liver disease in a double-blind, randomized, placebo-controlled trial. Patients were considered to have NAFLD on the basis of an ultrasonography and an alanine aminotransferase value >60 U/L. Those who received a probiotic were compared with a placebo group and followed for 28 weeks.
In this study, rather than a probiotic, technically, the treatment group received a synbiotic because it contained fructooligosaccharides (FOS) which are non digestible oligosaccharides in addition to a probiotic mixture. FOS can stimulate the growth of intestinal bacteria. The probiotic mixture included Lactobacillus case, Lactobacillus rhamnosus, Streptococcus thermopiles, Bificobacterium breve, Lactobacillus, acidophilus, B. longum, and Lactobacillus bulgaricus.
- There were improvements in ALT values and in baseline mild fibrosis (estimated by Fibroscan).
- There were decreased levels of circulating TNF-α and decreased nuclear transcription factor κβ in circulating mononuclear leukocytes –both consistent with decreased systemic inflammation
- Study did not include liver histology (biopsy). In addition, in nearly all subjects, the fibroscans were near normal, both before and after the intervention. Thus, the reduction in liver stiffness is not clear cut.
- Small number of participants.
- Short study period.
Bottomline: This study along with several others points towards a potential role for modulating the microbiome to improve NAFLD along with metabolic syndrome more broadly.
Related blog posts:
A new study identifies a potential microbiome signature that is associated with colic (Pediatrics 2013; 131: e550-58). Thanks to Mike Hart for this reference.
With new technology, the microbiome’s role in many gastrointestinal conditions is being unraveled. For colic, there has been concern about that an abnormal microbiome has been contributing to the pathophysiology. On a personal level, I have had an interest in this subject for quite a while:
In this current study, the authors serially followed the microbiome infants in a prospective longitudinal project. Nine stool specimens were obtained from each infant. Four were obtained in the first month at day of life 2, 7, 14, and 28 days. Five were obtained at 3 to 5 months. All specimens were stored at -20°C until analyzed by the “Human Intestinal Tract Chip (HITChip).” The HITChip microarray was considered satisfactory if 2 independent hybridizations had >95% correlation. These microarrays “showed a dynamic range of >10,000-fold and >200 independent microarray readouts were used.”
Of 160 healthy term infants, the authors identified colic in 25% who averaged >180 minutes of crying per day over a four-day period. Then, the authors selected the 12 infants who cried the most and compared them to the 12 infants who cried the least.
- In the infants who were highly similar, the “infants with colic showed a significantly reduced microbiota diversity at 14 and 28 days of life.” Proteobacteria, including Enterobacter aerogenes, Escherichia coli, and Klebsiella pneumoniae, were increased with more than a doubled abundance. Bifidobacteria and lactobacilli were decreased in colicy infants.
- At about 3-4 months of age, the colic group had a similar microbiome as the non-colicy group.
- The authors speculate that proteobacteria might cause inflammation and may displace helpful bacteria. Certain butyrate-producing bacteria like Butyrivibrio crossotus and Coprococcus estates were more commonly present in the non-colicy group. The authors note that butyrate reduces the pain sensation in adults.
These results could explain why administration of probiotics (and possibly antibiotics) can result in a decrease in colic symptoms.
Related blog entries:
- -Pediatrics 2010; 126: e526. Double-blind randomized placebo-controlled trial of Lactobacillus reuteri.
- -J Pediatr 2009; 155:823. Increased calprotectin in colicy infants. n=36. editorial pg 772.
- -J Pediatr 2009; 154: 514-20. Colic and reflux. (Orenstein et al), & 475 (editorial -Putnam). PPIs (lansoprazole) do not help colicy Sx in infants c GERD. n=162. Increased resp infections in pts on PPIs. 44% response in Rx & control group.
- -J Pediatr 2008; 152: 801. Probiotic helped reduce colic sx in 30 preterm infants, Lactobacillus reuteri
- -Pediatrics 2007; 119; e124. Probiotics reduced colic in breastfed babies more than simethicone. n=83, lactobacillus reuteri, 10-8th power per day. Decreased crying 18 minutes per day at 1 week compared to simethicone & by 94 minutes/day at 4 weeks (95% response vs 7% of simethicone)
- -Pediatrics 2005; 116: e709. Low-allergen maternal diet was helpful.
- -Arch Pediatr Adol Med 2002; 1183 &1172. lack of sequelae on maternal mental health.
- -Arch Pediatr Adol Med 2002; 156: 1123-1128. colic 24% of infants, breastfeeding did not help.
- -Pediatrics 2002; 109: 797-805. carbohydrate malabsorption with breath testing in colicy infants, n=30. 2 hour fasting period.
- -Arch Dis Child 2001; 84: 138-41. Lack of benefit (vs placebo) of chiropractic manipulation for colic, n=100. 86 completed study. 70% improved vs 60% in placebo.
- -JPGN 2001; 33:110-111. Lack of assoc c GER
- -Pediatrics 2000; 106: 1349. Use of hydrolysate decreased crying by 63mins/day
- -Pediatrics 2001; 108; 878-882. No assoc between colic and markers of atopy/asthma/allergy.