Incredible claims of some probiotic manufacturers are sometimes not backed up by incredible evidence. It is particularly a problem when such products are targeted at vulnerable groups. This includes children with autism. A recent review says it is now time to run proper trials on the efficacy of probiotics in autistic children. Let’s take a look.

Max’s Note: In the interests of full disclosure, approximately 10 years ago I was involved in some research about gut microbes and autism. This was during my PhD at The University of Reading. We had some success. Here is the paper.

Incredible Claims

Selling probiotics to treat medical conditions is a big, controversial business. The claims (particularly the ones plastered on the internet) are often not backed up by evidence, the prices can be wild and if swallowing a capsule containing “250 billion cfu of 25 scientifically proven probiotic bacteria and 88 clinically proven herbal extracts” is enough to replace actual drugs that do work, then this business can even be dangerous.

Where it gets very interesting is when medical science fails to provide answers on causes and treatments for diseases. This is exactly the position we find ourselves in with autism. Effective treatments are few and far between which means ‘alternative’ therapies are everywhere. One ‘treatment’ being flogged to rightfully desperate parents of autistic kids is probiotics. Gastrointestinal complaints are commonly reported in autism and probiotics are then often recommended (often by other parents or so-called experts). Assuming that the product is safe, this probably won’t do much harm except perhaps act as a self imposed taxation.

Scientific Consensus?

There is however a growing scientific consensus that there might, might be something in this approach. Writing in the open access journal “Gastroenterology Research and Practice” a group of scientists from the US, The Netherlands and UK now suggest it’s time to run clinical studies to validate whether probiotics can help with gastrointestinal symptoms in autistic children. They suggest that the prevalence of these symptoms in autism and the point that a significant minority of US physicians recommend giving probiotics to autistic children really now warrants proper investigation.

Reviewing the evidence

Setting out the evidence J. William Critchfield and team suggest that there are now four strands of evidence on which to base clinical studies. The first is the similarity between gastrointestinal symptoms in autism and IBS and that in certain circumstances certain strains of bifidobacteria have shown promise in clinical trials for IBS symptoms. Next, there is growing evidence (although there would appear to be more work needed here) that gut microbiota composition could be altered in autistic children and that probiotics might be useful here. Finding evidence of widespread issues with the way the gut actually functions in autistic children is a controversial topic (think vaccines and autism). Wisely, they suggest that proper studies should be performed to firmly establish this. Interestingly they point towards some studies suggesting that probiotics might stabilise the gut mucosa (this is the surface of the gut wall), stimulate immunity and improve gut barrier function. We should be cautious here as most of these studies are in animals or in a dish.

Behaviour Bugs

Finally, the review raises a significant point about gut microbes and behaviour. It suggests that compounds from the diet and from cell walls of bacteria can cross from the gut into the central nervous system and finally the brain… affecting behaviour. This is a story that has been around for a while. What is significant now is that in the past few years studies have shown that administration of probiotics can affect various parameters of behaviour in both rats and humans. Specifically we learn of a study revealing a positive effect of a probiotic drink on mood and cognition. Whilst it is unclear how this works, this has to be of some significance to autism. The concluding comments of the authors says it all:

“There is a range of indications that alterations in the intestinal microbiota in the gut might contribute to the disorder in a substantial number of individuals. Probiotics can be useful to restore the microbial balance in the intestine, to relieve gastrointestinal problems and to attenuate immunological abnormalities. Whether the use of probiotics by children with autism can lead to improvements in behaviors needs to be established in well-controlled trials with sufficient group sizes. Important for these trials is the choice of the bacterial strains, as effects of probiotic bacteria can be highly strain specific.”

Now, I am not suggesting with all this that there is concrete proof that probiotics have a proven role to play in helping autistic children and their symptoms. This is all about specific strands of evidence that point towards a possible role in the future. That is why we need clinical studies.

Trials?

The obvious thing to shout now is “Max… surely you can tell us where the clinical studies are!” Sorry to disappoint you, but there are none registered in any clinical trial database. The only strand of hope I can bring you is that two of the authors are from Winclove Bio Industries in Amsterdam who make probiotics for various applications. I have contacted them to ask for more details and I will update the post if I hear back. I can guarantee you that if clinical studies are planned we will not be seeing any proven probiotic product for gastrointestinal symptoms in autism for sale any time soon. Trials take time.

There is also one more intriguing point to all this. In 2006, Prof Glenn Gibson of The University of Reading reported that a trial of probiotics with autistic was so successful that the trial ultimately failed. Dropouts from the study were so high that no meaningful conclusion could be drawn. Now, when you hear about high dropout rates in studies we should normally get worried as it could mean patients are getting worse because of the study. Not in this case. The reports that emerged suggested that parents were withdrawing their children because the effects were so good in terms of mood and behaviour. They realised that their child was on the active treatment and then refused to cross over to the comparison placebo. Through no fault of the scientists, the trial became unblinded as a result of its success. The only option… another study. And, that has never happened to date. You can read the study report here… or at least the abstract on page 3.

Reference

Critchfield JW et al (2011). The potential role of probiotics in the management of childhood autism spectrum disorders. Gastroenterology Research and Practice 2011: 161358. DOI: 10.1155/2011/161358

A recently published review has again highlighted the progress being made in unravelling a possible relationship between gut microbes and obesity. Although controversial, there is a suggestion that specific species and groups of bacteria might be associated with being obese and that they might prove worthy targets for intervention (possibly via probiotics or prebiotics).

This begs a question – where are the clinical trials? We take a look at some studies that are ongoing or remain unpublished in this area.

First things first. Obesity is a complex topic. Apart from the obvious issue of body weight, obesity is also associated with numerous other complications including type II diabetes and systemic inflammation. It is a major concern as it has the potential to prematurely kill a lot of people, rates are on the rise and it is probably largely preventable.

Gut microbes have been in the picture for a while following a series of experiments showing that they are likely to be involved in fat storage, nutrient availability, and the low grade inflammation associated with obesity. If you want to get the full picture I point you directly to the review by Nathalie Delzenne and company.

Their conclusion? Numerous potential mechanisms and variations in gut microbe composition might be involved although it is too early to safely conclude which bacteria are involved. Also, any role for probiotics and prebiotics in the management of obesity and associated conditions still needs to be evaluated further. The few clinical studies that have been completed have turned up mixed results.

So, let’s have a look at the studies that are currently ongoing to see what we might learn in the coming years.

Weighty question

Let’s start with probiotics and the most obvious target – body weight. That’s the aim of this study, led by Dr Angelo Tremblay at Laval University, Canada. According to the trial registration:

“The primary objective of the clinical trial is to investigate whether a consumption of a probiotic formula when compared to an intake of a placebo, is associated with a greater reduction of body weight.”

Given that the study involves Nestlé (or at least they are mentioned as a “collaborator”), there is clearly some sort of an agenda of potential health claims here. There is no mention of which specific strain of probiotic is being used (although we’re guessing it might be one of the ones (here or here) used in the animal studies that preceded the study) or what might count as a clinically relevant weight reduction. Nevertheless, the study is fairly large so it should give us a good idea of whether this approach has a future. It will certainly be interesting to see what comes out of it. The end date for the study is December 2011 so we should expect reporting towards the end of next year, if not 2013.

A question of risk

Next, Denmark and we move on from weight management to obesity related metabolic problems and satiety. The University of Copenhagen has four studies of interest here (two on metabolic issues and two on satiety, see below). The first study involves feeding a probiotic or flax seed fibre notionally to alter the composition of gut microbes, then to look at the effects on a variety of markers related to insulin resistance (risk factor for heart disease), inflammation, body composition (i.e. amount of fat) and then try to find correlations with the changes in the composition of gut microbes. Unravelling whether the changes are a cause or consequence of each other will be hard and it’s a fair bet that the work will generate a hell of a discussion. Nevertheless, we should expect to see this appear well into 2013 at the earliest.

The next study, again at The University of Copenhagen, and this time supported by the Danish Medical Research Council and Danisco, has the following hypothesis:

“This study will examine if overweight in children is associated with a different intestinal microbiota and if a change in microbiota caused by probiotics can modify inflammation and risk factors for the metabolic syndrome.”

According to the registration, the study was due for completion in January 2011 although, as of November 2011, it appears to not have been published. Given the number of analyses described this is perhaps not a surprise. It will certainly be interesting if a positive result comes from this. For the moment though, we just have to quietly wait.

Staying with modifying gut microbe composition and risk factors for metabolic syndrome, there is a similar study currently running at the German Diabetes Center looking at probiotic supplementation and insulin resistance. It’s due for completion in December 2011 and we should expect these findings to be reported sometime next year.

Meanwhile, the microbe used in Yakult is under investigation at the Medical University of Graz, Austria in a small study looking at its effects on various markers associated with obesity and metabolic syndrome. With a completion date of December 2010 we should expect publication of the results of this study hopefully soon.

Yet another study, this time running in Chile, reports it is looking at the effects of a combination of probiotics and prebiotics on low grade gut inflammation in people with obesity. Again, with an expected completion date of December 2010, we might get to hear about those results sometime soon.

Feeling full

Obesity is often associated with over-eating. So, if it’s possible to modify when we feel full after a meal (satiety) then this could be of interest in tackling obesity. That’s exactly the aim of the ProSat study at The University of Copenhagen (again) which is, or was, all about feeding probiotic capsules to see if appetite fell. It would appear that it has taken place in two stages with this part being completed in 2009 and this part (the bigger one) apparently being completed June 2011. It is not entirely clear what the status of these studies are and they do not appear to be reported yet. Again we will just have to wait and see. It will be an interesting result if there is a positive effect. How they will explain the mechanism will be very interesting indeed.

…and finally

Most of the studies ongoing are about probiotics being used to alter gut microbe composition and induce a variety of effects related to obesity. We are certain that with the eventual publication of these studies we will learn more about whether any of this is worth following. It is perhaps surprising that prebiotics have not featured more in the ongoing studies. The evidence for effects in animal models suggests it might be time to try it all out in humans. Well, some studies have been published with mixed results. And… there is one study looking at a prebiotic in over-weight adults and risk factors for metabolic syndrome. It completed in December 2010 and it would not surprise me if we hear about that study fairly soon.

A large clinical study looking at the prevention of fever episodes by prebiotics (specific ingredients to help selected ‘healthy’ microbes grow in your gut) in infant formula has turned up a negative result. The addition of prebiotics made no difference to the number of episodes of kids getting sick. The obvious thing to do right now would be to ask some searching questions of the infant formula companies that continue to claim that their products will prevent your children getting sick.

We are not going to go there though. National and European Authorities have already had their substantial say about that sort of thing. No, this study points to a different future, informing us just how difficult it is to prove that a food is ‘functional’ and just why we should expect to see very, very few health claims cruising the aisles of the local supermarket in the next few years. Here’s the analysis.

Fever Episodes

Clinical studies tend to be worth taking note of when they have study populations that are big. This is because they tend to give answers that are rather more believable about effects that are in fact quite small. This is the case with a recently published study that looked at the effect of prebiotics added to infant formula on fever episodes at five different study centres from across Europe. Involving well over 1000 infants, we learn that the number of episodes of fever did not differ between infants receiving formula with prebiotics (specifically in this case a blend called Immunofortis, which appears in numerous brands owned by Nutricia/ Danone), and infants receiving formula with nothing added. Comparing this to infants receiving exclusive breast milk, the study also reveals that breast-feeding resulted in a very, very small higher risk of a fever. Note this though >

If you are breast-feeding, please do not stop based on that result. The evidence for the benefits of breast feeding for your child is absolutely overwhelming. Small deviations from the scientific body of evidence do happen and it’s the whole lot of evidence you should take note of… not just one result (which incidentally has some specific weaknesses). 

Healthy Infants

Now, where this study becomes very interesting is when the scientists start to discuss the meaning of the study. Led by Margriet van Stuijvenberg of UMC Groningen, The Netherlands, the team reveal in their analysis a number of intriguing points. For example, they note that their study was performed with healthy infants. This contrasts with the previous studies they quote that involved infants at increased risk of respiratory infections (and that found an effect). The study also revealed a very low rate of fever episodes which they suggest could mean that the study was under-powered.

When we talk about the ‘power’ of a study, that refers, statistically speaking, to the chances that the study actually had to detect the effect it was looking for. A hat-tip to the authors for suggesting this is really deserved. It’s worth looking at why they think this. According to van Stuijvenberg (and company):

“… if much more participants would be necessary to find an effect, the clinical relevance would be very small.”

You can probably draw your own conclusions on that statement. They go on. Finding a very small difference in favour of prebiotic intake after the first 6 months, they then rip into the finding both statistically and noting that it is probably clinically irrelevant. Then, looking at the variation between study centres, they note differences in fever episodes and specifically that breast feeding at some centres seemed to result in higher rates (please, please see my note above). Even more intriguingly they suggest that if your kid is white or happens to be a boy, you might expect a higher number of fever episodes. Happily, if you have a furry pet in the house, you might expect a lower rate of fever episodes. All that appears to be irrelevant of whether prebiotics are given or not.

Do take some care with these observations. I have previously picked over the difference between correlation and causation. I suspect this might be true here too.

What all this highlights is that running a study in the wild (i.e. making it as realistic as possible) also makes finding effects of an intervention, in this case with infant formula containing prebiotics, very hard to pin down. The real difficulty is that, at least in Europe, the regulatory authority, EFSA, has started to ask for proof that a product making a claim works in a general population. This study was performed in a general population and as far as I can tell, was performed very well.

Detecting effects that are small in healthy populations is very hard. Making health claims that are based on such evidence and that stick in the mind of the consumer is really very, very hard. As of next year (2012), it will be illegal in Europe to make health claims that are not substantiated and approved.

A final thought

Where it will get interesting is when the general population becomes an unhealthy population. I’m looking at the US and obesity… and there is some really fascinating work going on about that little issue… microbes are probably involved.

Probiotic yoghurts are popular mainly because they come labelled with claims that they are good for you. Whether this is true or not is the focus of a considerable number of studies that are ongoing at the moment. A small study, published yesterday, suggests that the impact of these yoghurt cultures on the composition of bacteria already present in our gut is negligible, they do not replace them, but that they might induce microbial activities involved in digestion. Let’s discuss.

Max’s Note: Before I get into this rather complicated little story, I have to note that the research paper is behind a paywall at Science Translational Medicine. If you want to read it you will need to hand over $15 or find a friend with access. For that reason I am going to keep this fairly short and explain just the topline implications of the research.

In one short sentence: Go and read this article by Ed Yong in Nature News about the research. There is no point in me re-inventing the news wheel for you.  Go, go…

Whistles…

Ok, seen any issues yet? Let’s dissect this carefully. The major result indicates that the composition of gut microbiota does not change very much when probiotic bacteria are ingested but that their activity might be altered. That is a very interesting finding.

The actual mechanisms by which probiotics affect health have actually remained a bit of a mystery for many years despite considerable research efforts. Various hypotheses exist (i.e. changing composition to a more “beneficial” mix, competitive inhibition of pathogens, modulation of innate immunity and the list go on). So, this result might add just that bit more to our understanding of the function of gut microbes and the properties of probiotics.

Next, the design of the study is worth looking at in more detail as it could open the door to more significant discoveries later on. The authors report they used identical twins (in the test and control groups) which would remove confounding factors from the analysis such as human genetic variation. This makes the result more robust. Next, verifying the result in mice with identical genetic makeup and very defined microbial communities in their guts under tightly controlled laboratory conditions adds more weight to the evidence. Then, using some very advanced molecular approaches (variously described under the monikers of microbiomics and more specifically metatranscriptomics, should you care to know) to unravel what are very complex communities of bacteria and their metabolic pathways and come to the conclusion they have is an impressive technical achievement. The finding would have been very difficult to verify without these checks, controls and techniques.

But, what of the result? What are the implications? Naturally there has been some caution towards the result, including from the lead author and other scientists and these all seem to be fair. The study is small and appears to have some technical limitations (given the comments of Dusko Ehrlich in Yong’s piece). What is interesting is the impact that this approach might have on future studies and on the perennial difficulty of manufacturers of probiotic products to prove their claims. Verifying effects seen in humans in animal models is on the face of it a bit backwards, but in this case it seems appropriate because of the super complexity of the systems involved. We can expect to see studies like this appear again in the future.

It is interesting, but not surprising, that this work has been partly funded by Danone, who are one of the largest manufacturers of probiotic products. They would naturally have an interest in any new approaches or technologies that help boost the strength of evidence behind probiotic bacteria. What is unclear is whether such evidence and approaches will be sufficiently robust to meet the strict requirements of regulatory authorities such as EFSA (who have recently been very critical of the quality of evidence used to support probiotic claims in general). We will have to wait and see on that tricky little area.