probiotics in irritable bowel syndrome

Both irritable bowel syndrome (IBS) and chronic idiopathic constipation (CIC) are very common (prevalence 5-20%) and without great medical therapies: "no therapy has been proven to alter the natural history of either condition in the long term". hence, the attraction of probiotics, which might alter the GI microbiome fundamentally. A recent meta-analysis looked at the use of probiotics (live or attenuated microorganisms that might be beneficial), prebiotics (food ingredients that remain undigested which might stimulate growth of beneficial microorganisms), and synbiotics (combinations of prebiotics and probiotics) -- see irritable bowel probiotics AJG 2014​ in dropbox, or doi:10.1038/ajg.2014.202​. they reviewed 43 articles. there were 35 randomized controlled trials (RCTs) of probiotics, 2 of synbiotics and 1 of prebiotics for IBS.  for CIC there were only 3 of probiotics, 2 of synbiotics, and 1 of prebiotics. results:

--probiotics for IBS (35 RCTs): 3452 patients, 14 studies at low risk of bias, 19 used combinations of probiotics, and 8 with only lactobacillus, 3 with bifidobacterium, 2 e.coli, one each of streptococcus, saccharomyces.  results:
    --23 RCTs comparing probiotics to placebo (2575 patients). 777 (55.8%) on probiotics NOT helped vs 865 (73.1%) on placebo -- ie, there was more improvement in those on probiotics (RR 0.79 (0.70-0.89), yielding NNT (number-needed-to-treat) of 7. no difference in efficacy of probiotics if look at the 12 RCTs with a low risk of bias. combination probiotics assessed in 12 trials (1197 patients) -- also with significant effect on symptoms (RR for no improvement of 0.82 (0.69-0.98) and NNT of 8. for single probiotics: (many with small numbers of patients), lactobacillus (422 patients) with nonsignificant RR of 0.75 (0.54-1.04). of those, the 3 trials with L. plantarum DSM 9843 (314 patients) did have a significant symptomatic improvement, where RR for no improvement was 0.67 (0.51-0.87). the 2 trials with e. coli (n=418 patients) found significant benefit only with e. coli DSM 17252, with RR 0.86 (0.79-0.93). S. faecium (1 trial, 54 patients) also superior to placebo with RR of 0.72 (0.53-0.99).
    --24 trials (2001 patients) looked specifically at the effects of probiotics on global IBS scores or abdominal pain. Lactobacillus (6 trials, 420 patients) and Bifidobacterium (3 trials, 501 patients) were not better than placebo (the Bifidobacterium did just reach significance at p=0.05, but clinically not very impressive). no benefit for L plantarum DSM 9843. for combination probiotics (15 trials, 1038 patients), there was a significant improvement in IBS symptom score, though they were unable to determine if any of the specific combinations tested were effective.  In terms of individual symptom scores -- bloating scores and flatulence  were reduced with probiotics, with no effect on symptoms of urgency.
    --adverse effects of probiotics: found in 16.5% of 1215 patients on probiotics vs 13.8% of 1192 patients on placebo, RR=1.21 (1.02-1.44), yielding a number-needed-to-harm of 35.
--data on synbiotics and prebiotics suffered from low numbers of patients and significant heterogeneity of results, so difficult to make informed decision about them.
--for CIC: small number of patients with prebiotics, poor quality trial, with no positive effects. for probiotics: again, small number of patients, high in heterogeneity, but there was a significant increase in mean number of stools per week with RR=1.49 (1.02-1.96)

so, what does this all mean?  the biggest issue to me is that the healthy intestinal microbiome is comprised of lots of beneficial bacteria, that there may well be a synergy between these different bacteria, that trying to isolate the few bacteria that are likely to be beneficial is too reductionist an approach, and the most important way to develop a healthy microbiome is to eat well, exercise and minimize antibiotic exposure. it is certainly a complex task to figure out scientifically what is the best combination of potentially beneficial bacteria. that being said, there does seem to be some benefit from probiotics for IBS and possibly CIC, especially with lactobacillus plantarum DSM 9843.  There was another double blind study of 214 patients with IBS given this bacterium alone and finding after 4 weeks that abdominal pain severity and frequency as well as bloating improved, with 78.1% considering the symptomatic effect as "excellent or good" vs 8.1% with placebo (see irritable bowel L plantarum WJG 2012 in dropbox, or World J Gastroenterol. Aug 14, 2012; 18(30): 4012–4018).  As an interesting aside, there does seem to be a genetic component to the microbiome above and beyond that associated with food, exercise, medications, etc. Lots of studies have linked differences in the microbiome composition to obesity. But a recent study looked at the complex interplay of host genetics and the microbiome. This was a twin study in the UK and found that there were abundant microbial taxa influenced by host genetics. If the microbiome of obese people was then transplanted into germ-free mice​, the mice gained weight. but if they also added​ some bacteria associated with a low BMI (specifically, Christensenellaceae), there was much less weight gain in the mice (see microbiome and genetics Cell 2014 or doi.org/10.1016/j.cell.2014.09.053). so, though there seems to be some microbiome determination by genetics, the specific content of that microbiome seems to be alterable and potentially lead to changes in weight, for example.