Thread by @JothamSuez

There is a considerable knowledge gap between what we think is "known" regarding probiotics and where the science actually stands. Our new review in @NatureMedicine attempts to bring order into a chaotic field (1/9)
@Elinav_Lab @EranElinav @segal_eran nature.com/articles/s4159…

Clinical efficacy: existing, but not as robust as we think for some of the most common conditions associated with probiotics, including gastroenteritis and C. difficile associated diarrhea. Meta-analyses should not replace high quality bias-free RCTs (2/9)

Can heterogeneity in efficacy relate to heterogeneity in colonization? data from our group and others point to inter-individual difference in both long-term colonization and even during supplementation. Why is this important? (3/9)

Mechanistically (check out my beloved figure 2): some probiotic mechanisms of action require approximation and adhesion to the intestinal mucosa (unsurprising given what we know of host-microbiome interactions), though others may not. No colonization = no clinical effect? (4/9)

Precision (figure 1): this needs to be further studied. Some evidence point to differential activity of probiotics relating to both host and microbiome factors, including no/reduced effect when probiotics don't colonize. (5/9)

Effect on the microbiome: may not be "a thing". Little evidence to support an effect on the microbiome without prior perturbation, no conclusive signature. No evidence that any clinical outcome may be related to an effect on the microbiome. What about after antibiotics? (6/9)

While extremely popular, evidence supporting facilitated post-antibiotics recovery of the microbiome with probiotics are weak and limited by technical biases. We and others demonstrated that the opposite (delayed recovery) may be true. Is this clinically relevant? (7/9)

Safety: the long-term effects of microbiome inhibition by probiotics are unknown and should be studied. While safe in healthy individuals, probiotics may result in serious complications in multiple sub-populations and this should not be taken lightly. (8/9)

Future: more RCTs free from industry bias to address safety & efficacy, use same tools available in microbiome research to study mechanisms, address importance of colonization, precision aspects, increase pool of available therapeutic strains, tailor strain to condition (9/9)