role, enabling researchers to generate testable hypotheses for those larger studies. In addition, well-designed and small sample studies can contribute to building a mechanistic understanding of clinical observations obtained in larger studies.
- With respect to fecal sampling, participants expressed concern that making inferences about what is happening inside the gut based on what is detected in feces can be dangerous given that the microbiome is a dynamic, complex system that is highly individual and easily perturbed. Jeremy Nicholson said, “For me, it is like trying to sniff an exhaust pipe of a Ferrari and tell you what color [the car] is. You have this very complex ecology which you have compressed into a piece of feces…. I think we need to develop new technologies to be able to study the microbes in situ and what they are doing locally.” Again, however, given its noninvasive nature, fecal sampling has been the only choice in many of these early studies.
- There were some calls for more mechanistic research. Even when sequencing data are complemented with functional annotation, purported functions are just that—purported. They still need to be validated with mechanistic study—thus, the importance of animal models or even non-animal models.
- Some workshop participants also called for more longitudinal studies as a way to examine causality. Much of what is being learned about diet-microbiome-health relationships is correlational, not causational (e.g., that a particular microbial strain or microbial metabolite is associated with a disease risk, but with no clear understanding of which came first).
Dietary interventions intended to have an impact on host biology via their impact on the microbiome are being developed, and the market for those products is seeing tremendous success. However, the current regulatory framework threatens to slow industry interest and investment.
- Much of this early research on the microbiome focuses on associations between the microbiome and disease, not health, and most dietary interventions intended to have an impact on host biology via their influence on the microbiome (e.g., probiotics) are being studied for their potential to prevent disease, not promote health. However, current regulatory constraints on food claims prohibit communicating to consumers many of the effects that studies focused on disease prevention demonstrate. Some workshop participants noted the challenges and value of conducting more studies in healthy populations versus changing the regulatory landscape to accommodate the science.