Personalized responses to diet

Personalized nutrition, personalized medicine, personalized probiotics. These are all trendy concepts, which now include a microbiome component, with exciting objectives but limited practical applications. Now add responses of your microbiome to your diet to this list of what can be personalized.

Briefly, this new study conducted a longitudinal assessment of daily dietary records and daily shotgun DNA sequencing of microbes from feces of 34 healthy subjects for 17 days. The complex and voluminous data presented in this paper are way beyond my ability to critically review. But I still wanted to share some interesting conclusions from the study.

  • Dietary diversity did not correlate with microbiota composition diversity but did correlate with microbiota stability. Perhaps it is not so important to health what microbes we harbor – many healthy people harbor very different microbes – but that our microbial communities can support our health by being resilience to stressors.
  • A given food may lead to different microbiota responses in different people. Within an individual, the study found that daily changes in foods predicted daily changes in relative abundance of specific species, but such a predictive model for one person did not extend to others. This observation provides a rationale for individualized response to probiotics.
  • A stable diet did not lead to microbiota stability. Two subjects consumed meal-replacement liquid shakes, almost exclusively for the 17 days. Results showed that although the diet did not vary, the microbiota did. The stability of the microbiota of these 2 subjects was not distinct from the other subjects in the study. Researchers speculated that microbiota stability may depend more on components of the microbial community rather than diet: they identified microbial species that were associated with stability.
  • Bacterial function (as predicted by DNA sequences) did not correlate with diet.
  • Current food databases have significant limitations. When interested in understanding potential interactions with our microbiota, food databases do not include minor but potentially biologically active compounds (e.g., phytochemicals) in foods, nonnutritive components of food such as preservatives (e.g., nitrates), additives (e.g., emulsifiers), and are not sufficiently granular (e.g., “fiber” is treated as a single substance in the database, yet it is not a single substance, chemically or functionally). The authors propose a novel tree-based analytical approach for grouping foods as complex admixtures of similar components.

Studies such as this are focused on understanding how our microbiota responds. No health endpoints were tracked. Further, all results come from fecal samples, which have some limitations. We can speculate that dietary approaches that encourage microbiota homeostasis will encourage health. Long-standing dietary recommendations that have encouraged eating a diversity of wholesome, minimally processed foods may accomplish this.