Gut microbiome study maps two opposing lifestyle axes for health and disease

A Nature Medicine paper published 5 November 2026 reports that the gut microbiome may mediate links between everyday lifestyle and both general health and multiple chronic diseases.

Study type and data

• Type: large cross-sectional multi-omics analysis with mediation modelling (observational; not a randomised lifestyle trial), plus limited longitudinal and external validation.
• Discovery cohort: fecal shotgun metagenomes from 6,803 adults in Japan’s 4D project (mean age about 60.9; physician-confirmed diagnoses).
• Outcomes included general health (n = 2,119 by strict criteria) and 15 chronic diseases with ≥100 cases each.
• Subsets had fecal short-chain fatty acids (n = 1,713) and plasma cytokines (n = 1,740). Prediction scores (MADE) were checked in 13 non-Japanese cohorts (n = 3,095).

Main findings (authors’ framing)

Researchers describe two opposing microbiome-mediated axes:

• a health-oriented axis tied to plant-based, soy and dairy intake and to butyrate-linked, less inflammatory host profiles;
• a disease-oriented axis enriched in oral-derived bacteria and pathobionts, linked to refined carbohydrates and sedentary behaviour.

Starter-culture microbes from fermented foods were detected after eating those foods but did not show statistically significant mediation of the observed health effects. Alcohol showed mixed health- and disease-oriented mediation patterns. MADE scores separated health vs disease in the discovery set and in most external cohorts.

Limitations

Cross-sectional mediation cannot prove that changing diet or activity will change disease risk through the microbiome; reverse causation and residual confounding remain possible. Diet came from food-frequency questionnaires; discovery was Japanese, so generalisability needs care despite external checks. This is not personal medical advice — talk to your doctor about diet or gut-health concerns.