Machine learning identifies microbiome and clinical predictors of sustained weight loss following prolonged fasting

Abstract Background Prolonged fasting may improve metabolic health, but controlled data in healthy adults with longer follow-up and multi-omics profiling are limited. We investigated the immediate and 12-week follow-up effects of a 5-day fasting intervention on body composition, gut microbiome, and circulating and fecal metabolites, and assessed whether baseline characteristics predict individual weight-loss response. Methods In a randomized, waitlist-controlled trial, 38 healthy adults completed a 5-day fasting intervention with 12-week follow-up (LEANER study). Outcomes included body mass index and body composition, gut microbiome composition, and plasma and fecal metabolites. Changes over time and between groups were evaluated using regression-based models and paired non-parametric tests, as appropriate. Additionally, permutation-based multivariate testing was performed on microbiome and metabolome data. Twelve-week body weight response was predicted using data-driven machine learning with cross-validation, followed by external validation in three independent cohorts undergoing prolonged fasting protocols. Results Fasting reduced body mass index acutely, predominantly driven by loss of fat mass, and these improvements partially persisted at 12 weeks. Fasting induced marked shifts in gut microbiome composition and in plasma and fecal metabolites. Post-fasting and longer-term changes in microbial diversity were associated with baseline microbiome diversity. A model combining baseline microbiome and clinical variables predicted body mass index response at 12 weeks; prominent predictors included an unclassified Faecalibacterium species, Oscillibacter sp. 50_27, low-density lipoprotein cholesterol, and systolic blood pressure. The model generalized to three independent cohorts, including individuals with metabolic syndrome, patients with multiple sclerosis exposed to repeated fasting, and healthy volunteers fasting for 6–12 days. Conclusions In healthy adults, a 5-day prolonged fasting intervention produces robust short-term metabolic changes with partial persistence and consistent remodeling of the gut microbiome and metabolite profiles. Baseline microbiome and clinical characteristics can help stratify expected longer-term responses, supporting the development of individualized fasting-based interventions. Trial registration ClinicalTrials.gov, NCT04452916. Prospectively registered on June 29, 2020.

Authors

Publication Details

Journal
Genome Medicine
Published
2026-09-16
DOI
https://doi.org/10.1186/s13073-026-01765-0
Primary Topic
Dietary Effects on Health
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
article

Machine learning identifies microbiome and clinical predictors of sustained weight loss following prolonged fasting

Wolfram Gronwald, Nicola Wilck, Friederike Gutmann, Alexander Krannich et al.
Genome Medicine
Dietary Effects on Health
article

Machine learning identifies microbiome and clinical predictors of sustained weight loss following prolonged fasting

Wolfram Gronwald, Nicola Wilck, Friederike Gutmann, Alexander Krannich, Sylvia Bähring, Alma Zernecke, Till Schütte, Georg Zeller, Robin Mesnage, Friedemann Paul, Gabriele Rahn, Ulrike Löber, Nadja Siebert, Sakshi Kamboj, Theda U.P. Bartolomaeus, Matanat Mammadli, Quinten R. Ducarmon, Sofia K. Forslund-Startceva, Victoria McParland, Lina S. Bahr, Peter J. Oefner, Dominik N. Müller, Kristin Schütte, Lena Braun, Hendrik Bartolomaeus, Lajos Markó, Gelsomina Kaufhold
article en

Abstract

Abstract Background Prolonged fasting may improve metabolic health, but controlled data in healthy adults with longer follow-up and multi-omics profiling are limited. We investigated the immediate and 12-week follow-up effects of a 5-day fasting intervention on body composition, gut microbiome, and circulating and fecal metabolites, and assessed whether baseline characteristics predict individual weight-loss response. Methods In a randomized, waitlist-controlled trial, 38 healthy adults completed a 5-day fasting intervention with 12-week follow-up (LEANER study). Outcomes included body mass index and body composition, gut microbiome composition, and plasma and fecal metabolites. Changes over time and between groups were evaluated using regression-based models and paired non-parametric tests, as appropriate. Additionally, permutation-based multivariate testing was performed on microbiome and metabolome data. Twelve-week body weight response was predicted using data-driven machine learning with cross-validation, followed by external validation in three independent cohorts undergoing prolonged fasting protocols. Results Fasting reduced body mass index acutely, predominantly driven by loss of fat mass, and these improvements partially persisted at 12 weeks. Fasting induced marked shifts in gut microbiome composition and in plasma and fecal metabolites. Post-fasting and longer-term changes in microbial diversity were associated with baseline microbiome diversity. A model combining baseline microbiome and clinical variables predicted body mass index response at 12 weeks; prominent predictors included an unclassified Faecalibacterium species, Oscillibacter sp. 50_27, low-density lipoprotein cholesterol, and systolic blood pressure. The model generalized to three independent cohorts, including individuals with metabolic syndrome, patients with multiple sclerosis exposed to repeated fasting, and healthy volunteers fasting for 6–12 days. Conclusions In healthy adults, a 5-day prolonged fasting intervention produces robust short-term metabolic changes with partial persistence and consistent remodeling of the gut microbiome and metabolite profiles. Baseline microbiome and clinical characteristics can help stratify expected longer-term responses, supporting the development of individualized fasting-based interventions. Trial registration ClinicalTrials.gov, NCT04452916. Prospectively registered on June 29, 2020.

Genome Medicine
Openalex Percentile: Top 11%
Dietary Effects on Health
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.