Machine Learning Identifies Microbiome and Clinical Predictors of Sustained Weight Loss Following Prolonged Fasting

Aug 5, 2026

Prolonged fasting may support metabolic health, yet evidence in healthy individuals is limited. A randomized, waitlist-controlled LEANER study enrolled 38 participants who completed a 5-day fasting intervention with 12-week follow-up. Fasting acutely lowered BMI via fat mass reduction and partially sustained effects afterward. It reshaped gut microbiome composition and shifted plasma and fecal metabolites, while baseline microbiome features predicted long-term BMI response using machine learning.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages47

Reading Time1 h 58 min

AccessFree to download

Abstract

Introduction

medRxiv preprint doi: [https://doi.org/10.1101/2025.06.26.25330331](https://doi.org/10.1101/2025.06.26.25330331); this version posted June 26, 2025. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. It is made available under a CC-BY-NC-ND 4.0 International license . 1 Machine Learning Identifies Microbiome and Clinical Predictors of Sustained Weight Loss Following 2 Prolonged Fasting 3 4 Gelsomina N. Kaufhold1,2,3,*, Theda U. P. Bartolomaeus1,2,3,4, *, Kristin Kräker1,2,3,4, Till Schütte1,2,3,4, 5 Sakshi Kamboj5, Ulrike Löber1,2,3,4, Gabriele Rahn1, Victoria McParland1,3,6, Lena Braun1, 6 Lajos Markó1,2,3,4, Matanat Mammadli1,2,3, Alexander Krannich7, Lina S. Bahr1,2,3, 7 Friederike Gutmann1,2,3,4, Friedemann Paul1,2,3, Nicola Wilck1,3,4,6, Alma Zernecke8, Peter J. Oefner5, 8 Wolfram Gronwald5, Dominik N. Müller1,2,3,4, Sofia K. Forslund-Startceva1,2,3,4,9, Sylvia Bähring1,2,3,†, 9 Hendrik Bartolomaeus3,4,8,†, Nadja Siebert1,2,3,† 10 11 1 Experimental and Clinical Research Center (ECRC), a cooperation of Charité - Universitätsmedizin 12 Berlin and Max Delbrück Center for Molecular Medicine, Berlin, Germany 13 2Charité-Universitätsmedizin Berlin, corporate member of Freie Universität Berlin and Humboldt- 14 Universität zu Berlin, Berlin, Germany. 15 3 Max Delbrück Center for Molecular Medicine in the Helmholtz Association (MDC), Berlin, Germany 16 4 DZHK (German Centre for Cardiovascular Research), partner site Berlin, Germany 17 5 Institute of Functional Genomics, University of Regensburg, Regensburg, Germany 18 6 Department of Nephrology and Internal Intensive Care Medicine, Charité –Universitätsmedizin 19 Berlin, Germany. 20 7 BioStats GmbH, Nauen, Germany 21 8 Institute of Experimental Biomedicine, University Hospital Würzburg, Germany. 22 9Structural and Computational Biology Unit, European Molecular Biology Laboratory (EMBL), 23 Heidelberg, Germany. 24 25 *contributed equally 26 †jointly supervised the work 27 Correspondence: bartolomae [h@ukw.de](h@ukw.de) NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice. medRxiv preprint doi: [https://doi.org/10.1101/2025.06.26.25330331](https://doi.org/10.1101/2025.06.26.25330331); this version posted June 26, 2025. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. It is made available under a CC-BY-NC-ND 4.0 International license . 28 Abstract 29 Prolonged fasting may benefit metabolic health, but data in healthy individuals remain 30 limited. We performed a randomized, waitlist-controlled study (LEANER study), with 38 31 healthy participants completing a 5-day-fasting intervention with 12-week follow-up. Fasting 32 acutely lowered body mass index (BMI), via fat mass loss. These changes partially persisted 33 at follow-up. Fasting altered the gut microbiome composition and induced metabolite shifts 34 in plasma and feces. Changes to gut microbiome alpha diversity after fasting correlated with 35 baseline microbiome diversity. Long-term BMI response at follow-up could be predicted 36 through machine learning (ML) using baseline microbiome and clinical data, highlighting an 37 unknown Faecalibacterium sp., Oscillibacter sp. 50_ 27, LDL cholesterol, and systolic blood 38 pressure as key predictors. This ML model was validated in independent patient cohorts with 39 metabolic syndrome and multiple sclerosis. These findings support prolonged fasting as an 40 effective metabolic intervention and demonstrate that individual responses to fasting 41 interventions can be predicted using pre-intervention features. 42 43 Trial [registration: ClinicalTrials.gov](registration: ClinicalTri

FAQs

What study design evaluated the fasting intervention?

The LEANER study used a randomized, waitlist-controlled design with 38 healthy participants completing a 5-day fasting intervention and a 12-week follow-up.

How did prolonged fasting affect BMI and body composition?

Fasting acutely lowered BMI through fat mass loss, and some of these changes partially persisted at follow-up.

What predictors were identified for sustained weight loss response?

A machine learning model using baseline microbiome and clinical data highlighted Faecalibacterium sp. and Oscillibacter sp. 50_, LDL cholesterol, and systolic blood pressure as key predictors.

Aug 5, 2026

Upload by Valentina