About

Matthew Odenwald, MD, PhD, is an expert hepatologist specializing in diagnosing and treating the full spectrum of liver diseases. He provides personalized care for conditions such as hepatitis, fatty liver disease, liver cancer, and cirrhosis, and manages patients before and after liver transplants. His work focuses on offering comprehensive care, including symptom management, slowing disease progression, and preventing complications, with a goal to tailor treatment plans to each patient's unique needs and circumstances.

Research topics

• Biology
• Bioinformatics
• Medicine
• Microbiology
• Immunology
• Internal medicine
• Biochemistry
• Chemistry
• Cell biology
• Genetics
• Gastroenterology
• Endocrinology
• Oncology
• Physiology
• Intensive care medicine

Selected publications

• P-1569. Microbiome Derived Stool Metabolites Predict E. faecium Expansion in Hospitalized Patients

  Open Forum Infectious Diseases · 2026-01-01

  articleOpen access

  Abstract Background Vancomycin resistant Enterococcus faecium (VRE) poses a distinct threat to hospitalized patients.(CDC, 2022) Expansion of VRE within the gut microbiome is closely linked with invasive infections in multiple hosts.(Lehmann et al., 2024; Taur et al., 2012) Stool microbial metabolite measurement offers a new diagnostic avenue to rapidly identify VRE colonization. (Lehmann et al., 2024) We developed a machine learning model using stool metabolite measurements to predict E. faecium expansion in patients.Table 1.Patient Characteristics.Demographic information of the patients whose samples were utilized to train and test this model.Figure 1.Overall performance of the Elastic Net Model.A. Receiver operator characteristic plot showing true positive vs false positive rates. This demonstrates an area under the curve of 0.919. B. Table displaying a variety of model performance metrics. Notably, this model demonstrates a high F1 score of 0.935 as well as positive and negative predictive values of 0.915 and 0.804 respectively. C. Confusion Matrix of model performance on unseen test data. Methods Using qualitative stool metabolite concentrations as measured by targeted GC and LC-MS analysis paired with microbiota composition obtained via shotgun metagenomic sequencing, we generated, tuned, and evaluated an elastic net-based machine learning model to predict stool expansion of E. faecium over 30% relative abundance.(Kuhn, 2008) Collinearity and multicollinearity analysis of the metabolites was performed using Spearman’s Rank-Order method, and Variance Inflation Factors respectively. Following hyperparametric tuning of this model, test data was used to evaluate the model’s performance. The selected classification cutoff for the model maximized the F1 score based on precision and recall.Figure 2.Collinearity Assessment of Stool Metabolites.A. Top 10 pairwise Spearman Correlations are listed with both respective metabolites and the absolute value of their correlation. B. All metabolites that demonstrate high multicollinearity (VIF>10) are shown with their respective variance inflation factor. Results 2738 stool samples from 1504 patients were included from 6 observational clinical studies on liver disease, liver transplant, heart transplant, critical care, internal medicine, and leukemia patients.(Dela Cruz et al., 2023; Lehmann et al., 2024; Odenwald et al., 2023; Stutz et al., 2022) (Table 1) The model predicted E. faecium expansion with an accuracy was 0.894, sensitivity of 0.949, precision of 0.919 and AUC of 0.917. (Figure 1.) Conclusion This work provides a proof of concept that stool metabolites can identify pathogen expansion in the gut. The approach could aid in early diagnosis leading to better outcomes. It could also identify key inhibitory metabolites as future therapeutic candidates. Future work will identify the metabolites with highest predictive value and apply this method to other gut pathogens such as Klebsiella pneumonia. Disclosures Bhakti Patel, MD, CHEST: Board review course director|Merck: Wrote medical chapters

  PublisherDOI

• Longitudinal clinical monitoring reveals widespread effects of medications on gut microbiome dynamics

  Zenodo (CERN European Organization for Nuclear Research) · 2026-04-03

  datasetOpen access

  Abstract Medications exert strong effects on the gut microbiome, impacting host physiology and affecting health outcomes. Studies have characterized associations between medications and the gut microbiome, primarily through in vitro analyses and, when conducted in humans, either using cross-sectional designs or focusing on specific medication classes. However, how medications directly reshape the microbiome within individual patients over time, especially in clinical settings, remains poorly understood. Here, we leverage longitudinal stool metagenomics, metabolomics, and electronic health records from 3,469 samples representing 1,122 patients at the University of Chicago Medical Center between 2020 and 2024. Using a longitudinal design where each patient serves as their own control, we compare patient microbiome composition before and after starting a medication while controlling for the effects of medical procedures, diagnoses, demographic variables, and other drugs taken simultaneously. In total, we identify 36,637 associations relating 138 of the most commonly given medications, such as oxycodone and pantoprazole, with 106 microbial genera, 204 microbial species, 627 microbial pathways, and 35 gut metabolites. We find, for example, that oral prednisone is associated with decreases in Parabacteroides and Enterobacter abundance and increases in microbial ergosterol and long-chain fatty acid biosynthesis pathways. Furthermore, the selective serotonin reuptake inhibitor sertraline is associated with increases in Alistipes abundance, increases in microbial pathways related to dopamine degradation, and increases in fecal concentration of the related metabolite kynurenic acid. Critically, we discover that medication effects are time-dependent, with microbiome recovery to baseline occurring over days to weeks. Our findings represent the first systematic characterization of medication-microbiome dynamics within individual patients in a clinical setting, establishing temporal patterns and mechanistic pathways that inform microbiome-guided therapeutic strategies. Description This repository contains contains data files and code to run the models and generate the figures in this paper, which should take about 5 minutes to run on a typical desktop computer. To generate the figures in this paper, extract the zip file and run:`python3 build_figures.py` Most data files are stored in the highly efficient Apache parquet format, which can be read in Python using the read_parquet function in pandas or in R using the read_parquet function in arrow. MRNs (medical records numbers) in data files have also been anonymized. Code Description build_figures.py Script to build figures used in this paper.paperfigures.py Contains code to build each of the figures in this paper and output them as PDFs.icd10.py Builds a table that maps ICD-10 codes to their text descriptions.main.py Runs the linear mixed models used in our analysis pipeline. pathways.py Contains functions to perform pathway enrichment analysisplots.py Helper functions for plotting.util.py Helper functions for use in the analysis pipeline. File Descriptions out/ output figures and supplementary tables produced by build_figures.py output/ Results from the linear mixed models, with columns indicating medication, feature, p-value, coefficient, significance (FDR < 0.1), estimated standard error, medication class, model duration, sample level type, model type, and whether the medication was a duration effect or not. output/results_2_to_10_days.csv Presence/absence results from our 2-10 day models. output/results_10_to_20_days.csv Presence/absence results from our 10-20 day models.output/results_20_to_30_days.csv Presence/absence results from our 20-30 day models.output/results_duration.csv Results from our duration models. data/ Intermediate data files generated from running our pipeline. data/table_demographics.parquet Anonymized demographics data with patient sex and birth year. data/table_microbes_clr.csv Data table containing CLR-transformed microbiome abundance data.data/table_pathways.parquet Data table containing microbial pathway abundances.data/table_pathways_general.parquet Data table containing microbial pathway abundances.data/metacyc_pathway_hierarchy.json Pathway hierarchy for humann3 metacyc pathways.data/hospital_visits.parquet List of all patient hospital visits, with accompanying clinical data.data/study_intervals.parquet List of all pairs of stool collections, with accompanying clinical data.data/table_microbes.csv Data table containing microbiome counts.data/table_microbes_relabund.parquet Data table containing microbiome relative abundances.data/medication_classes.parquet List of all medication classes.data/table_genus_metab_correl.parquet Correlations between microbial genera and metabolites.data/medications.parquet List of all medications included.data/samples.parquet De-identified table containing list of stool samples included in the study with collection dates.data/icd10cm-tabular-April-2024.xml ICD-10 codes reference table Dependencies Package Version python 3.10 numpy 1.23.5 pandas 1.5.3 matplotlib 3.6.3 scikit-learn 1.2.1 scikit-bio 0.5.8 seaborn 0.12.2 scipy 1.10.0 statsmodels 0.13.5 matplotlib 3.6.3 networkx 3.0

  PublisherDOI

• 76. Hospitalized Patients Have Microbiome Dysbiosis Linked to Post-Discharge Infection

  Open Forum Infectious Diseases · 2026-01-01

  articleOpen access

  Abstract Background A healthy gut microbiome offers protection against pathogenic bacteria, such as multidrug-resistant organisms (MDROs), Enterococcus and Enterobacterales. Expansion of these organisms is a harbinger for invasive infection and could contribute to the spread of MDROs. Acute care hospitals account for the majority of MDRO infections nationally. To test whether hospitalized patients develop pathogenic expansion within the microbiome, we examined the gut microbiome and metabolome of hospitalized patients and compared them to post-discharge infection. Methods In a prospective, observational study we collected stool samples from hospitalized patients on a single unit that approximates a general inpatient population. We performed shotgun metagenomic sequencing and quantitative GC and LC-MS metabolite measurement on each sample. We examined the differences in microbiota diversity and composition to healthy donors and compared discharge microbiome composition to post-discharge infection. Results A total of 71 samples were collected from 46 hospitalized patients (Table 1). Compared to 32 healthy donors, hospitalized patients had lower diversity, compositional changes and expansion of both Enterobacterales and Enterococcus in 17 (23%) and 6 (8%) samples, respectively (Figure 1). Beneficial microbe-derived short chain fatty acids and secondary bile acids were also reduced (Figure 2). Among hospitalized patients that developed post-discharge infection, the discharge microbiome had 9 (64%) of the infecting organisms (Figure 3). For infections caused by Enterobacterales, 7 (88%) were detected in the discharge microbiome. E.coli expanded &amp;gt;2.5% in 5 (71%) patients with E.coli infection. Candida was found in 2 (66%) discharge stool samples that developed a Candida infection. Conclusion Hospitalized patients have a reduction in microbiome diversity, beneficial microbial metabolites, and pathologic expansion of Enterobacterales and Enterococcus species. Expansion at the time of discharge was associated with post-discharge infection. Gut microbiome expansion in this population could be responsible for the rise of MDROs. Therapies to restore microbiome composition and function could be used to combat MDRO infections. Disclosures All Authors: No reported disclosures

  PublisherDOI

• Fecal butyrate and deoxycholic acid quantitation for rapid assessment of the gut microbiome

  PLoS ONE · 2026-01-09

  articleOpen access

  The intestinal microbiome is composed of myriad microbial species with impacts on host health that are mediated by the production of metabolites. While loss of bacterial species and beneficial metabolites from the fecal microbiome is associated with development of a range of diseases and medical complications, there are currently no clinical diagnostic tests that rapidly identify individuals with microbiome deficiencies. This method aims to rapidly quantify fecal concentrations of butyrate and deoxycholic acid, as depletion of these two metabolites are associated with adverse clinical outcomes and result from the loss of a subset of health-associated bacterial species. We present a rapid diagnostic screen based on 3-nitrophenylhydrazine derivatization and ultrahigh-performance liquid chromatography-mass spectrometry that measures fecal butyrate and deoxycholic acid concentrations as markers of microbiome function. A matrix-matched calibration curve was developed using a simulated fecal mixture to optimize accuracy and facilitate adherence to clinical laboratory regulations. The assay resulted in an analytical measurement range from 4.30-3030 µM (LLOQ = 3.71 µM) for butyrate and from 0.9-64.9 µM (LLOQ = 0.7 µM) for deoxycholic acid. Precision evaluation demonstrated a coefficient of variation <15% at all quality control levels tested. The rapid liquid chromatography-mass spectrometry screen can be performed in under an hour from extraction to provision of quantitative results, enabling the rapid identification of patients with defective microbiome function.

  PublisherDOI

• Longitudinal clinical monitoring reveals widespread effects of medications on gut microbiome dynamics

  Zenodo (CERN European Organization for Nuclear Research) · 2026-04-03

  datasetOpen access

  Abstract Medications exert strong effects on the gut microbiome, impacting host physiology and affecting health outcomes. Studies have characterized associations between medications and the gut microbiome, primarily through in vitro analyses and, when conducted in humans, either using cross-sectional designs or focusing on specific medication classes. However, how medications directly reshape the microbiome within individual patients over time, especially in clinical settings, remains poorly understood. Here, we leverage longitudinal stool metagenomics, metabolomics, and electronic health records from 3,469 samples representing 1,122 patients at the University of Chicago Medical Center between 2020 and 2024. Using a longitudinal design where each patient serves as their own control, we compare patient microbiome composition before and after starting a medication while controlling for the effects of medical procedures, diagnoses, demographic variables, and other drugs taken simultaneously. In total, we identify 36,637 associations relating 138 of the most commonly given medications, such as oxycodone and pantoprazole, with 106 microbial genera, 204 microbial species, 627 microbial pathways, and 35 gut metabolites. We find, for example, that oral prednisone is associated with decreases in Parabacteroides and Enterobacter abundance and increases in microbial ergosterol and long-chain fatty acid biosynthesis pathways. Furthermore, the selective serotonin reuptake inhibitor sertraline is associated with increases in Alistipes abundance, increases in microbial pathways related to dopamine degradation, and increases in fecal concentration of the related metabolite kynurenic acid. Critically, we discover that medication effects are time-dependent, with microbiome recovery to baseline occurring over days to weeks. Our findings represent the first systematic characterization of medication-microbiome dynamics within individual patients in a clinical setting, establishing temporal patterns and mechanistic pathways that inform microbiome-guided therapeutic strategies. Description This repository contains contains data files and code to run the models and generate the figures in this paper, which should take about 5 minutes to run on a typical desktop computer. To generate the figures in this paper, extract the zip file and run:`python3 build_figures.py` Most data files are stored in the highly efficient Apache parquet format, which can be read in Python using the read_parquet function in pandas or in R using the read_parquet function in arrow. MRNs (medical records numbers) in data files have also been anonymized. Code Description build_figures.py Script to build figures used in this paper.paperfigures.py Contains code to build each of the figures in this paper and output them as PDFs.icd10.py Builds a table that maps ICD-10 codes to their text descriptions.main.py Runs the linear mixed models used in our analysis pipeline. pathways.py Contains functions to perform pathway enrichment analysisplots.py Helper functions for plotting.util.py Helper functions for use in the analysis pipeline. File Descriptions out/ output figures and supplementary tables produced by build_figures.py output/ Results from the linear mixed models, with columns indicating medication, feature, p-value, coefficient, significance (FDR < 0.1), estimated standard error, medication class, model duration, sample level type, model type, and whether the medication was a duration effect or not. output/results_2_to_10_days.csv Presence/absence results from our 2-10 day models. output/results_10_to_20_days.csv Presence/absence results from our 10-20 day models.output/results_20_to_30_days.csv Presence/absence results from our 20-30 day models.output/results_duration.csv Results from our duration models. data/ Intermediate data files generated from running our pipeline. data/table_demographics.parquet Anonymized demographics data with patient sex and birth year. data/table_microbes_clr.csv Data table containing CLR-transformed microbiome abundance data.data/table_pathways.parquet Data table containing microbial pathway abundances.data/table_pathways_general.parquet Data table containing microbial pathway abundances.data/metacyc_pathway_hierarchy.json Pathway hierarchy for humann3 metacyc pathways.data/hospital_visits.parquet List of all patient hospital visits, with accompanying clinical data.data/study_intervals.parquet List of all pairs of stool collections, with accompanying clinical data.data/table_microbes.csv Data table containing microbiome counts.data/table_microbes_relabund.parquet Data table containing microbiome relative abundances.data/medication_classes.parquet List of all medication classes.data/table_genus_metab_correl.parquet Correlations between microbial genera and metabolites.data/medications.parquet List of all medications included.data/samples.parquet De-identified table containing list of stool samples included in the study with collection dates.data/icd10cm-tabular-April-2024.xml ICD-10 codes reference table Dependencies Package Version python 3.10 numpy 1.23.5 pandas 1.5.3 matplotlib 3.6.3 scikit-learn 1.2.1 scikit-bio 0.5.8 seaborn 0.12.2 scipy 1.10.0 statsmodels 0.13.5 matplotlib 3.6.3 networkx 3.0

  PublisherDOI

• Fecal Butyrate and Deoxycholic Acid Concentrations Correlate With Mortality in Patients With Liver Disease

  Gastro Hep Advances · 2025-01-01 · 5 citations

  articleOpen access1st authorCorresponding

  Background and Aims: The intestinal microbiome produces metabolites, including short chain fatty acids (SCFAs) and secondary bile acids (BAs), that impact host physiology. Loss of intestinal microbiome diversity is associated with cirrhosis progression, but the impact of microbiome-associated metabolites on liver disease remains largely undefined. We aimed to correlate fecal metabolite concentrations with the severity and progression of liver disease. Methods: In this cross-sectional study, fecal samples from patients hospitalized with liver disease were analyzed by shotgun metagenomic sequencing to determine microbiome compositions and targeted mass spectrometry to quantify SCFAs and BAs. Random survival forest and logistic regression models identified clinical, metagenomic, and metabolomic features associated with rehospitalization and survival. Results: . Conclusion: Mass spectrometry rapidly identifies patients with low fecal butyrate and DCA concentrations who are at increased risk of 30-day mortality. These findings set the stage for clinical trials of microbiome reconstitution with butyrate and DCA-producing bacterial species.

  PublisherDOI

• Practical Considerations for the Use of Semaglutide in MASH

  The American Journal of Gastroenterology · 2025-10-23

  article1st authorCorresponding
  PublisherDOI

• Fecal metabolite profiling identifies critically ill patients with increased 30-day mortality

  Science Advances · 2025-06-06 · 5 citations

  articleOpen access

  Critically ill patients admitted to the medical intensive care unit (MICU) have reduced intestinal microbiota diversity and altered microbiome-associated metabolite concentrations. Metabolites produced by the gut microbiota have been associated with survival of patients receiving complex medical treatments and thus might represent a treatable trait to improve clinical outcomes. We prospectively collected fecal specimens, defined microbiome compositions by shotgun metagenomic sequencing, and quantified microbiota-derived fecal metabolites by mass spectrometry from 196 critically ill patients admitted to the MICU for non-COVID-19 respiratory failure or shock to correlate microbiota features and metabolites with 30-day mortality. Microbiota compositions of the first fecal sample after MICU admission did not independently associate with 30-day mortality. We developed a metabolic dysbiosis score (MDS) that uses fecal concentrations of 13 microbiota-derived metabolites, which predicted 30-day mortality independent of known confounders. The MDS complements existing tools to identify patients at high risk of mortality by incorporating potentially modifiable, microbiome-related, independent contributors to host resilience.

  PublisherDOI

• Aspirin Use and Mediterranean Diet Adherence as Factors Associated With Overall Mortality in Patients With MASLD in NHANES

  Hepatology Research · 2025-11-30 · 2 citations

  article

  AIM: Full approval for new non-cirrhotic metabolic dysfunction-associated steatohepatitis (MASH) therapies requires demonstrating improved mortality. As cardiovascular disease (CVD) is the leading cause of death in metabolic dysfunction-associated steatotic liver disease (MASLD), we identified factors of all-cause mortality in a nationally representative cohort, focusing on aspirin and diet. METHODS: We analyzed 4541 adults (40-79 years) with MASLD from the National Health and Nutrition Examination Survey (NHANES) 2011-2018. Key exposures were aspirin use and adherence to the Mediterranean diet, assessed by the Mediterranean diet score (MDS). Advanced fibrosis was defined as FIB-4 > 2.67 and included as a covariate in all multivariable Cox models. To adjust for confounding, propensity scores were estimated using multivariable logistic regression that incorporated the complex NHANES survey design. RESULTS: In a cohort of 2175 patients with MASLD followed for a mean of 4.5 years, 124 deaths occurred. Among adults > 65 years without prior history of CVD, aspirin use was associated with higher all-cause mortality (hazard ratio [HR] 9.82; 95% confidence interval [CI] 1.15-83.69); this association was weaker at higher MDS (interaction p = 0.04). In adults with a history of CVD, aspirin use was not associated with survival, whereas higher MDS was independently associated with lower mortality (HR 0.61; 95% CI 0.41-0.93). No material associations were observed in younger adults. MDS was inversely related to inflammatory markers in older patients. CONCLUSIONS: Aspirin use is independently associated with a several-fold increased mortality risk in older MASLD patients without prior CVD, an effect partly mitigated by high adherence to the Mediterranean diet. Aspirin for primary CVD prevention should be used with caution in this population. Aspirin use and diet are important, often uncontrolled factors associated with mortality in MASLD clinical trials.

  PublisherDOI

• Author Correction: Dietary- and host-derived metabolites are used by diverse gut bacteria for anaerobic respiration

  Nature Microbiology · 2025-05-16

  erratumOpen access
  PublisherOA PDFDOI

Recent grants

• Cytoskeletal mechanisms of epithelial morphogenesis

  NIH · $124k · 2014–2017

Frequent coauthors

• Jerrold R. Turner

  Harvard University

  33 shared

• Thomas G. Cotter

  The University of Texas Southwestern Medical Center

  13 shared

• Gurminder Singh

  North Dakota State University

  11 shared

• Juanmin Zha

  First Affiliated Hospital of Soochow University

  11 shared

• Eric G. Pamer

  University of Chicago

  11 shared

• Wei‐Ting Kuo

  National Taiwan University

  11 shared

• Raphaël Méheust

  10 shared

• Christopher Lehmann

  8 shared