About

John Harley Warner is the Avalon Professor of the History of Medicine, Professor of History and of American Studies, and Chair of the Section of the History of Medicine at Yale University’s School of Medicine. He received his Ph.D. in 1984 from Harvard University in the History of Science and was a Postdoctoral Fellow at the Wellcome Institute for the History of Medicine in London from 1984 to 1986. Since joining Yale in 1986, he has focused his research on the cultural and social history of medicine in 19th and 20th century America, as well as comparative history involving Britain and France. His scholarly interests include clinical practice, orthodox and alternative healing, the multiple meanings of scientific medicine, and the interactions among identity, narrative, and aesthetics in the grounding of modern medicine. Warner’s work explores the transformation of medical practices and ideas over time, emphasizing the cultural and social contexts that shape medical knowledge and professional identity.

Research topics

• Internal medicine
• Psychology
• Biology
• Medicine
• Oncology
• Neuroscience
• Physical medicine and rehabilitation
• Pathology
• Genetics
• Biochemistry

Selected publications

• From genes to trajectories: mapping genetic influences on Huntington’s disease progression

  Bioinformatics · 2026-02-13

  articleOpen access

  MOTIVATION: There are many diseases with established genetic factors, such as Huntington's disease (HD), that are characterized by variable rates of progression. However, beyond the contribution of the known genetic factors - in this case the Huntingtin (HTT) gene - the impact of the full human genome on the natural progression of such diseases throughout a patient's life remains largely unknown. The increased availability of genome wide association (GWA) data in HD gene expansion carriers (HDGECs), combined with the clinical assessment scores on the same set of patients, has provided a perfect opportunity to assess the potentially broader genetic impact on the natural progression of HD. RESULTS: We present a genetics-driven, probabilistic disease progression model designed to identify and investigate the ways in which a range of genetic factors affect the natural progression of HD. When applied to a clinico-genomic HD dataset, our model identified several single nucleotide polymorphisms (SNPs) with previously unreported effects on disease progression that act at distinct stages and with varying magnitudes. This discovery may shed light on the potential mechanistic impact of previously unidentified genes on HD that may have implications for clinical management. As increasing amounts of GWA data become available more generally, we anticipate that this modeling framework will be broadly applicable to other diseases with strong genetic components. AVAILABILITY AND IMPLEMENTATION: The source code for IHDPM is available at https://github.com/BiomedSciAI/IHDPM.

  PublisherDOI

• Comparison of Partial Volume Correction methods for Huntington’s disease integrated PET/MR imaging analysis

  Research Square · 2026-05-12

  preprintOpen access
  PublisherOA PDFDOI

• PET imaging with [¹¹C]CHDI-00485180-R, designed as radioligand for aggregated mutant huntingtin, in people with Huntington’s disease

  European Journal of Nuclear Medicine and Molecular Imaging · 2025-06-17 · 4 citations

  article
  PublisherDOI

• F010 Shield-HD: a longitudinal natural history study with implications for interventional trials

  2024-09-01

  articleOpen access

  <h3>Background</h3> Shield-HD is a longitudinal natural history study well-positioned to answer questions about how biomarkers, clinical assessments, and volumetric MRI data interrelate over the course of disease progression. Participants were recruited and assigned to three cohorts based on CAP score (&lt;290, 290–400, &gt;400). Ultimately, N = 68 subjects passed screening and were followed for up to 120 weeks. <h3>Aims</h3> Our goals are threefold: 1) To explore the role of the Huntington’s Disease Integrated Staging System (HD-ISS) in Shield and its adoption in future studies. 2) To estimate departures from baseline for common endpoints and to compare these results to those from previous studies. 3) To estimate correlations between outcomes, fully leveraging Shield’s high-resolution multimodal data. <h3>Methods</h3> Analyses are based on regression techniques for correlated error terms (such as mixed models). Subject-specific effects are extracted from these models in order to compute correlations between multiple longitudinal variables. Results are augmented with data from Enroll PDS6 and Track-HD/TrackOn-HD. <h3>Results</h3> We mention several results pertaining to the second goal above. Putamen and Caudate volumes declined at estimated rates of 1.9% and 6.5% per year respectively. Concentrations of plasma NfL increased slightly (1.0 ng/L per year) while no analogous changes were found in the CSF. Clinical composite scores showed either no significant changes (HD-CAB) or slight decline (cUHDRS; .3 points per year). These statistically detectable changes from baseline were generally smaller than expected. <h3>Conclusions</h3> Shield-HD is particularly valuable in guiding future interventional trials. We emphasize this perspective throughout, especially with respect to recruitment and the selection of endpoints.

  PublisherOA PDFDOI

• F100 A review of how data from Enroll-HD has been used to advance our understanding of HD

  2024-09-01

  review

  <h3>Background</h3> Huntington’s disease (HD) is a hereditary neurodegenerative disorder characterised by progressive functional impairment. Mathematical models have facilitated the advancement of HD research by enhancing understanding of the disease’s progression.(1–6) Objectives: Review mathematical models developed over past five years in collaboration among CHDI Foundation, University of Iowa and RAND Corporation that characterize HD. <h3>Methods</h3> Models were developed using data from Enroll HD, IMAGE HD, PREDICT HD, REGISTRY, and TRACK HD studies. Objectives of the models, statistical methodologies, and its application to HD research was evaluated. <h3>Results</h3> Survival analysis was used to develop a CAP score predictive of age at clinical diagnosis. CAP is a prognostic score used to index HD progression and enrich for clinical trial population selection, which can be impactful in the context of the HD Integrated Staging System. Linear mixed modelling was used to develop a Phenotype Atlas (EHDPA), summarising HD phenotypes across motor, cognitive, behavioural, and functional domains. EHDPA provides categorised assessment measurements and visual/statistical resources for HD phenotypes that illustrates how the variables in HD progress over time. Logistic regression analysed comorbidity frequency in PwHD versus controls. The comorbidity analysis showed higher frequencies of sleep disorders, falls, and orthopaedic complications in PwHD; specific conditions vary across HD-ISS stages. Propensity score weighting assessed the impact of non-genetic factors on HD progression. Propensity score weighting corrects for biases in observational studies by weighting observations so that covariates are balanced between treatment and control groups. <h3>Conclusion</h3> These studies provide models that enhance the characterisation and understanding of HD phenotypes and comorbidities, aiding clinical management and research. <h3>References</h3> Long JD, Mills JA. Joint modeling of multivariate longitudinal data and survival data in several observational studies of Huntington’s disease. <i>BMC Med Res Methodol</i>. 2018;<b>18</b>(1). doi:10.1186/s12874-018-0592-9 2.Warner JH, Long JD, Mills JA, <i>et al.</i> Standardizing the CAP Score in Huntington’s Disease by Predicting Age-at-Onset. <i>J Huntingtons Dis</i>. 2022;<b>11</b>(2):153-171. doi:10.3233/JHD-210475 Griffin BA, Booth MS, Busse M,<i> et al</i>. Estimating the causal effects of modifiable, non-genetic factors on Huntington disease progression using propensity score weighting. <i>Parkinsonism Relat Disord.</i> 2021;<b>83</b>:56-62. doi:10.1016/j.parkreldis.2021.01.010 Mills J, Long J, Vaidya J, Sampaio C, Sathe S. Comorbidities in Huntington’s disease: An Enroll-HD analysis [abstract]. Movement Disorders. Published 2022. Accessed April 23, 2024. https://www.mdsabstracts.org/abstract/comorbidities-in-huntingtons-disease-an-enroll-hd-analysis/ Langbehn DR, Sathe SS, Loy C, Sampaio C, Mccusker EA. A Phenotypic Atlas for Huntington Disease Based on Data From the Enroll-HD Cohort Study. <i>Neurol Genet</i>. 2023;<b>9</b>(6). doi:10.1212/nxg.0000000000200111 Tabrizi SJ, Schobel S, Gantman EC, et al. A biological classification of Huntington’s disease: the Integrated Staging System. <i>Lancet Neurol</i>. 2022;<b>21</b>(7):632-644. doi:10.1016/S1474-4422(22)00120-X

  PublisherDOI

• Study protocol for the iMarkHD study in individuals with Huntington's disease

  Journal of Huntington s Disease · 2024-10-08 · 2 citations

  article

  Background: Huntington's disease (HD) is still often defined by the onset of motor symptoms, inversely associated with the size of the CAG repeat expansion in the huntingtin gene. Although the cause of HD is known, much remains unknown about mechanisms underlying clinical symptom development, disease progression, and specific clinical subtypes/endophenotypes. Objective: In the iMarkHD study, we aim to investigate four discrete molecular positron emission tomography (PET) tracers and magnetic resonance imaging (MRI) markers as biomarkers for disease and symptom progression. Methods: Following MRI optimization in five healthy volunteers (cohort 1), we aim to recruit 108 participants of whom 72 are people with HD (PwHD) and 36 healthy volunteers (cohort 2). Pending interim analysis, these numbers could increase to 96 PwHD and 48 healthy controls. Participants will complete a total of 10 study visits, consisting of a screening visit followed by a clinical and MRI visit and PET visits at baseline, year 1, and year 2. PET targets include the cannabinoid 1, histamine 3, and serotonin 2A receptors, and phosphodiesterase 10A, whereas MRI will be multimodal, including, but not limited to, the assessment of cerebral blood flow, functional connectivity, and brain iron. Results: Recruitment is currently active and started in September 2022. Conclusions: By combining PET and multi-modal MRI assessments we expect to provide a comprehensive examination of the molecular, functional, and structural framework of HD progression. As such, the iMarkHD study will provide a solid base for the identification of treatment targets and novel outcome measures for future clinical trials.

  PublisherDOI

• A004 A markov chain-based bayesian population model for somatic instability in human white blood cells

  2024-09-01

  articleOpen access1st authorCorresponding

  <h3></h3> Using a database of HTT CAG-length distributions determined by amplicon sequencing in White Blood Cells (WBCs) collected from 1447 participant-visits from the Registry and Enroll-HD Studies, we build a Bayesian population model based on the assumption that CAG length distributions evolve over time according to a continuous time Markov chain in which changes in CAG length occur in single CAG length jumps. These Markov chains have jump rates that increase with CAG length, resulting in both expansions and contractions with a bias toward expansion. The effects of PCR-induced slippage were corrected using single molecule data, a novel multinomial logit model and a novel algorithm based on convex programming. Models were fit to the PCR slippage-corrected data using a fast BFGS algorithm in R. This algorithm minimizes both least squares and Kullback-Liebler loss functions and uses analytic gradients of the matrix exponential function. Participant-level models are shown to fit very well and predict CAG length distributions at the second visit 2 from distributions at visit 1. Model parameters for jump rate and expansion bias are correlated with inherited CAG length, CAP score, age-at-baseline, and time- between-visits. Finally, a Bayesian population-based model (implemented in the Stan Statistical Software Package) was used to model the precision of the participant-level model fits using a novel application of the Dirichlet distribution. Implications for studies of somatic instability in other cell types are discussed as are the uses of the statistical methodology described here in biomarker development.

  PublisherOA PDFDOI

• Predicting Huntington’s disease state with ensemble learning &amp; sMRI: more than just the striatum

  medRxiv · 2023-07-27 · 3 citations

  preprintOpen access

  Abstract Developing effective treatments for Huntington’s disease (HD) requires reliable markers of disease progression. Striatal atrophy has been the hallmark of HD progression, but volumetric anomalies are also found in other brain regions. Little is known about the potential increase in predictive biomarking accuracy when volumetric scores from multiple brain regions are combined to predict the HD status of individual participants. We used cross-sectional structural MRI data from 184 HD gene-positive participants to a) test a novel ensemble machine learning model in classifying participants in one of four HD progression states (PreHD A; PreHD B; HD1; HD2), and (b) identify the brain regions that carry HD biomarking signal from 15 regions. We used 5-fold cross validation and backward feature elimination to find the optimal predictors and investigated the stability of the findings through repeated analyses. The ensemble predictive model systematically matched or outperformed the accuracy of nine standard machine learning models, reaching 55.3%±6.1 balanced accuracy in 4-group classification. The accuracy was higher for binary classifications (PreHD vs HD: 83.3%±6.3; PreHD A vs PreHD B: 76.7%±8.0; PreHD B vs HD1: 75.9%±8.5; HD1 vs HD2: 70.9%±9.4). Striatal structures (caudate and putamen) were systematically found to be top predictors. However, the accuracy increased substantially when we included other regions in the model (e.g., occipital cortex, lateral ventricles, cingulate, temporal lobe). Optimal models frequently included 2-7 brain regions from different areas. Overall, the accuracy of classifications remained stable across repetitions but the list of selected brain regions could vary, likely due to collinearities in volumetric scores. This is the first study to demonstrate the improvement of classification accuracy when predicting HD progression with a stacked ensemble model. Our findings indicate that HD progression is marked not only by striatal atrophy but also by volumetric changes outside the striatum, without which biomarking models cannot achieve optimal results. The robust methods applied here exposed instability in the selection of brain regions despite the sizeable sample size (n=184); such instabilities could lead to different conclusions in different studies when single analyses are applied on smaller sample sizes. From a translational perspective, our study informs on the selection of candidate endpoints or target regions for therapeutic intervention in future clinical trials.

  PublisherOA PDFDOI

• PCR252 Early Steps Towards a Huntington's Disease Burden-of-Illness Model: Estimating Out-of-Pocket Costs for Patients Showing Clinical Symptoms

  Value in Health · 2023-12-01

  articleOpen access
  PublisherOA PDFDOI

• The temporal event-based model: Learning event timelines in progressive diseases

  Imaging Neuroscience · 2023-08-01 · 17 citations

  articleOpen access

  Abstract Timelines of events, such as symptom appearance or a change in biomarker value, provide powerful signatures that characterise progressive diseases. Understanding and predicting the timing of events is important for clinical trials targeting individuals early in the disease course when putative treatments are likely to have the strongest effect. However, previous models of disease progression cannot estimate the time between events and provide only an ordering in which they change. Here, we introduce the temporal event-based model (TEBM), a new probabilistic model for inferring timelines of biomarker events from sparse and irregularly sampled datasets. We demonstrate the power of the TEBM in two neurodegenerative conditions: Alzheimer’s disease (AD) and Huntington’s disease (HD). In both diseases, the TEBM not only recapitulates current understanding of event orderings but also provides unique new ranges of timescales between consecutive events. We reproduce and validate these findings using external datasets in both diseases. We also demonstrate that the TEBM improves over current models; provides unique stratification capabilities; and enriches simulated clinical trials to achieve a power of 80% with less than half the cohort size compared with random selection. The application of the TEBM naturally extends to a wide range of progressive conditions.

  PublisherOA PDFDOI

Recent grants

• NIH Grant R01AA011312

  NIH · $123k · 2000

• NIH Grant R01LM005013

  NIH · $41k · 1993

Frequent coauthors

• Sarah J. Tabrizi

  University College London

  68 shared

• Cristina Sampaio

  CHDI Foundation

  61 shared

• Alexandra Dürr

  Sorbonne Université

  54 shared

• Blair R. Leavitt

  British Columbia Children's Hospital

  53 shared

• Beth Borowsky

  Tris Pharma (United States)

  51 shared

• Allan J. Tobin

  Sorbonne Université

  50 shared

• Chris Becker

  50 shared

• Howard Schulman

  Panorama Research (United States)

  50 shared

Labs

• Program in the History of Science and MedicinePI