Related Experiment Video
Updated: Sep 23, 2025
Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
Using machine learning to predict COVID-19 infection and severity risk among 4510 aged adults: a UK Biobank cohort
Auriel A Willette1,2,3, Sara A Willette4, Qian Wang5
1Department of Food Science and Human Nutrition, Iowa State University, 2302 Osborn Drive, Ames, IA, 50011-1078, USA. Awillett@iastate.edu.
Insights
Prior host immunity, including antibody titers to pathogens like cytomegalovirus, significantly predicts severe COVID-19 infection risk and hospitalization in older adults.
Area of Science:
- Immunology
- Epidemiology
- Biostatistics
Background:
- Numerous risk factors for coronavirus disease 2019 (COVID-19) have been identified, but their collective predictive power for infection and severe outcomes remains unclear.
- Understanding these factors is crucial for risk stratification, particularly in vulnerable populations such as older adults.
Purpose of the Study:
- To investigate the collective predictive value of demographic, biochemical, anthropometric, and immunological factors for COVID-19 infection and hospitalization risk in UK Biobank participants.
- To determine if pre-existing host immunity, indicated by antibody titers to common infectious diseases, can predict current COVID-19 susceptibility and severity.
Main Methods:
- Utilized UK Biobank data from 4510 older adults, including baseline biomedical data and COVID-19 testing results.
- Employed permutation-based linear discriminant analysis and receiver operating characteristic curves to assess predictive models for COVID-19 risk and hospitalization.
- Analyzed a subset of 80 participants with available antibody titers against 20 common infectious diseases.
Main Results:
- Predictive models for the full cohort showed marginal performance.
- A "best-fit" model using antibody titers, immune markers, lipids, and demographic data in the subset achieved excellent discrimination for COVID-19 risk (AUC 0.969).
- The hospitalization risk model in the subset, primarily based on serology titers, demonstrated a more modest but significant predictive capability (AUC 0.803).
Conclusions:
- Pre-existing host immunity, particularly antibody titers to pathogens such as human cytomegalovirus, is a strong predictor of COVID-19 infection risk and hospitalization in older adults.
- Accurate risk profiles can be developed using readily available clinical and serological data.
- Further research is warranted to explore the link between prior and current host immunity in the context of COVID-19.
Abstract:
Many risk factors have emerged for novel 2019 coronavirus disease (COVID-19). It is relatively unknown how these factors collectively predict COVID-19 infection risk, as well as risk for a severe infection (i.e., hospitalization). Among aged adults (69.3 ± 8.6 years) in UK Biobank, COVID-19 data was downloaded for 4510 participants with 7539 test cases. We downloaded baseline data from 10 to 14 years ago, including demographics, biochemistry, body mass, and other factors, as well as antibody titers for 20 common to rare infectious diseases in a subset of 80 participants with 124 test cases. Permutation-based linear discriminant analysis was used to predict COVID-19 risk and hospitalization risk. Probability and threshold metrics included receiver operating characteristic curves to derive area under the curve (AUC), specificity, sensitivity, and quadratic mean. Model predictions using the full cohort were marginal. The "best-fit" model for predicting COVID-19 risk was found in the subset of participants with antibody titers, which achieved excellent discrimination (AUC 0.969, 95% CI 0.934-1.000). Factors included age, immune markers, lipids, and serology titers to common pathogens like human cytomegalovirus. The hospitalization "best-fit" model was more modest (AUC 0.803, 95% CI 0.663-0.943) and included only serology titers, again in the subset group. Accurate risk profiles can be created using standard self-report and biomedical data collected in public health and medical settings. It is also worthwhile to further investigate if prior host immunity predicts current host immunity to COVID-19.
More Related Videos
12:18A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
09:38Generalized Psychophysiological Interaction PPI Analysis of Memory Related Connectivity in Individuals at Genetic Risk for Alzheimer's Disease
Published on: November 14, 2017
Related Concept Videos
01:18Steps in Outbreak Investigation
01:29Bias in Epidemiological Studies
01:25Statistical Methods for Analyzing Epidemiological Data
01:17Classification of Illness
An illness is a response to a disease in which the person's level of functioning is changed compared with a previous level. The general classification of illness includes acute and chronic.
Acute illness is severe...