Related Experiment Video
Updated: Aug 5, 2026
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Using machine learning to predict COVID-19 infection and severity risk among 4,510 aged adults: a UK Biobank cohort
Auriel A Willette1,2,3, Sara A Willette3, Qian Wang1
1Department of Food Science and Human Nutrition, Iowa State University, Ames, IA, USA.
Insights
Predicting COVID-19 infection and hospitalization risk is possible using existing health data. Prior immune status, including antibody titers to common pathogens, significantly aids in predicting novel coronavirus disease (COVID-19) outcomes.
Area of Science:
- Immunology
- Epidemiology
- Biostatistics
Background:
- Novel 2019 coronavirus disease (COVID-19) has numerous identified risk factors.
- The collective predictive power of these factors for COVID-19 infection and severe outcomes (hospitalization) remains unclear.
Approach:
- Utilized UK Biobank data from 4,510 aged adults, including baseline demographics, biochemistry, body mass, and antibody titers for 20 infectious diseases.
- Employed permutation-based linear discriminant analysis and receiver operating characteristic curves to assess predictive models for COVID-19 infection and hospitalization risk.
Key Points:
- A predictive model incorporating age, immune markers, lipids, and serology titers demonstrated excellent discrimination for COVID-19 risk (AUC=0.969).
- Factors included serology titers to common pathogens like human cytomegalovirus.
- A separate model for hospitalization risk, based solely on serology titers, showed more modest discrimination (AUC=0.803).
Conclusions:
- Standard self-report and biomedical data are sufficient for creating accurate COVID-19 risk profiles.
- Further research is warranted to explore the predictive relationship between prior host immunity and current immunity to COVID-19.
Background:
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).
Methods:
Among aged adults (69.3 ± 8.6 years) in UK Biobank, COVID-19 data was downloaded for 4,510 participants with 7,539 test cases. We downloaded baseline data from 10-14 years ago, including demographics, biochemistry, body mass, and other factors, as well as antibody titers for 20 common to rare infectious diseases. 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.
Results:
The "best-fit" model for predicting COVID-19 risk 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.
Conclusions:
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.
