Sporadic, late-onset, and multistage diseases.

Area of Science:

• Epidemiology
• Biostatistics
• Genetics

Background:

• Multistage disease processes often exhibit a linear relationship between the log of incidence rates and the log of age.
• This relationship is observed in various conditions, including cancers and non-malignant diseases, often linked to somatic mutations.

Purpose of the Study:

• To model and characterize age-related disease onset using parametric models in UK Biobank data.
• To differentiate between disease progression models with ordered versus unordered stages.
• To investigate the impact of risk factors on age-related disease risk.

Main Methods:

• Utilized Weibull distribution to model diseases with ordered stages.
• Employed an alternative model for diseases where stages can occur in any order.
• Analyzed UK Biobank data for over 450 diseases to assess model fit and infer disease characteristics.
• Introduced "relative aging rates" to quantify the modification of age-related risk by risk factors.

Main Results:

• Both ordered and unordered multistage models accurately described the incidence of over 450 diseases, indicating that model fit alone cannot infer the sequence of disease stages.
• Parametric models provided insights into age-related disease not obtainable from conventional relative risk studies.
• Introduced "relative aging rates" to distinguish between "sporadic" and "late-onset" diseases based on how disease risk increases with age.
• Found that common risk factors significantly modify the effective age-at-risk for sporadic diseases.

Conclusions:

• The findings suggest that a substantial burden of sporadic diseases could be delayed or prevented through early lifestyle interventions.
• Relative aging rates offer a valuable tool for risk stratification, enabling the identification of equivalent-risk ages across different exposure groups.
• Multistage disease modeling, while informative about age-related risk, does not inherently elucidate the specific sequence of events in disease development.