Joint Spatio-Temporal Shared Component Model with an Application in Iran Cancer Data

Area of Science:

• Epidemiology
• Biostatistics
• Geographic Information Systems (GIS)

Background:

• Shared component models are increasingly popular for joint disease mapping.
• Integrating temporal aspects into spatial models enhances disease data inference.
• Seven prevalent cancers in Iran, accounting for 50% of all cancer cases, were analyzed.

Purpose of the Study:

• To combine multivariate shared components with spatio-temporal modeling for joint disease mapping.
• To apply the model to incidence rates of seven prevalent cancers in Iran.
• To identify geographical and temporal variations and shared risk factors for these cancers.

Main Methods:

• Development of a joint disease mapping model incorporating multivariate shared components and spatio-temporal trends.
• Each component is shared by different subsets of diseases, with spatial and temporal trends estimated for each.
• Estimation of the relative weight of these trends for each component and disease.

Main Results:

• Identified high-risk provinces for specific cancers (e.g., Northern for esophagus/stomach, Northwest for bladder/lung, specific provinces for colorectal, prostate, and breast cancers).
• Determined the geographical distribution and impact of shared risk factors: smoking (esophagus, stomach, bladder, lung), overweight/obesity (esophagus, colorectal, prostate, breast), and low physical activity (colorectal, breast).
• Quantified the differential importance of risk factors across diseases (e.g., smoking for stomach vs. esophagus, overweight/obesity for colorectal vs. esophagus).

Conclusions:

• The proposed spatio-temporal joint disease mapping model effectively captures geographical and temporal variations among diseases.
• The model offers valuable insights into shared risk factors and their influence across different cancer types.
• This approach provides benefits over existing joint disease mapping models for epidemiological research.