Impact of the interaction between major cardiovascular disease risk factors on the models predicting cardiovascular disease risk
Approved Research ID: 64670
Approval date: September 16th 2020
A number of risk factors are known to increase the risk of having a myocardial infarction (heart attack) or stroke. These include high blood pressure and cholesterol and smoking. It has been widely assumed that one can predict the risk of having a stroke or myocardial infarction based on knowledge of a range of risk factors using what are called multiplicative models. However, there is no underlying biological rationale for making this assumption. This project aims to test this assumption by taking a variety of different statistical modelling approaches to see if multiplicative models perform better than others within the UK Biobank population. In technical terms we will be examining whether statistical interactions are evident in different models from multiplicative to additive. The findings of the study may have important implications for individual disease prediction as well as for estimating how far disease levels in populations can be attributed to various different risk factors.