The onset and progression of noncommunicable diseases (such as cardiovascular diseases, renal diseases, metabolic related diseases, cancer, etc.) is associated with lifestyle, environmental factors, and genetic factors. The main purpose of the study is to clarify the causal effects of these factors in the onset and progression of diseases, providing relevant data-based evidence for disease prevention and health guidance by using causal inference analyses methods. We will apply propensity score methods (matching/ weighting/ doubly robust model/ propensity score method with machine learning algorithms), instrumental variable model, causal mediation analysis methods to indicate the pathway of causal inference between factors and outcomes (onset or progression of noncommunicable diseases). Meanwhile, we will also verify the applicability of the causal inference models we have constructed.
Additionally, we will attempt to construct prediction model by using machine learning methods, as well as transfer learning or deep learning methods based on the identified factors.
Our study will span 36 months, and we plan to utilize the whole cohort dataset.
This study will provide accurate data-based evidence/suggestion for disease prevention and improving people’s status/level, and we also establish precise prediction models to support disease forecasting.
