A Joint Model was Used to Analyze the Influence of Dynamic Changes in Cerebrospinal Fluid Biomarkers on the Time of Dementia Transformation

Abstract

Dementia, specifically Alzheimer’s disease (AD), presents a grave public health hazard on a worldwide scale, with more than 13 million individuals in China and a 10% yearly growth prediction. The cerebrospinal fluid biomarkers like Aβ42 and p-Tau181 are critical to early detection, but static models cannot capture the movements of sickness processes, which drops down the exactness of forecasts. This study aims to investigate how changes over time in CSF biomarkers connect to the amount of time until someone gets dementia. Using a combined model that includes Linear Mixed-Effects Models (LMMs) for biomarker patterns and Cox Models with Time-Varying Covariates for conversion times, we looked at data from cognitively normal people and MCI patients. Results showed that faster Aβ42 drops and steeper p-Tau181 rises linked to quicker dementia conversion, showing the joint model’s worth for better early AD risk sorting. It gives a more accurate and helpful tool for doctors to check each person’s dementia advancement danger, helping them give more targeted AD treatment.