Multivariate modeling of two associated cognitive outcomes in a longitudinal study

Area of Science:

• Neuroscience
• Psychology
• Biostatistics

Background:

• Longitudinal Alzheimer's disease studies track cognitive decline and its predictors.
• Cognitive function comprises multiple processes that may decline independently and be influenced by distinct factors.
• Understanding the interrelations between declining cognitive processes is crucial.

Purpose of the Study:

• To introduce and illustrate the application of multivariate growth models for analyzing longitudinal cognitive data.
• To assess the associations between different cognitive processes (e.g., memory, executive function) during decline.
• To identify shared predictors influencing the decline of multiple cognitive functions.

Main Methods:

• Utilized multivariate growth models, an extension of random effects models.
• Modeled separate random effects and predictors for each cognitive process simultaneously.
• Applied the methodology to longitudinal data on memory and executive function in Alzheimer's disease.

Main Results:

• Demonstrated the capability of multivariate growth models to estimate parameters and variability for individual cognitive processes.
• Enabled simultaneous assessment of associations between the decline trajectories of different cognitive functions.
• Provided a framework to determine if predictors of decline in one process are also associated with decline in another.

Conclusions:

• Multivariate growth models offer a robust approach to simultaneously analyze the decline of multiple cognitive processes in Alzheimer's disease.
• This methodology enhances the understanding of how different cognitive functions interrelate and respond to various predictors over time.
• The technique facilitates a more nuanced investigation into the complex nature of cognitive decline in neurodegenerative diseases.