Multilevel Conditional Autoregressive models for longitudinal and spatially referenced epidemiological data

Abstract

Multilevel Conditional Autoregressive (CAR) models help to explain the spatial effect in epidemiological studies, where subjects are nested within geographical units. This paper has two goals. Firstly, it further develops the multilevel models for longitudinal data by adding existing random effects with CAR structures that change over time. We name these models MLM tCARs. We compare the MLM tCARs to the classical multilevel growth model via simulation studies. We observe a better performance of the MLM tCARs, to retrieve the true regression coefficients and with better fit. Secondly, it provides a comprehensive decision tree for analysing data in epidemiological studies with spatially nested structure: we also consider the Multilevel CAR models (MLM CARs) for cross-sectional studies in simulation studies. We apply the models comparatively on the analysis of the association between greenness and depression in the longitudinal Heinz Nixdorf Recall Study. The results show negative association between greenness and depression.

Keywords: Conditional Autoregressive; Cross-sectional; Decision tree; Longitudinal; Multilevel; Spatial effect.