Abstract
This paper has both methodological and substantive application for mental-health researchers. Methodologically, it presents the latent growth curve (LGC) technique within a structural equation modelling (SEM) framework as a powerful tool to analyse change in depressive symptoms and potential correlates of such changes. The rationale for LGC analysis and subsequent elaboration of this statistical approach are presented. The limitations of traditional analytical methods are also addressed. Substantively, the paper considers socio-contextual factors as correlates of change in symptoms, and examines the dynamic systematic relationship with the degree of economic integration of south-east Asian immigrants in Canada over time. Using the LGC technique, this study also investigated how the longitudinal course of subclinical depression places individuals at risk for developing full-blown major depression. The LGC results provided strong evidence for the reciprocal influence between economic integration and subclinical depression of immigrants. The initial level of economic integration negatively influenced the rate of change in subclinical depression whereas the initial level of subclinical depression negatively influenced the rate of change in economic integration. Both initial level and the rate of change in subclinical depression placed individuals at risk for full-blown major depression. However, traditional auto-regressive models were not capable of revealing these dynamic associations. Thus, an investigation of within-individual change in symptoms and potential correlates of such changes is necessary to understand the process that results in full-blown mental disorder.