Recent Changes Leading to Subsequent Changes: Extensions of Multivariate Latent Difference Score Models

Abstract

Latent difference score models (e.g., McArdle & Hamagami, 2001) are extended to include effects from prior changes to subsequent changes. This extension of latent difference scores allows for testing hypotheses where recent changes, as opposed to recent levels, are a primary predictor of subsequent changes. These models are applied to bivariate longitudinal data collected as part of the Baltimore Longitudinal Study of Aging on memory performance, measured by the California Verbal Learning Test, and lateral ventricle size, measured by structural MRIs. Results indicate that recent increases in the lateral ventricle size were a leading indicator of subsequent declines in memory performance from age 60 to 90.

Keywords:

• latent difference score
• dynamic
• longitudinal
• growth
• memory
• brain

Notes

¹This sample size is generally small for LDS modeling; however, the measures generally have good measurement properties and there are several assessments. Determining an appropriate sample size for this type of modeling depends on several issues including the size of measurement error, effect size, attrition, number, and spacing of measurement occasions.

^aSelected model.

²When dynamic parameters are estimated in univariate and bivariate latent difference score models, the sign and magnitude of the mean of the constant change parameter varies as a function of magnitude and sign of the scale (of raw scores) because this estimated parameter functions as an intercept in the change equations. For example, adding 10 to all observed scores results in a new estimate for the mean of the constant change component; however, the estimates for all dynamic parameters are identical. Thus, interpreting the mean of the constant change component can be difficult.

^aSelected model.