In the structural equation modeling framework, latent curve models have gained popularity for modeling change over time. Much work has focused on the use of covariates, whether time-invariant or time-varying, to predict the growth factors. Comparatively little work has focused on the use of growth factors as independent variables themselves. This project evaluated the performance of models where growth factors were used as main-effects predictors of a distal outcome; this main-effects-only model was expanded to include the interaction between the growth factors as a predictor. My results demonstrate the bias present when a main-effects-only model is fit to data where an interaction effect truly exists. These results provide motivation for researchers who employ growth factors as predictors of a distal outcome to test for an interaction effect in order to more clearly understand the role of starting point and rate of change over time, taken together, as predictors.