Time lags and time interactions in mixed effects models impacted longitudinal mediation effect estimates.

Area of Science:

• Statistics
• Biostatistics
• Longitudinal Data Analysis

Background:

• Mixed effects models are versatile for longitudinal data analysis.
• They can estimate various effects including contemporaneous and lagged effects.
• Little attention has been given to time lags and interactions in longitudinal mediation analysis using these models.

Purpose of the Study:

• To demonstrate the impact of time lags and time interactions on longitudinal mediation effect estimates.
• To highlight the importance of considering temporal dynamics in mediation analysis.
• To improve the interpretation of mediation effects in longitudinal studies.

Main Methods:

• Utilized mixed effects models for longitudinal mediation analysis.
• Incorporated time lags and time interactions between determinant, mediator, and outcome variables.
• Applied the methods to a data example from the Amsterdam Growth and Health Longitudinal Study.

Main Results:

• The selection of time lags significantly influenced the magnitude and interpretation of mediation effect estimates.
• Time interactions allowed for modeling linear and nonlinear developmental trajectories of mediation effects.
• The inclusion of temporal dynamics alters the understanding of mediation pathways.

Conclusions:

• Time lags and time interactions are essential considerations in mixed effects models for longitudinal mediation.
• Accounting for these temporal aspects enables the estimation of lagged and time-dependent mediation effects.
• This approach enhances the accuracy and interpretability of longitudinal mediation analyses.