cosinoRmixedeffects

cosinoRmixedeffects models longitudinal periodic data using mixed-effects cosinor models to estimate and predict linear and non-linear circadian parameters (MESOR, amplitude, acrophase) and assess their interactions with covariates.

Key Features:

• Mixed-Effects Framework: Supports inclusion of both fixed and random effects for longitudinal data to model individual variation.
• Covariate Integration: Incorporates covariates and their interactions with non-linear circadian parameters, including time-varying covariates and factors with multiple categories.
• Syntax Compatibility with emmeans: Leverages the syntax and functions of the emmeans package to obtain estimated marginal means and contrasts.
• Bootstrapping for Hypothesis Testing: Employs bootstrapping techniques for estimation and hypothesis testing involving non-linear circadian parameters.

Scientific Applications:

• Circadian Rhythm Analysis: Analysis of physiological data collected over 24-hour periods, such as heart rate variability (HRV), to evaluate circadian patterns.
• Longitudinal Studies: Assessment of changes in circadian rhythms over time while accommodating factors with any number of categories and complex interactions.
• Health Research: Modeling HRV among healthcare workers to assess differences in circadian patterns across genders, BMI categories, and during SARS-CoV2 infection.

Methodology:

Implements mixed-effects cosinor models with fixed and random effects to estimate and predict MESOR, amplitude, and acrophase; models interactions between circadian parameters and covariates or factors with any number of categories; uses bootstrapping for hypothesis testing and emmeans functions for estimated marginal means and contrasts.