A linear mixed model approach to compare the evolution of multiple biological rhythms

Area of Science:

• Chronobiology
• Biostatistics
• Immunology

Background:

• Comparing multiple biological rhythms is essential for understanding health and disease.
• Functional alterations can lead to chronodisruption or internal desynchronization.
• Existing methods may not adequately handle correlated biorhythms and longitudinal data.

Purpose of the Study:

• To propose a multivariate linear mixed model for jointly analyzing multiple biorhythms.
• To provide a statistical framework for assessing relationships between biorhythms under functional alterations.
• To facilitate pairwise comparisons of biorhythms, including testing for identity or opposition.

Main Methods:

• A multivariate linear mixed model approach is employed.
• The model jointly analyzes functions of several biorhythms longitudinally.
• It accounts for correlations between biorhythms, within-subject measurements, and between-subject heterogeneity using random effects.

Main Results:

• The proposed method effectively models correlated biorhythms and longitudinal data.
• Pairwise comparisons using contrasts can identify identical or opposing biorhythm patterns.
• The approach was validated using simulated data and real biological data from lymphocyte profiles.

Conclusions:

• The developed multivariate linear mixed model offers a robust approach for comparing biorhythms.
• This method enhances the understanding of chronodisruption and internal desynchronization.
• The findings have potential applications in clinical practice, particularly in analyzing immune system modulation in diseases like lung cancer.