Phenotypic Heterogeneity of Movement Behaviors and Its Association With Cardiometabolic Health Using Data From the...

Area of Science:

• Human movement science
• Biostatistics
• Cardiovascular health research

Background:

• Movement behavior is complex and hard to measure with single indicators.
• Dimensionality reduction is needed to capture multiple behavioral aspects.
• Identifying distinct movement behavior phenotypes is crucial for understanding health links.

Purpose of the Study:

• To identify distinct movement behavior phenotypes using a novel approach.
• To investigate the relationship between movement behavior phenotypes and cardiometabolic health.
• To uncover key indicators of movement behavior variation.

Main Methods:

• Utilized a cross-sectional study of 1009 Luxembourg residents (aged 25-79) with accelerometry data.
• Applied reversed graph embedding to reduce movement behavior complexity into a 2D tree structure.
• Employed linear regression to analyze associations between tree dimensions, movement indicators, and health biomarkers.

Main Results:

• Identified four distinct phenotypes: inactive, irregularly active, regularly active, and regularly low active.
• Found favorable associations between both tree dimensions and cardiometabolic health outcomes.
• Determined that average 24-hour acceleration and total step count were key indicators of phenotypic variation.

Conclusions:

• A novel dimensionality reduction method effectively captures movement behavior complexity and identifies key indicators.
• Distinct movement behavior phenotypes linked to cardiometabolic health were successfully distinguished.
• Findings offer new insights into behavioral heterogeneity and its connection to cardiometabolic well-being.