Development of sex differences in the human brain.

Area of Science:

• Neuroscience
• Developmental Neuroscience
• Brain Anatomy

Background:

• Existing research on sex differences in brain anatomy lacks consistent findings.
• Previous studies often used binary classifications, potentially oversimplifying complex variations.

Purpose of the Study:

• To investigate age-related changes in sex differences in brain structure using a continuous machine learning classifier.
• To determine if brain sex differences become more pronounced during childhood and adolescence.

Main Methods:

• Applied a machine learning classifier to estimate 'Brain Sex' on MRI data from 162 boys and 185 girls (ages 5-18).
• Utilized a sliding window approach to analyze changes in estimated sex differences across age groups.
• Calculated effect sizes (Cohen's d) to quantify the magnitude of sex differences.

Main Results:

• The machine learning classifier demonstrated high performance (80.4% accuracy, 0.897 AUC).
• A significant increase in the magnitude of sex differences in brain structure was observed with increasing age.
• Effect sizes for sex differences grew from d = 1.2 in childhood to d = 1.6 by age 17.

Conclusions:

• Systematic sex differences in brain structure are present from childhood.
• These differences become substantially more pronounced during adolescence, supporting the hypothesis of diverging neurodevelopmental trajectories.