Sex-specific associations of gestational age at birth and birth size with early life within-network brain connectivity: an exploratory study

Abstract Gestational age at birth and birth size are major risk factors for early life behavioral/cognitive problems, but their impact on functional brain network dynamics during this period is not understood. Our objective was to conduct an exploratory study to evaluate associations of birth measures with longitudinal early life functional connectivity. The Baby Connectome Project used resting-state functional magnetic resonance imaging to assess connectivity within seven canonical brain networks (Yeo atlas): dorsal attention, salience, limbic, frontoparietal, default mode, visual, and sensorimotor. For 254 children <3 years-old (contributing 583 observations), birth weight, birth length, and gestational age at birth were self-reported or abstracted from medical records, and we calculated weight-to-length ratio. Using covariate-adjusted multiple linear mixed models, we evaluated overall and sex-specific associations of birth measures as continuous variables and in tertiles with each network, which were Fisher r-to-z-transformed. Most children (54% female) were born to non-Hispanic White (80%) and college-educated (83%) mothers, were delivered ≥37 weeks gestation (97%), and had birth weights ≥2.5 kg (98%). Only birth size measures were associated with brain network connectivity. Compared to tertile 2, frontoparietal network connectivity was higher in birth weight tertile 1 (β: 0.02; 95% CI: 0.00, 0.04) and tertile 3 (β: 0.03; 95% CI: 0.01, 0.05). Also, compared to tertile 2, birth size tertile 3 was associated with higher limbic (birth length β: 0.03; 95% CI: 0.00, 0.07) and default mode (birth length β: 0.02; 95% CI: 0.00, 0.03), but decreased sensorimotor (birth weight β: -0.03; 95% CI: -0.05, 0.00; birth length β: -0.03; 95% CI: -0.05, 0.00) network connectivity. Compared to tertile 2, birth size tertile 1 was associated with lower limbic (birth weight β: -0.04; 95% CI: -0.08, 0.00) and default mode (weight-to-length ratio β: -0.02; 95% CI: -0.04, 0.00). In sex stratified models, birth size was associated with frontoparietal and default mode networks in both sexes; sensorimotor, limbic, and dorsal attention networks in males; and salience and visual networks in females. Associations followed a U-shaped pattern in females, whereas those in males appeared at only the lowest or highest tertile. In this non-clinical sample, birth size was sex-specifically associated with early life brain network dynamics. This may have implications for later neurodevelopment.