Causal effects of physical activity and screen time on childhood intelligence via Mendelian randomization: The mediating role of intracranial volume

Abstract

Growing evidence suggests that physical activity and screen time affect intelligence (IQ) during childhood, a critical period for brain development, yet the relationship between these factors remains controversial. Using bidirectional Mendelian randomization (MR), we investigated these associations while accounting for potential reverse causality. Our two-sample MR analysis revealed a positive causal effect of moderate to vigorous physical activity (PA) on childhood IQ (β = 0.42, 95 % confidence interval (CI): [0.12, 0.72], p = 6.26 × 10⁻³), whereas leisure screen time (LST) exhibited a negative causal effect (β = −0.35, 95 % CI: [ −0.60, −0.10], p = 5.59 ×10⁻³). Reverse MR analysis found no evidence of causations. A two-step MR mediation framework further suggested that the intracranial volume (ICV) mediated 21.69 % (95 % CI: [15.25 %, 28.13 %]) of the negative effect of LST on childhood IQ. These MR-derived findings demonstrate that PA positively influences childhood IQ, whereas LST negatively impacts it, partly through reduced ICV. By leveraging genetic instruments, this study strengthens causal inference and highlights the potential of PA promotion and screen time reduction to support cognitive development. Further research is needed to elucidate the mechanisms underlying these associations and their long-term cognitive consequences.