Bayesian causal network modeling suggests adolescent cannabis use accelerates prefrontal cortical thinning

Area of Science:

• Neuroscience
• Developmental Psychology
• Psychiatry

Background:

• Cannabis use during adolescence is linked to brain development differences, but causality remains unclear.
• Existing evidence is largely correlational, hindering definitive conclusions about cannabis's neurodevelopmental impact.
• Bayesian Causal Network (BCN) modeling offers a method to infer causal relationships from correlational data.

Purpose of the Study:

• To investigate the potential causal relationship between adolescent cannabis use and accelerated prefrontal cortical thinning.
• To determine if cannabis use causally influences neurodevelopment in adolescents.
• To leverage BCN modeling to analyze neurodevelopmental trajectories in relation to cannabis exposure.

Main Methods:

• Employed Bayesian Causal Network (BCN) modeling on data from 637 cannabis-naïve adolescents at age 14 from the IMAGEN study.
• Incorporated variables including cannabis use, prefrontal cortical thickness, demographics, psychopathology, childhood adversity, and other substance use.
• Utilized conditional probabilities within BCNs to estimate directional associations between variables.

Main Results:

• BCN algorithms consistently indicated a directional relationship from adolescent cannabis use to accelerated prefrontal cortical thinning.
• The findings suggest cannabis use may causally affect neurodevelopmental processes.
• Results align with previous correlational findings and animal studies on cannabis and brain development.

Conclusions:

• BCN modeling provides evidence supporting a causal link between adolescent cannabis use and accelerated prefrontal cortical thinning.
• While not definitive proof, the results strengthen the hypothesis that cannabis adversely impacts adolescent brain development.
• Further research is warranted to confirm these causal inferences and understand the mechanisms involved.