Modeling internalizing symptoms and substance use comorbidity using causal network methodology

The cooccurrence of internalizing and substance use disorders can be explained from several perspectives, yet most are modeled in isolation and limited to correlational evidence. Advancements on two common etiological perspectives - i) internalizing symptoms predisposing toward substance use disorder symptoms, ii) transdiagnostic factors predisposing toward both symptom domains - are therefore hampered by scarce multivariate and causal evidence. The present study addressed this gap by modeling direct (causal) links from internalizing symptoms to substance use behaviors and (causal) links from the transdiagnostic factors cognitive control, locus of control (LOC), and meaning in life. We analyzed data from wave 7 (2019-2023; N = 1725), the 15-year follow-up of the ongoing longitudinal cohort study the Netherlands Study of Depression and Anxiety, using a correlational network and a causal network based on the fast causal inference algorithm. The correlational network showed that depression and social anxiety symptoms were positively related to alcohol-related problems and negatively related to alcohol consumption. Cognitive control and external LOC correlated positively with alcohol consumption. Meaning in life was negatively associated with depression and social anxiety symptoms, alcohol consumption, alcohol problems, and drug use. The causal network showed causal relations of external LOC with depression and social anxiety symptoms. Although our correlational findings gave some support to both perspectives on internalizing and substance use comorbidity, the present findings do not support a causal, etiological interpretation of these perspectives.