, new research suggests.
Aside from sociodemographic parameters, risk factors for substance use initiation include prenatal exposure to substances, peer use of alcohol and nicotine, and problematic school behavior, among other things, the study showed.
The results show certain modifiable risk factors may play a role in preventing youth from starting to use substances, said study author ReJoyce Green, PhD, research assistant professor, Department of Psychiatry and Behavioral Sciences, Medical University of South Carolina, Charleston.
“If we’re designing, say, a prevention program or an early intervention program, these are things that could make a difference, so let’s make sure we’re bringing them into the conversation.”
The findings were presented at the annual meeting of the American Psychiatric Association American Psychiatric Association (APA) and published online in The American Journal of Psychiatry.
Critical Risk Factors
Use of alcohol, tobacco, and cannabis often begins during adolescence. One recent survey showed that 23% of 13-year-olds reported using alcohol, 17% reported vaping nicotine, and 8% reported vaping cannabis. Other research links younger age at substance use initiation to a more rapid transition to substance use disorders and higher rates of psychiatric disorders.
Previous studies examining predictors of substance use initiation in the Adolescent Brain Cognitive Development (ABCD) Study dataset focused primarily on self-reported measures, but the current study also looked at models that include hormones and neurocognitive factors as well as neuroimaging.
This study included 6829, 9- and 10-year-olds from the ABCD Study who had never tried substances and were followed for 3 years.
A sophisticated statistical approach was used to examine 420 variables as predictors of substance use initiation. Initiation was defined as trying any nonprescribed substance by age 12 years. “That’s including a single sip of alcohol or puff of a cigarette,” said Dr. Green.
In addition to alcohol, nicotine, and cannabis, researchers looked at initiation of synthetic cannabinoids, cocaine, methamphetamine, and ketamine, among other substances.
Self-reported measures included demographic characteristics, self and peer involvement with substance use, parenting behaviors, mental and physical health, and culture and environmental factors.
The analytical approach used machine-learning algorithms to compare the ability of domains to identify the most critical risk factors. Magnitudes of coefficients were used to assess variable importance, with positive coefficients indicating greater likelihood of substance initiation and negative coefficients indicating lower likelihood of initiation.
By age 12 years, 14.4% of the children studied reported substance initiation. Alcohol was the substance most commonly initiated (365 individuals), followed by nicotine (94 individuals) and cannabis (40 individuals), with few or no children initiating other substances.
Both those who did and did not initiate substances were similarly aged, and most participants identified as White and non-Hispanic. But the substance-use group had a lower percentage of girls and higher percentage of White participants compared with the no-substance-use group.
The model with only self-reported data had similar accuracy in predicting substance use initiation (area under the curve [AUC], 0.67) as models that added resource-intensive measures such as neurocognitive tests and hormones (AUC, 0.67) and neuroimaging (AUC, 0.66).