The most important predictor for SUD later in life is early initiation of substance use during adolescence, thus prevention programs must target youth in delaying the onset of age of initiation. However, the interaction between both risk and protective factors on early initiation of substance use is not well studied. Therefore, this cross sectional study used two different types of modeling to examine these associations, the multivariate regression analysis and multiple correspondence analysis (MCA). The MCA allows for testing of interdependence between risk factors and protective factors while the multivariate regression model analyzes the effects of one variable on another. Using Communities that Care Youth Survey, data from 1,272 adolescents aged 12 to 19 years of age were analyzed with risk factors and protective factors grouped into the following three categories: family, school, and community dynamics.
The MCA found high risk and low protection factors grouped closer to early initiation while low risk and high protection factors grouped closer to late initiation of substance use. Clusters were stronger around risk factors than for protective factors. When examining the data separately among family, social, and community dynamics, social dynamics were the most homogeneous and linear for risk and protective factors while community dynamics were less linear among all three dynamics. According to the distance analysis, the top five factors associated with early initiation were: perceived availability of drugs (RF), laws and norms favorable to drug use (RF), rewards for prosocial involvement in the community (PF), negative relationships with peers at school (RF), and low commitment to school (RF), suggesting that prevention should focus on social and community dynamics. In contrast, family dynamics were among the least associated factors with early initiation and composed of family attachment (PF), family history of antisocial behavior (RF), and opportunities for prosocial involvement in the family (PF). Note: effect sizes cannot be determined from this distance analysis.
In the multiple linear regression analysis, the strongest factors associated with early initiation were related to family dynamics. The five factors with the largest effects were parental positive attitudes towards drug use (RF), perceived availability of drugs (RF), family conflict (RF), family history of antisocial behavior (RF), and poor family management (RF).
The risk factor, perception of availability of drugs, was found to be a factor for early initiation in both models suggesting this should be of high priority for health policy. This perception of availability of drugs is influenced at various levels such as policing and should be taken into account when discussing legalization and decriminalization.
The two models differed in their results with the MCA model suggesting that early initiation is associated with social and community dynamics while the multivariate model suggests family dynamics play a larger role. The researchers indicate social and community dynamics may be associated with early initiation as adolescents may seek identity outside of the family and thus be more prone to such influences. In contrast, research on adolescents in Colombia have found that parental supervision and family members not using drugs were associated with resistance to initiate drug use. It is important to take the interdependence of various factors into play when targeting early initiation of substance use.
Trujillo, C. A., Obando, D., & Trujillo, A. (2019). An examination of the association between early initiation of substance use and interrelated multilevel risk and protective factors among adolescents. PLoS ONE, 14(12). https://doi.org/10.1371/journal.pone.0225384