Predicting Later Adolescent Risk Behaviors: Which Proximal Outcomes Matter?

Area of Science:

• Adolescent Health
• Behavioral Science
• Public Health Interventions

Background:

• Prevention programs often target adolescent risk behaviors, but effectiveness is typically measured by long-term outcomes.
• There is a need for evidence on proximal outcomes that predict later adolescent risk behaviors.
• Relationship education programs aim to influence adolescent development before risk behaviors emerge.

Purpose of the Study:

• To analyze longitudinal data to identify proximal outcomes that predict later adolescent sexual behavior and substance use.
• To evaluate the effectiveness of relationship education programs by examining intermediate outcomes.

Main Methods:

• Longitudinal survey data from ~1,100 adolescents (grades 9-12) who participated in a relationship education program.
• Regression models assessed associations between 30 posttest proximal outcomes and nine later behavioral outcomes (sexual behavior, substance use) over six months.
• Exploratory analyses focused on effect-size magnitude, directional consistency, and theoretical coherence.

Main Results:

• Twenty-two proximal outcomes significantly predicted later behavioral outcomes.
• Risky social media behavior was a consistent cross-domain risk factor.
• Self-efficacy in relationship skills and self-awareness were protective for sexual behavior; intentions to have sex showed unexpected protective associations.

Conclusions:

• Proximal outcomes, such as self-efficacy and self-awareness, can serve as useful predictors of adolescent risk behaviors.
• Findings support the use of proximal outcome assessment in evaluating prevention programs when distal outcomes are difficult to measure.
• Risky social media behavior is a critical area for intervention and future research in adolescent risk prevention.