LONGITUDINAL HEALTH ENDANGERING BEHAVIOR RISK AMONG RESILIENT AND NONRESILIENT EARLY ADOLESCENTS

Purpose: To explore the relative likelihood of engaging in new health- endangering behaviors among a group of resilient early adolescents com pared to a sample of nonresilient peers and a sample of normal low-ris k peers in a nonclinical, school-based setting. Methods: Resilient ado lescents and their peer groups were identified by way of a multiple li near regression equation in which age, family structure (single or ste pparent family), gender, self-injurious behaviors, and emotional risk were used to predict propensity to initiate risky health behaviors. Th e resilient sample consisted of those adolescents who were predicted t o be above the standardized mean, yet actually scored below it. The no nresilient population included those who were predicted to and actuall y scored above the standardized mean. The normal, low-risk population consists of adolescents who were predicted to and scored below the sta ndardized mean. The mean age for all populations was 13.78 years. All students completed a Health Behavior Questionnaire and the Rosenberg S elf-Esteem Inventory. Results: Odds ratios with 95% confidence interva ls revealed that in the year following identification as resilient, no nresilient, or normal, the resilient adolescents were less likely than the nonresilient adolescents to initiate a variety of risky behaviors . At the same time, the resilient adolescents were more likely than th eir normal, not at-risk peers to have initiated those same risky behav iors. The resilient adolescents have modestly higher mean self-esteem than the nonresilient peers (t = 2.47, p < 0.05) but lower self-esteem than their normal, not at-risk peers (t = 3.66, p < 0.01). Conclusion s: Determination of resilience status by way of multiple linear regres sion yielded identifiable groups which conformed to expected elevated risk of initiating new risky behaviors relative to normal, not at-risk peers but lowered risk relative to nonresilient peers. Differences we re most notable with reference to new reports of substance use. The lo wer rate of initiating new risky behaviors among resilient relative to nonresilient peers is seen as a reflection of behavioral competence i n an adverse context. However, the elevated rate of initiating new ris ky behaviors among resilient relative to normal, not at-risk peers is seen as a reflection of the continuing, negative impact of that advers e context. (C) Society for Adolescent Medicine, 1998