This study used a longitudinal design from age six through age 19 (N = 1206 (603 girls)) to examine the associations between anxious-withdrawal and reactive aggression during childhood, distinct profiles of peer victimization from kindergarten to grade 11 and victimization in college or at work in emerging adulthood. In particular, it was tested whether the predictive effect of personal characteristics on victimization in emerging adulthood would be mediated via chronic peer victimization experiences during the school years. Teachers evaluated children’s personal characteristics, whereas peer nominations and self-reports were used to assess victimization. Control variables included sex, parent-reported harsh parenting and SES. Longitudinal latent profile analysis revealed four distinct profiles of peer victimization during the school years: Consistently-Low (39.7%), Low-Moderate (42.8%), High-Decreasing (8.8%) and High-Increasing–Decreasing (8.7%). A subsequent 3-step regression-based path analysis supported the mediation hypothesis – albeit differently for different profiles of peer victimization. Specifically, compared to a Consistently-Low profile of peer victimization in school, a High-Decreasing profile was predicted by reactive aggression, but not anxious-withdrawal. In contrast, a High-Increasing–Decreasing profile was predicted by reactive aggression and anxious-withdrawal. In turn, elevated peer victimization profiles were associated with higher levels of later victimization in college or at work. The indirect effects linking the childhood behaviors to later victimization in college or at work – via elevated peer victimization profiles during childhood and adolescence – were significant. These results highlight the need for tailored interventions to optimize reactively aggressive or anxious-withdrawn children’s response strategies to challenging and potentially threatening peer interactions.

本研究使用了从 6 岁到 19 岁的纵向设计（N = 1206 (603 个女孩)) 来检查童年时期焦虑退缩和反应性攻击之间的关联，从幼儿园到 11 年级的同龄人受害的不同概况以及成年后在大学或工作中的受害情况。特别是，测试了个人特征对新兴成年期受害的预测作用是否会通过学年期间的长期同伴受害经历来介导。教师评估儿童的个人特征，而同伴提名和自我报告则用于评估受害情况。控制变量包括性别、父母报告的严厉养育方式和 SES。纵向潜在特征分析揭示了学年期间同伴受害的四种不同特征：持续低 (39.7%)、低-中 (42.8%)、高-减少 (8. 8%) 和高增长 - 下降 (8.7%)。随后的基于 3 步回归的路径分析支持了中介假设——尽管对于不同的同伴受害情况有所不同。具体来说，与学校中同伴受害的持续低配置相比，反应性攻击预测了高减少配置，但不是焦虑退缩。相比之下，反应性攻击和焦虑戒断预测了高增长 - 下降的特征。反过来，较高的同伴受害情况与后来在大学或工作中受害程度较高有关。将童年行为与后来在大学或工作中受害联系起来的间接影响 - 通过在童年和青春期提高同伴受害情况 - 是显着的。