Purpose: This study aims at identifying the sex-stratified associations of involvement in traditional bullying during middle and high school years and in cyberbullying during college years with multiple health risk behaviors in undergraduate students.

Design: This cross-sectional analysis draws on the data of the second wave of the LATO study (Lifestyle & Attitudes in a Student Population) in Greece.

Methods: During November and December 2013, 812 second-year undergraduate students (mean age = 19.3 years; girls = 66.1%) provided data on substance use (smoking, alcohol abuse or drunkenness, illegal drug use including marijuana, hashish, and cannabis) and sexual risk taking (paying for sex and not using condoms) and completed the Cyberbullying and its Effects and the Retrospective Bullying Questionnaires. Logistic regression models performed were adjusted for potential confounders.

Findings: Both male and female late adolescents who were victims of bullying during middle and high school were less likely to use condoms during college years when compared to uninvolved students. Among males, being a bully or victim at school doubled the odds for past month drunkenness and tripled the odds of paying for sex. Greater likelihood to pay for sex was also evident in bullying victims. Cyberbully or cybervictim male students were more likely to report smoking. In female bullying victims, alcohol abuse associations were somewhat conflicting, with decreased lifetime but increased past month likelihood for drunkenness.

Conclusions: Engagement in bullying and cyberbullying is associated with the manifestation of gender-specific health risk behaviors for the different involvement groups in college students.

Clinical relevance: Involvement in bullying and cyberbullying is a major public health concern due to the associations with multiple health risk behaviors. Nurses and healthcare professionals should adopt multifaceted prevention interventions tailored according to bullying status and gender that extend through all educational levels.