Past 6-Year Trends in Current Alcohol Use among Cyberbullied Adolescents

ABSTRACT

Background

Most of the research on alcohol use and being cyberbullied has focused on the correlation between the two. Less is known about the recent epidemiology of adolescent current alcohol use. Objective: To describe the trends in current alcohol use among adolescence that self-report being cyberbullied across sex and racial/ethnic groups. Methods: The data from the present study came from the national Youth Risk Behavior Surveys (YRBS) from 2011 to 2017. We estimated the prevalence of current alcohol use across years by biological sex and race/ethnicity. Further we tested for linear and quadratic trends. Results: We observed statistically significant decreases in alcohol use cyberbullied (b= −0.12, p < 0.00), and non-cyberbullied individuals (b=-0.15, p < 0.00). Further, we found significant decreases for cyberbullied males (b= −0.10, p < 0.01), females (b= −0.17, p < 0.01) blacks (b= −0.39, p < 0.00), and Hispanics (b= −0.17, p < 0.01). Whites and other races were did not have a significant change. We also found significant decreases for cyberbullied white males and females, black males (b=-0.46, p < 0.03) and females (b= −0.37, p < 0.02), Hispanic males (b=-0.33, p < 0.00). White males and females and other males and females did not have significant changes in alcohol use prevalence. Conclusions: Consistent with national trends, alcohol use among adolescents is decreasing. The decrease is occurring within those that are being cyberbullied. Further research with different data are necessary to further validate these results.

KEYWORDS:

• Cyberbullying
• alcohol use
• adolescents

Notes

1 To handle missing data, the models took advantage of the multi-stage cluster design (CDC, 2017a). This includes weights that handle non-response to the items (i.e., missing data). This is consistent with Schneider et al. (2018) use of these data conducting a trend analysis of cocaine use.

2 The authors understand that statistical significance is more likely given the sample size. As one reviewer pointed out, discussing the effect sizes might be more fruitful. Our use of logistic regression is only designed to determine the shape of the trend, and not a substantive interpretation of the relationship between the variables. We, thus, chose not to interpret the effect sizes to stay consistent with the CDC’s (2017a) guidelines for performing trend studies with their data.