Suicidal Thoughts and Behaviors Among High School Students — Youth Risk Behavior Survey, United States, 2021

Methods

Data Source

This report includes data from the 2019 (N = 13,677) and 2021 (N = 17,232) YRBS, a cross-sectional, school-based survey conducted biennially since 1991. Each survey year, CDC collects data from a nationally representative sample of public and private school students in grades 9–12 in the 50 U.S. states and the District of Columbia. Additional information about YRBS sampling, data collection, response rates, and processing is available in the overview report of this supplement (5). The prevalence estimates for suicidal thoughts and behaviors for the overall study population and by sex, race and ethnicity, grade, and sexual identity are available at https://nccd.cdc.gov/youthonline/App/Default.aspx. The full YRBS questionnaire, data sets, and documentation are available at https://www.cdc.gov/healthyyouth/data/yrbs/index.htm. This activity was reviewed by CDC and was conducted consistent with applicable federal law and CDC policy.*

Measures

Four questions about suicidal thoughts and behavior are the focus of this report. The first asked, “During the past 12 months, did you ever seriously consider attempting suicide?”; the second, “During the past 12 months, did you make a plan about how you would attempt suicide?”; the third, “During the past 12 months, how many times did you actually attempt suicide?”; and the fourth, “If you attempted suicide during the past 12 months, did any attempt result in an injury, poisoning, or overdose that had to be treated by a doctor or nurse?” All questions had response options of yes or no except for the question about how many times a student attempted suicide, which was recoded to reflect whether a student had attempted suicide: yes (1 time, 2 or 3 times, 4 or 5 times, or 6 or more times) versus no (0 times). The question about needing medical treatment for a suicide attempt had an additional response option of “I did not attempt suicide during the past 12 months,” which was recoded to no.

Demographic variables included sex, measured as female or male and grade, measured as 9, 10, 11, or 12. Race and ethnicity was coded as a composite of two questions. First, students were asked, “Are you Hispanic or Latino?” This question measured as yes versus no; regardless of how students responded to the race question, those who answered yes were coded as Hispanic or Latino (Hispanic). Second, students were asked, “What is your race? Check all that apply” and coded as American Indian or Alaska Native (AI/AN), Asian, Black or African American (Black), Hispanic, Native Hawaiian or other Pacific Islander (NH/OPI), and White. Students who selected more than one race were coded as multiracial. (Persons of Hispanic origin might be of any race but were categorized as Hispanic; all racial groups were non-Hispanic.) Sexual orientation was measured by sexual identity and sex of sexual contacts. Sexual identity, measured as heterosexual; lesbian, gay, bisexual, questioning, or other (LGBQ+), had new response options in 2021 and was not comparable to the sexual identity measure from 2019. Sex of sexual contacts (“During your life, with whom have you had sexual contact?”) was measured as: “I have never had sexual contact,” “females,” “males,” or “females and males.” Responses were compared with the student’s sex to create the following categories: no sexual contact, opposite sex only (e.g., female students who have only ever had sexual contact with males), same sex only, or both sexes (e.g., female students who reported contact with females only or females and males).

Analysis

Prevalence estimates and 95% CIs were calculated for each of the four suicidal measures, stratified by sex (historically, female youths are more likely to have suicidal thoughts and attempts, whereas male youths are more likely to die by suicide) (6), for the years 2019 and 2021. Sex-stratified prevalence estimates were further stratified by race and ethnicity, grade, sexual identity, and sex of sexual contacts. Using unadjusted logistic regression models with a statement to generate predicted marginal proportions, prevalence difference (PD) and prevalence ratios (PRs) were calculated to detect a difference in prevalence of a suicidal behavior for 2019 as compared with 2021 within a stratum (e.g., AI/AN female students). Using 2021 data only, PR were calculated to detect a difference in the prevalence of a suicidal behavior between subgroups across a demographic characteristic as compared with a referent group. A p value of <0.05 and 95% CI that did not cross the null value of 1.0 were used to determined statistical significance. The absolute value of the prevalence difference is presented. Prevalence estimates with a denominator <30 were considered statistically unreliable and therefore were suppressed (5). All analyses were conducted using SUDAAN (version 11.0.3; RTI International).