Table 1 includes demographics stratified by sexual identity across study variables. LGB participants were 11% of the sample. Among LGB participants, 74% identified as bisexual. Most (73%) LGB participants were assigned female sex at birth. LGB participants tended to be older and in higher grades compared to heterosexual participants. The majority of participants were Black or Latina/o/x in both groups. There was a greater proportion of White and Asian participants in the heterosexual than LGB group.

Figure 1 shows the trends in past-year suicidal ideation and attempt from 2009 to 2019 across race/ethnicity and sexual identity. During this period, the weighted overall prevalence of suicidal ideation and attempt among LGB participants was 22% and 32%, respectively, and 11% and 6% among heterosexual participants, respectively. Adjusted Odds Ratios (AOR) for suicidal ideation and attempt across race and ethnicity and sexual identity are included in Table 2.

Yearly suicidal ideation and attempt means by race/ethnicity and sexual identity

For suicidal ideation among LGB participants, Black and Latina/o/x LGB participants showed significant linear increases, with a quadratic increase and nonsignificant quadratic change, respectively. White LGB participants showed a significant linear decrease with a quadratic increase. Asian LGB participants showed a nonsignificant linear change. For suicide attempts among LGB participants, all racial/ethnic groups showed a significant linear decrease except for Black LGB participants, for whom there was a linear and quadratic increase. Latina/o/x and White LGB participants showed quadratic decreases. 

For suicidal ideation among heterosexual participants, all ethnic/racial heterosexual groups showed linear increases, though linear increases among Black and Asian heterosexual participants were 4–5 times greater than those for White and Latina/o/x heterosexual participants. Increases among Asian and Latina/o/x heterosexual participants showed a quadratic increase, while those among White heterosexual participants showed a quadratic decrease. For suicide attempts among heterosexual participants, all racial/ethnic heterosexual groups showed significant linear increases except for Latina/o/x heterosexual participants, for whom there was a decrease. All heterosexual groups showed significant quadratic increases except for White heterosexual participants, for whom there was a quadratic decrease.

The weighted overall prevalence for bullying at school and e-bullying among LGB participants was 22% and 21%, respectively, and 12.5% and 10.7% among heterosexual participants, respectively. The disaggregated mean trends for bullying at school and e-bullying are depicted in Fig. 2.

Yearly bullying at school and e-bullying means by race/ethnicity and sexual identity

For bullying at school among LGB participants, Black (AOR = 1.06, SE = 0.004, p < 0.001) and Latina/o/x (AOR = 1.01, SE = 0.001, p < 0.001) LGB participants showed significant linear increases with a quadratic increase (AOR = 0.996, SE = 0.00, p = 0.007) and decrease (AOR = 1.01, SE = 0.001, p < 0.001), respectively. White LGB participants showed a significant linear decrease (AOR = 0.98, SE = 0.007, p < 0.001) with a quadratic decrease (AOR = 0.99, SE = 0.001, p = 0.001). Asian LGB participants showed a nonsignificant linear change (AOR = 1.00, SE = 0.008, p = 0.78). For e-bullying among LGB participants, Black (AOR = 1.04, SE = 0.004, p < 0.001) and Latina/o/x (AOR = 1.01, SE = 0.003, p < 0.001) LGB participants showed significant linear increases with a quadratic increase (AOR = 0.99, SE = 0.001, p < 0.001) and decrease (AOR = 1.01, SE = 0.001, p < 0.001), respectively. White LGB participants (AOR = 0.94, SE = 0.007, p < 0.001) and Asian LGB participants (AOR = 0.91, SE = 0.008, p < 0.001) showed significant linear decreases with a nonsignificant quadratic change (AOR = 1.01, SE = 0.003, p = 0.08) and a quadratic decrease (AOR = 0.97, SE = 0.002, p < 0.001), respectively.

For bullying at school among heterosexual participants, Black (AOR = 1.08, SE = 0.002, p < 0.001), Latina/o/x (AOR = 1.02, SE = 0.001, p < 0.001), and White (AOR = 1.05, SE = 0.002, p < 0.001) heterosexual participants showed significant linear increases. Black (AOR = 0.997, SE = 0.001, p < 0.001), Latina/o/x (AOR = 0.995, SE = 0.001, p < 0.001), and White (AOR = 0.993, SE = 0.001, p < 0.001) heterosexual participants also showed significant quadratic decreases. Asian heterosexual participants showed linear (AOR = 1.03, SE = 0.002, p < 0.001) and quadratic (AOR = 1.01, SE = 0.001, p < 0.001) increases. For e-bullying among heterosexual participants, Black (AOR = 1.04, SE = 0.002, p < 0.001), Latina/o/x (AOR = 1.01, SE = 0.001, p < 0.001), White (AOR = 1.01, SE = 0.002, p < 0.001), and Asian (AOR = 1.06, SE = 0.002, p < 0.001) heterosexual participants showed significant linear increases. Black (AOR = 1.01, SE = 0.001, p < 0.001), Latina/o/x (AOR = 1.01, SE = 0.001, p < 0.001), and Asian (AOR = 1.01, SE = 0.001, p < 0.001) heterosexual participants showed quadratic increases, but White (AOR = 0.998, SE = 0.002, p = 0.23) heterosexual participants showed a nonsignificant quadratic change. For the models that examined associations between suicidal ideation and attempt and bullying at school and e-bullying, both forms of bullying were positively associated with suicidal ideation and attempt across all groups (Table 3).