Prevalence and Correlates of Suicidal Ideation and Deliberate Self-harm in Canadian Adolescents During the Coronavirus Disease 2019 Pandemic

Objective In light of recent evidence that the coronavirus disease 2019 (COVID-19) pandemic has resulted in marked increases in depression, anxiety, substance use, and other mental health concerns among Canadian adolescents, we investigated the rates of self-harm thoughts and behaviours in this population. Specifically, this study explored: (1) the demographic and geographic distributions of suicidal ideation (SI) and deliberate self-harm (DSH), and (2) the associations of mental health and substance use with SI and DSH. Method A total of 809 Canadian adolescents, aged 12–18 years, completed an online survey between June 17, 2020 and July 31, 2020. Results 44% of adolescents reported experiencing SI since the pandemic began, while 32% reported engaging in DSH. SI and DSH were more common among youth who: identified as transgender, non-binary or gender fluid; who did not reside with both parents; and who reported psychiatric concerns or frequent cannabis use. Conclusion Canadian adolescents appear to be experiencing higher rates of self-harm thoughts and behaviours relative to before the COVID-19 pandemic. It is important for adults who are likely to interact with distressed youth to be aware of potential warning signs that a youth is struggling with self-harm, and to refer youth to specialty mental health services where appropriate.


Results
Our sample closely matched known demographics of Canadian youth, with similar proportions of youth identifying as visible minorities (26% vs 27% 2 ), living in two-parent households (70% in both 2 ), reporting any history of alcohol use (42% vs 44% 2 ) and accessing mental health supports (17% in both 2 ). Higher rates of psychiatric concerns in this sample (e.g. major depression: 50%) versus representative pre-pandemic surveys (11% 2 ) may be attributable to use of a screening measure 3 here versus prior diagnoses. 2 Table 1 summarizes results. One-quarter of adolescents (26%, n = 212) reported both SI and DSH, one in six (17%, n = 139) reported SI only, and one in 20 (5%, n = 43) reported DSH only. There were no regional or racial/ethnic differences in SI or DSH. Transgender/non-binary youth reported more SI and DSH than cis-gender youth, and cis-gender females reported more DSH than cis-gender males. Youth who lived with both parents reported less SI and DSH than youth who lived with a single parent, divided time between multiple households, or did not live with parents.
Psychiatric concerns accounted for 43% and 33% of the variance in SI and DSH, respectively. Depression, anxiety, and conduct problems were uniquely associated with SI and DSH, and social anxiety was uniquely associated with DSH. Youth with at least one clinically significant psychiatric concern had 6-9 times greater odds of SI and DSH, versus those who scored below all clinical cutoffs.
Frequency of substance use accounted for 9% and 4% of the variance in SI and DSH, respectively. Only cannabis use was uniquely associated with SI and DSH, controlling for other substances. Youth with any recent alcohol, cannabis, or illicit drug use had 40% greater odds of SI and DSH than youth who reported no alcohol or drug use. Compared to youth who used alcohol only, youth who used cannabis or illicit drugs reported 50% higher odds of SI.

Discussion
Identifying demographic features associated with self-harm can inform targeted assessment and intervention, while identifying modifiable risk factors such as psychiatric concerns can inform what types of supports may be beneficial. Our results identified gender minority youth and youth in singleparent homes as being at especially high risk of SI and DSH, suggesting a need for additional screening of these potentially vulnerable youth. Moreover, we identified depression, anxiety, conduct problems, and frequent cannabis use as associated with risk of SI and DSH, suggesting the potential value of coupling targeted suicide prevention (e.g. safety planning, lethal means restriction) with interventions for internalizing and externalizing concerns.
Several limitations merit consideration. Firstly, convenience samples remain susceptible to self-selection biases and caution must be applied to any survey with unobserved variables influencing participation. Secondly, this study's cross-sectional design prevented comparisons across phases of the pandemic. Thirdly, we did not differentiate DSH with versus without suicidal intent, and so were unable to investigate rates of non-suicidal self-injury. Finally, we note that available data suggest suicide deaths remained steady or slightly declined in the early stages of the pandemic, 5 and trends for suicidal thoughts, non-fatal self-harm, and fatal selfharm often diverge. Ongoing monitoring of population-level indicators across this spectrum of risk is warranted as the pandemic and its aftermath unfold. This study nonetheless augments literature on the well-being of Canadian adolescents during the COVID-19 pandemic and bolsters calls to enhance support for youth in crisis.