Purpose: Suicide is the third leading cause of mortality among adolescents worldwide. Recent research documents that rates of suicidal ideation (SI) and those who develop a plan (SP) to carry out a
suicide attempt, are on the increase. The objective of this study was to examine the social and demographic factors associated with selfreported SI and SP among adolescents in a middle-income sub Saharan
country.

Methods: Data were obtained from the Global School Health Survey. In-school surveys were completed by 1,432 secondary students aged 11-17 in the Republic of the Seychelles. Participants were divided
into three groups: those not reporting a suicidal gesture (Control group, N1190 ); those reporting SI but not SP (SI group; N89), and those reporting both SI and SP (SIP group; N141) each with a
12-month period of recall. We explored sociodemographic variables (age, gender, economic deprivation), mental health status (anxiety, depression, loneliness), relational variables (having close friends,
close relationship with parents) and substance use (tobacco, cannabis, alcohol) to discern whether they were associated with suicidal gestures in the SI and SIP group compared with controls. A multinomial
logistic regression using the control group as the reference category was used while adjusting for covariates to compare groups. Results are reported as Relative Risk Ratios (RRR) with 95% confidence
intervals.

Results: Within the recall period, 16.2% of adolescents reported having had a suicidal gesture whether with SP (10%) or without (6.2%). Compared to controls, those reporting SI were younger
(RRR: 0.80 [0.68/0.94]) and more likely to feel depressed (RRR: 1.80 [1.12/2.91]), lonely (RRR: 3.54 [2.05/6.11]) and to smoke (RRR: 2.18 [1.23/3.85]). Having at least one close friend was a
protective factor (RRR: 0.34 [0.14/0.84]) Compared to controls, SIP were more likely to be males (RRR: 2.11 [1.36/3.27]), anxious (RRR: 3.18 [1.99/5.07]), depressed (RRR: 2.66 [1.78/3.97]) or lonely
(RRR: 1.87 [1.15/3.04]). They were also more likely to be smokers (RRR: 2.38 [1.47/3.85]). Having at least one friend (RRR: 0.34 [0.16/0.73]) and a close relationship with parents (RRR: 0.48 [0.31/
0.75]) were protective factors.

Conclusions: Our results indicate that the prevalence rates of SI and SP are comparable to those in high-income country settings. Our findings indicate that although gender does not seem to have an
effect on SI, SIP is mainly a male behavior and emphasizes the importance of parents and friends as protective factors. These findings underscore a general need for more attention to the mental health
needs of adolescents in middle-income settings and for preventive frameworks.

Sources of Support: This survey was supported by the Ministry of Health of Seychelles. Further partial support to the study came from the World Health Organization and from some other nonprofit organizations.