Quarshie et al. BMC Psychiatry (2023) 23:169 BMC Psychiatry https://doi.org/10.1186/s12888-023-04646-7 R E S E A R C H Open Access Adolescent suicidal behaviour in Namibia: a cross-sectional study of prevalence and correlates among 3,152 school learners aged 12–17 years Emmanuel Nii-Boye Quarshie1* , Nutifafa Eugene Yaw Dey1 and Kwaku Oppong Asante1,2 Abstract Background While adolescent suicidal behaviour (ideation, planning, and attempt) remains a global public health concern, available county-specific evidence on the phenomenon from African countries is relatively less than enough. The present study was conducted to estimate the 12-month prevalence and describe some of the associated factors of suicide behaviour among school-going adolescents aged 12–17 years old in Namibia. Methods Participants (n = 4531) answered a self-administered anonymous questionnaire developed and validated for the nationally representative Namibia World Health Organization Global School-based Student Health Survey conducted in 2013. We applied univariate, bivariable, and multivariable statistical approaches to the data. Results Of the 3,152 analytical sample, 20.2% (95% confidence interval [CI]: 18.3–22.2%) reported suicidal ideation, 25.2% (95% CI: 22.3–28.4%) engaged in suicide planning, and 24.5% (95% CI: 20.9–28.6%) attempted suicide during the previous 12 months. Of those who attempted suicide, 14.6% (95% CI: 12.5–16.9%) reported one-time suicide attempt, and 9.9% (95% CI: 8.1–12.1%) attempted suicide at least twice in the previous 12 months. The final adjusted multivariable models showed physical attack victimisation, bullying victimisation, loneliness, and parental intrusion of privacy as key factors associated with increased likelihood of suicidal ideation, planning, one-time suicide attempt, and repeated attempted suicide. Cannabis use showed the strongest association with increased relative risk of repeated attempted suicide. Conclusion The evidence highlights the importance of paying more attention to addressing the mental health needs (including those related to psychological and social wellness) of school-going adolescents in Namibia. While the current study suggests that further research is warranted to explicate the pathways to adolescent suicide in Namibia, identifying and understanding the correlates (at the individual-level, family-level, interpersonal-level, school context and the broader community context) of adolescent suicidal ideations and non-fatal suicidal behaviours are useful for intervention and prevention programmes. *Correspondence: Emmanuel Nii-Boye Quarshie enquarshie@ug.edu.gh; enquarshie@gmail.com Full list of author information is available at the end of the article © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Quarshie et al. BMC Psychiatry (2023) 23:169 Page 2 of 12 Keywords Adolescents, Attempted suicide, Global school-based student health survey, Ideation, Namibia, Suicidal behaviour, Suicide Background divorce, physical abuse victimisation, conflict between The World Health Organization (WHO) defines suicidal parents, child marriage, parental/family poverty) [8, behaviour as “a range of behaviours that include think- 10, 12], interpersonal-level (e.g., lack of peer support, ing about suicide (or ideation), planning for suicide, romantic relationships problems, breakups, sexual abuse attempting suicide and suicide itself” [1]. Suicidal ide- victimisation) [8, 10, 12], school context (bullying victimi- ation, suicidal planning, and (repeated) attempted suicide sation, peer support at school, poor school climate, peer represent important risk factors for suicide mortality in suicide or attempted suicide, and poor academic perfor- the general population [1–3]. Globally, 703,000 suicidal mance, truancy) [8, 12], and the broader community-level deaths are recorded annually, representing more than one – e.g., community violence or war, poverty [10, 12–14]. in every 100 deaths in 2019 [4]. Among young persons The multi-layered and multi-contextual nature of the fac- aged 15–19 years, suicide was the third leading cause of tors associated with adolescent suicidal behaviour can be death among girls (after maternal conditions) and the understood within the socio-ecological model. The socio- fourth leading cause of death in boys, after tuberculosis ecological model provides a helpful framework to under- – as at the end of 2019 [4]. About 79% of the world’s sui- standing and preventing (adolescent) suicidal behaviour cides are recorded in low- and middle-income countries in that the model considers an integration of population- (LMICs), with the Africa region recording the highest specific and general risks and protective factors [1, 15, rate – 11.2 per 100,000 people [4, 5]. However, national- 16]. level representative data on suicidal ideation, planning, Within Southern sub-Saharan Africa, South Africa and attempts, and their associated factors are still less remains the only country with relatively large data on than enough to support research-informed intervention adolescent self-harm and suicidal behaviour [10, 17–19]. and prevention efforts and programmes, particularly in Hence, towards contributing evidence to addressing the LMICs, including those in Africa [1]. dearth of research on adolescent suicidal behaviour in Evidence from a recent global systematic review and Southern sub-Saharan African countries (apart from meta-analysis suggests varying 12-month prevalence esti- South Africa), several scholars have explored and pub- mates of suicidal behaviour among adolescents: suicide lished evidence from the WHO-GSHS data accessed ideation = 14.2% (95% confidence interval [CI] = 11.6– from some countries in the sub-region: Botswana [20], 17.3%), suicidal planning = 7.5% (CI = 4.5–12.1%), and Eswatini [14], Malawi [21], Mauritius [22], Mozambique suicide attempt = 4.5% (95% CI = 3.4–5.9%) [6]. Com- [23, 24], Zambia [25], and Zimbabwe [26]. Thus far, it is paratively, pooled regional rates drawing on published only the published study by Peltzer and Pengpid (2017) data from the World Health Organization Global School- drawing on the Namibian WHO-GSHS data that reports based Student Health Survey (WHO-GSHS) indicate evidence on suicide attempt [27]. In other words, no that the 12-month prevalence estimates of adolescent peer-reviewed publication is available on the secondary suicidal behaviour in LMICs are higher in the African analysis of the country-specific prevalence and corre- region: ideation = 21% (95%CI = 20.1–21.0%), suicidal lates of suicidal ideation, planning, and (repeated) suicide plan = 23.7% (95%CI = 19.1–28.3%), and attempt = 16.3% attempt in the nationally representative school-going (95% CI = 8.4–29%) [7–9]. Notably, the 12-month median adolescents sample (aged 17 years and younger) of the prevalence estimate of adolescent self-harm in Southern Namibian WHO-GSHS data. It must be noted that the sub-Saharan Africa – where Namibia is located – (16.5% 2013 Namibian WHO-GSHS data is the latest available interquartile range [IQR] = 10.9–24.0%) is comparable to dataset from the country’s participation in the survey. the overall estimate across the sub-Saharan region (16.9% Namibia has a youthful population, as persons aged 17 [IQR] = 11.5–25.5%) [10]. years and younger constitute 43% of the general popula- Factors associated with suicidal behaviour among tion [28]. The mean years of schooling in Namibia is 7.2 adolescents in LMICs, including those in (sub-Saharan) years [29]. Primary to middle school education, which Africa, have been found to exist at the individual-level stretches from grade 1 through 7, is compulsory for all (e.g., [female] gender, low self-worth, hopelessness, HIV/ children between the ages of 6 and 16 years, while sec- AIDS and other chronic medical conditions in adoles- ondary education remains optional [30]. Namibia is an cents, alcohol and drug [mis]use, depression, and anxiety English-speaking Southern sub-Saharan African country and other untreated psychiatric conditions) [7, 8, 10, 11], classified as an upper-middle-income country [31], with a family-level (e.g., parental understanding and support, Medium Human Development Index rank of 139 [29]. In food insecurity [hunger], conflict with parents, parental 2019, the country’s age-standardised, all ages suicide rate Quarshie et al. BMC Psychiatry (2023) 23:169 Page 3 of 12 was 13.5 per 100,000 people, higher than both the Africa complete. A total of 4531 students aged 13-18years par- rate (11.2 per 100,00 people) and global average (9.0 per ticipated in the Namibia GSHS. The school response rate 100, 000 people) [4]. was 100%; that of students was 89%, while the overall Thus, the present study seeks to analyse the 2013 response rate was 89%. The reporting of the current study Namibian WHO-GSHS data on non-fatal suicidal behav- was guided by the recommendations of Strengthening iours to address the following research aims: the Reporting of Observational Studies in Epidemiology 1]. Estimate the 12-month prevalence of suicidal (STROBE) [34]. behaviour (ideation, planning, and attempt) among school-going adolescents 12–17 years in Namibia. Measures 2]. Describe some of the commonly reported factors Outcome variables at the personal/lifestyle-, family-, school-, and Three domains of suicidal behaviour namely suicide ide- interpersonal relationship-levels associated with ation, planning and attempt were considered as the out- suicide ideation, planning, and (repeated) suicide come variables for this study. The three domains were attempt among school-going adolescents (aged each assessed using a single-item question. Suicide ide- 12–17 years) in Namibia. Considering the nature ation was measured with, “during the past 12 months, of the data analysed, his study considers suicidal did you ever seriously consider attempting suicide?”, behaviour to comprise suicidal ideation, planning, suicide planning with, “during the past 12 months, did and attempt – excluding suicidal death. you make a plan about how you would attempt suicide?” and attempt with “during the past 12 months, how many Methods times did you actually attempt suicide?”. The responses Context and data source for suicide ideation and plan were, “Yes = 1” and “No = 0” This study drew on data from the 2013 Namibia WHO- while that of attempt required students to indicate the GSHS. The survey was conducted by WHO and the Cen- number of times with “0”, “1”, “2 or 3”, “4 or 5”, and “6 or ters for Disease Control and Prevention (CDC) of the more times”. Guided by the GSHS recoding procedures United States in collaboration with the Government of [33], the suicide attempt variable was treated as a binary Namibia [32]. The data is publicly available and has been variable assigning “0” to no attempt and “1” to one or accessed freely for the current study from the WHO web- more attempts. For the purposes of examining factors site [32]. responsible for repeated suicide attempts, the attempt variable was subsequently reclassified into three catego- Study design and sampling ries namely, no attempt, one-time attempt, and repeated The WHO-GSHS is a national representative cross- attempts. Therefore, responses of adolescents who chose sectional survey conducted to assess behavioural health “1” for suicide attempt was recoded into the one-time factors amongst school-going adolescents (mainly aged attempt category and that of adolescents who chose “2 13 to 17) in participating WHO member countries [33]. and above” into the repeated attempts category. Coding Data was collected using a validated self-administered of variables included in this study (i.e., demographic vari- questionnaire consisting of items assessing a wide range ables, exposure factors, and missing data) are presented of personal lifestyle (e.g., alcohol and drug use), family as supplementary material (e-Table 1). relationships (e.g., parental supervision and understand- ing), and school environment variables (e.g., truancy Exposure variables and bullying) [33]. Prior to the data collection, ethical Participants’ demographic characteristics, mental health approval was sought from relevant authorities as well as and lifestyle factors, interpersonal factors, school-level consent from parents/guardians of adolescents. The 2013 factors as well as family-level factors were selected as Namibia GSHS targeted students in grades 7–12 which is exposure variables in this study. Besides performing typically attended by students aged 13–17 years. A two- bivariable analyses to assess the relationships between stage sampling approach was used for data collection. In the exposure and outcomes variables, the selection of the first stage, a cluster of schools were randomly selected these exposure variables was based on evidence from from a list of all schools in Namibia using a probability previous studies within sub-Saharan Africa drawing data proportionate to enrollment size method. This process from the WHO-GSHS [14, 27, 35–37]. Examples of the resulted in a list of eligible and nationally representative specific factors include age, school grade, gender, can- schools. The second stage involved the random sampling nabis use, loneliness, anxiety, and alcohol use. Complete of classrooms from these eligible schools, making all stu- details of the variables, their groupings, survey ques- dents in the selected classes eligible to participate. Stu- tions and coding can be found in supplementary material dents who volunteered to participate were then handed (e-Table 1). an anonymised computer scannable questionnaire to Quarshie et al. BMC Psychiatry (2023) 23:169 Page 4 of 12 Statistical analyses Table 1 12-month prevalence estimates of suicidal behaviour Reporting of the statistical analyses plan in this study is Suicidal behaviour Total sample Male Female informed by the Statistical Analyses and Methods in the % (95%CI) % (95% CI) % (95%CI) Published Literature (SAMPL) guidelines [38]. We lim- Suicide ideation 20.2 20.0 (17.0-23.4) 20.3 (18.3–22.2) (18.4–22.4) ited the analysis to participants aged 12 to 17 years for Suicide plan 25.2 25.7 (21.8–30.1) 24.8 two reasons: first, most of the participants were within (22.3–28.4) (22.0-27.9) this age range, and second, data on the precise ages below One-time suicide 14.6 15.2 (12.6–18.4) 14.0 and above this age range were not available. Age and gen- attempt (12.5–16.9) (11.9–16.5) der distribution across the analytical sample is presented Repeated suicide 9.9 (8.1–12.1) 10.9 (8.8–13.6) 9.1 as supplementary material (e-Table 2). Statistical analyses attempts (7.2–11.4) involving univariate, bivariate and multivariate testing Suicide attempt 24.5 26.2 (21.6–31.4) 23.1 were conducted in Stata 14.0 statistical software (Stata- (overall) (20.9–28.6) (19.7–26.9) Corp LP, College Station, Texas, USA). Prior to conduct- ing these tests, clusters, stratification, and sample weights or more hours a day engaged in leisure time or sedentary characteristic of data collected with complex designs behaviour. Additionally, about 29% and 5% had use alco- were adjusted to account for possible analytical errors hol and cannabis in the past month, respectively. and make appropriate inferences [39]. Following from this, univariate analysis computing frequencies, propor- Prevalence estimates tions and relevant 95% confidence intervals of all study Table  1 presents comparable 12-month prevalence esti- variables was done. Chi-square tests (χ2) of independence mates of suicidal ideation (20.2% [95% CI = 18.3–22.2%]), were performed to examine the association between the plan (25.2% [95% CI = 22.3–28.4%]), one-time attempt exposure and outcomes variables. Multivariate analy- (14.6% (95% CI = 12.5–16.9%]) and repeated attempt ses with logistic regression and multinomial logistic (9.9% (95% CI = 8.1–12.1%]) across the total analytic sam- regression were then conducted in two steps. Step 1: ple. Overall, 24.5% (95% CI = 20.9–28.6%) of the analytic Logistic regression was performed to assess the sociode- sample reported attempted suicide during the previous mographic factors and exposures variables associated 12 months. Similarly, the 12-month prevalence estimates with suicidal ideation, planning and attempt. Step 2: We of suicidal ideation, plan, one-time and repeated attempt performed a multinomial logistic model assessing the were comparable between boys and girls, as the CIs of factors associated with repeated suicide attempt (keep- these estimates overlapped. ing the base at ‘no attempt’). Given the presence of sparse data only exposure variables that reached statistical sig- Bivariate associations nificance in the binary logistic regression models were As shown in Table 2, most of the exposure variables were included in the multinomial logistic model. Age, gender significantly related to suicidal behaviour. Of the two and school grade were included as covariates in the mul- demographic factors (gender, and school grade) included tivariable analyses. Statistical significance was set at an in the study, only, school grade was significantly related alpha of 0.05. Adjusted odds ratios (AOR) are reported to suicidal behaviour. A high proportion of students for each logistic model and adjusted relative risk ratio in grade 6–9 (compared to students in grades 10–12) (ARRR) for the multinomial logistic model. Given the reported suicide ideation, plan and attempt during the arbitrary nature of the alpha (p < 0.05), the significance of previous 12 months. Among the mental health and life- each analysis was also determined based on CIs and their style factors, loneliness, anxiety, alcohol use and canna- clinical importance [40, 41]. Missing responses or incom- bis use were all significantly related to suicidal behaviour. plete data on key variables were excluded from the final Leisure-time sedentary behaviour was significantly analysis (e-Table 1). related to only suicidal ideation (χ2 = 4.6, df = 1, p = 0.01). Under the interpersonal factors, being sexually active and Results physical fight were related to suicide behaviour. Having Participant characteristics a close friend was significantly related to suicide plan. A total of 3152 in-school adolescents comprising 1380 Concerning school-level factors, a significant propor- (44.3%) male and 1772 (55.7%) females were included tion of students who were physically attacked, truant and as analytical sample in the secondary analysis for this were victims of bullying reported experiences of suicidal study. The average age of these adolescents was M = 15.1 behaviour. Some significant associations were also found (SD = 1.4) and most of them were in grade 10–12 (60.5%). among the family-related factors (e.g., intrusion of pri- Regarding mental health and lifestyle factors, approxi- vacy by parents, food insecurity, and parental supervi- mately 14 and 13% of adolescents had experienced loneli- sion) and suicidal behaviour. Notably, food insecurity and ness and anxiety respectively and about 36% had spent 3 Quarshie et al. BMC Psychiatry (2023) 23:169 Page 5 of 12 Table 2 Bivariate analysis of the factors associated with suicidal behaviour Study Variable Total Suicidal ideation Suicidal plan Suicidal attempt sample N = 3152 Yes No χ2 (p) Yes No χ2 (p) Yes No χ2 (p) n (%) n (%) n (%) n (%) n (%) n (%) n (%) Demographics Gender 0.04 0.33 (0.61) 4.0 (0.08) (0.87) Male 1380 (44.3) 278 (20.0) 1102 (80.0) 362 (25.7) 1018 (74.3) 365 (26.2) 1015 (73.8) Female 1772 (55.7) 365 (20.3) 1407 (79.7) 457 (24.8) 1315 (75.2) 429 (23.2) 1343 (76.8) School grade 11.1 47.2 68.8 (0.001) (< 0.001) (< 0.001) Grade 6–9 1918 (60.5) 423 (22.1) 1495 (77.9) 576 (29.3) 1342 (70.7) 576 (29.6) 1342 (70.4) Grade 10–12 1183 (39.5) 209 (17.2) 974 (82.8) 224 (18.4) 959 (81.6) 202 (16.5) 981 (83.5) Mental health and lifestyle factors Loneliness 48.3 42.7 44.7 (< 0.001) (< 0.001) (< 0.001) Yes 449 (14.3) 146 (32.3) 303 (67.7) 171 (37.7) 278 (62.3) 170 (37.0) 279 (63.0) No 2673 (85.8) 488 (18.1) 2185 (81.9) 643 (23.2) 2030 (76.8) 614 (22.3) 2059 (77.7) Worry/anxiety 38.7 50.0 51.4 (< 0.001) (< 0.001) (< 0.001) Yes 435 (13.8) 138 (31.3) 297 (68.8) 169 (39.0) 266 (61.0) 167 (38.2) 268 (61.8) No 2699 (86.2) 499 (18.3) 2200 (81.7) 647 (23.1) 2052 (76.9) 620 (22.2) 2079 (77.8) Alcohol use 8.6 4.9 (0.02) 3.6 (0.01) (0.04) Yes 860 (29.3) 204 (23.4) 656 (76.6) 247 (28.0) 613 (72.0) 234 (26.6) 626 (73.4) No 2120 (70.7) 400 (18.7) 1720 (81.3) 531 (24.2) 1589 (75.8) 511 (23.3) 1609 (76.7) Leisure-time sedentary behaviour 4.6 5.2 (0.06) 1.5 (0.29) (0.01) Yes 1138 (36.4) 248 (21.9) 890 (78.1) 320 (27.4) 818 (72.6) 301 (25.5) 837 (74.5) No 1966 (63.6) 377 (18.7) 1589 (81.3) 483 (23.8) 1483 (76.2) 475 (23.6) 1491 (76.4) Cannabis use 47.4 25.9 60.4 (< 0.001) (0.001) (< 0.001) Yes 134 (4.6) 58 (42.4) 76 (57.6) 59 (42.8) 75 (57.2) 72 (50.8) 62 (49.2) No 2909 (95.4) 552 (18.7) 2357 (81.3) 715 (23.8) 2194 (76.2) 668 (22.3) 2241 (77.7) Interpersonal factors Being sexually active 17.9 16 (0.004) 10.4 (0.001) (0.01) Yes 958 (33.5) 228 (24.0) 730 (76.0) 456 (22.2) 1495 (77.8) 260 (25.6) 698 (73.4) No 1951 (66.5) 349 (17.4) 1602 (82.6) 279 (29.1) 679 (70.9) 431 (21.2) 1520 (78.8) Close friends 2.1 (0.20) 7.4 (0.03) 0.2 (0.71) None 403 (13.0) 93 (22.7) 310 (77.3) 125 (30.6) 278 (69.5) 105 (25.1) 298 (74.9) One or more 2701 (87.0) 536 (19.7) 2165 (80.3) 677 (24.2) 2024 (75.8) 670 (24.2) 2031 (75.8) Physical fight 28.3 67.7 127.7 (< 0.001) (< 0.001) (< 0.001) Yes 1074 (34.4) 274 (25.5) 800 (74.5) 378 (34.1) 696 (65.9) 402 (36.4) 672 (63.6) No 2065 (65.6) 366 (17.4) 1699 (82.6) 438 (20.6) 1627 (79.4) 358 (18.1) 1680 (81.9) School-level factors Physical attack 51.3 81.2 184.6 (< 0.001) (< 0.001) (< 0.001) Yes 1253 (39.8) 327 (26.5) 926 (73.5) 429 (33.8) 824 (66.2) 472 (37.4) 781 (62.6) No 1859 (60.2) 308 (16.0) 1551 (84.0) 379 (19.5) 1480 (80.5) 313 (16.0) 1546 (84.0) Truancy 23.0 65.5 84.7 (0.001) (< 0.001) (< 0.001) Yes 747 (24.0) 193 (26.2) 554 (73.8) 273 (36.5) 474 (63.5) 279 (36.9) 468 (36.1) No 2355 (76.0) 436 (18.1) 1919 (81.9) 535 (21.7) 1820 (78.3) 497 (20.3) 1858 (79.7) Peer support 0.1 (0.78) 2.1 (0.27) 0.5 (0.52) Quarshie et al. BMC Psychiatry (2023) 23:169 Page 6 of 12 Table 2 (continued) Study Variable Total Suicidal ideation Suicidal plan Suicidal attempt sample Yes 853 (27.1) 179 (20.6) 674 (79.4) 239 (26.9) 614 (73.1) 223 (25.2) 630 (74.8) No 2270 (72.9) 459 (20.0) 1811 (80.0) 570 (24.4) 1700 (75.6) 558 (24.0) 1712 (76.0) Bullying victimisation 36.5 60.5 144.0 (< 0.001) (< 0.001) (< 0.001) Yes 1311 (44.7) 326 (25.3) 985 (74.7) 415 (31.8) 896 (68.2) 448 (34.1) 863 (65.9) No 1577 (55.4) 264 (16.3) 1313 (83.7) 319 (19.2) 1258 (80.8) 247 (15.0) 1330 (85.0) Family-level factors Parental supervision 1.3 (0.40) 23.1 17.5 (0.001) (0.01) Yes 1369 (42.9) 290 (21.0) 1079 (79.0) 412 (29.3) 957 (70.7) 389 (28.1) 980 (72.0) No 1743 (57.1) 342 (19.3) 1401 (80.7) 392 (21.8) 1351 (78.2) 389 (21.5) 1354 (78.5) Parental understanding 4.9 (0.07) 2.3 (0.26) 0.1 (0.71) Yes 1288 (40.9) 238 (18.2) 1050 (81.8) 314 (23.6) 974 (76.4) 315 (23.8) 973 (76.2) No 1794 (59.1) 389 (21.4) 1405 (78.6) 482 (26.0) 1312 (74.0) 450 (24.4) 1344 (75.6) Parental monitoring 1.0 (0.33) 0.1 (0.86) 6.3 (0.03) Yes 1104 (35.1) 214 (19.1) 890 (80.9) 290 (24.8) 814 (75.2) 253 (21.7) 851 (78.3) No 2018 (64.9) 420 (20.6) 1598 (79.4) 516 (25.2) 1502 (74.8) 529 (25.8) 1489 (74.3) Parental intrusion of privacy 16.0 19 (0.002) 41.1 (0.002) (< 0.001) Yes 1764 (56.7) 313 (17.5) 1451 (82.5) 401 (22.1) 1363 (77.9) 363 (20.0) 1401 (80.0) No 1356 (43.3) 319 (23.3) 1037 (76.7) 405 (29.0) 951 (71.0) 416 (30.0) 940 (70.0) Food insecurity 13.9 22.3 38.3 (0.001) (0.001) (< 0.001) Yes 310 (9.9) 87 (28.3) 223 (71.7) 113 (36.3) 197 (63.7) 119 (38.8) 191 (61.2) No 2796 (90.1) 546 (19.3) 2250 (80.7) 692 (24.0) 2104 (76.0) 659 (22.8) 2137 (77.2) Note: Statistically significant results are in bold face parental intrusion of privacy were both related to all the behaviour outcomes, none of these associations reached domains of suicidal behaviour examined. the desired threshold of statistical significance (Table 3). Multivariate associations Multinomial logistic regression Tables 3 and 4 show the findings of the adjusted logistic As shown in Table 4, five factors (physical attack victimi- regression models and multinomial models respectively. sation, bullying victimisation, loneliness, parental intru- sion of privacy, and grade in school) were associated with Logistic regression the likelihood of both one-time attempt and repeated As presented in Table  3, the most important exposure attempted suicide during the previous 12 months. Spe- factors contributing to the increased odds of all the cifically, physical attack victimisation, bullying victimisa- domains of suicide included grade in school, loneliness, tion, loneliness, and parental intrusion of privacy were physical attack and parental intrusion of privacy. Other associated with increased likelihood of both one-time variables including age, anxiety, leisure-time sedentary attempt and repeated attempted suicide. However, ado- behaviour, cannabis use, physical fight, bullying victimi- lescents in grades 10–12 (compared to those in grades sation, and parental monitoring were related to at least 6–9) had reduced relative risk of reporting one-time one form of suicide behaviour. For instance, sedentary attempt (ARRR = 0.50, 95% CI: 0.36, 0.70, p < 0.001) behaviour was associated with increased the odds of or repeated attempt (ARRR = 0.51, 95% CI:0.35, 0.75, only suicidal ideation (AOR = 1.36, 95% CI:1.08, 1.72, p = 0.001). It is also worthy of mention that while the like- p = 0.009), while parental monitoring was related to lihood association between cannabis use and one-time reduced odds of attempted suicide (AOR = 0.73, 95% CI: suicide attempt did not reach the desired threshold of sta- 0.57, 0.94, p = 0.001). Interestingly, although gender, alco- tistical significance, comparatively, the likelihood associ- hol use, being sexually active, number of close friends, ation between cannabis use and repeated suicide attempt truancy, peer support, parental understanding, and showed the strongest statistical and clinical importance food insecurity were associated with each of the suicidal (ARRR = 3.72, 95% CI: 2.08, 6.65, p < 0.001) between the two statistical models. Put differently, students who used Quarshie et al. BMC Psychiatry (2023) 23:169 Page 7 of 12 Table 3 Multivariate logistic regression predicting factors on suicidal behaviour Variable Suicide Ideation (N = 2199) Suicide Plan (N = 2199) Suicide Attempt (N = 2199) β AOR [95% CI] p-value β AOR [95% CI] p-value β AOR [95 CI] p-value Demographics Gender: Female 1 1 1 Male -0.15 0.86 [0.64, 1.15] 0.309 -0.19 0.82 [0.66, 1.03] 0.088 -0.09 0.91 [0.77, 1.08] 0.269 Age 0.12 1.13 [1.04, 1.22] 0.004 0.10 1.11 [0.99, 1.23] 0.061 0.10 1.10 [1.00, 1.21] 0.047 School grade: Grade 6–9 1 1 1 Grade 10–12 -0.41 0.66 [0.52, 0.85] 0.002 -0.64 0.52 [0.40, 0.70] < 0.001 -0.72 0.49 [0.35, 0.69] < 0.001 Mental health and lifestyle factors Loneliness 0.54 1.71 [1.22, 2.39] 0.003 0.54 1.71 [1.27, 2.29] 0.001 0.47 1.60[1.17, 2.18] 0.005 Worry/anxiety 0.40 1.49[1.13, 1.96] 0.007 0.44 1.56 [1.11, 2.18] 0.013 0.31 1.36 [0.94, 1.97] 0.104 Alcohol use 0.15 1.16 [0.81, 1.66] 0.389 0.17 1.19 [0.91, 1.54] 0.189 0.05 1.05 [0.76, 1.46] 0.769 Leisure-time sedentary behaviour 0.32 1.36 [1.08, 1.72] 0.009 0.22 1.25 [0.96, 1.63] 0.090 0.10 1.11 [0.83, 1.49] 0.479 Cannabis use 0.35 1.42 [0.84, 2.42] 0.182 0.48 1.61 [0.99, 2.64] 0.057 0.78 2.18[1.29, 3.67] 0.005 Interpersonal factors Being sexually active 0.20 1.22 [0.93, 1.60] 0.147 0.30 1.34 [0.99, 1.81] 0.054 0.18 1.20 [0.95, 1.53] 0.125 Close friends 0.08 1.08 [0.72, 1.63] 0.694 -0.14 0.87 [0.58, 1.31] 0.480 0.06 1.06 [0.73, 1.54] 0.745 Physical fight 0.18 1.19 [0.84, 1.70] 0.309 0.30 1.35 [1.00, 1.83] 0.048 0.21 1.24 [0.93, 1.65] 0.141 School-level factors Physical attack 0.42 1.53 [1.15, 2.03] 0.005 0.41 1.51 [1.15, 1.99] 0.005 0.69 1.99 [1.56, 2.56] < 0.001 Truancy -0.05 0.96 [0.68, 1.35] 0.788 0.19 1.21 [0.88, 1.66] 0.236 0.17 1.19 [0.83, 1.70] 0.338 Peer support 0.07 1.08 [0.81, 1.44] 0.609 0.25 1.28 [0.96, 1.72] 0.091 0.20 1.22 [0.96, 1.54] 0.098 Bullying victimisation 0.18 1.20 [0.93, 1.54] 0.157 0.35 1.42 [1.12, 1.79] 0.005 0.61 1.83 [1.39, 2.42] < 0.001 Family-level factors Parental supervision 0.07 1.06 [0.82, 1.40] 0.619 0.38 1.46 [1.12, 1.90] 0.006 0.26 1.29 [0.98, 1.70] 0.069 Parental understanding -0.26 0.77 [0.56, 1.06] 0.106 -0.21 0.81 [0.59, 1.12] 0.202 0.16 1.17 [0.92, 1.48] 0.187 Parental monitoring -0.07 0.93 [0.76, 1.14] 0.478 -0.04 0.96 [0.72, 1.27] 0.751 -0.32 0.73 [0.57, 0.94] 0.016 Parental intrusion of privacy 0.29 1.34 [1.04, 1.73] 0.024 0.25 1.29 [1.00, 1.65] 0.047 0.44 1.55 [1.21, 1.99] 0.001 Food insecurity 0.23 1.26 [0.89, 1.78] 0.181 -0.07 0.93 [0.71, 1.23] 0.602 0.15 1.16 [0.82, 1.65] 0.392 McFadden’s R2 0.054 0.08 0.110 McKelvey and Zavoina’s R2 0.090 0.14 0.182 Hosmer–Lemeshow GOF test (sig.) 2072 (0.30) 2064 (0.35) 2011 (0.67) Overall percentage correctly classified 81.40% 76.40% 79.45% Note. AOR = adjusted odds ratio; CI = Confidence Interval; statistically significant results are in bold face cannabis were at about three times increased relative risk attack victimisation, bullying victimisation, loneliness, of repeated attempted suicide during the previous 12 and parental intrusion of privacy were associated with months, compared to students who did not use cannabis. increased likelihood of suicidal ideation, planning, one- time suicide attempt and repeated attempted suicide. Discussion Adolescents in higher grades (grades 10–12), compared This cross-sectional study sought to describe the to those in lower grades (grades 6–9), had reduced rela- 12-month prevalence estimates of suicidal behav- tive risk of reporting one-time attempt or repeated iour (ideation, planning, and [one-time and repeated] attempts at suicide. Cannabis use was associated with attempts) and associated factors among school-going increased relative risk of repeated attempted suicide. adolescents aged 12–17 years in Namibia, drawing on the The 12-month prevalence estimates of suicidal behav- 2013 Namibian WHO-GSHS. In summary, the current iour found in the current study are comparable to those study has two key findings. First, comparable estimates found generally among school-going adolescents within of suicidal behaviour were found between boys and girls, the (sub-Saharan) African region, where estimates of sui- with about 2 in 10 adolescents reporting suicidal ideation, cidal ideations, planning and attempt range from 20.1 to planning, or attempt in the previous 12 months. Approxi- 29% [7–9]. Beyond the similarity of the estimates of this mately, 1 in 20 students reported repeated attempted sui- study to those reported earlier from other Southern Afri- cide during the previous 12 months. Secondly, physical can countries – e.g., Eswatini, Malawi, Mozambique, and Quarshie et al. BMC Psychiatry (2023) 23:169 Page 8 of 12 Table 4 Multinomial logistic regression predicting factors on repeated suicide attempt (N = 2660) Variable One-time attempt Repeated attempts β ARRR [95% CI] p-value β ARRR [95 CI] p-value Demographics Gender: Female 1 1 Male -0.02 0.98 [0.80, 1.20] 0.849 0.11 1.12 [0.86, 1.45] 0.395 Age 0.12 1.13 [1.03, 1.24] 0.013 0.10 1.10 [0.96, 1.27] 0.164 School grade: Grade 6–9 1 1 Grade 10–12 -0.69 0.50 [0.36, 0.70] < 0.001 -0.67 0.51 [0.35, 0.75] 0.001 Mental health and lifestyle factors Loneliness 0.43 1.54 [1.05, 2.25] 0.028 0.72 2.06 [1.27, 3.35] 0.005 Cannabis use 0.26 1.29 [0.65, 2.58] 0.455 1.31 3.72 [2.08, 6.65] < 0.001 School-level factors Physical attack 0.65 1.92 [1.47, 2.50] < 0.001 1.07 2.92 [2.28, 3.74] < 0.001 Bullying victimisation 0.52 1.68 [1.27, 2.22] 0.001 0.96 2.60 [1.83, 3.70] < 0.001 Family-level factors Parental monitoring -0.22 0.80 [0.61, 1.05] 0.108 -0.20 0.82 [0.57, 1.18] 0.269 Parental intrusion of privacy 0.35 1.41 [1.13, 1.76] 0.003 0.58 1.78 [1.30, 2.44] 0.001 Note. ARRR = adjusted relative risk ratios; CI = Confidence Interval; statistically significant results are in bold face. South Africa – [14, 19, 21, 23, 24], the evidence regarding Further, within the lens of the socio-ecological model, comparable estimates between boys and girls have also our finding supports the multi-factorial, multi-layered been reported from other Southern African countries and multi-contextual nature of the factors associated and sub-Saharan Africa in general [10, 21]. The cross- with suicidal behaviour among adolescents [15, 16]. In national similarity of the estimates is to be expected, con- broader terms, the identified key associated factors of sidering that the data are drawn from the WHO-GSHS suicidal behaviour among adolescents in the current conducted in the respective countries using the same study (i.e., physical attack victimisation, bullying vic- measures and definitions of included variables. The lack timisation, and parental intrusion of privacy) support of significant gender differences in the estimates of sui- evidence in the literature that exposure to (longstand- cidal behaviour could be pointing to the possibility that, ing) interpersonal social adversities and relational dif- perhaps, the factors presenting as risks are comparably ficulties occurring in the family, school or within peer difficult for both school-going boys and girls in Namibia. relationships contribute to suicidal behaviour among This finding could also be supporting the emerging evi- young people [10, 45, 46]. Besides being a common phe- dence that self-harm and suicidal behaviour are not nomenon, bullying victimisation has a strong positive neatly differentiated between boys and girls within sub- association with involvement physical fighting and other Saharan Africa [10]. Taken together, this evidence could interpersonal adversities among school-going adoles- be underscoring the importance of universal intervention cents in Namibia [47, 48]. Considered as an antecedent of and prevention efforts focused on suicidal behaviour in suicidal thoughts and behaviour, social adversities (e.g., both school-going adolescent boys and girls in Namibia. physical attack victimisation, and bullying victimisation) We found physical attack victimisation, bullying victi- result in internalising problems, often leading to lowered misation, loneliness, and parental intrusion of privacy to self-esteem, self-blame, and self-dislike, which in turn be significantly associated with increased likelihood of heighten the vulnerability to self-harming thoughts and suicidal ideation, planning, one-time suicide attempt and behaviours in adolescents [49]. repeated attempted suicide. Cannabis use was associated Recent literature is replete with evidence that in both with increased relative risk of (repeated) attempted sui- high-income countries and LMICs, lifestyle factors, par- cide. Global and regional systematic reviews and meta- ticularly health risk behaviours (e.g., cannabis smoking, analyses [7, 8, 10, 12, 42] and recent primary studies alcohol use) and – untreated – mental health problems drawing data mainly from the WHO-GSHS [14, 21, 23, (e.g., loneliness, anxiety, depression) have strong associa- 36, 37, 43, 44] have also identified these factors to be crit- tion with suicidal behaviour among adolescents[11, 42, ical in school-going adolescent suicidal tendencies and 44, 50–52]. Although cannabis possession and use are behaviours. illegal in Namibia, national-level estimates suggest that among school-going adolescents 6.6% boys and 4% girls Quarshie et al. BMC Psychiatry (2023) 23:169 Page 9 of 12 report having used cannabis during the previous month clinical significance of their associations with school- [53] – a clear indication that there are lapses and prob- going adolescent suicidal behaviour in Namibia. lems in the implementation of the law [54]. What is not This study has identified the key factors associated readily clear from the current study is whether the par- with adolescent suicidal behaviours to exist mainly at ticipants’ cannabis use was a coping strategy for their the individual level/ (e.g., loneliness, anxiety, cannabis adverse mental health experiences (e.g., loneliness or use), within school (e.g., bullying victimisation, school anxiety) or for another purpose. Nonetheless, there is grade), family (e.g., intrusion of privacy by parents, evidence to suggest that substance (cannabis or alcohol) parental monitoring), and community context (e.g., phys- use has negative – mental – health outcomes among ical attack victimisation). The implication of the multi- adolescents. Besides underlying increased susceptibility ecological nature of the key evidence for practice is that to brain damage in the long-term, cannabis use impairs intervention and prevention efforts must be designed judgement and memory, increases impulsivity and nega- with a multi-sectoral and multi-layered orientation. For tive mood, and potentially complicates the natural course example, while the initiation (or improvement of existing) of loneliness, anxiety, and depression – which in turn ele- anti-alcohol and substance use and anti-bullying polices vate the risk for suicidal thoughts and behaviour among are needed to enhance school social climate, community- adolescents [51, 55–58]. level training for supportive parenting can be designed Our finding that school-going adolescents in higher for families with adolescents. Similarly, while parents grades, compared to those in lower grades, have reduced and significant others living with adolescents need to be relative risk of reporting one-time or repeated attempts at observant regarding the identification of warning signs suicide supports earlier evidence from South Africa [59], of potential adolescent suicidal behaviours, the Namib- but it is inconsistent with a recent finding from Ghana, ian Ministry of Education could consider including where no significant association was observed between mental health literacy and help-seeking lessons in the school grade and suicidal behaviour [60]. Whereas expla- curriculum for high schools – this would contribute to nations for this school grade difference are not read- improving help-seeking and self-care behaviours among ily clear from the Southern African context, perhaps, in school-going adolescents at risk of self-harm and suicidal Namibia, curriculum-based functional mental health lit- tendencies and other emotional crises, including loneli- eracy could be suggested. Among other aims, the Namib- ness, anxiety and depression. ian school curriculum seeks “to foster the highest moral, ethical and spiritual values such as integrity, responsibil- Strengths and Limitations ity, equality, and reverence for life” [30]. Maybe, the value A critical significance of this study is that it contributes of ‘reverence for life’ – which essentially proscribes and to addressing the problem of dearth research on sui- eschews suicidal thoughts and actions [61] – might have cidal behaviour among adolescents in Namibia. Data been more actionably consolidated in students in upper on adolescent mental health (outcomes) are still insuffi- school grades than those in lower grades. Beyond this cient to inform policy, intervention efforts, and preven- speculation, further studies are needed to understand tion programmes in Namibia [62]. In particular, data on this differentiation of suicidal thoughts and behaviours in self-harm and suicidal behaviours among (both school- terms of school grade in Namibia but also within the gen- going and out-of-school) adolescents in Namibia are less eral (Southern) sub-Saharan African context. than enough [10, 63]. Additionally, this study contributes While the factors showing significant associations with broadly to advance our knowledge and understanding of suicidal behaviour (including repeated attempted sui- the prevalence and associated factors of suicidal behav- cide) are identified, it is also imperative to comment on iour among in-school adolescents in Namibia in that the the (exposure) variables reported to be important in the study draws on a relatively large data accessed from a literature but showed no significant associations with the nationally representative sample. outcomes in the current study. Interestingly, in the cur- Beyond these strengths, there are noteworthy limita- rent study, although alcohol use, being sexually active, tions of this study. The findings of the study should be number of close friends, truancy, peer support, parental generalised with caution, as the key evidence may not understanding, and food insecurity were associated with necessarily apply to out-of-school adolescents. While each of the suicidal behaviour outcomes, none of these the study sample excludes students who were absent on associations reached the desired threshold of statistical the day of the survey, evidence suggests that the average significance. While we suspect sparse data to account for annual dropout rate in Namibia ranges between 3 and the lack of statistical significance, we believe future stud- 10.4% [64]. The one-time cross-sectional survey design ies involving relatively large responses to each of these used for the WHO-GSHS implies that the outcome and data items will contribute to clarifying the statistical and exposure variables were measured at the same time point, making it impossible to identify sequence, temporal link, Quarshie et al. BMC Psychiatry (2023) 23:169 Page 10 of 12 and causal relationship between the exposure and out- AbbreviationsCDC Centers for Disease Control and Prevention come variables. Thus, consistent with recommendations GSHS G lobal School-based Student Health Survey by recent studies from of other countries within the LMICs Low- and middle-income countries continent [10, 13, 65], future studies using more robust WHO World Health Organization approaches, including longitudinal designs and carefully designed qualitative studies are needed for contextually Supplementary Information nuanced understanding of self-harm and suicidal behav- The online version contains supplementary material available at https://doi. iours among adolescents in Namibia and (sub-Saharan) org/10.1186/s12888-023-04646-7. Africa generally. Supplementary Material 1 Notably, in the current study, it is not clear why the esti- mates of suicidal planning and attempt are higher than Acknowledgements suicidal ideation. Similar evidence has been reported We also thank the Namibian Ministries of Health and Social Services and from other sub-Saharan African countries drawing on the Education, and WHO and its partners for making freely available the data from WHO-GSHS data – e.g., Benin, Ghana, Liberia, Malawi, the 2013 Namibian WHO-GSHS. Lastly, but more importantly, we also thank all the students who contributed data for this survey. and Sierra Leone [21, 37, 66–68]. However, this evidence is unconventional, counterintuitive, and not in keeping Authors’ contributions with the suicide process or pathway model [69, 70], which ENBQ, NEYD, and KOA conceived, designed and organised the study. ENBQ, and NEYD curated the data and performed the statistical analysis; and ENBQ superlatively suggests that, typically, estimates of ideation and KOA contributed to the interpretation of the data. ENBQ and NEYD are expected to be highest, followed by estimates of plan- drafted the manuscript, and KOA critiqued the manuscript for important ning, then estimates of attempt. Perhaps, the use of a intellectual content. All authors read and approved the final version of the manuscript. ENBQ serves as guarantor for the contents of this paper. single-item measures in the WHO-GSHS to assess these outcomes could account for this unconventional finding. Funding As cautioned elsewhere, the use of single-item measure The authors received no financial support or specific grant from any funding agency in the public, commercial or not-for-profit sectors, for the research, to assess suicidal ideation, planning, and attempt must authorship, and/or publication of this article. be interpreted cautiously, as single-item measures typi- cally result in misclassification of suicidal behaviours and Data Availability The datasets used and/or analysed during the current study are freely higher estimates [71]. Specially, we also suspect that the available from the WHO website: https://extranet.who.int/ncdsmicrodata/ lower estimate of suicidal ideation – relative to suicidal index.php/catalog/478. The 2013 Namibia GSHS questionnaire is also available attempt – in the current study may be due to impulsiv- freely on the WHO’s website: https://extranet.who.int/ncdsmicrodata/index. php/catalog/478#metadata-questionnaires. ity. There is evidence from high-income countries to sug- gests that impulsivity may result in the onset of suicidal Declarations attempt among adolescents, even in the absence of prior suicidal ideation [72, 73]. Competing interests ENBQ is an Associate Editor of BMC Psychiatry. The rest of the authors declare that they have no competing interests. Conclusion The evidence of this study adds to the literature on Ethics approval and consent to participate non-fatal suicidal behaviour among school-going ado- The 2013 Namibian WHO-GSHS was approved by the Institutional Review Boards of the Namibian Ministry of Health and Social Services. The study was lescents in (sub-Saharan) Africa, but also underscores supported by WHO and the US Centers for Disease Control and Prevention. the marked estimates of suicide ideation, planning, and Policies laid out by the Ministry of Education regarding consent procedures (repeated) suicide attempt among school-going adoles- for participation in students surveys were followed including detachment of identifier information. Official written permissions were obtained from cents (aged 12–17 years) in Namibia. The relatively high Namibian Ministry of Health and Social Services, and the Ministry of Education, prevalence estimates and multi-layered correlates (at the the selected schools, and classroom teachers. Written informed consent were individual-level, family-level, interpersonal-level, school obtained from students, while an additional written parental consent was obtained from parents of participants aged younger than 18 years. All methods context and the broader community context) contribute were carried out in accordance with relevant guidelines and regulations. to our understanding of adolescent suicide in Namibia. The evidence highlights the importance of paying more Consent for publication Not applicable. attention to addressing the mental health needs (includ- ing those related to psychological and social wellness) Author details 1 of school-going adolescents in Namibia. While the cur- Department of Psychology, University of Ghana, P.O. Box LG 84, Accra, Ghana rent study suggests that further research is warranted to 2Department of Psychology, University of the Free State, Bloemfontein, explicate the pathways to adolescent suicide in Namibia, South Africa identifying and understanding the correlates of adoles- cent suicidal ideations and non-fatal suicidal behaviours Received: 29 September 2022 / Accepted: 1 March 2023 are useful for intervention and prevention programmes. Quarshie et al. BMC Psychiatry (2023) 23:169 Page 11 of 12 References 24. Seidu A-A, Amu H, Dadzie LK, Amoah A, Ahinkorah BO, Ameyaw EK, 1. WHO. Preventing suicide: a global imperative. Geneva, Switzerland: WHO; Acheampong HY, Kissah-Korsah K. Suicidal behaviours among in-school 2014. adolescents in Mozambique: cross-sectional evidence of the prevalence and 2. Demesmaeker A, Chazard E, Hoang A, Vaiva G, Amad A. Suicide mortality predictors using the Global School-Based Health Survey data. PLoS ONE. after a nonfatal suicide attempt: a systematic review and meta-analysis. 2020;15(7):e0236448. Australian & New Zealand Journal of Psychiatry. 2022;56(6):603–16. 25. Muula AS, Kazembe L, Rudatsikira E, Siziya S. Suicidal ideation and associ- 3. Fazel S, Runeson B. Suicide. N Engl J Med. 2020;382(21):266–74. ated factors among in-school adolescents in Zambia. Tanzan J Health Res. 4. WHO. Suicide worldwide in 2019: global health estimates. Geneva: World 2007;9(3):202–6. Health Organization; 2021. 26. Rudatsikira E, Siziya S, Muula AS. Suicidal ideation and associated factors 5. Suicide. : one person dies every 40 seconds [[https://www.who.int/news/ among school-going adolescents in Harare, Zimbabwe. J Psychol Afr. item/09-09-2019-suicide-one-person-dies-every-40-seconds]] 2007;17(1–2):93–7. 6. Lim K-S, Wong CH, McIntyre RS, Wang J, Zhang Z, Tran BX, Tan W, Ho CS, 27. Peltzer K, Pengpid S. Lifestyle and mental health among school-going adoles- Ho RC. Global lifetime and 12-month prevalence of suicidal behavior, cents in Namibia. J Psychol Afr. 2017;27(1):69–73. deliberate self-harm and non-suicidal self-injury in children and adolescents 28. UNICEF. Young People in Namibia: an analysis of the 2011 Population & Hous- between 1989 and 2018: a meta-analysis. Int J Environ Res Public Health. ing Census. Windhoek: UNICEF Namibia; 2014. 2019;16(22):4581. 29. UNDP. Human Development Report 2021/2022. Uncertain times, unsettled 7. Biswas T, Scott JG, Munir K, Renzaho AM, Rawal LB, Baxter J, Mamun AA. lives: shaping our future in a transforming world. New York: UNDP; 2022. Global variation in the prevalence of suicidal ideation, anxiety and their 30. UNESCO. International Bureau of Education: World data on education: correlates among adolescents: a population based study of 82 countries. Namibia. Geneva, Switzerland: UNESCO-IBE; 2010. EClinicalMedicine. 2020;24:100395. 31. World Bank Country and Lending Groups. Country classifica- 8. Li L, You D, Ruan T, Xu S, Mi D, Cai T, Han L. The prevalence of suicidal behav- tion [[https://datahelpdesk.worldbank.org/knowledgebase/ iors and their mental risk factors among young adolescents in 46 low-and articles/906519-world-bank-country-and-lending-groups]]] middle-income countries. J Affect Disord. 2021;281:847–55. 32. Global School-Based Student Health Survey 2013. : Namibia, 2013 [[https:// 9. Uddin R, Burton NW, Maple M, Khan SR, Khan A. Suicidal ideation, suicide extranet.who.int/ncdsmicrodata/index.php/catalog/478]] planning, and suicide attempts among adolescents in 59 low-income and 33. Global school-based. student health survey [https://www.who.int/ middle-income countries: a population-based study. The Lancet Child & teams/noncommunicable-diseases/surveillance/systems-tools/ Adolescent Health. 2019;3(4):223–33. global-school-based-student-health-survey] 10. Quarshie EN, Waterman MG, House AO. Self-harm with suicidal and non- 34. Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke suicidal intent in young people in sub-saharan Africa: a systematic review. JP, Initiative S. The strengthening the reporting of Observational Studies in BMC Psychiatry. 2020;20(1):1–26. Epidemiology (STROBE) statement: guidelines for reporting observational 11. Orri M, Scardera S, Perret LC, Bolanis D, Temcheff C, Séguin JR, Boivin M, studies. Ann Intern Med. 2007;147(8):573–7. Turecki G, Tremblay RE, Côté SM. Mental health problems and risk of suicidal 35. Pengpid S, Peltzer K. Prevalence and correlates of suicidal behaviour among ideation and attempts in adolescents.Pediatrics2020, 146(1). adults in Malawi: a nationally representative cross-sectional survey in 2017. 12. Ati NA, Paraswati MD, Windarwati HD. What are the risk factors and protec- Int J Mental Health Syst. 2021;15(1):1–8. tive factors of suicidal behavior in adolescents? A systematic review. J Child 36. Quarshie EN-B, Andoh-Arthur J. Suicide attempts among 1,437 adolescents Adolesc Psychiatric Nurs. 2021;34(1):7–18. aged 12–17 years attending junior high schools in Ghana. Crisis; 2020. 13. Orri M, Ahun MN, Naicker S, Besharati S, Richter LM. Childhood factors associ- 37. Quarshie EN-B, Onyeaka HK, Oppong Asante K. Suicidal behaviours among ated with suicidal ideation among south african youth: a 28-year longitudinal adolescents in Liberia. BMC Psychiatry. 2020;20(1):1–12. study of the birth to twenty plus cohort. PLoS Med. 2022;19(3):e1003946. 38. Lang TA, Altman DG. Basic statistical reporting for articles published in 14. Quarshie EN-B, Atorkey P, García KPV, Lomotey SA, Navelle PL. Suicidal Behav- biomedical journals: the “Statistical analyses and methods in the published iors in a Nationally Representative Sample of School-Going Adolescents Literature” or the SAMPL Guidelines. Int J Nurs Stud. 2015;52(1):5–9. Aged 12–17 Years in Eswatini.Trends in Psychology2021:1–30. 39. West BT, Sakshaug JW, Aurelien GAS. How big of a problem is analytic error in 15. Cramer RJ, Kapusta ND. A social-ecological framework of theory, assessment, secondary analyses of survey data? PLoS ONE. 2016;11(6):e0158120. and prevention of suicide.Frontiers in Psychology2017:1756. 40. Greenland S, Senn SJ, Rothman KJ, Carlin JB, Poole C, Goodman SN, Altman 16. Shagle SC, Barber BK. A social-ecological analysis of adolescent suicidal DG. Statistical tests, P values, confidence intervals, and power: a guide to ideation. Am J Orthopsychiatry. 1995;65(1):114–24. misinterpretations. Eur J Epidemiol. 2016;31(4):337–50. 17. James S, Reddy SP, Ellahebokus A, Sewpaul R, Naidoo P. The association 41. Sterne JA, Smith GD. Sifting the evidence - what’s wrong with significance between adolescent risk behaviours and feelings of sadness or hopelessness: tests? Phys Ther. 2001;81(8):1464–9. a cross-sectional survey of south african secondary school learners. Psychol 42. Abio A, Owusu PN, Posti JP, Bärnighausen T, Shaikh MA, Shankar V, Lowery Health Med. 2017;22(7):778–89. Wilson M. Cross-national examination of adolescent suicidal behavior: a 18. Reddy S, James S, Sewpaul R, Sifunda S, Ellahebokus A, Kambaran NS, Omar- pooled and multi-level analysis of 193,484 students from 53 LMIC countries. dien RG. Umthente uhlaba usamila: the 3rd south african national youth risk Soc Psychiatry Psychiatr Epidemiol. 2022;57:1603–13. behaviour survey 2011. Cape Town: South African Medical Research Council; 43. Pandey AR, Bista B, Dhungana RR, Aryal KK, Chalise B, Dhimal M. Factors 2013. associated with suicidal ideation and suicidal attempts among adolescent 19. Shilubane HN, Ruiter RA, van den Borne B, Sewpaul R, James S, Reddy PS. students in Nepal: findings from Global School-based Students Health Suicide and related health risk behaviours among school learners in South Survey. PLoS ONE. 2019;14(4):e0210383. Africa: results from the 2002 and 2008 national youth risk behaviour surveys. 44. Tetteh J, Ekem-Ferguson G, Swaray SM, Kugbey N, Quarshie EN-B, Yawson AE. BMC Public Health. 2013;13(1):1–14. Marijuana use and repeated attempted suicide among senior high school 20. Swahn MH, Bossarte RM, Elimam DM, Gaylor E, Jayaraman S. Prevalence and students in Ghana: Evidence from the WHO Global School-Based Student correlates of suicidal ideation and physical fighting: A comparison between Health Survey, 2012.General Psychiatry2020, 33(6). students in Botswana, Kenya, Uganda, Zambia and the United States. Public 45. Fortune S, Stewart A, Yadav V, Hawton K. Suicide in adolescents: health yearbook, 2010. edn. Edited by Merrick J:Nova Biomedical Books; using life charts to understand the suicidal process. J Affect Disord. 2012:pp. 233–245. 2007;100(1–3):199–210. 21. Shaikh MA, Lloyd J, Acquah E, Celedonia KL, Wilson L. Suicide attempts 46. Hawton K, Saunders KE, O’Connor RC. Self-harm and suicide in adolescents. and behavioral correlates among a nationally representative sample of The Lancet. 2012;379(9834):2373–82. school-attending adolescents in the Republic of Malawi. BMC Public Health. 47. Davis LE, Abio A, Wilson ML, Shaikh MA. Extent, patterns and demographic 2016;16(1):1–8. correlates for physical fighting among school-attending adolescents in 22. Hoogstoel F, Samadoulougou S, Lorant V, Kirakoya-Samadoulougou F. A Namibia: examination of the 2013 Global School-based Health Survey. PeerJ. latent class analysis of health lifestyles in relation to suicidality among adoles- 2020;8:e9075. cents in Mauritius. Int J Environ Res Public Health. 2021;18(13):6934. 48. Rudatsikira E, Siziya S, Kazembe LN, Muula AS. Prevalence and associated 23. Pengpid S, Peltzer K. Suicide attempt and associated factors among in-school factors of physical fighting among school-going adolescents in Namibia. Ann adolescents in Mozambique. J Psychol Afr. 2020;30(2):130–4. Gen Psychiatry. 2007;6(1):1–5. Quarshie et al. BMC Psychiatry (2023) 23:169 Page 12 of 12 49. Page RM, West JH. Suicide ideation and psychosocial distress in sub-saharan 63. Ministry of Health and Social Services. National Multisectoral Strategic african youth. Am J Health Behav. 2011;35(2):129–41. Plan for Prevention and Control of Non-Communicable Diseases (NCDs) in 50. Carvalho AF, Stubbs B, Vancampfort D, Kloiber S, Maes M, Firth J, Kurdyak PA, Namibia 2017/18–2021/22. Windhoek, Namibia: Ministry of Health and Social Stein DJ, Rehm J, Koyanagi A. Cannabis use and suicide attempts among Services; 2017. 86,254 adolescents aged 12–15 years from 21 low-and middle-income 64. Nekongo-Nielsen H, Mbukusa NR, Tjiramba E. Investigating factors that countries. Eur Psychiatry. 2019;56(1):8–13. lead to school dropout in Namibia. The Namibia CPD Journal for Educators. 51. Fresán A, Dionisio-García DM, González-Castro TB, Ramos-Méndez M, Castillo- 2015;2(1):99–118. Avila RG, Tovilla-Zárate CA, Juárez-Rojop IE, López-Narváez ML, Genis-Men- 65. Quarshie EN, Waterman MG, House AO. Adolescent self-harm in Ghana: a doza AD, Nicolini H. Cannabis smoking increases the risk of suicide ideation qualitative interview-based study of first-hand accounts. BMC Psychiatry. and suicide attempt in young individuals of 11–21 years: A systematic review 2020;20(1):1–14. and meta-analysis.Journal of Psychiatric Research2022. 66. Asante KO, Kugbey N, Osafo J, Quarshie EN-B, Sarfo JO. The prevalence 52. Tetteh J, Ekem-Ferguson G, Quarshie EN-B, Swaray SM, Ayanore MA, and correlates of suicidal behaviours (ideation, plan and attempt) among Seneadza NAH, Asante KO, Yawson AE. Marijuana use and suicidal behaviours adolescents in senior high schools in Ghana. SSM-population Health. among school-going adolescents in Africa: assessments of prevalence and 2017;3:427–34. risk factors from the Global School-Based Student Health Survey.General 67. Randall JR, Doku D, Wilson ML, Peltzer K. Suicidal behaviour and related Psychiatry2021, 34(4). risk factors among school-aged youth in the Republic of Benin. PLoS ONE. 53. Peltzer K, Pengpid S. Cannabis and amphetamine use and associated factors 2014;9(2):e88233. among school-going adolescents in nine african countries. J Child Adolesc 68. Asante KO, Quarshie EN-B, Onyeaka HK. Epidemiology of suicidal behaviours Subst Abuse. 2018;27(2):112–8. amongst school-going adolescents in post-conflict Sierra Leone. J Affect 54. Cannabis in Namibia – Laws, Use, and History, [[. https://sensiseeds.com/en/ Disord. 2021;295:989–96. blog/countries/cannabis-in-namibia-laws-use-history/]] 69. Klonsky ED, May AM, Saffer BY. Suicide, suicide attempts, and suicidal ide- 55. Borges G, Loera CR. Alcohol and drug use in suicidal behaviour. Curr Opin ation. Annu Rev Clin Psychol. 2016;12(1):307–30. Psychiatry. 2010;23(3):195–204. 70. Millner AJ, Lee MD, Nock MK. Describing and measuring the pathway to 56. Paruk S, Burns JK. Cannabis and mental illness in adolescents: a review. South suicide attempts: a preliminary study. Suicide and Life-Threatening Behavior. Afr Family Pract. 2016;58(sup1):18–S21. 2017;47(3):353–69. 57. Volkow ND, Baler RD, Compton WM, Weiss SR. Adverse health effects of 71. Hom MA, Joiner TE Jr, Bernert RA. Limitations of a single-item assessment of marijuana use. N Engl J Med. 2014;370(23):2219–27. suicide attempt history: implications for standardized suicide risk assessment. 58. Sideli L, Quigley H, La Cascia C, Murray RM. Cannabis use and the risk for Psychol Assess. 2016;28(8):1026. psychosis and affective disorders. J Dual Diagnosis. 2020;16(1):22–42. 72. Anestis MD, Soberay KA, Gutierrez PM, Hernández TD, Joiner TE. Reconsider- 59. Shilubane HN, Ruiter RA, Bos AE, van den Borne B, James S, Reddy PS. ing the link between impulsivity and suicidal behavior. Personality and Social Psychosocial determinants of suicide attempts among black south african Psychology Review. 2014;18(4):366–86. adolescents: a qualitative analysis. J Youth Stud. 2012;15(2):177–89. 73. May AM, Klonsky ED. Impulsive” suicide attempts: what do we really mean? 60. Quarshie EN-B, Odame SK. Suicidal ideation and associated factors among Personality Disorders: Theory Research and Treatment. 2016;7(3):293. school-going adolescents in rural Ghana. Current Psychology 2021(advance online publication):1–14. 61. Gyekye K. African cultural values. An introduction. Accra, Ghana: Sankofa Publisher’s Note publishing company; 2003. Springer Nature remains neutral with regard to jurisdictional claims in 62. WHO. Adolescent health in Namibia. Windhoek, Namibia: WHO Regional published maps and institutional affiliations. Office for Africa; 2018.