Table 1.
Characteristics of the included papers
| References | Country | Data collection | Study design | N (or case/controls) | % female | Age range at enrolment (mean/median) | Groups of covariatesA | Quality assessment score |
|---|---|---|---|---|---|---|---|---|
| Almansour and Siziya [50] | Swaziland | 2013 | Cross-sectional | 3680 | 51.2 | < 11 to > 18 | Age, gender, social, mental health, substances, family, behaviour | 6/10 |
| Asante et al. [46] | Ghana | 2012 | Cross-sectional | 1973 | 45.7 | < 14 to > 18 | Age, mental health, behaviour, substances, family | 8/10 |
| Bailey et al. [73] | US | 2006–10 | Cohort | 20,536/1,132,116g | NR | 11–17 | None | 9/13 |
| Bjarnason and Thorlindsson [71] | Iceland | 1992 | Cross-sectional | 7018 | 50 | 10th grade (15.3) | School, social, family, substances, suicide suggestion | 6/10 |
| Borowsky et al. [43] | USa | NR | Cross-sectional | 11,666 | 52 | 12–18 | None | 6/10 |
| Borowsky et al. [61] | US | 1994–6 | Cohort | 13,110 | NR | 7th–10th grade | Age, family, social | 10/13 |
| Brunner et al. [69] | 11 European countriesb | 2009–10 | Cross-sectional | 12,068 | 55.7 | (14.9) | Age, gender, social, family, mental health, behavioural, substances | 8/10 |
| Chen et al. [47] | Malaysia | 2001 | Cross-sectional | 4500 | 53.6 | 12–19 (15.3) | Gender, school, substances, family, social, behavioural, mental health (only urban students in multivariate analysis) | 6/10 |
| Cheng et al. [51] | China | 2003 | Cross-sectional | 9015 | 51 | 11–17 | Age, gender, family, social | 9/10 |
| Choquet and Menke [53] | France | 1988 | Cross-sectional | 1519 | 45 | 13–16 | None | 4/10 |
| Cwik et al. [42] | USc | 2006–2009 | Cross-sectional | 71 | 64.8 | 10–19 (16) | None | 3/10 |
| Davaasambuu et al. [48] | Mongolia | 2013 | Cross-sectional | 5393 | 51.7 | 12–17 | Gender, substances, behaviour, social, mental health | 8/10 |
| De Man et al. [59] | Canada | NR | Cross-sectional | 558 | 51.3 | 11–18 (14) | Social, family, school, mental health, (unclear which were included in multivariate model) | 5/10 |
| Donath et al. [63] | Germany | 2007–2008 | Cross-sectional | 44,610 | 48.7 | 9th grade (15.3) | Age, gender, family, social, substances, mental health | 8/10 |
| Epstein and Spirito [52] | US | 2005 | Cross-sectional | 12,095 | 50 | 12 to 18+ | Behaviour, social | 6/10 |
| Evren et al. [68] | Turkey | 2012 | Cross-sectional | 4957 | 47.3 | (15) | None | 4/10 |
| Fergusson et al. [49] | New Zealand | 1992–1995 | Cohort | 1063 | 49.8 | 15–21 | Not clear what is adjusted for | 9/13 |
| Kandel et al. [54] | US | 1986 | Cross-sectional | 593 | NR | 9th and 11th grade | None | 4/10 |
| Larsson and Sund [70] | Norway | 1998–2000 | Cohort | 2464 | 50.8 | 12–15 (13.9) | None | 9/13 |
| Lau et al. [45] | Chinad | 2008 | Cross-sectional | 3324 | 55.7 | 61.7% < 15 | Gender, age, social, impact of earthquake, mental health, family | 9/10 |
| Lewinsohn et al. [66] | US | 1987–1990 | Cross-sectional | 1710 | 52.9 | 14–18 | Gender and then depression | 7/10 |
| Lewinsohn et al. [67] | US | 1987–1989 | Cohort | 1508 | NR | 14–18 | Gender | 10/13 |
| Lyon et al. [44] | USe | NR | Case control | 38/76 | 82 | 12–17 (14.8) | Age, gender, social, mental health, substances | 9/13 |
| Noble et al. [72] | US | NR | Case control | 638/638 | 11–19 (14.92) | Gender, age, social, behaviour, school factors | 10/13 | |
| Pages et al. [65] | France | NR | Cross-sectional | 11,718 | 51.9 | 11–21 (16.6) | Age (in logistic regression model: behaviour, mental health, substances) | 7/10 |
| Peltzer and Pengpid [57] | 7 SE Asian countriesf | 2007–2013 | Cross-sectional | 30,284 | 51.2 | 13–15 (14.1) | Gender, age, social, mental health, substances, physical health | 8/10 |
| Pillai et al. [62] | India | 2006 | Cross-sectional | 1594 | 51.4 | 16–25 (19.4)h | Gender, age, social, family | 8/10 |
| Randall et al. [56] | Benin | 2009 | Cross-sectional | 2690 | 33.1 | < 12 to 16+ | Age, gender, mental health, social, substances | 9/10 |
| Sharma et al. [55] | Peru | 2014 | Cross-sectional | 903 | 53.6 | (15) | Family, social | 8/10 |
| Taliaferro and Muehlenkamp [60] | US | 2010 | Cross-sectional | 70,722 | 50 | 9th and 12th grade | Age, ethnicity, social, family, behaviour, mental health, substances | 8/10 |
| Wilson et al. [58] | Republic of Seychelles | NR | Cross-sectional | 1432 | 52 | 11–17 (14) | None | 7/10 |
| Xin et al. [64] | China | NR | Cross-sectional | 11,880 | 50.5 | 10–20 (14.62) | Age, social | 8/10 |
AEach group listed here may comprise several variables, NR not reported
aAmerican Indian and Alaska native youth
bAustria, Estonia, France, Germany, Hungary, Ireland, Israel, Italy, Romania, Slovenia and Spain
cApaches (American Indians) who attempted suicide within the past 90 days
dIn Chengdu, China, after the Sichuan Earthquake
eAfrican American population only, cases and controls taken from clinical population
fCambodia, Indonesia, Malaysia, Myanmar, Philippines, Thailand and Vietnam
gPerson years with truancy/person years without truancy (denominator is derived from the whole base population)
hAge 16–25 (19.4) years, but absence measured in younger subsample