Methods: 1559 middle-school adolescents from north-eastern France (mean age 13.5, SD 1.3) completed in class a questionnaire including DM (Kandel scale), factors studied, and their occurring/persisting period. Data were analyzed using Cox models taking into account risk factors which came before and persisted until DM occurring.
Findings: Lifetime-DM affected 7.6% of boys and 19.1% of girls. Among girls factors with significant crude hazard ratios (HR) were: parent divorce/separation (1.57), insufficient income (1.95), repeating a school-year (1.95), victim of violence (2.99) or sexual abuse (4.96), and lack of supports (4.08 for score 1-2 and 7.74 for score 3+, vs. score 0) while tobacco use was close to significance (1.77). Taking all factors into account retained only repeating a school-year (adjusted HR 1.87), victim of violence (2.50) or sexual abuse (4.02), and lack of supports (3.83 and 7.28). Among boys, the significant factors were: hard-drug use (6.01), victim of violence (2.88), and lack of supports (2.27 and 4.78) while insufficient income (1.75), low parent education (1.60), and victim of sexual abuse (3.58) were close to significance. Full model retained only victim of violence (adjusted HR 2.29) and lack of supports (2.16 and 4.18).
Conclusion: Taking chronology in consideration revealed strong gender disparities for DM and its associations with family characteristics, school difficulties, unhealthy behaviours, violence, and social supports that may be prevention targets.