Abstract / Summary
Abstract Background Road traffic injuries remain a major cause of preventable death and disability among children and young people, yet multi-country evidence on behavioural correlates among African school-going adolescents is limited. We estimated the prevalence and social patterning of motor-vehicle-related serious injury and examined its association with current alcohol, cigarette, cannabis, and polysubstance use across participating African Global School-based Student Health Survey (GSHS) datasets. Methods We harmonised 20 GSHS public-use survey files from 12 African countries. Seventeen surveys from nine countries contained a usable injury-cause outcome, yielding 44,303 adolescents with valid road traffic injury data and 2,520 unweighted cases. Road traffic injury identified motor-vehicle involvement as the major cause of the respondent's most serious injury during the previous 12 months. Pooled prevalence and associations were estimated using within-survey normalised weights, globally unique strata and primary sampling units, and Taylor-linearised variance estimation. Country-specific estimates used original survey weights. Survey-weighted logistic regression adjusted for age, sex, country, and survey period. Prespecified secondary analyses assessed substance-specific associations, polysubstance dose-response, sex interaction, country-specific estimates with random-effects meta-analysis, leave-one-country-out robustness, and restriction to seriously injured adolescents. Results The pooled weighted prevalence of road traffic injury was 5.8% (95% CI 5.5 to 6.2%). Prevalence was higher among males than females, and male sex remained independently associated with injury (aOR 1.63, 95% CI 1.47 to 1.80). Any current substance use was associated with higher odds of road traffic injury (aOR 1.78, 95% CI 1.50 to 2.12), as were current alcohol use (1.74, 1.52 to 1.99), cigarette smoking (2.11, 1.80 to 2.47), cannabis use (3.34, 2.63 to 4.24), and polysubstance use (2.36, 1.83 to 3.04). Injury prevalence increased from 4.2% among adolescents using no substance domain to 9.8% among those using all three domains, with an aOR of 1.52 (95% CI 1.36 to 1.69) per additional domain. Country-specific associations were heterogeneous, with a random-effects pooled aOR of 1.82 (95% CI 1.08 to 3.07; I2 = 81.8%). Conclusions Current substance use identifies an important behavioural risk cluster for road traffic injury among school-going adolescents in participating African GSHS surveys. The consistency of the pooled exposure gradients, alongside substantial country heterogeneity, supports integrated adolescent substance-use and road-safety prevention strategies tailored to local contexts. The cross-sectional design and differing exposure and outcome recall periods preclude causal interpretation.