Journal of Epidemiology and Public Health
ISSN: 3065-9078 | DOI: 10.64030/3065-9078
Yem Sokha, Kem Sokunthea and Tun Sreypeov
Background Teenage pregnancy remains a significant public health challenge in Cambodia, with implications for maternal and child health outcomes, educational attainment, and socioeconomic development.
Objective To determine the prevalence of teenage pregnancy among Cambodian women aged 15-19 years and identify its individual and community-level determinants using multilevel analysis of data from the 2021-22 Cambodia Demographic and Health Survey (CDHS).
Method We analyzed data from 5,783 adolescent women aged 15-19 years from the nationally representative CDHS 2021-22. Given the hierarchical structure of DHS data (individuals nested within 709 clusters), we employed multilevel logistic regression models. Four nested models were fitted, including a null model with random intercept only, an individual-level model including age, education, marital status, wealth, contraceptive use, and internet use, a community-level model including residence and distance to health facilities, and a full model combining both levels. Intraclass correlation coefficient (ICC) and likelihood ratio tests assessed clustering effects. We reported adjusted odds ratios (AOR) with 95% confidence intervals.
Results The weighted prevalence of teenage pregnancy was 28.08% (95% CI: 26.48-29.75). The ICC indicated 10.07% of variance in teenage pregnancy was attributable to cluster-level factors, strongly justifying multilevel modeling (LR test comparing multilevel to standard logistic regression: χ²=52.69, df=1, p<0.001). In the full multilevel model, older adolescent age (18- 19 vs 15-17 years: AOR=2.73, 95% CI: 1.80-4.14, p<0.001), current marriage (AOR=4,611.27, 95% CI: 1,929.44-11,020.73, p<0.001), and former marriage (AOR=434.84, 95% CI: 132.85-1,423.27, p<0.001) were significantly associated with increased odds of teenage pregnancy. Higher education showed protective effects (AOR=0.06, 95% CI: 0.01-0.60, p=0.017). Wealthier households demonstrated lower odds (richest vs poorest: AOR=0.22, 95% CI: 0.11-0.46, p<0.001). Current contraceptive use was associated with higher odds (AOR=4.69, 95% CI: 2.85-7.71, p<0.001), reflecting post-conception initiation rather than primary prevention. After accounting for individual-level factors, community-level variables (rural residence, distance to facilities) were not independently significant.
Conclusions More than one in four Cambodian adolescent girls experience teenage pregnancy, with substantial between-cluster variation (ICC=10.07%). Early marriage is the dominant pathway to teenage pregnancy. Multilevel interventions addressing child marriage prevention, educational retention, socioeconomic empowerment, and youth-friendly reproductive health services are urgently needed.