Table 3 Results of bivariable and multivariable multilevel logistic regression analyses of factors associated with adolescent childbearing in Burundi
Variables/categories | 2016–17 BDHS (weighted sample, N = 3859) | |||
|---|---|---|---|---|
Bivariable analysis | Multivariable analysis | |||
uOR (95%CI) | P value | aOR (95% CI) | P value | |
Individual-level variables | ||||
Adolescent’s age | ||||
15–17 years (RC) | 1.00 | 1.00 | ||
18–19 years | 13.87 (9.93–19.38) | < 0.001 | 5.85 (3.54–9.65) | < 0.001 |
Education | ||||
Secondary /High (RC) | 1.00 | 1.00 | ||
Primary | 2.71 (2.02–3.64) | < 0.001 | 2.58 (1.54–4.25) | < 0.001 |
No education | 5.79 (3.80–8.80) | < 0.001 | 4.18 (1.88–9.30) | < 0.001 |
Currently Working | ||||
No (RC) | 1.00 | 1.00 | ||
Yes | 3.17 (2.34–4.29) | < 0.001 | 1.43 (0.86–2.39) | 0.170 |
Household Wealth Index | ||||
Poorest (RC) | 1.00 | 1.00 | ||
Poorer | 0.56 (0.38–0.84) | 0.005 | 1.18 (0.56–2.46) | 0.660 |
Middle | 0.50 (0.34–0.75) | 0.001 | 0.75 (0.34–1.12) | 0.467 |
Richer | 0.56 (0.38–0.83) | 0.004 | 0.61 (0.55–1.07) | 0.052 |
Richest | 0.46 (0.31–0.70) | < 0.001 | 0.52 (0.45–0.87) | 0.007 |
Religion | ||||
No religion (RC) | 1.00 | 1.00 | ||
Catholic | 0.21 (0.04–1.05) | 0.057 | 2.78 (0.02–495.32) | 0.698 |
Protestant | 0.28 (0.05–1.41) | 0.122 | 2.98 (0.02–531.20) | 0.679 |
Adventist | 0.60 (0.11–3.30) | 0.559 | 3.25 (0.02–653.55) | 0.663 |
Muslim | 0.40 (0.07–2.27) | 0.303 | 10.55 (0.06–2016.49) | 0.379 |
Othersa | 0.44 (0.07–2.73) | 0.381 | 4.22 (0.02–960.45) | 0.603 |
Access to Mass Media | ||||
No Access (RC) | 1.00 | 1.00 | ||
Has Access | 0.65 (0.51–0.84) | 0.001 | 0.67 (0.43–1.06) | 0.090 |
Age at First Marriage | ||||
18–19 years (RC) | 1.00 | 1.00 | ||
≤ 17 years | 9.66 (3.57–26.11) | < 0.001 | 9.28 (3.11–27.65) | < 0.001 |
Never in union (NA) | 0.00 (0.00–0.01) | < 0.001 | 0.01 (0.00–0.02) | < 0.001 |
Knowledge of Any Contraceptive Methods | 1.00 | |||
Has Knowledge (RC) | 1.00 | 1.00 | ||
No Knowledge | 9.86 (3.82–25.44) | < 0.001 | 5.33 (1.48–19.16) | 0.010 |
Modern Contraceptive Use | ||||
Yes (RC) | 1.00 | 1.00 | ||
No | 35.17 (21.03–58.79) | < 0.001 | 24.48 (9.80–61.14) | < 0.001 |
Community-Level Variables | ||||
Place of Residence | ||||
Rural (RC) | 1.00 | |||
Urban | 1.17 (0.77–1.78) | 0.456 | ||
Health Regions | ||||
Bujumbura Mairie | 1.00 | 1.00 | ||
North | 0.96 (0.54–1.70) | 0.887 | 0.46 (0.14–1.44) | 0.182 |
Central East | 0.56 (0.31–1.02) | 0.058 | 0.34 (0.11–1.05) | 0.062 |
West | 0.83 (0.44–1.54) | 0.546 | 0.26 (0.08–0.86) | 0.027 |
South | 0.44 (0.24–0.81) | 0.008 | 0.31 (0.10–0.96) | 0.041 |
Community-level Education | ||||
Low (RC) | 1.00 | |||
High | 0.88 (0.60–1.27) | 0.491 | ||
Community-level Poverty | ||||
Low (RC) | 1.00 | 1.00 | ||
High | 1.31 (0.97–1.78) | 0.083 | 2.19 (1.03–4.64) | 0.042 |
Random effect | Empty model | Full model | ||
Community variance (SE) | 0.83 (0.20) | 0.62 (0.14) | ||
ICC (%) | 20.2 | 15.8 | ||
MOR | 2.37 | 2.11 | ||
Model fitness | ||||
Log likelihood | − 1079.61 | − 452.85 | ||
Deviance | 2159.22 | 905.70 | ||
AIC | 2163.22 | 955.71 | ||
BIC | 2175.74 | 1112.16 | ||
Clusters | 553 | 553 | ||