Table 2 Multilevel logistic regression models for individual and household/community predictors of adolescent pregnancy in Nigeria
Variables | Model 0 | Model I | Model II | Model III |
|---|---|---|---|---|
| Â | aOR [95% CI] | aOR [95% CI] | aOR [95% CI] | |
Fixed effects results | ||||
 Individual-level variables |  |  |  |  |
Adolescent current age | ||||
 15 |  | RC |  | RC |
 16 |  | 1.08 [0.66-1.77] |  | 1.09 [0.67-1.78] |
 17 |  | 0.94 [0.59-1.48] |  | 0.95 [0.60-1.50] |
 18 |  | 1.24 [0.81-1.92] |  | 1.25 [0.81-1.94] |
 19 |  | 1.06 [0.68-1.67] |  | 1.10 [0.70-1.74] |
Age at sexual debut | ||||
 Less than 15 years |  | RC |  | RC |
 Between 15 and 19 |  | 1.50** [1.16-1.92] |  | 1.49** [1.161.92] |
level of education | ||||
 No Education |  | RC |  | RC |
 Primary |  | 1.27 [0.91-1.76] |  | 1.28 [0.92-1.80] |
 Secondary & above |  | 1.27 [0.93-1.75] |  | 1.43* [1.00-2.03] |
Marital status | ||||
 Never married |  | RC |  | RC |
 Currently Married |  | 70.71*** [44.12-113.32] |  | 67.00*** [41.27-108.76] |
 Cohabitating |  | 79.48*** [40.04-157.74] |  | 69.35*** [34.16-140.80] |
 Previously Married |  | 7.85** [1.77-34.71] |  | 7.91** [1.78-35.19] |
Working status | ||||
 Not working |  | RC |  | RC |
 Working |  | 0.72** [0.57-0.91] |  | 0.69** [0.55-0.88] |
Ethnicity | ||||
 Hausa |  | RC |  | RC |
 Yoruba |  | 1.72 [0.92-3.22] |  | 1.38 [0.53-3.57] |
 Igbo |  | 1.68 [0.92-3.04] |  | 3.73* [1.04-13.30] |
 Others |  | 1.18 [0.88-1.58] |  | 1.31 [0.93-1.84] |
Religion | ||||
 Christianity |  | RC |  | RC |
 Islam |  | 1.29 [0.87-1.91] |  | 1.37 [0.89-2.10] |
 Traditionalist & others |  | 2.17 [0.50-9.40] |  | 2.00 [0.45-8.80] |
Exposure to media | ||||
 No |  | RC |  | RC |
 Yes |  | 0.83 [0.66-1.05] |  | 0.86 [0.67-1.09] |
 Household-level |  |  |  |  |
Place of residence | ||||
 Urban |  |  | RC | RC |
 Rural |  |  | 1.10 [0.82-1.48] | 0.87 [0.63-1.20] |
Wealth index | ||||
 Poorest |  |  | RC | RC |
 Poorer |  |  | 1.36* [1.06-1.75] | 1.08 [0.83-1.41] |
 Middle |  |  | 1.09 [0.80-1.49] | 1.05 [0.75-1.47] |
 Richer |  |  | 0.94 [0.63-1.42] | 0.87 [0.55-1.37] |
 Richest |  |  | 0.37** [0.18-0.77] | 0.47 [0.21-1.04] |
Sex of household head | ||||
 Male |  |  | RC | RC |
 Female |  |  | 0.48*** [0.33-0.70] | 0.89 [0.58-1.35] |
Region | ||||
 North Central |  |  | RC | RC |
 North East |  |  | 1.47* [1.05-2.07] | 1.93 [0.64-1.37] |
 North West |  |  | 2.18*** [1.59-2.98] | 1.11 [0.74-1.66] |
 South East |  |  | 1.03 [0.60-1.75] | 0.52 [0.14-1.96] |
 South South |  |  | 1.26 [0.73-2.17] | 1.15 [0.60-2.20] |
 South West |  |  | 1.43 [0.83-2.45] | 1.73 [0.71-4.21] |
Community literacy level | ||||
 Low |  |  | RC | RC |
 Medium |  |  | 0.46*** [0.35-0.61] | 0.94 [0.68-1.30] |
 High |  |  | 0.24*** [0.16-0.38] | 0.88 [0.52-1.48] |
Community socioeconomic status | ||||
 Low |  |  | RC | RC |
 Medium |  |  | 0.43 [0.16-1.12] | 0.56 [0.68-1.30] |
 High |  |  | 0.85 [0.55-1.32] | 0.90 [0.55-1.48] |
Random effects results | ||||
 PSU Variance (95% CI) | 1.09 [0.79-1.51] | 0.03 [0.00-10.36] | 0.27 [0.12-0.60] | 0.03 [0.00-16.42] |
 ICC | 0.25 | 0.01 | 0.08 | 0.01 |
 LR Test | χ2 = 84.61, p < 0.001 | χ2 = 0.12, p > 0.05 | χ2 = 8.53, p < 0.01 | χ2 = 0.10, p > 0.05 |
 Wald χ2 | Reference | 525.05*** | 261.32*** | 523.29*** |
Model fitness | ||||
 Log-likelihood | − 1840.77 | − 1207.84 | − 1680.80 | − 1199.78 |
 AIC | 3685.54 | 2453.68 | 3395.60 | 2467.56 |
 BIC | 3699.61 | 2587.42 | 3515.26 | 2706.88 |
 Number of clusters | 1363 | 1363 | 1363 | 1363 |