Factors associated with childhood mortality identified by multilevel multivariate logistic regression models
| Variable | Model 1* | Model 2† | Model 3‡ | Model 4§ |
|---|---|---|---|---|
| Fixed effects | aOR (CrI) | aOR (CrI) | aOR (CrI) | aOR (CrI) |
| Individual-level factors | ||||
| Male (vs female) | 1.20 (1.11–1.30) | 1.20 (1.11–1.30) | 1.20 (1.11–1.30) | |
| Age of mothers in years | ||||
| 15–24 | 1 (reference) | 1 (reference) | 1 (reference) | |
| 25–34 | 0.89 (0.81–0.98) | 0.90 (0.82–0.98) | 0.90 (0.81–0.98) | |
| 35+ | 1.03 (0.93–1.15) | 1.04 (0.93–1.15) | 1.04 (0.93–1.16) | |
| Education attainment of mother | ||||
| No education | 1 (reference) | 1 (reference) | 1 (reference) | |
| Primary | 0.98 (0.86–1.09) | 0.96 (0.85–1.08) | 0.96 (0.85–1.09) | |
| Secondary | 0.80 (0.70–0.91) | 0.78 (0.67–0.91) | 0.79 (0.67–0.91) | |
| Higher | 0.69 (0.51–0.89) | 0.68 (0.49–0.90) | 0.68 (0.50–0.88) | |
| Occupation of mother | ||||
| Not working | 1 (reference) | 1 (reference) | 1 (reference) | |
| Manual | 0.92 (0.65–1.24) | 0.93(0.66–1.26) | 0.93 (0.67–1.27) | |
| White collar | 1.04 (0.92–1.16) | 1.05(0.94–1.16) | 1.05 (0.94–1.17) | |
| Others | 1.08 (0.98–1.26) | 1.09(0.99–1.26) | 1.09 (0.99–1.27) | |
| Wealth index of family | ||||
| Richest | 1 (reference) | 1 (reference) | 1 (reference) | |
| Richer | 1.30 (1.08–1.54) | 1.23 (1.00–1.54) | 1.23 (1.04–1.46) | |
| Middle | 1.36 (1.11–1.63) | 1.19 (0.96–1.48) | 1.20 (0.99–1.45) | |
| Poorer | 1.91 (1.55–2.31) | 1.53 (1.22–2.01) | 1.55 (1.25–1.93) | |
| Poorest | 1.79 (1.44–2.16) | 1.40 (1.09–1.86) | 1.41 (1.12–1.79) | |
| Neighbourhood-level factors | ||||
| Rural (vs urban) | 1.24 (1.09–1.40) | 1.25 (1.08–1.43) | ||
| High (vs low) poverty rate | 1.25 (1.07–1.46) | 1.22 (1.05–1.43) | ||
| High (vs low) illiteracy rate | 1.04 (0.72–1.09) | 1.03 (0.71–1.09) | ||
| High (vs low) unemployment rate | 1.07 (0.96–1.19) | 1.05 (0.94–1.17) | ||
| State-level factors | ||||
| High (vs low) poverty rate | 1.20 (0.93–1.60) | |||
| High (vs low) illiteracy rate | 0.90 (0.64–1.16) | |||
| High (vs low) unemployment rate | 1.09 (0.79–1.43) | |||
| Measures of variation | ||||
| State level | ||||
| Variance (SE) | 0.116 (0.062–0.204) | 0.046 (0.020–0.092) | 0.045 (0.016–0.091) | 0.040 (0.020–0.094) |
| Explained variation (%) | Reference | 60.3 | 61.2 | 65.5 |
| Intrastate correlation (%) | 3.3 | 1.3 | 1.3 | 1.2 |
| MOR | 1.38 | 1.25 | 1.22 | 1.14 |
| Neighbourhood level | ||||
| Variance (SE) | 0.165 (0.114–0.218) | 0.127 (0.084–0.182) | 0.122 (0.075–0.169) | 0.120 (0.077–0.167) |
| Explained variation (%) | Reference | 23.0 | 26.1 | 27.3 |
| Intraneighbourhood correlation (%) | 7.9 | 5.0 | 4.8 | 4.6 |
| MOR | 1.47 | 1.40 | 1.39 | 1.39 |
| Model fit statistics | ||||
| Bayesian DIC | 18 878 | 18 785 | 18 779 | 18 767 |
*Model 1 is the empty model, a baseline model without any predictor variable.
†Model 2 is adjusted for sex of child, age, education, occupation and wealth status of mother.
‡Model 3 is additionally adjusted for neighbourhood-level socioeconomic factors.
§Model 4 is additionally adjusted for state-level socioeconomic factors.
aOR, adjusted OR; CrI, credible interval; DIC, deviation information criterion; MOR, median OR.