Table 5.
Multilevel logistic models for categorical boys' and girls' smoking outcomes after imputation, adjusting for age, FAS, SIMD, School type and rurality, MCMCa estimation, odds ratios and credible intervals.
| Fixed effects | Tried smoking | Current smoking | Weekly smoking | Daily smoking |
|---|---|---|---|---|
| Boys | ||||
| Deprivation (Ref: SIMDb 5 (least deprived)) | ||||
| SIMD 4 | 1.22 (0.85, 1.77) | 1.23 (0.76, 1.99) | 1.07 (0.62, 1.84) | 0.92 (0.47, 1.78) |
| SIMD 3 | 1.13 (0.77, 1.67) | 1.64 (0.99, 2.74) | 1.89 (1.07, 3.32)* | 1.72 (0.92, 3.21) |
| SIMD 2 | 2.25 (1.50, 3.38)* | 2.06 (1.19, 3.56)* | 1.94 (1.06, 3.53)* | 1.86 (0.97, 3.59) |
| SIMD 1 (most deprived) | 1.58 (0.97, 2.56) | 1.69 (0.87, 3.29) | 1.87 (0.90, 3.86) | 1.82 (0.82, 4.03) |
| Random effects | ||||
| Level 1 (child) variancec | 1 | 1 | 1 | 1 |
| Level 2 (school) variance | 0.178 (0.110) | 0.430 (0.173) | 0.179 (0.161) | 0.220 (0.209) |
| Level 3 (Education authority) variance | 0.137 (0.098) | 0.079 (0.098) | 0.240 (0.177) | 0.324 (0.228) |
| Girls | ||||
| Family Affluence Scale (Ref: High FAS) | ||||
| Medium FAS | 1.22 (0.95, 1.57) | 1.45 (1.05, 1.99)* | 1.47 (1.02, 2.10)* | 1.53 (1.01, 2.31)* |
| Low FAS | 1.26 (0.97, 1.64) | 1.63 (1.18, 2.25)* | 1.66 (1.15, 2.39)* | 1.56 (1.03, 2.35)* |
| Deprivation (Ref: SIMDb 5) | ||||
| SIMD 4 | 1.26 (0.90, 1.78) | 1.18 (0.75, 1.83) | 1.27 (0.77, 2.10) | 1.43 (0.79, 2.58) |
| SIMD 3 | 1.49 (1.05, 2.12)* | 1.60 (1.03, 2.49)* | 1.85 (1.10, 3.10)* | 2.02 (1.12, 3.65)* |
| SIMD 2 | 1.68 (1.12, 2.54)* | 1.73 (1.06, 2.82)* | 1.91 (1.07, 3.42)* | 2.51 (1.32, 4.80)* |
| SIMD 1 (most deprived) | 1.77 (1.15, 2.73)* | 2.33 (1.34, 4.04)* | 3.08 (1.69, 5.62)* | 3.98 (2.06, 7.69)* |
| School type (ref: State school) | ||||
| Independent school | 1.46 (0.72, 2.96) | 2.42 (1.17, 5.02)* | 1.45 (0.62, 3.39) | 0.08 (0.01, 1.11) |
| Rurality (Ref: 4 Cities) | ||||
| Other urban | 1.30 (0.84, 2.01) | 1.33 (0.80, 2.24) | 1.02 (0.58, 1.78) | 1.25 (0.68, 2.28) |
| Accessible towns | 1.64 (1.01, 2.66)* | 1.21 (0.64, 2.26) | 1.03 (0.51, 2.05) | 1.49 (0.69, 3.19) |
| Remote towns | 2.07 (1.19, 3.59)* | 1.34 (0.69, 2.58) | 0.91 (0.43, 1.91) | 1.50 (0.68, 3.30) |
| Accessible rural | 1.28 (0.82, 1.99) | 1.45 (0.82, 2.54) | 1.21 (0.65, 2.26) | 1.97 (1.02, 3.80)* |
| Remote rural | 1.73 (1.04, 2.87)* | 1.99 (1.08, 2.87)* | 1.67 (0.90, 3.11) | 2.16 (1.13, 4.11)* |
| Random effects | ||||
| Level 1 (child) variancec | 1 | 1 | 1 | 1 |
| Level 2 (school) variance | 0.347 (0.108) | 0.328 (0.136) | 0.306 (0.166) | 0.128 (0.135) |
| Level 3 (Education authority) variance | 0.066 (0.070) | 0.034 (0.046) | 0.056 (0.071) | 0.081 (0.087) |
∗Significant at 95% level.
a
Via Markov chain Monte Carlo (MCMC); estimates are based on a chain of length of 50,000 following a burn-in of 15,000.
b
SIMD income domain.
c
Variance at the child level is constrained to 1.