Table 4.
Longitudinal repeated measures mixed effects model showing the associations between predictors and paternal depressive symptoms in the Millennium Cohort Study across all time points a
| Predictors from all sweeps | Coefficient (95 % CI)b | p |
|---|---|---|
| MCS wave | ||
| Wave 1 | reference group | |
| Wave 2 | -0.49 (-0.61 - -0.37) | <0.001 |
| Wave 3 | -0.44 (-0.56 - -0.31) | <0.001 |
| Wave 4 | -0.38 (-0.49 - -0.22) | <0.001 |
| Family context factors | ||
| Maternal depressive symptoms | 0.08 (0.06 - 0.10) | <0.001 |
| Marital conflict | 0.84 (0.77 - 0.91) | <0.001 |
| Socioeconomic factors | ||
| Paternal education | 0.522 | |
| No qualifications | reference group | |
| School level, Degree or higher | -0.10 (-0.40 - 0.20) | |
| Ethnic background | ||
| White | reference group | |
| Indian, Pakistani, & Bangladeshi | 0.56 (0.05 - 1.07) | 0.031 |
| Afro-Caribbean & British afro-Caribbean | -0.74 (-1.23 - 0.25) | 0.003 |
| Mixed, other | 0.76 (-0.01 - 1.53) | 0.053 |
| Fathers employment | ||
| Employed | reference group | |
| Unemployed | 1.13 (0.70 - 1.56) | <0.001 |
| Housing tenure | 0.001 | |
| Council and private renting, % | reference group | |
| home owner, % | -0.41 (-0.66 - -0.17) | |
| Family income (per £1000 p.a.) | -0.31 (-0.58 - -0.06) | 0.022 |
a Finite population correction factor could not be fitted onto model, but all other required weights (accounting for clustering, attrition/non-response and oversampling for ethnic minorities/disadvantaged participants) were incorporated
b Sample size 4766