Table 4.
Multivariate adjusted linear regression: change in carer negative affect (n = 159) and change in carer risk of depression (n = 133) between baseline and follow-up
| Explanatory variable |
Model 1
|
Model 2
|
Model 3
|
|---|---|---|---|
| ∆ Negative affect 1 β (95% CI) | ∆ Negative affect 1 β (95% CI) | ∆ Risk of depression 2 β (95% CI) | |
|
Community ID
|
|
|
|
| 1 |
ni |
-0.28 (-1.13, 0.58) |
ns |
| 2 |
ni |
-0.74 (-1.52, 0.03) |
ns |
| 3 |
ni |
0.17 (-0.50, 0.84) |
ns |
| 4 |
ni |
0.81 (-0.31, 1.92) |
ns |
| 5 |
ni |
1.78 (1.00, 2.56) |
ns |
| 6 |
ni |
-0.15 (-1.13, 0.84) |
ns |
| 7 |
ni |
-0.82 (-1.65, 0.01) |
ns |
| 8 (median change in negative affect) |
ni |
0.0 |
ns |
| 9 |
ni |
0.35 (-1.06, 1.77) |
ns |
| 10 |
ni |
1.26 (-0.21, 2.72) |
ns |
|
Carer age at baseline
|
|
|
|
| <20 yrs |
ns |
ns |
-9.70 (-15.58, -3.83) |
| 20-34 yrs |
ns |
ns |
0.0 |
| 35 + yrs |
ns |
ns |
0.76 (-2.28, 3.80) |
|
Change in carer ran out of money last year
|
|
|
|
| Other |
0.0 |
ns |
ns |
| From had money to ran out of money last year |
-0.62 (-1.18, -0.05) |
ns |
ns |
|
Change in child day care attendance
|
|
|
|
| No change |
ns |
ns |
0.0 |
| Change from No to Yes |
ns |
ns |
-4.03 (-6.37, -1.68) |
| Change from Yes to No |
ns |
ns |
-1.66 (-9.55, 6.23) |
|
∆ in number of people carer could get help from not in house
|
|
|
|
| Other |
0.0 |
0.0 |
ns |
| From none to able to get help from one or more |
0.86 (0.26, 1.45) |
0.65 (0.05, 1.26) |
ns |
|
Change in carer’s has clan/kin in community
|
|
|
|
| Other |
ns |
ns |
0.0 |
| Had clan/kin in community to did not |
ns |
ns |
-3.55 (-5.46, -1.64) |
|
NLES: ∆ in knowing someone who had a serious accident
|
|
|
|
| Other |
0.0 |
0.0 |
0.0 |
| From yes to no |
1.18 (0.59, 1.78) |
0.98 (0.48, 1.47) |
9.38 (3.44, 15.31) |
|
NLES: ∆ in knowing someone sent to jail
|
|
|
|
| Other |
0.0 |
ns |
ns |
| From yes to no |
1.13 (0.53, 1.73) |
ns |
ns |
| Adjusted R 2 | 22.6% | 35.1% | 22.1% |
1 Carer Negative affect balance (range: -4 to 4); 2 Carer Risk of depression (range: -30 to 30).
A positive Beta coefficient indicates the variable is associated with an improvement in mental health. Bold font indicates the variable was significant at p ≤ 0.05; ni – not applicable; ns – not significant.