Table 2. Population-Based Analyses on the Association Between Parental Death and Eligibility for Attending Secondary Education and Mean School Grades Among 908 064 Swedish Childrena.
| Variable | No. | School outcomes in the final year of compulsory school at age 15-16 y | ||||
|---|---|---|---|---|---|---|
| Ineligibility for upper secondary education | Mean grade z score, β coefficient (95% CI) | |||||
| No. (%) | Risk ratio (95% CI) | |||||
| Unadjusted | Adjusted risk ratio | Unadjusted | Adjusted | |||
| Parental death before finishing school | ||||||
| Parental death | 22 634 | 3570 (15.8) | 1.74 (1.68 to 1.80)b | 1.36 (1.32 to 1.41)b | −0.30 (−0.32 to −0.29)b | −0.19 (−0.21 to −0.18)b |
| No parental death | 885 430 | 80 241 (9.1) | 1 [Reference] | 1 [Reference] | 0 [Reference] | 0 [Reference] |
| Maternal death before finishing school | ||||||
| Maternal death | 6712 | 927 (13.8) | 1.52 (1.43 to 1.63)b | 1.31 (1.23 to 1.39)b | −0.23 (−0.26 to −0.20)b | −0.17 (−0.19 to −0.15)b |
| No parental death | 885 430 | 80 241 (9.1) | 1 [Reference] | 1 [Reference] | 0 [Reference] | 0 [Reference] |
| Paternal bereavement before finishing school | ||||||
| Paternal death | 15 922 | 2643 (16.6) | 1.83 (1.76 to 1.90)b | 1.38 (1.33 to 1.43)b | −0.33 (−0.35 to −0.31)b | −0.20 (−0.22 to −0.19)b |
| No parental death | 885 430 | 80 241 (9.1) | 1 [Reference] | 1 [Reference] | 0 [Reference] | 0 [Reference] |
a
Poisson regression was used for eligibility, and ordinary least square linear regression was used for mean grade z score. Cluster robust estimation for standard error was used to account for associations between siblings. In both regression models, adjusted estimates were obtained from models described in the Methods section.
b
Statistical significance at the P < .05 level.