Mendelian randomisation causal effect estimates for depression on employment outcomes
| Outcome | Method | OR | 95% CI | P value |
| Not in paid employment | IVW fixed effects | 1.158 | 1.064 to 1.26 | 6.7E-04 |
| Sick/disabled | IVW fixed effects | 1.563 | 1.344 to 1.819 | 7.4E-09 |
| Caring for home/family | IVW fixed effects | 1.164 | 0.977 to 1.386 | 9.0E-02 |
| Retired | IVW fixed effects | 1.078 | 0.958 to 1.212 | 2.1E-01 |
| Unemployed | IVW fixed effects | 1.031 | 0.840 to 1.266 | 7.7E-01 |
| Outcome | Method | Beta | 95% CI | P value |
| Townsend Deprivation Index | IVW random effects | 0.095 | −0.0744 to 0.264 | 2.7E-01 |
| Hours worked | IVW random effects | −0.130 | −0.685 to 0.424 | 6.5E-01 |
| Outcome | Method | OR | 95% CI | P value |
| Highest educational attainment | IVW random effects | 1.092 | 0.938 to 1.273 | 2.6E-01 |
| Household income | Egger random effects | 1.239 | 0.817 to 1.881 | 3.2E-01 |
All outcome effects were adjusted for age, sex, study assessment centre, and genetic principal components. Household income level effect was additionally adjusted for number in household winsorised to 12. We report the estimate of the MR method selected via the Rücker Model selection framework.