Table 2.
Association of socioeconomic indicators at age 50 years with subsequent mortality
| HR (95% CI) | pnon-linearity | ||
|---|---|---|---|
| Education | .. | 0·13 | |
| High | 1 (ref) | .. | |
| Medium | 1·22 (0·98–1·52) | .. | |
| Low | 1·13 (0·92–1·38) | .. | |
| Education scale* | 1·09 (0·89–1·32) | .. | |
| Occupation | .. | 0·93 | |
| High | 1 (ref) | .. | |
| Medium | 1·24 (1·04–1·49) | .. | |
| Low | 1·57 (1·18–2·09) | .. | |
| Occupation scale* | 1·57 (1·19–2·07) | .. | |
| Literacy† | .. | 0·37 | |
| High | 1 (ref) | .. | |
| Medium | 1·00 (0·82–1·22) | .. | |
| Low | 1·19 (0·96–1·48) | .. | |
| Literacy scale* | 1·02 (0·93–1·12) | .. | |
611 deaths occurred in 6425 participants. Analyses were done using proportional hazards regression with Weibull distribution and inverse probability weighting. Models were adjusted for age, sex, race, marital status, and birth cohort. HR=hazard ratio.
*
Categories of socioeconomic status indicators were recoded (0, 0·5, and 1·0) so that when entered as a continuous variable the reported HR corresponded to the increase in risk in the lowest socioeconomic group compared with the highest socioeconomic group.
†
Assessed using the Mill Hill vocabulary test.