TABLE 3.
Effect of Long Working Hours and Low Job Control on Self-Rated Health∗
Long Working Hours (−) | Long Working Hours (+) | OR for Long Working Hours (−) versus Long Working Hours (+) Within Strata of Job Control | |
OR (95% CI): P | OR (95% CI): P | OR (95% CI): P | |
High job control | reference | 1.24 (1.13–1.35): <0.001 | 1.24 (1.13–1.35): <0.001 |
Low job control | 1.04 (0.97–1.13): 0.252 | 1.47 (1.33–1.62): <0.001 | 1.40 (1.27–1.55): <0.000 |
OR for low job control (0) versus low job control (1) Within strata of long working hours | 1.04 (0.97–1.13): 0.252 | 1.18 (1.06–1.31): <0.001 | |
Measure of interaction on additive scale: RERI | 0.18 (0.02–0.34): 0.027 | ||
Measure of interaction on multiplicative scale: ratio of ORs | 1.13 (0.99–1.28): 0.061 |
Long working hours (−): within the legal limit: 36 ≤ working hours ≤ 52.
Long working hours (+): more than the legal limit: 52 < working hours.
CI, confidence interval; OR, odds ratio; RERI, relative excess risk due to interaction.
*The model was adjusted for age, sex, education, income, occupation, smoking, and alcohol consumption. ORs and ratios of ORs were estimated using the linear combination command, and RERI was estimated using the nonlinear combination command after multiple survey logistic analysis (linear combination and nonlinear combination are post-estimation commands for the combination of effects in Stata).