Table 2.
Univariable and multivariable Cox proportional hazards models for incident cancer
| Univariable | Multivariable | |||
|---|---|---|---|---|
| HR (95% CI) | P | HR (95% CI) | P | |
| Surgery (yes vs. no) | 0.60 (0.43–0.84) | 0.003 | 0.63 (0.44–0.89) | 0.008 |
| Age (years) | 1.07 (1.04–1.10) | <0.001 | 1.07 (1.04–1.10) | <0.001 |
| Sex (men vs. women) | 1.04 (0.75–1.46) | 0.801 | 0.89 (0.61–1.31) | 0.558 |
| Alcohol consumption (g/day) | 1.01 (0.98–1.03) | 0.560 | 1.01 (0.98–1.03) | 0.664 |
| Daily smoking (yes vs. no) | 1.28 (0.88–1.86) | 0.204 | 1.46 (0.99–2.15) | 0.057 |
| Sagittal diameter (cm) | 1.03 (0.98–1.08) | 0.230 | 1.05 (1.00–1.10) | 0.075 |
| Serum insulin (pmol/L) | 1.00 (0.99–1.01) | 0.506 | 1.01 (1.00–1.01) | 0.154 |
| University education | 0.95 (0.59–1.53) | 0.834 | 0.89 (0.56–1.42) | 0.625 |
HRs are expressed as “vs.” for dichotomous variables and per unit for continuous variables. Men coded as 1 and women as 0. Yes coded as 1 and no as 0.