Table 2.
Multivariable Cox regression of all-cause mortality in cancer patients and additive interaction.
| Mortality Risk by exposure, HR (95% CI) | Additive Interaction (95% CI) | ||||||
|---|---|---|---|---|---|---|---|
| NO CKD or DM (n = 78,665) |
CKD (n = 1734) |
DM (n = 19,530) |
DM and CKD (n = 1755) |
RERI | AP | SI | |
| Events | 8726 | 504 | 4264 | 649 | |||
| Person-years | 401,652 | 7366 | 87,339 | 6624 | |||
| Model 1 | 1 (Reference) | 3.07 (2.81, 3.36) | 2.21 (2.13, 2.30) | 4.36 (4.02, 4.72) | 0.07 (-0.36, 0.51) | 0.02 (-0.08, 0.12) | 1.02 (0.9, 1.17) |
| Model 2 | 1 (Reference) | 1.53 (1.39, 1.68) | 1.25 (1.2, 1.3) | 1.99 (1.84, 2.17) | 0.22 (0.01, 0.42) | 0.11 (0.01, 0.21) | 1.28 (1.01, 1.62) |
| Sex | |||||||
| Male (n = 50,975) | 1 (Reference) | 1.46 (1.32, 1.63) | 1.22 (1.16, 1.27) | 1.84 (1.67, 2.03) | 0.16 (-0.06,0.39) | 0.09 (-0.03, 0.2) | 1.24 (0.91, 1.67) |
| Female (n = 50,709) | 1 (Reference) | 1.66 (1.37, 2) | 1.33 (1.23, 1.44) | 2.63 (2.23, 3.11) | 0.65 (0.14, 1.16) | 0.25 (0.08, 0.41) | 1.66 (1.11, 2.48) |
| P for interaction | P = 0.53 | ||||||
| Age | |||||||
| < 60 (n = 55,757) | 1 (Reference) | 2.17 (1.54, 3.07) | 1.42 (1.31,1.53) | 2.58 (1.92, 3.48) | -0.01 (-1.07, 1.06) | 0 (-0.42, 0.41) | 1 (0.51, 1.95) |
| ≥ 60 (n = 45,927) | 1 (Reference) | 1.86 (1.7, 2.05) | 1.25 (1.2, 1.31) | 2.26 (2.08, 2.47) | 0.15 (-0.1, 0.4) | 0.07 (-0.04, 0.17) | 1.13 (0.92, 1.4) |
| P for interaction | P = 0.49 | ||||||
| BMI | |||||||
| < 25 kg/m2 (n = 67,718) | 1 (Reference) | 1.51 (1.35, 1.69) | 1.23 (1.17, 1.29) | 2.04 (1.84, 2.26) | 0.3 (0.04, 0.56) | 0.15 (0.03, 0.26) | 1.4 (1.04, 1.89) |
| ≥ 25 kg/m2 (n = 33,966) | 1 (Reference) | 1.5 (1.27, 1.77) | 1.24 (1.16, 1.33) | 1.85 (1.6, 2.13) | 0.1 (-0.24, 0.45) | 0.06 (-0.13, 0.24) | 1.14 (0.74, 1.76) |
| P for interaction | P = 0.51 | ||||||
*Abbreviations: AP Attributable Proportion due to interaction, CI Confidence Interval, CKD Chronic Kidney Disease, DM Diabetes Mellitus, HR Hazard Ratio, RERI Relative Excess Risk due to Interaction, SI Synergy Index.
*Null hypothesis for each interaction is RERI = 0, AP = 0, and SI = 1.
*Model 1 is the unadjusted model; Model 2 is adjusted for age, sex, BMI, smoking status, alcohol consumption, history of hypertension, and cancer types.
*P for interaction was tested by three-way interaction.