Figure 1.

All-cause and Cause-specific Adjusted Hazard Ratio for Diabetes Over Time in HEMO Study Patients§§.§§ (a) All-cause mortality HR estimate using the Cox’s proportional hazards (Cox-PH) is 2.11 (p = 0.003), HR for DM using Cox-TVC model is 1.57 (p = 0.11) for main effects and 1.13 (p = 0.02) for the interaction term, Gray’s model estimates are 1.41 to 2.21 (p <0.001). (b) Cardiac mortality HR estimate using the Cox-PH is 1.81 (p = 0.16), HR for DM using Cox-TVC model is 1.22 (p = 0.68) for main effects and 1.18 (p = 0.07) for the interaction term, Gray’s model estimates are 1.70 to 2.46 (p = 0.003). (c) Cardiovascular mortality HR estimate using the Cox-PH is 2.00 (p = 0.06), HR for DM using Cox-TVC model is 1.28 (p = 0.55) for main effects and 1.21 (p = 0.02) for the interaction term, Gray’s model estimates are 1.58 to 2.46 (p = 0.001). (d) Infectious mortality HR estimate using the Cox-PH is 2.01 (p = 0.22), HR for DM using Cox-TVC model is 1.48 (p = 0.53) for main effects and 1.12 (p = 0.27) for the interaction term, Gray’s model estimates are 1.17 to 2.21 (p = 0.21).

Sattar et al. BMC Nephrology 2012 13:130   doi:10.1186/1471-2369-13-130
Download authors' original image