Table 7.
Most parsimonious Cox models of independent predictors of all-cause death in the FDS2 type 2 diabetes cohort with CAN category defined (n = 830) and valid CARTs variables as continuous variables added and retained if statistically significant (up to n = 1101)
| HR (95% CI) | P-value | |
|---|---|---|
| CAN category defined and valid CART results | ||
| All-cause mortality (n = 827) | ||
| Age (increase of 1 year) | 1.06 (1.04, 1.08) | < 0.001 |
| Male sex | 1.55 (1.12, 2.15) | 0.008 |
| Aboriginal descent | 8.95 (3.26, 24.6) | < 0.001 |
| Other European background | 0.41 (0.19, 0.90) | 0.026 |
| Antidepressant therapy | 3.40 (1.44, 8.00) | 0.005 |
| Angiotensin receptor blocker use | 1.38 (1.001, 1.89) | 0.049 |
| eGFR < 30 mL/min/1.73m2 | 3.03 (1.51, 6.06) | 0.002 |
| Distal symmetrical polyneuropathy | 1.15 (0.96, 1.39) | 0.130 |
| Cerebrovascular disease | 1.70 (1.04, 2.77) | 0.033 |
| HF | 2.31 (1.36, 3.93) | 0.002 |
| MCR (increase of 1.0) | 0.98 (0.97, 0.99) | 0.001 |
| Time-varying covariates | ||
| Aboriginal descent | 0.43 (0.24, 0.76) | 0.004 |
| Antidepressant therapy | 0.57 (0.38, 0.98) | 0.043 |
| Participants of all ages and valid CART results | ||
| All-cause mortality (n = 1091) | ||
| Age (increase of 1 year) | 1.02 (0.998, 1.05) | 0.070 |
| Male sex | 1.45 (1.13, 1.87) | 0.004 |
| Aboriginal descent | 2.61 (1.58, 4.31) | < 0.001 |
| Antidepressant therapy | 1.47 (1.08, 2.01) | 0.016 |
| Ln(uACR)a | 1.10 (1.004, 1.21) | 0.042 |
| eGFR < 30 mL/min/1.73m2 | 3.20 (1.88, 5.44) | < 0.001 |
| Distal symmetrical polyneuropathy | 1.31 (1.09, 1.57) | 0.003 |
| Peripheral arterial disease | 1.42 (1.09, 1.85) | 0.010 |
| HF | 1.67 (1.11, 2.50) | 0.014 |
| MCR (increase of 1.0) | 0.99 (0.98, 0.998) | 0.021 |
| Time-varying covariates | ||
| Age (increase of 1 year) | 1.03 (1.02, 1.05) | < 0.001 |
Hazard ratios (HR) and 95% confidence intervals (95% CI) are shown
aA 2.72-fold increase in urinary albumin:creatinine ratio corresponds to an increased risk of 1 in ln (urinary albumin:creatinine)