Table 2.

Details of the Fine and Gray competing risks proportional hazards model for the estimation of major vascular events

Variable	Coefficienta	sHR	95% CI	p value
Glucose treatment allocation (intensive vs standard)	−0.0992	0.90	0.83,0.98	0.021
Sex (women vs men)	−0.3548	0.70	0.63, 0.77	<0.001
Age (per 1 year)	0.0099	1.01	1.00, 1.02	0.009
Duration of diabetes (per 1 year)	0.0279	1.03	1.02, 1.04	<0.001
Systolic BP if untreated (per 1 mmHg)	0.0039	1.00	1.00, 1.01	0.000
Systolic BP if treated (per 1 mmHg)	0.0050	1.01	1.00, 1.01	<0.001
Non-HDL-cholesterol (per 1 mmol/l)	0.0429	1.04	1.01, 1.08	0.022
HbA1c (per 1%)	0.1189	1.13	1.10, 1.16	<0.001
UACR (per 1 mg/mmol loge)	0.1625	1.18	1.14, 1.22	<0.001
eGFR (per 1 ml/min increase)	−0.0255	0.97	0.96, 0.99	<0.001
eGFR squared (per 1 ml/min2)	0.0001	1.00	1.00, 1.00	0.010
Waist circumference (per 1 cm)	−0.0003	1.00	1.00, 1.00	0.880
Smoking (current vs never or former)	0.0032	1.00	0.88, 1.14	0.960
History of microvascular disease (yes vs no)	0.4738	1.62	1.42, 1.86	<0.001
History of macrovascular disease (yes vs no)	0.3180	1.38	1.26, 1.51	<0.001
Educational attainment (≥16 years at completion of education)	−0.2309	0.79	0.72, 0.87	<0.001
Region
Eastern Europe	0.0344	1.04	0.90, 1.19	0.610
Asia	0.2879	1.34	1.19, 1.51	<0.001
BP treatment allocation (perindopril/indapamide vs placebo)	−0.0822	0.92	0.84, 1.00	0.056

5 year major vascular event risk (%) = (1 − S₀(5)^{exp(A-1.6641)}) × 100%. Where S₀(5) = 0.8363 (the 5 year baseline survival) and A is the sum, over all variables in the model, of the patient’s specific value × the corresponding coefficient

^aCoefficients were penalised by a shrinkage factor of 0.979 to increase external validity, whereas unbiased HRs and statistics were derived from an unpenalised Fine and Gray model

To convert values for HbA_1c in % to mmol/mol, subtract 2.15 and multiply by 10.929

sHR, subdistribution hazard ratio