|
Parameter estimates of the risk score for heart failure hospitalisation in the training dataset. |
||||||
| Univariate Analysis |
Multivariate Analysis |
|||||
|
|
||||||
| Variables† |
Hazard ratio (95% CI) |
P value |
Estimates of β values |
S.E.M |
Hazard ratio (95% CI) |
P value |
|
|
||||||
| Variables used in the predicting model |
||||||
| Age, per year |
1.10(1.08–1.12) |
<.0001 |
0.0709 |
0.0107 |
1.07(1.05–1.10) |
<.0001 |
| Body mass index, per kg/m2 |
1.01(0.97–1.05) |
0.6568 |
0.0627 |
0.0286 |
1.07(1.01–1.13) |
0.0282 |
| Glycated haemoglobin, per % |
1.16(1.07–1.25) |
0.0003 |
0.1363 |
0.0550 |
1.15(1.03–1.28) |
0.0132 |
| Log10(1+spot urinary albumin to creatinine ratio, per mg/mmol) |
4.50(3.65–5.54) |
<.0001 |
0.9915 |
0.1409 |
2.70(2.32–3.55) |
<.0001 |
| Blood haemoglobin, per g/dL |
0.61(0.55–0.68) |
<.0001 |
-0.3606 |
0.0684 |
0.70(0.61–0.80) |
0.0001 |
| Coronary heart disease during follow-up (1. yes, 0. no) |
3.61(2.24–5.81) |
<.0001 |
0.8161 |
0.2675 |
2.26(1.26–4.05) |
0.0061 |
| Variables not used in the predicting model |
||||||
| Duration of diabetes, per year |
1.07(1.05–1.09) |
<.0001 |
Not selected |
|||
| Glomerular filtration rate, per ml min-1 1.73 m-2 |
0.97(0.96–0.97) |
<.0001 |
Not selected |
|||
| Systolic blood pressure, per mmHg |
1.03(1.02–1.04) |
<.0001 |
Not selected |
|||
| Diastolic blood pressure, per mmHg |
1.01(0.99–1.02) |
0.5640 |
Not selected |
|||
| Low density lipoprotein cholesterol, per mmol/L |
1.12(0.94–1.33) |
0.2155 |
Not selected |
|||
| High density lipoprotein cholesterol, per mmol/L |
0.80(0.48–1.34) |
0.3976 |
Not selected |
|||
| Triglyceride, per mmol/L |
1.04 (0.96–1.12) |
0.3665 |
Not selected |
|||
| Smoking status |
Not selected |
|||||
| Current smokers |
0.62(0.37–1.02) |
0.0592 |
||||
| Ex-smokers |
1.96(1.28–3.01) |
0.0020 |
||||
| Never |
1.0 |
|||||
| Use of oral anti-diabetic drugs at baseline |
1.04(0.73–1.47) |
0.8471 |
Not selected |
|||
| Use of anti-hypertensive drugs at baseline |
3.98(2.79–5.67) |
<.0001 |
Not selected |
|||
| Use of LLD at baseline |
1.88(1.16–3.09) |
0.0112 |
Not selected |
|||
| Use of ACEI or ARB at baseline |
2.44(1.66–3.59) |
<.0001 |
Not selected |
|||
| Use of insulin at baseline |
3.59(2.52–5.10) |
<.0001 |
Not selected |
|||
|
†, Sex-stratified Cox model was used to obtain estimates of β values and hazard rato of predctors for heart failure. χ2 test for likelihood ratio test = 234.60, Degree of freedom = 6, p < 0.0001 and the shrinkage of the model is = (234.60-6)/274.66 = 0.9744; Not selected indiates not selected by the backward algorithm at the significant levelt of 0.05. CI, confidence interval; LLD, lipid lowering drugs; ACEI, Angiotensin-converting enzyme inhibitors; ARB, angiotensin II receptor blockers | ||||||
Yang et al. Cardiovascular Diabetology 2008 7:9 doi:10.1186/1475-2840-7-9 |
||||||