Table 3.

Proportional-Hazards Models of the Effect of Time-Dependent Covariates on the Risk of Cardiovascular Disease and of the Effect of the Treatment Group after Adjustment for the Time-Dependent Covariate.

Time-Dependent Covariate	Effect of Time-Dependent Covariate*		Treatment Group Adjusted for Time-Dependent Covariate*
	Hazard Ratio (95% CI)	P Value	Hazard Ratio (95% CI)	P Value
None	—	—	0.53 (0.34-0.83)	0.005
Renal disease (yes vs. no)†	2.99 (0.72-12.5)	0.20	0.54 (0.34-0.84)	0.006
Microalbuminuria (yes vs. no)‡	2.93 (1.85-4.65)	<0.001	0.62 (0.39-0.97)	0.04
Albuminuria (yes vs. no) §	2.57 (1.36-4.88)	0.009	0.58 (0.37-0.91)	0.02
Mean glycosylated hemoglobin value¶
Per 10% increase	1.25 (1.10-1.43)	<0.001	0.84 (0.43-1.64)	0.61
Per 10% decrease	0.80 (0.70-0.91)	<0.001

^*All models were adjusted for the glycosylated hemoglobin value, age, cholesterol level, and smoking status at baseline in the DCCT.

^†Renal disease was defined by a serum creatinine level of at least 2 mg per deciliter, a history of kidney transplantation, or the implementation of dialysis.

^‡Microalbuminuria was defined by a history of microalbuminuria or renal disease.

^§Albuminuria was defined by a history of albuminuria or renal disease.

^¶The log mean glycosylated hemoglobin value was used so that the hazard ratio per c-fold change in risk is c^2.26144, where 2.26144 is the estimated regression coefficient; a c of 1.1 corresponds to a 10 percent increase in the mean glycosylated hemoglobin value, and a c of 0.9 to a 10 percent decrease.