We know that the DCCT nad UKPDS both showed reduction and diabetic complications especially the microvascular complications in both type 1 and 2 diabetes. However, I just want to stress herehow important is stricter glycemic control: (quoted from: http://www.medscape.com/viewarticle/551055)

What is very interesting is what happens to patients who achieve very good control, and then are followed after the clinical trial ends. One of the examples is the ongoing follow-up of the data from DCCT. The patients who ended the DCCT in 1993 have now been followed for about 12 years. At the end of the DCCT, the A1C difference between the 2 groups was about 2%: 9% vs. 7%. However, 1 year after the study ended, the difference between the two groups actually became far less, as happens in any clinical trial. People who were in the intensive group got tired of taking four shots of insulin and measuring their blood glucose six times a day. And their A1C started to go up. And those in the conventional treatment group actually saw the light, and they immediately started improving their glycemic control and their A1C improved. So within the first year, the curves became closer together. Subsequently, in the several years that we followed these people, the A1C levels between the two groups have remained more or less the same; no difference between the two groups--about 7.8%/7.9%.
In spite of the fact that the original intensive group's A1C levels have now gone up and are now no different than the conventional group, the amazing observation is that the difference in the risk of complications remains the same.

Looking at retinopathy as an endpoint over the course of 8 years after the end of the DCCT, those who were originally in the intensive group continued to have much less progression of retinopathy compared with those who were in the original conventional control. That just suggests that perhaps there is some kind of a metabolic memory of the benefit from good glycemic control that persists for a long time, even when the glycemic control subsequently may not stay as good as it was in the past.
Thus, this study has shown us the benefits of glycemic control early on with persistently long-term benefits.