SA JOURNAL OF DIABETES & VASCULAR DISEASE
REVIEW
VOLUME 7 NUMBER 3 • SEPTEMBER 2010
103
control as mitochondrial function or detailed vascular health are
not easily or routinely assessed – and indeed, if irreversible, not
particularly helpful for the individual. Perhaps carotid artery intimal
thickness or vascular reactivity and resistance should be measured
and followed to help identify good and poor risk patients.
The accumulation of AGEs, which are formed during periods
of hyperglycaemia and persist for many years, may be one of the
important factors in metabolic memory. AGEs are a heterogenous
group of chemical moieties occurring as a result of a nonenzymatic
reaction of glucose with proteins, lipids and nucleic acids. The role
of AGEs in the progression and complications of diabetes has been
reviewed recently.
3
AGEs act directly to induce cross-linking of
long-lived proteins such as collagen to promote vascular stiffness,
and, thus, alter vascular structure and function. AGEs can also
interact with certain receptors, to induce intracellular signalling
that leads to enhanced oxidative stress and elaboration of key
proinflammatory and prosclerotic cytokines. AGE modification of
mitochondrial proteins may be irreversible and may result in decline
of mitochondrial function with excess formation of reactive species.
4
Important support for the concept of the deleterious effects of the
accumulation of AGEs (glycated collagen and carboxymethyllysine)
has been obtained from skin biopsies in the DCCT and the EDIC
study
5
with the prediction of risk of progression of retinopathy and
nephropathy even after adjustments for mean HbA
1c
. Indeed the
predictive effect of HbA
1c
was abolished after adjustment for the
two AGEs measured.
Other theories on the mechanisms of metabolic memory have
been reviewed recently
6
and include the idea that overproduction
of oxidative stress persists after normalisation of glucose levels,
7
that there is long-lasting activation of epigenetic changes in the
promoter of a key inflammatory marker by transient spikes of
hyperglycaemia in mice
8
and that insulin, not only suppressing
glucotoxicity and lipotoxicity, has important anti-inflammatory
effects.
9
Clinical studies
The importance of good glycaemic control in the early stages of
diabetes with long-term benefits has been reviewed in this journal.
10
Examples in type 2 diabetes include UKPDS,
11
the Veterans Affair
Diabetes Trial,
12
and Steno-2 study.
13
In the UKPDS the benefit of
reduction in microvascular disease, largely laser therapy, seen in
the sulphonylurea and insulin group, but not in the metformin
group, was sustained in the 10-year post trial period despite rapid
convergence of HbA
1c
. It is also possible that the period of exposure
to hyperglycaemia before the diagnosis of type 2 diabetes may
contribute to hyperglycaemic memory.
14
Another important finding
in type 2 diabetes in UKPDS was the absence of a significant
legacy effect of blood pressure as the differences in end-points
disappeared after about 2 years of the blood pressures coming
together. In other words, blood pressure pathology is more related
to current than to past levels of blood pressure whereas glycaemic
pathology is more lasting.
Equivalent data on the importance of good glycaemic control
early in type 1 diabetes are limited and rely almost entirely on the
DCCT and the EDIC study. However, astute observations were first
made by diabetologists over 40 years ago
15
though their data were
limited by difficulties in documenting overall control of diabetes
before the era of HbA
1c
measurements. Based on glycosuria
percentage, the frequency of retinopathy in patients with initial
good control and later poor control resembled that of patients with
good control at all times (36– 38%) while in patients with initial poor
control and later good control it resembled that in patients with
poor control throughout (55–60%). good glycaemic control in the
first five years of diabetes resulted in protection from proliferative
retinopathy. One of the earlier, long-term, prospective studies using
HbA
1c
16
in type 1 diabetes started in 1979 and the baseline HbA
1c
measured in 1980 onwards was one of the predictors for the future
development of albuminuria (relative risk 1.18) in addition to blood
pressure, height and being male.
Epidemiology of Diabetes Intervention and
Complications study
The first DCCT follow up showed that the risk of retinopathy
remained significantly reduced in the intensive compared with the
conventional treatment groups in the first four years after the end of
the trial, despite similar HbA
1c
levels over this period.
17
Perhaps more
convincing are the results after 10 years EDIC follow-up in which
the HbA
1c
levels had converged completely and the follow-up was
longer. The former intensive treatment group still had lower rates
of progression of retinopathy and proliferative retinopathy (Table 1)
but the risk reductions at 10 years were attenuated compared with
the first 4 years of follow up.
18
Diabetic nephropathy was reported
for the 8 years follow up and showed in the former intensively
treated group a reduction of new microalbuminuria, a reduction in
clinical albuminuria, and fewer cases of hypertension, and patients
needing renal replacement therapy.
19
Diabetic neuropathy was also reported at 8 years follow-up with
a reduction in neuropathic symptoms and signs.
20
During the 17
years of follow up in DCCT and EDIC, intensive therapy reduced
the risk of any cardiovascular disease and the risk of non-fatal
myocardial infraction, stroke or death from cardiovascular disease.
21
A beneficial effect of former intensive diabetic control was noted
Table 1.
Epidemiology of Diabetes Intervention and Complications (EDIC)
study
Complication (years of follow-up)
% reduction in
former intensive
treatment group
Retinopathy (10 years EDIC)
progression
24
proliferative
59
Nephropathy (8 years EDIC)
new microalbuminuria
59
clinical albuminuria
84
Neuropathy (8 years EDIC)
symptoms
51
signs
43
Cardiovascular disease (17 years DCCT
+
EDIC)
any
42
non-fatal myocardial infarct, stroke or CVD death
57
CVD
=
cardiovascular disease; DCCT
=
Diabetes Control and Complications
Trial
