P O S I T I O N S T A T E M E N T

Implications of the United Kingdom

D
iabetes is a metabolic disorder pri- approached the normal range. Improved 9.4%) significantly reduced the microvas-
marily characterized by elevated glycemic control was also associated with cular complications of diabetes. The ex-
blood glucose levels and by micro- reduced cardiovascular events in the tent of the risk reduction in this Japanese
vascular and cardiovascular complications DCCT, but the difference was not statisti- study was similar to that in the DCCT,
that substantially increase the morbidity cally significant. Perhaps this was because thereby supporting the hypothesis that
and mortality associated with the disease the population studied was young adults glycemic control is important in both
and reduce the quality of life. Type 1 diabe- and therefore the event rate was very low. types of diabetes.
tes is characterized by total reliance on ex- Many of the observational studies also The third trial in type 2 diabetes was a
ogenous insulin for survival and comprises support a correlation between glycemic pilot study that randomized 153 men to
⬃10% of all cases of diabetes. The more control and diabetic complications in pa- intensive or conventional therapy (6). De-
prevalent form of diabetes, called type 2, tients with type 2 diabetes, but until now, spite a 2% absolute HbA1c difference in
comprising 90% of all people with diabetes, there have been only three randomized con- glycemic control between the two groups,
is characterized by insulin deficiency and/or trolled trials attempting to test the benefit of the trial reported no significant difference
insulin resistance. lowering blood glucose on the incidence of in cardiovascular events (when adjusted
An association between the complica- complications. The first of these studies was for baseline characteristics) in a follow-up
tions of diabetes and elevated blood glu- the University Group Diabetes Program period of only 27 months.
cose levels was postulated in the early part (UGDP), which showed no benefit of glyce- With this background, we now have
of this century. However, only in the last 3 mic control in new-onset type 2 diabetic pa- the results of the largest and longest study
decades has a substantial body of animal tients (4). However, in the UGDP, there on type 2 diabetic patients that has ever
experimental studies and human observa- were only 200 subjects in each treatment been performed (7–10). The United King-
tional studies and clinical trials directly group, HbA1c was not available as a reliable dom Prospective Diabetes Study (UKPDS)
linked hyperglycemia with the develop- method for measuring chronic glycemia, recruited 5,102 patients with newly diag-
ment of diabetic complications (1). Some and the difference in glucose control be- nosed type 2 diabetes in 23 centers within
of these studies have also demonstrated tween the most intensively treated group the U.K. between 1977 and 1991. Patients
that treatment that lowers blood glucose and the other treatment groups was only a were followed for an average of 10 years to
reduces the risks of diabetic retinopathy, fasting plasma glucose of ⬃30 mg/dl (1.7 determine 1) whether intensive use of phar-
nephropathy, and neuropathy. mmol/l). Of note, a major concern emanat- macological therapy to lower blood glucose
Notable are the results of the Diabetes ing from the UGDP was the observation that levels would result in clinical benefits (i.e.,
Control and Complications Trial (DCCT) the sulfonylurea agent (tolbutamide) and a reduced cardiovascular and microvascular
(2) and the similarly designed but smaller biguanide (phenformin) used to reduce hy- complications) and 2) whether the use of
Stockholm Diabetes Intervention Study (3). perglycemia were associated with increased various sulfonylurea drugs, the biguanide
These studies showed unequivocally in type cardiovascular mortality. The suspicion that drug metformin, or insulin have specific
1 diabetes that lowering blood glucose de- glucose lowering with oral agents in pa- therapeutic advantages or disadvantages. In
layed the onset and slowed the progression tients with type 2 diabetes may actually be addition, patients with type 2 diabetes who
of microvascular complications. Risk re- harmful has persisted since publication of were also hypertensive were randomized to
ductions for various outcomes ranged from the UGDP data in 1970. “tight” or “less tight” blood pressure control
35 to 75%. Secondary analyses in these The second controlled trial in type 2 to ascertain the benefits of lowering blood
studies showed strong relationships be- diabetes was only recently reported (5). pressure and to ascertain whether the use of
tween the risks of developing these compli- This small study conducted in 110 lean an ACE inhibitor (captopril) or ␤-blocker
cations and glycemic exposure over time. Japanese subjects showed that multiple (atenolol) offered particular therapeutic ad-
Moreover, there was no discernable glucose insulin injections resulting in better gly- vantages or disadvantages.
threshold, i.e., there was a continuous re- cemic control (HbA1c ⫽ 7.1%) compared
duction in complications as glycemic levels with conventional treatment (HbA1c ⫽ SUMMARY OF THE MAIN
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● RESULTS AND CONCLUSIONS
This statement was prepared by a work group consisting of Saul Genuth, MD (chair); Richard Eastman, MD; OF THE UKPDS
Richard Kahn, PhD; Ronald Klein, MD, MPH; John Lachin, DSc; Harold Lebovitz, MD; David Nathan, MD;
and Frank Vinicor, MD. The paper was approved by the Executive Committee of the Board of Directors of the ● The UKPDS results establish that reti-
American Diabetes Association, September 1998.
Abbreviations: DCCT, Diabetes Control and Complications Trial; FPG, fasting plasma glucose; UGDP, nopathy, nephropathy, and possibly
University Group Diabetes Program; UKPDS, United Kingdom Prospective Diabetes Study. neuropathy are benefited by lowering

S28 DIABETES CARE, VOLUME 25, SUPPLEMENT 1, JANUARY 2002

blood glucose levels in type 2 diabetes ● Epidemiological analysis showed a con- dized. Thus, combination therapy was
with intensive therapy, which achieved tinuous relationship between the risk of used, mixing insulin or metformin with
a median HbA1c of 7.0% compared all the above outcomes and systolic blood sulfonylureas, as well as crossing over pa-
with conventional therapy with a me- pressure. There was no evidence of a tients into the alternate pharmacological
dian HbA1c of 7.9%. The overall micro- threshold for these complications above a treatment groups. The final main “inten-
vascular complication rate was de- systolic blood pressure of 130 mmHg. tion to treat” comparison was between in-
creased by 25%. ● The results of the UKPDS and its impli- tensive therapy, which now included all
● These results materially increase the ev- cations for the treatment of type 2 dia- patients originally assigned to insulin and
idence that hyperglycemia causes, or is betes will now be discussed in the sulfonylurea drugs, and conventional
the major contributor, to these compli- format of pertinent questions and an- therapy, which included all patients orig-
cations. Epidemiological analysis of the swers. This discussion also reflects the inally randomized to diet treatment.
UKPDS data showed a continuous rela- position of the American Diabetes As- It must also be noted that the diet
tionship between the risks of microvas- sociation on the UKPDS findings. group could likewise not be kept pure.
cular complications and glycemia, such When patients in this group exceeded an
that for every percentage point decrease QUESTION 1: How was the FPG level of 15 mmol/l (270 mg/dl), they
in HbAlc (e.g., 9 to 8%), there was a 35% UKPDS conducted? were also treated with the same pharma-
reduction in the risk of complications. cological agents used in the other groups.
● The results demonstrate that the risks The design of the UKPDS was explicit and Ultimately, 80% of the patients in the diet
of complications can be significantly was directed by prespecified protocols group required one or more of the same
lowered even in the range of hypergly- with appropriate randomization of pa- pharmacological agents for which this
cemia where HbAlc levels are ⬍8.0%. tients. Laboratory and clinical tests were group was to serve as the control. Thus,
There was no evidence of any glycemic performed by accepted methodologies, all effects of each of the pharmacological
threshold for any of the microvascular and all end points were satisfactorily doc- treatment groups were confounded by
complications above normal glucose umented. The large number of patients crossovers, making it difficult to discern
levels (i.e., HbAlc ⬎6.2%). and their heterogeneity provide great as- specific drug effects. For example, of the
● These results confirm previous conclu- surance that the results should apply to total person-years of treatment in the con-
sions that lowering blood glucose the U.S. population of men and women trol group, 58% were spent on the original
would be beneficial based on observa- with type 2 diabetes. diet treatment only, 25% on a sulfonylurea
tional studies, pathological studies, and The UKPDS was originally designed alone or in combination with other drugs,
on three randomized clinical trials: the as a straightforward randomized clinical 12% on metformin, and 15% on insulin. In
DCCT, the Stockholm Diabetes Inter- trial comparing the effects of an “intensive the obese subgroup, metformin was admin-
vention Study, and the Japanese study. treatment policy” with four pharmacolog- istered alone or in combination with other
● No significant effect of lowering blood ical monotherapies, versus a diet control drugs for 82% of person-years to those as-
glucose on cardiovascular complica- group, on the cardiovascular and micro- signed to it and for 10% of person-years to
tions was observed. A 16% reduction vascular complications of type 2 diabetes. those in the obese control group.
(which was not statistically significant, The three main original monotherapies to On the one hand, these drug cross-
P ⫽ 0.052) in the risk of combined which all patients were randomized were overs make the ultimate differences in ef-
fatal or nonfatal myocardial infarction chlorpropamide, glyburide, and insulin. ficacy observed between the intensive and
and sudden death was observed. In the subgroup of overweight subjects, conventional groups all the more impres-
● Epidemiological analysis showed a metformin as monotherapy was com- sive. On the other hand, the prevalence of
continuous association between the pared with the control group and to the treatment crossovers and additions re-
risk of cardiovascular complications other three pharmacological agents. duces our confidence in the differences
and glycemia, such that for every per- The primary and major question of observed, or not detected, among the var-
centage point decrease in HbAlc (e.g., 9 the study was whether lowering blood ious pharmacological agents. Of note, the
to 8%), there was a 25% reduction in glucose was beneficial. Therefore, the main analysis performed by the UKPDS is
diabetes-related deaths, a 7% reduction treatment goal in all the intensive phar- identical to the analysis performed in the
in all-cause mortality, and an 18% re- macotherapy groups was a fasting plasma DCCT. In both studies, intensive treat-
duction in combined fatal and nonfatal glucose (FPG) level ⬍6.0 mmol/l (108 ment designed to achieve near normal gly-
myocardial infarction. Again, no glyce- mg/dl), and the treatment goal in the con- cemia with whatever means necessary was
mic threshold for these complications ventional diet control group was an FPG compared with conventional therapy.
above normal glucose levels was evident. level ⬍15 mmol/l (270 mg/dl). These
● The highest average annual incidence widely different treatment targets were QUESTION 2: What has the
of major hypoglycemic events was meant to insure attainment of adequate UKPDS contributed to our
2.3% of patients per year in those re- glycemic separation to test the main hy- understanding of the biology
ceiving insulin therapy. pothesis. However, it became apparent of diabetic complications?
● The study showed that lowering blood that none of the oral pharmacological
pressure to a mean of 144/82 mmHg monotherapies were capable of maintain- The UKPDS results confirm and extend
significantly reduced strokes, diabetes- ing the intensive treatment goal, and previous evidence supporting the hy-
related deaths, heart failure, microvas- therefore adequate glycemic separation pothesis that hyperglycemia and its
cular complications, and visual loss. from the control group might be jeopar- sequelae are a major cause of the micro-

DIABETES CARE, VOLUME 25, SUPPLEMENT 1, JANUARY 2002 S29

vascular complications of diabetes. The ously recommended by the American Di- served with either insulin or sulfonylurea
risk gradient in the UKPDS for late micro- abetes Association and others. drugs—this despite the fact that both
vascular events was very similar to that agents led to greater weight gain and
seen in the DCCT for early microvascular QUESTION 4: What are the higher plasma insulin levels than in the
events. This indicates that the presence of risks of aggressive glucose conventional group. This observation
hyperglycemia is a toxic state whether it or blood pressure control? gives some additional reassurance that in-
occurs early or late in life and irrespective sulin should not be held culpable for ath-
of its underlying cause. In the UKPDS, at the glycemic levels erosclerotic events and that sulfonylurea
The UKPDS also demonstrated by ep- achieved, a small fraction of patients had a drugs should not be held responsible for
idemiological analysis that cardiovascular major episode of hypoglycemia regardless lethal cardiovascular toxicity. In that re-
outcomes were consistently associated of the pharmacological therapy used, and gard, physicians who have been greatly
with hyperglycemia in a manner similar one patient died from hypoglycemia over concerned with the possibility of these se-
to the relationship between microvascular the 27,000 patient-years of intensive ther- rious adverse effects should feel consider-
complications and hyperglycemia. None- apy. Thus, the risk of hypoglycemia ably less constrained in using these agents
theless, the UKPDS did not prove defini- should not discourage attaining HbAlc for their valuable blood glucose-lowering
tively that intensive therapy that lowered levels approaching normal. benefits, such as saving vision.
blood glucose levels reduced the risk of As mentioned above, the UGDP study
cardiovascular complications compared raised concerns that treatment with sulfo- QUESTION 6: What were the
with conventional therapy. Thus, the role nylurea agents may increase cardiovascu- role and results of metformin
of hyperglycemia in cardiovascular com- lar events or death. However, the UKPDS therapy in the UKPDS?
plications is still unclear. found no increase in the rates of myocar-
dial infarction or diabetes-related deaths This topic requires individual discussion
QUESTION 3: What level of when participants treated intensively because of the different ways in which
glucose or blood pressure with sulfonylurea drugs were compared metformin was used and the conflicting
control do the UKPDS results with those treated conventionally. results observed in each instance. The ini-
suggest should be achieved Some but not all previous observa- tial UKPDS design included assignment
in patients with type 2 tional studies have suggested that plasma of obese patients to metformin as well as
diabetes? insulin levels may be associated with in- to the other intensive therapies and con-
creased cardiovascular disease risk. How- ventional therapy. Patients initially as-
The median HbAlc levels achieved in both ever, the UKPDS showed no increase in signed to intensive therapy with
the conventionally treated group (7.9%) cardiovascular events or mortality in pa- metformin had decreased risks of com-
and the intensively treated group (7.0%) tients assigned insulin therapy, even bined diabetes-related end points, diabe-
are lower than the present average HbAlc though their fasting plasma insulin levels tes-related deaths, all-cause deaths, and
in type 2 diabetic patients in the U.S. were higher than those of the convention- myocardial infarction compared with the
(⬃8.5–9.0). These differences most likely ally treated patients. Thus, the beneficial conventionally treated patients. These
reflect the fact that patients were enrolled in effects of intensive glucose control with risks were significantly reduced by about
the UKPDS on diagnosis, whereas the HbAlc insulin or sulfonylurea agents outweigh one-third (P ⬍ 0.0023–0.017). The
levels usually achieved in the general popu- their purported risks. beneficial effect on cardiovascular disease
lation reflect diabetes of mixed durations. is in contrast to the failure of insulin or
A significant reduction in complica- QUESTION 5: What sulfonylurea treatment to reduce cardio-
tions was achieved with intensive therapy differences were observed vascular outcomes when compared with
that lowered HbAlc levels to a median of between the various forms conventionally treated obese patients.
7.0% over 10 years when compared with of intensive therapy? This difference between drugs may possi-
conventional treatment that achieved a bly relate to the absence of weight gain with
median HbAlc of 7.9%. Furthermore, on In the main trial, there were no significant metformin and/or to some beneficial effects
epidemiological analysis, there was no ev- differences with regard to diabetic com- of metformin on the insulin resistance syn-
idence of any glycemic threshold above a plications or adverse cardiovascular drome. Surprisingly, no significant decrease
normal HbAlc level of 6.2%. Therefore, events between therapy with insulin and in microvascular complications was ob-
the results of the UKPDS mandate that with sulfonylurea drugs. The strength of served with intensive metformin therapy or
treatment of type 2 diabetes include ag- this conclusion is somewhat attenuated with combined insulin/sulfonylurea inten-
gressive efforts to lower blood glucose by the frequency with which drug cross- sive therapy in the obese subjects. This in-
levels as close to normal as possible. overs and additions occurred for the pur- consistency, along with drug crossovers
The results of the UKPDS blood pres- pose of achieving blood glucose targets. between the treatment groups and the lesser
sure study also indicate that aggressive Patients randomized to sulfonylurea numbers of patients in the subgroups, cre-
treatment of even mild-to-moderate hy- drugs remained on them ⬃80% of the ates uncertainty regarding the overall bene-
pertension is beneficial. Moreover, con- time (with a fraction treated with alternate ficial effect of metformin on obese patients.
tinued reduction of blood pressure into therapies), and patients randomized to in- Late in the study, 537 obese and nor-
the normal range resulted in fewer com- sulin remained on it ⬃75% of the time. It mal-weight patients originally assigned to
plications. Hence, blood pressure should is again worth noting that no increase in sulfonylurea therapy who failed to main-
be kept below 130/85 mmHg, as previ- cardiovascular events or death was ob- tain blood glucose in the designated target

S30 DIABETES CARE, VOLUME 25, SUPPLEMENT 1, JANUARY 2002

range were randomly assigned to con- The UKPDS also compared antihy- But since patients in the UKPDS entered
tinue sulfonylurea therapy alone or to pertensive treatment with an ACE inhibi- the study at an age equivalent to the aver-
have metformin added. In this substudy, tor to that with a ␤-blocker. Both drugs age age of diagnosis in the U.S., vigorous
an intention-to-treat analysis showed that were about equally effective in lowering treatment to prevent the complications of
the group assigned to combined met- blood pressure, although patients on diabetes from the time of diagnosis is
formin/sulfonylurea therapy had a 96% ␤-blockers had slightly better blood pres- clearly warranted given the UKPDS re-
increase in diabetes-related deaths (P ⬍ sure control (a 1-mmHg systolic and sults. In addition, because the UKPDS
0.039) and a 60% increase in all-cause 2-mmHg diastolic improvement). Nei- confirmed that a substantial proportion of
death (P ⬍ 0.041), compared with the ther drug was superior to the other in any newly diagnosed patients (⬃50%) al-
patients assigned to continue maximal outcome measured, including diabetes- ready have some early evidence of dia-
doses of sulfonylurea drugs alone. How- related deaths, myocardial infarction, and betic complications, such disturbing
ever, to keep the FPG level ⬍15 mmol/l all microvascular end points. Also, there findings should provide an even greater
(270 mg/dl) and the patients asymptom- were no significant differences in mi- impetus to aggressive intervention.
atic, metformin was given to 25% of the croalbuminuria or proteinuria. However, Finally, it is important to note that
patients initially randomized to receive because of the low prevalence of ne- kidney failure is much more common in
sulfonylurea drugs alone. Additionally, phropathy in the population studied, it is the U.S. than in the U.K., because the U.S.
the lack of a placebo control and the in- unclear whether there were sufficient has larger minority populations who de-
ability to employ masking in this sub- events to observe a protective effect of ei- velop type 2 diabetes earlier in life and
study also call these detrimental effects ther drug on the progression of nephrop- have worse average blood glucose con-
into question. athy. We conclude that both drugs used trol. There may also be a genetic suscep-
In an attempt to reconcile the benefi- to reduce hypertension are equally effec- tibility to kidney disease in some ethnic
cial effects observed when metformin was tive and safe, and either can be used with populations in the U.S. Because the UKPDS
used in obese subjects, with the detrimen- great benefit to treat uncomplicated hy- reported significant reductions in the risk of
tal effects observed when metformin was pertension in patients with type 2 diabetes. developing kidney disease and in the risk of
added to maximum sulfonylurea therapy, Of note, conventionally and intensively its progression, the results are particularly
the results of the two substudies were treated blood glucose study patients had important for the prevention of diabetic re-
combined in a meta-analysis. We believe equal benefit from blood pressure lowering. nal disease in the U.S. population.
this analytical approach does not resolve Likewise, the tightly and less tightly con-
the discrepancy in the results with met- trolled blood pressure study patients had QUESTION 9: Are the results
formin, that is, it does not provide assur- equal benefit from blood glucose lowering. of the UKPDS achievable for
ance that the combination is safe or prove Thus, both hyperglycemia and hyperten- most people with diabetes?
that it is unsafe. sion should be vigorously treated when they
If there is some specific mechanism of occur together with an expectation that re- In theory and practice the answer is yes.
adverse interaction between metformin ductions in microvascular and cardiovascu- Although the UKPDS was a clinical re-
and sulfonylurea drugs, this can only be lar outcomes will be additive. search trial, subjects were enrolled and
definitely determined in a new, appropri- managed in a wide variety of community
ately designed, randomized, placebo- QUESTION 8: Is tight control clinics. The professionals conducting the
controlled trial. Until such a trial is contraindicated in any group study and directing care were very knowl-
concluded, we do not recommend any of type 2 diabetic patients? edgeable about diabetes, but the overall
change in the current guidelines for the treatment program was not unusually so-
use of metformin as monotherapy or in Patients with type 2 diabetes are usually phisticated or complex. Patients enrolled
combination with sulfonylurea drugs. diagnosed at an age when the likelihood in the UKPDS started with an HbAlc level
of having comorbid conditions increases. of 9.1%. The conventional treatment
QUESTION 7: What were the Competing comorbidities such as hyper- group achieved a 10-year median HbAlc
effects of blood pressure tension and dyslipidemia can lead pa- level of 7.9%, and the group treated in-
control? tients and physicians toward emphasizing tensively with readily available glucose-
the treatment of one problem over an- lowering agents achieved a median HbAlc
“Tight blood pressure control,” as other. In most cases, however, this does level of 7.0%. Perhaps the most important
achieved in the UKPDS, significantly re- not have to occur, and the results of the ingredient leading to therapeutic success
duced the risks of virtually all cardiovas- UKPDS at least support the premise that was persistence.
cular and microvascular outcomes, with equal attention to diabetes and hyperten- It should be acknowledged that there
risk reductions ranging from 24 to 56%. A sion can be accomplished. is an increase in blood glucose levels with
21% reduction seen in myocardial infarc- Older type 2 diabetic patients have a increasing duration of type 2 diabe-
tion was not significant (P ⫽ 0.13). The shorter life expectancy from time of dia- tes—an upward trend noted in both the
type and number of adverse effects seen betes diagnosis by virtue of their age and conventional and intensive treatment
with “tight blood pressure control” (mean their risk or presence of cardiovascular groups of the UKPDS. The ability to pre-
144/82 mmHg) or “less tight blood pres- disease. Thus, it has been argued that the vent or at least retard this rise may be fa-
sure control” (mean 154/87 mmHg) were benefits of glycemic control that are real- cilitated by recently approved glucose-
not different from those generally re- ized over time may be preempted by ear- lowering drugs that were not available to
ported in the literature. lier adverse outcomes of other diseases. the UKPDS. Both the success and the dif-

DIABETES CARE, VOLUME 25, SUPPLEMENT 1, JANUARY 2002 S31

ficulties of the UKPDS should stimulate Other studies would also be desirable to Research Group: The effect of intensive
further development of new treatments establish the risks and benefits of glucose treatment of diabetes on the development
and combinations that can successfully control in patients who already have car- and progression of long-term complica-
manage hyperglycemia in type 2 diabetes. diovascular disease. Such risks and bene- tions in insulin-dependent diabetes mel-
litus. N Engl J Med 329:977–986, 1993
fits may well be different from those
3. Reichard P, Nilsson BY, Rosenqvist V: The
QUESTION 10: Are there characteristic of earlier stages in the natu- effect of long-term intensified insulin
other major unanswered ral history of diabetes. treatment on the development of micro-
questions in the treatment vascular complications of diabetes melli-
of diabetes? CONCLUSIONS — The UKPDS has tus. N Engl J Med 329:304 –309, 1993
provided strong support for the American 4. University Group Diabetes Program: A
The UKPDS and the DCCT have an- Diabetes Association’s position that vigor- study of the effects of hypoglycemic
ous treatment of diabetes can decrease the agents on vascular complications in pa-
swered the question of whether blood
morbidity and mortality of the disease by tients with adult-onset diabetes. Diabetes
glucose control is beneficial for people 19 (Suppl. 2):747– 830, 1970
with type 1 and type 2 diabetes. It defi- decreasing its chronic complications. The
results show that lowering blood glucose 5. Ohkubo Y, Kishikawa H, Araki E, Miyata
nitely is. However, both trials enrolled pa- T, Isami S, Motoyoshi S, Kojima Y, Fu-
tients before serious microvascular reduces the incidence of microvascular
ruyoshi N, Shichiri M: Intensive insulin
complications had developed. complications in type 2 diabetes as it does therapy prevents the progression of dia-
The benefits of achieving normal in type 1 diabetes. In addition, lowering betic microvascular complications in Jap-
blood glucose levels are not known in blood pressure reduces the incidence of anese patients with non-insulin-dependent
those who already have more advanced cardiovascular complications as it does in diabetes mellitus: a randomized prospective
complications. Also, the risk of hypogly- nondiabetic individuals, and also leads to 6-year study. Diabetes Res Clin Pract
cemia may be greater in such patients. further reduction in the severity of micro- 28:103–117, 1995
vascular complications. Although the 6. Abraira C, Colwell J, Nuttall F, Sawin CT,
Thus, additional studies would be desir- Henderson W, Comstock JP, Emanuele
able on the risks and benefits in people UKPDS did not directly establish any ef-
fect of lowering blood glucose on cardio- NV, Levin SR, Pacold I, Lee HS: Cardio-
with diabetes who have more advanced vascular events and correlates in the Vet-
complications. vascular complications, the use of insulin,
erans Affairs Diabetes Feasibility Trial:
Neither study gave a definitive answer sulfonylureas, or metformin (and perhaps
Veterans Affairs Cooperative Study on
to the question of whether glucose control metformin in combination with sulfonyl- Glycemic Control and Complications in
reduces the risk of cardiovascular disease. ureas) does not appear to increase the risk Type II Diabetes. Arch Intern Med 157:
Both studies showed trends in reducing of cardiovascular events. Therefore, noth- 181–188, 1997
the risk of cardiovascular events, but ing should stop practitioners from pursu- 7. UK Prospective Diabetes Study Group:
these trends were not statistically signifi- ing the American Diabetes Association’s Intensive blood-glucose control with sul-
cant. Several observational studies, in- goals for glycemia and blood pressure. phonylureas or insulin compared with
The UKPDS is another landmark diabetes conventional treatment and risk of com-
cluding the results of the epidemiologic plications in patients with type 2 diabetes
analyses of UKPDS data, have shown study proving the value of metabolic con-
trol. It is time for all health professionals (UKPDS 33). Lancet 352:837– 853, 1998
strong associations between blood glu- 8. UK Prospective Diabetes Study Group:
cose control and the risk of cardiovascular to treat diabetes aggressively. It is also
Effect of intensive blood-glucose control
disease morbidity (heart attacks, strokes, time for patients to take their diabetes
with metformin on complications in over-
and amputations) and all-cause mortality. with utmost seriousness. And it is incum- weight patients with type 2 diabetes (UK-
However, these studies do not prove that bent upon the health care system to pro- PDS 34). Lancet 352:854 – 865, 1998
high blood glucose causes these compli- vide the necessary resources for both to be 9. UK Prospective Diabetes Study Group:
cations and that intensive treatment to successful. Compromise or acceptance of Tight blood pressure control and risk of
lower glucose would reduce the risk. a disadvantageous and dangerous status macrovascular and microvascular com-
Fortunately, the patients in both the quo in people with diabetes should not be plications in type 2 diabetes (UKPDS 38).
tolerated any longer. BMJ 317:703–713, 1998
DCCT and the UKPDS are enrolled in 10. UK Prospective Diabetes Study Group:
longterm follow-up studies that will de- Efficacy of atenolol and captopril in re-
termine whether the initial period of in- References ducing risk of both macrovascular and
tensive treatment prevents cardiovascular 1. Genuth S: A case for blood glucose con- microvascular complications in type 2 di-
complications by comparison with the trol. Adv Int Med 40:573– 623, 1995 abetes (UKPDS 39). BMJ 317:713–720,
initial period of conventional treatment. 2. Diabetes Control and Complications Trial 1998

S32 DIABETES CARE, VOLUME 25, SUPPLEMENT 1, JANUARY 2002