Long-Term Results of the Kumamoto

Study on Optimal Diabetes Control in

Type 2 Diabetic Patients

Motoaki Shichiri, MD, PHD

Hideki Kishikawa, MD, PHD
Yasuo Ohkubo, MD, PHD
Nakayasu Wake, MD

OBJECTIVE To examine whether intensive glycemic control could decrease the

frequency or severity of diabetic microvascular complications, an 8-year prospective study
of Japanese patients with type 2 diabetes was performed.

RESEARCH DESIGN AND METHODS A total of 110 patients with type 2 diabetes
(55 with no retinopathy [the primary prevention cohort] and 55 with simple retinopathy [the
secondary intervention cohort]) were randomly assigned to multiple insulin injection
therapy (MIT) groups and administered three or more daily insulin injections or assigned to
conventional insulin injection therapy (CIT) groups and administered one or two daily
intermediate-acting insulin injections. Worsening of microvascular complications was
regularly assessed during 8 years. Two or more steps up in the 19 stages of the modified
Early Treatment of Diabetic Retinopathy Study classification in retinopathy and one or
more stages up among three stages in nephropathy (normoalbuminuria, microalbuminuria,
and albuminuria) were defined as worsening of complications.

RESULTS In both primary prevention and secondary intervention cohorts, the

cumulative percentages of worsening in retinopathy and nephropathy were significantly
lower (P < 0.05) in the MIT group than in the CIT group. In neurological tests after 8 years,
the MIT group showed significant improvement (P < 0.05) in the median nerve conduction
velocities (motor and sensory nerves), whereas the CIT group showed significant
deterioration (P < 0.05) in the nerve conduction velocities and vibration threshold. From
this study, the glycemic threshold to prevent the onset and progression of diabetic
microvascular complications was as follows: HbA1c <6.5%, fasting blood glucose
concentration <110 mg/dl, and 2-h postprandial blood glucose concentration <180 mg/dl.

CONCLUSIONS Intensive glycemic control can delay the onset and progression of the
early stages of diabetic microvascular complications in Japanese patients with type 2
diabetes.

Diabetes Care 23 (Suppl. 2):B21B29, 2000

Diabetic microvascular complications are now major problems for diabetic patients. The
Diabetes Control and Complications Trial (DCCT) and European studies have
demonstrated the efficacy of intensive insulin treatment in preventing the onset and/or
progression of microvascular complications such as retinopathy, nephropathy, and
neuropathy in patients with type 1 diabetes (15).

The Kumamoto Study was a randomized clinical trial designed to compare intensive insulin
therapy using multiple insulin injections with conventional insulin injection therapy to
evaluate the effects on the development and progression of microvascular complications in
Japanese patients with type 2 diabetes. Two cohorts were designed to answer two different,
but related, questions: will intensive insulin therapy prevent the onset of diabetic
microvascular complications in type 2 diabetic patients with no retinopathy and a urinary
albumin excretion level of <30 mg/24 h (primary prevention cohort), and will intensive
insulin therapy reduce the progression of microvascular complications in type 2 diabetic
patients with simple retinopathy and a urinary albumin excretion level of <300 mg/24 h
(secondary intervention cohort)? In a previous 6-year study, we demonstrated in Japanese
patients with type 2 diabetes that establishing good glycemic control could help reduce the
development and progression of diabetic retinopathy, nephropathy, and neuropathy (6).

In this article, we present the results of the 8-year Kumamoto Study in Japanese patients
with type 2 diabetes.

RESEARCH DESIGN AND METHODS

Patients
Patients with type 2 diabetes who had been treated with once- or twice-daily injections of
intermediate-acting insulin in an outpatient clinic were selected on the basis of the
following criteria: had no retinopathy or simple retinopathy determined by clinical
funduscopic evaluation, had a urinary albumin excretion rate <300 mg/24 h and a serum
creatinine level <1.5 mg/dl, had no diabetic somatic or had autonomic neuropathy severe
enough to require treatment, was under 70 years of age, and was otherwise healthy. All
patients were diagnosed as having type 2 diabetes if they had no history of ketoacidosis,
had negative islet cell antibodies, and had a 24-h urinary C-peptide excretion rate of >20
g. Informed consent was obtained from each patient.

Study design
There were 110 patients divided into two cohorts: the primary prevention cohort (no
retinopathy and urinary albumin excretion rate <30 mg/24 h, n = 55) and the secondary
intervention cohort (simple retinopathy and urinary albumin excretion rate <300 mg/24 h, n
= 55). These patients were randomly assigned to either a conventional insulin injection
therapy (CIT) group (n = 55) or a multiple insulin injection therapy (MIT) group (n = 55).
The evaluation on the fundus photography was masked to the randomization status in both
primary prevention and secondary intervention cohorts. The CIT group was administered
once- or twice-daily injections of intermediate-acting insulin, whereas the MIT group was
administered short-acting insulin at each meal and intermediate-acting insulin at bedtime.
The clinical goal of the CIT group was to show no symptoms of hyperglycemia or
hypoglycemia and to have glycemic control, which meant having a fasting blood glucose
(FBG) level close to <140 mg/dl.

The goal of the MIT group was to maintain glycemic control, which meant having an FBG
level close to <140 mg/dl, a 2-h postprandial blood glucose concentration <200 mg/dl, an
HbA1c level <7.0%, and a mean amplitude of glycemic excursions (MAGE) <100 mg/dl
(7). Daily blood glucose profiles were measured during 1 day of hospitalization, and
frequent blood glucose measurements were also taken by the patients. Based on these
results, insulin doses were corrected to improve MAGE. To correct fasting blood glucose
levels, the doses of intermediate-acting insulin at bedtime were adjusted, whereas to correct
postprandial blood glucose levels, the doses of short-acting insulin before each meal were
adjusted. Adjustment doses of insulin were usually 24 U at each point. There was no
significant difference at entry in either the primary prevention cohort or the secondary
intervention cohort between CIT and MIT groups with regard to sex distribution, age,
duration of diabetes, BMI, insulin doses, urinary C-peptide excretion, HbA1c, blood
pressure, serum cholesterol, serum triglyceride, serum HDL cholesterol, degree of
retinopathy, urinary albumin excretion, and nerve conduction velocity (Table 1). After 8
years, 99 patients remained in the study, 5 patients died (2 in the MIT group and 3 in the
CIT group), 4 patients had moved to another city (2 in the MIT group and 2 in the CIT
group), and 2 patients had changed from conventional insulin injection therapy to multiple
insulin injection therapy. Follow-up examinations of glycemic control and diabetic
microvascular complications were performed at the third month and every 6 months after
the initiation of the study.
Evaluation of glycemic control
Metabolic control was estimated by HbA1c, FBG, mean blood glucose (MBG), M value of
Schlichtkrull, and MAGE calculated from diurnal blood glucose profiles (8). The mean
values for HbA1c, FBG, MBG, M, and MAGE during the study were calculated to evaluate
the glycemic control during a period from the third month to 8 years. Severe hypoglycemia
was defined as an event with symptoms consistent with hypoglycemia in which the patient
required the assistance of another person and which was associated with a blood glucose
level <50 mg/dl and a prompt recovery after intravenous glucose loading. Mild
hypoglycemia was defined as an event with symptoms consistent with hypoglycemia
(sweating, palpitations, hunger, or blurred vision) in which the patient did not require the
assistance of another person and which was associated with a blood glucose level <50
mg/dl by self-monitoring.

Evaluation of diabetic microvascular complications

Retinopathy. All of the patients had direct ophthalmoscopy every 6 months, with their
pupils dilated. Funduscopic findings were evaluated by at least two examiners (one by an
ophthalmologist and one by an internist) and were followed by color fundus photography
and fluorescein angiography. The degree of retinopathy for each patient was determined by
the two eye examiners using the modified Early Treatment of Diabetic Retinopathy Study
(ETDRS) classification with a scale of 19 stages (6,911). The development and
progression of retinopathy were defined as a change of at least two steps up from stage 1 in
the primary prevention cohort and as a change of two or more steps up from stages 25 in
the secondary intervention cohort. Once changes were found, the funduscopic examinations
were performed every month, and the worsening of retinopathy was confirmed when the
changes were observed in two consecutive examinations.

Nephropathy. The patients with nephropathy were divided into three stages depending on
their urinary albumin excretion: normoalbuminuria (<30 mg/24 h), microalbuminuria (30
300 mg/24 h), or albuminuria (>300 mg/24 h) (3,12). A change from any one stage to any
higher stage was defined as progression.

Neuropathy. Peripheral nerve functions were evaluated by median nerve conduction

velocity and by vibration threshold on the radial styloid process of the arm and the medial
malleolus of the leg on both sides. The coefficient of variation (CV) of R-R intervals on an
electrocardiogram and orthostatic hypotension were used to evaluate the autonomic nerve
functions. The CV of R-R intervals was evaluated during both bed rest and deep breathing.
In orthostatic hypotension, the decrease in systolic blood pressure caused by standing after
5 min of bed rest was measured. These neurological indexes at entry and after 8 years of the
study in both CIT and MIT groups were compared in the combined cohort consisting of
primary prevention and secondary intervention cohorts.

Comparison of the reduction in risk with intensive insulin therapy on the development
and progression of microvascular complications between the Kumamoto Study and
the DCCT
To compare the effects of intensive insulin therapy on the development and progression of
diabetic microvascular complications between type 2 and type 1 diabetes, the reduction in
risk for the development and progression of diabetic microvascular complications with
intensive insulin therapy in the Kumamoto Study was compared with those in the DCCT.

Relationship between the worsening of microvascular complications and glycemic

control indexes
The relationship between the worsening of retinopathy and nephropathy and glycemic
control indexes, expressed as HbA1c, FBG, and 2-h postprandial blood glucose
concentration, was evaluated in the combined cohort. The crude rate within the six grades
evaluated according to their glycemic control indexes was calculated and expressed as 100
patient-years. The regression lines estimated as the function of the log of mean glycemic
control indexes were calculated.

Evaluation of macrovascular complications

To evaluate the effect of intensive glycemic control on macrovascular complications in
patients with type 2 diabetes in the Kumamoto Study, the events of cardiovascular,
cerebrovascular, and peripheral vascular disease during 8 years in the MIT group were
compared with those in the CIT group. Cardiovascular events (angina pectoris or
myocardial infarction), cerebrovascular events (stroke), and peripheral vascular events
(intermittent claudication, gangrene, or amputation) were confirmed by coronary
angiography, electrocardiogram, brain computed tomography, or peripheral vascular
angiography.
Assays
All of the assays were standardized before initiation of the follow-up. Blood glucose
concentration was assayed using the glucose oxidaseperoxidase, 4-aminophenozone,
phenol method (Autoanalyzer; Miles, Tarrytown, NY). HbA1c value was assayed using
high-performance liquid chromatography (HLC 723 GHb type 2; Tosoh, Tokyo) (normal
range 4.86.4%). HbA1c results obtained from 1987 to 1994 were corrected according to
the recommendation of the Committee on an Interlaboratory Standardization of HbA 1c
Determination of the Japan Diabetes Society (13). Funduscopic findings were recorded
using the CR-45 camera (Canon, Tokyo). Urinary albumin excretion rate in a 24-h
specimen was measured by turbidimetric immunoassay (Automatic Analyzer 7150; Hitachi,
Tokyo). The mean of the values measured on two or three independent days within 4 weeks
were used (CV of the day-to-day variation: 30%). Nerve conduction velocity, vibration
threshold, and CV of R-R intervals were measured by the Neuromatic 2000 (Dantec,
Skovlunde, Denmark), the Vibratory Sensation Meter (TM-31A; Medic International,
Tokyo), and the Electro Cardiograph Analysis System 12 (Nihon Kohden, Tokyo),
respectively.

Statistical analysis
Wilcoxon's rank-sum test was used to compare the two treatment groups within the cohort
with regard to ordinal and numerical variables, and the 2 test was used for comparison of
categorical variables. The life-table method was used to estimate the cumulative incidence
of events, with adjustment for periodically timed assessments. The difference between two
cumulative incidence curves was tested by the Mantel (log-rank) test (14).

The relative risks of retinopathy and nephropathy were estimated as the crude relative risk
in the primary prevention cohort, the secondary intervention cohort, and the combined
cohort. The crude relative risk was calculated according to the following equation: 1
P0/P1, where P0 and P1 represent the cumulative percentage of worsening in CIT and MIT
groups, respectively.

In addition, the relative risks were also calculated using proportional-hazards analysis only
in the combined cohort. Event rates of complications were presented as the number of
events per 100 patient-years based on the ratio of observed number of events to the total
number of patient-years of exposure. All results were expressed as mean SD. P < 0.05
was considered to be statistically significant. All analyses were undertaken by intention to
treat.

RESULTS

Glycemic control
As shown in Fig. 1, near-normoglycemia was obtained in the MIT group during the third
month after the initiation of intensive insulin therapy and was maintained over the 8 years
of the study, whereas the glycemic control in the CIT group did not change significantly.
The resultant mean values of FBG, HbA1c, MBG, M, and MAGE over the 8 years were
significantly lower (P < 0.001) in the MIT group than in the CIT group (FBG 122 30 vs.
162 30 mg/dl, HbA1c 7.2 1.0 vs. 9.4 1.3%, MBG 155 34 vs. 221 40 mg/dl, M
value 13 9 vs. 44 17, MAGE 90 26 vs. 136 35 mg/dl). On the other hand, there was
a slight but not significant decrease in insulin doses from the baseline to 8 years in the MIT
group (MIT group 0.40 0.28 to 0.38 0.39 U/kg/24 h, CIT group 0.41 0.29 to 0.43
0.33 U/kg/24 h).

Figure 1Glycemic control as assessed by

FBG and HbA1c over an 8-year period in
patients with type 2 diabetes receiving intensive
( ) and conventional ( ) insulin injection
therapy. Mean values SD are shown. *P <
0.05 vs. conventional insulin injection therapy.

Retinopathy
Primary prevention cohort. The cumulative percentage of patients who developed
retinopathy after 8 years in the MIT group was significantly lower when compared with
that in the CIT group (15.4 vs. 47.8%; 1.9 vs. 6.0 events/100 patient-years, P = 0.022, Fig.
2A). During the 8-year period, no patient developed preproliferative or proliferative
retinopathy.
 Figure 2Cumulative incidences of a change in
retinopathy in patients with type 2 diabetes treated
by intensive ( ) and conventional ( ) insulin
injection therapy. A change in the severity of
retinopathy was defined as at least two steps up
from the baseline (ETDRS interim scale). The
primary prevention (A) and secondary intervention
(B) cohorts are shown. P values were tested using
the Mantel (log-rank) test.

Secondary intervention cohort. The cumulative percentage of patients who progressed to

retinopathy was 24.0% in the MIT group and 56.0% in the CIT group after 8 years (3.0 vs.
7.0 events/100 patient-years, P = 0.023, Fig. 2B). During the 8-year period, the progression
to preproliferative and proliferative retinopathy was lower in the MIT group than in the CIT
group (1.5 vs. 3.0 events/100 patient-years).

Nephropathy
Primary prevention cohort. The cumulative percentage of patients who developed
nephropathy after 8 years was significantly lower in the MIT group than in the CIT group
(11.5 vs. 43.5%; 1.4 vs. 5.4 events/100 patient-years, P = 0.029, Fig. 3A). During the 8-year
period, the progression to albuminuria was found only in the CIT group (1.1 and 0
events/100 patient-years for the CIT and MIT groups, respectively).
 Figure 3Cumulative incidences of change in
nephropathy in patients with type 2 diabetes
treated by intensive ( ) and conventional ( )
insulin injection therapy. A change in the severity
of nephropathy was defined as one or more stages
up among three stages (normoalbuminuria,
microalbuminuria, and albuminuria). The
primary prevention (A) and secondary
intervention (B) cohorts are shown. P values were
tested using the Mantel (log-rank) test.

Secondary intervention cohort. The cumulative percentage of patients who progressed to

nephropathy was significantly lower in the MIT group than in the CIT group (16.0 vs.
40.0%; 2.0 vs. 5.0 events/100 patient-years, P = 0.043), as shown in Fig. 3B. During the 8-
year period, the progression to albuminuria was found in the CIT group (2.0 and 0
events/100 patient-years for the CIT and MIT groups, respectively).

Neuropathy
Peripheral nerve functions were evaluated in the combined cohort at entry and after 8 years.
The median nerve conduction velocities (both motor and sensory nerves) significantly
increased after 8 years in the MIT group as compared with the values at entry, while the
median sensory nerve conduction velocities significantly decreased in the CIT group (Fig.
4A). There were significant differences in the motor and sensory nerve conduction
velocities between the two groups after 8 years of study (P < 0.05). The vibration
thresholds in the MIT group showed slight, but not significant, increases after 8 years,
while those in the CIT group significantly increased after 8 years (P < 0.05). There were
significant differences (P < 0.05) in vibration thresholds between the MIT and CIT groups
after 8 years of study. The MIT group showed only slight improvements in the orthostatic
hypotension and CV of R-R interval, whereas the CIT group showed slight deterioration in
both tests (Fig. 4B).

Figure 4Peripheral nerve functions and autonomic nerve

functions in the CIT and MIT groups at entry of the study
and after 8 years in the Kumamoto Study. Peripheral nerve
functions (A) and autonomic functions (B) are shown. In the
CIT group, the values at entry ( ) and after 8 years ( )
are indicated. In the MIT group, the values at entry ( ) and
after 8 years ( ) are indicated. Mean values SD are
shown. *P < 0.05 vs. CIT group; #P < 0.05 vs. at entry.

Development and progression of microvascular complications and reduction in risk

with intensive insulin therapy
Table 2 demonstrates the reduction in risk for the development and progression of diabetic
microvascular complications in the Kumamoto Study for type 2 diabetes. With intensive
insulin therapy, the risks of retinopathy, calculated by crude relative risk, were reduced by
68 and 57% in the primary prevention and secondary intervention cohorts, respectively.
The risk calculated by proportional-hazards analysis in the combined cohort was the same
as that in the DCCT (63%). In addition, intensive insulin therapy also reduced the risk of
microalbuminuria by 74% and albuminuria by 100% in the primary prevention cohort. It
also reduced the risk of microalbuminuria by 60% and albuminuria by 100% in the
secondary prevention cohort, respectively. The risk reduction in the combined cohort
calculated by proportional-hazards risk analysis was greater in the Kumamoto Study with
type 2 diabetes than in the DCCT with type 1 diabetes (74 vs. 3954%, respectively).

Relationship between the worsening in microvascular complications and glycemic

control indexes
By analyzing the relationship between the rate of worsening of retinopathy and
nephropathy and glycemic indexes, a continuously increasing risk of the worsening of
retinopathy and nephropathy was observed with increasing HbA1c, FBG, and 2-h
postprandial blood glucose concentration (Fig. 5). No worsening of retinopathy or
nephropathy was observed in patients whose HbA1c, FBG, and 2-h postprandial blood
glucose concentration were below 6.5%, 110 mg/dl, and 180 mg/dl, respectively.
 Figure 5Relationship between the rate of worsening of retinopathy and
nephropathy and glycemic control indexes expressed as HbA1c, FBG, and 2-h
postprandial blood glucose concentration in the combined cohort. , The crude
rate within the six grades evaluated according to their glycemic control
levels; each square corresponds to more than 100 patient-years. , Regression
lines estimated as a function of the log of the mean HbA1c, FBG, and 2-h
postprandial glucose concentration.

Macrovascular complications
The total events of cardiovascular, cerebrovascular, and peripheral vascular diseases in the
CIT group occurred twice as much as those in the MIT group (1.3 vs. 0.6 events/100
patient-years). In the MIT group, during the 8-year follow-up period, one patient died
suddenly (probably because of myocardial infarction), two patients developed angina
pectoris, and one other patient complained of intermittent claudication. On the other hand,
in the CIT group, three patients died of cerebral vascular disease, two patients developed
angina pectoris, and two patients complained of intermittent claudication.

Hypoglycemia, weight gain, blood pressure, and lipid profiles

Over the entire study period, episodes of mild hypoglycemic reactions in the MIT group
occurred 1.6 times more than those in the CIT group (35 vs. 22 events/100 patient-years),
but no patient experienced episodes of coma, seizure, or severe hypoglycemia that required
the assistance of another person. Also, no patient in either the MIT or CIT group showed
weight gain of >120% of the ideal body weight (MIT group 20.5 2.1 to 21.5 2.0 kg/m 2,
CIT group 20.3 2.8 to 21.3 2.6 kg/m2). There were no significant differences in blood
pressure levels and lipid profiles or in therapeutic regimens for antihypertensive therapy
(ACE inhibitors or -blockers) and lipid-lowering therapy (HMG-CoA reductase
inhibitors or nicotinic acid) between the MIT and CIT groups at entry (Table 1) and after 8
years of follow-up (systolic blood pressure 129 16 vs. 130 17 mmHg, diastolic pressure
70 12 vs. 72 11 mmHg, total cholesterol 205 27 vs. 207 27 mg/dl, triglycerides 93
35 vs. 102 38 mg/dl, HDL cholesterol 51 14 vs. 50 14 mg/dl, antihypertensive
therapy 14 vs. 19%, and lipid-lowering therapy 12 vs. 15%, for the MIT vs. CIT group,
respectively).

CONCLUSIONS In Japan, nonobese hypoinsulinemic insulin-requiring type 2 diabetic

patients have been treated with once- or twice-daily injections of intermediate-acting
insulin. However, this conventional insulin injection therapy seemed to be inappropriate for
the prevention or intervention of diabetic complications, since conventional insulin
injection therapy could not normalize postprandial hyperglycemia (15). Therefore, we
evaluated the effect of the intensive insulin therapy on the prevention of the onset and
progression of diabetic microvascular complications in Japanese patients with type 2
diabetes.

The results of 6 years of the Kumamoto Study and the superiority of multiple insulin
injection therapy were explained to the patients, and the selection of insulin treatment
regimens was left to them. Only two patients in the CIT group selected multiple insulin
injection therapy, and other patients in both the CIT and MIT groups wanted to adhere to
the same treatment regimens. Therefore, the follow-up study was continued by intention of
the patients. As a result, it was demonstrated that intensive glycemic control with multiple
insulin injection therapy effectively delayed the onset and progression of diabetic
retinopathy, nephropathy, and neuropathy in Japanese type 2 diabetic patients.

In the recent report from the U.K. Prospective Diabetes Study (UKPDS) group, newly
diagnosed patients with type 2 diabetes (BMI 2728 kg/m2) randomly assigned to
conventional treatment with diet or intensive treatment with sulfonylureas, metformin, or
insulin were followed over 10 years. As a result, mean HbA1c was 7.0% in the intensive
group as compared with 7.9% in the conventional group, and the risk of microvascular
complications in the intensive group was significantly lower than that in the conventional
group (16,17). The UKPDS group also reported that tight blood pressure control
significantly decreases the risk of microvascular complications in hypertensive patients
with type 2 diabetes (18). Together with the results from the Kumamoto Study, which
evaluates insulin-requiring nonobese type 2 diabetic patients, it was determined that strict
glycemic control, by whatever means, could be useful in preventing the onset and
progression of diabetic microvascular complications in type 2 diabetic patients, whether
they are obese or nonobese.

In the evaluation of retinal findings, the development and progression of retinopathy were
defined as a change of 2 steps up or more in a modified ETDRS scale of 19 stages in the
Kumamoto Study, 3 steps up or more in a ETDRS scale of 25 stages in the DCCT, and 2
steps up in a modified ETDRS scale of 21 stages in the UKPDS (2,6,911,16). Because the
patients with preproliferative retinopathy at the initiation of the study were not included in
the Kumamoto Study, this classification system of 19 stages allowed us to evaluate the
retinal findings efficiently. The cumulative percentage of patients showing the worsening in
retinal findings was significantly less in patients receiving intensive insulin treatment than
in those receiving conventional insulin treatment. With intensive insulin therapy, the risks
of retinopathy were reduced in the primary prevention, the secondary intervention, and the
combined cohorts. The risk calculated by proportional-hazards analysis in the combined
cohort was comparable to that in the DCCT. Intensive insulin therapy also reduced the risk
of severe nonproliferative or proliferative retinopathy and that of treatment of
photocoagulation. Analysis by Cox's proportional-hazards model in the Kumamoto Study
revealed that the relative risk of developing retinopathy decreased by 47% for every 1%
decrease in HbA1c. A similar result was obtained in the DCCT: a 10% lower HbA1c (e.g., 8
vs. 7.2%) was associated with a 43% lower risk of developing retinopathy in the intensive
group and a 45% lower risk in the conventional group (19). At the 34th Annual Meeting of
the European Association for the Study of Diabetes, held in 1998, the UKPDS group also
reported that for every 1% decrease in HbA1c, the relative risk of developing microvascular
complications decreased by 35%. Therefore, it was concluded that intensive glycemic
control was effective in delaying the onset and/or progression of retinopathy, both in type 1
and type 2 diabetes.

Intensive glycemic control was also effective in patients with diabetic nephropathy. The
development and progression of nephropathy was significantly lower in patients on
intensive insulin therapy than in those receiving conventional treatment. In the primary
prevention cohort, intensive insulin therapy reduced the risk of microalbuminuria by 74%
and albuminuria by 100%. It also reduced the risk of microalbuminuria by 60% and
albuminuria by 100% in the secondary intervention cohort. The risk reduction calculated by
proportional-hazards risk analysis was greater in the Kumamoto Study with type 2 diabetes
than that in the DCCT with type 1 diabetes (risk reduction in the combined cohort: 74 vs.
3954%, respectively). Also, in the UKPDS, risk reduction of microalbuminuria was the
highest among microvascular complications (34% for the difference of 0.9% in HbA1c)
(16). This result was comparable to our result in which risk reduction of microalbuminuria
was 74%, for the difference of 2.0% in HbA1c. The reason intensive glycemic control
significantly reduced the risk of progression to microalbuminuria or albuminuria in the
Kumamoto Study has not been explained thoroughly, but might possibly be the different
susceptibility to nephropathy between the types of diabetes (type 2 and 1) and between the
races (Caucasian and Asian). It has been known that Japanese patients with type 2 diabetes
are susceptible to nephropathy (20,21). Further studies are necessary to solve this question.

After an 8-year period, there were significant differences between the CIT and MIT groups
in the nerve conduction velocities and vibration threshold. The orthostatic hypotension and
CV of the R-R interval in the MIT group tended to improve after 8 years, whereas these
indexes in the CIT group tended to deteriorate. Therefore, we concluded that intensive
glycemic control would have a beneficial effect on both somatic and autonomic nerve
functions.

The effect of intensive glycemic control on macrovascular complications was also

evaluated in type 2 patients in the Kumamoto Study. The total number of macrovascular
events in the CIT group was twice that in the MIT group but was not statistically
significant, probably because of the small number of patients studied in the Kumamoto
Study. Jensen-Urstad et al. (22) reported that early atherosclerosis could be retarded by
improved glycemic control. In the present study, intensive glycemic control might have a
beneficial effect on the progression of macrovascular complications. This is another
question that remains to be solved.

It has been reported that intensive insulin treatment can induce side effects, such as
hypoglycemia or weight gain. The DCCT showed that the risk of severe hypoglycemia
(defined as an episode with symptoms consistent with hypoglycemia, in which the patient
required the assistance of another person and which was associated with a blood glucose
level <50 mg/dl or prompt recovery after oral carbohydrate or glucagon or intravenous
glucose) was three times higher with intensive insulin therapy than with conventional
insulin therapy (2,23). In the UKPDS, 2.3% of insulin-treated patients experienced major
hypoglycemic episodes (defined when third-party help or medical intervention was
necessary). However, in the present study, no patient experienced a severe hypoglycemic
attack, and any episodes of mild hypoglycemic reactions in the MIT group occurred only
1.6 times more than those in the CIT group (35 vs. 22/100 patient-years). This might be
explained by the difference in insulin doses among the UKPDS and Kumamoto Study
patients with type 2 diabetes, and the DCCT patients with type 1 diabetes (~0.4 [Kumamoto
Study], 0.2 for nonobese and 0.5 for obese [UKPDS], and 0.7 U/kg of body wt/24 h
[DCCT]. In intensive care groups, there was also a difference in clinic visits for patient
care. Routine clinic visits were every 3 or 4 months in the UKPDS and every month in the
DCCT, whereas in the Kumamoto Study, patients were assigned to visit every 2 weeks, and
insulin doses were adjusted according to their degree of glycemic control during the 8-year
follow-up study.

In the present study, the MIT group showed an improvement in glycemic control using the
same insulin doses as the CIT group, regardless of having the same instructions for diet and
exercise. This could be mainly explained by the usefulness of prandial supplementation
with additional doses of short-acting insulin, although the possibility of the improvement of
insulin sensitivity after long-term strict glycemic control should not be excluded. In a
previous report, we demonstrated that optimal daily profiles of free plasma insulin were
achieved during the periods with combined basal plus prandial insulin supplements and
reported that instead of the basal insulin supplement with once- or twice-daily injections of
intermediate-acting insulin, short-acting insulin supplements three times a day were
essential to achieve the optimal glycemic control in Japanese type 2 diabetic patients (15).
We also reported that in the MIT group, insulin sensitivity measured by the euglycemic
clamp technique was improved significantly after 13 months, but not in the CIT group
(24).

It is also notable that because most of the insulin-requiring type 2 diabetic patients in Japan
are less obese and not hyperinsulinemic, glycemic control could be achieved with relatively
smaller doses of insulin when compared with type 2 diabetic patients in Western countries.

There was a slight but not significant increase in BMI in both the MIT and CIT groups
from baseline to 8 years. This increase was also smaller than that reported from the DCCT
(2,25).

It is clinically important to indicate the glycemic threshold to prevent the development and
progression of diabetic microvascular complications. As shown in Fig. 5, no worsening of
retinopathy and nephropathy was observed in type 2 diabetic patients whose HbA 1c, FBG,
and 2-h postprandial blood glucose concentrations were below 6.5%, 110 mg/dl, and 180
mg/dl, respectively. Therefore, in the present study, the glycemic thresholds to prevent the
onset and progression of diabetic microvascular complications are indicated as follows: an
HbA1c level at least <6.5%, FBG <110 mg/dl, and 2-h postprandial blood glucose
concentration <180 mg/dl. These glycemic thresholds suggested by the Kumamoto Study
might possibly be oversimplified because of the analysis of the data with smaller numbers
of patients studied. Epidemiological analyses of the DCCT and UKPDS data showed a
continuous relationship between the risks of diabetic microvascular complications and
glycemia, and there was no evidence of any glycemic thresholds for any of the
microvascular complications above normal glucose levels (2,16,19). Therefore, the
epidemiological analyses suggested that by whatever means, intensive therapy with the goal
of achieving normal glycemia should be implemented as early as possible in as many type 1
and type 2 diabetic patients as safely possible.

In conclusion, intensive glycemic control can delay the onset and progression of the early
stages of diabetic retinopathy, nephropathy, and neuropathy in type 2 diabetic patients as
well as in type 1 diabetic patients. As the results show, intensive glycemic control could
improve the quality of life and save medical resources.

Acknowledgments The authors are grateful to Dr. H. Maruyama, Department of

Cardiovascular Disease, Minamata General Medical Center, Kumamoto; Dr. S. Ueno,
Department of Metabolic Medicine, Arao Citizen's Hospital, Kumamoto; and many other
staff members in the Departments of Metabolic Medicine and Ophthalmology for their
assistance throughout the follow-up study.

References

1. The DCCT Research Group: The Diabetes Control and Complications Trial (DCCT):
design and methodologic considerations for the feasibility phase. Diabetes 35:530545,
1986

2. The Diabetes Control and Complications Trial Research Group: The effect of intensive
treatment of diabetes on the development and progression of long-term complications in
insulin-dependent diabetes mellitus. N Engl J Med 329:977986, 1993

3. Feldt-Rasmussen B, Mathiesen ER, Jensen T, Lauritzen T, Deckert T: Effect of

improved metabolic control on loss of kidney function in type I (insulin-dependent)
diabetic patients: an update of the Steno studies. Diabetologia 34:164170, 1991

4. The Kroc Collaborative Study Group: Diabetic retinopathy after two years of intensified
insulin treatment. JAMA 260:3741, 1988
5. Brinchmann-Hansen O, Dahl-Jrgensen K, Sandvik L, Hanssen KF: Blood glucose
concentrations and progression of diabetic retinopathy: the seven year results of the Oslo
study. BMJ 304:1922, 1992

6. Ohkubo Y, Kishikawa H, Araki E, Miyata T, Isami S, Motoyoshi S, Kojima Y,

Furuyoshi N, Shichiri M: Intensive insulin therapy prevents the progression of diabetic
microvascular complications in Japanese patients with non-insulin-dependent diabetes
mellitus: a randomized prospective 6-year study. Diabetes Res Clin Pract 28:103117,
1995

7. Service FJ , Molner GD, Rosevear JW, Ackerman E, Gatewood LC, Taylor WF: Mean
amplitude of glycemic excursions, a measure of diabetic instability. Diabetes 19:644655,
1970

8. Schlichtkrull J, Munck O, Jersild M: The M-value, an index of blood-sugar control in

diabetics. Acta Med Scand 177:95102, 1965

9. Early Treatment Diabetic Retinopathy Study Research Group: Early Treatment Diabetic
Retinopathy Study: Manual of Operations. Bethesda, MD, Public Health Service, 1980

10. Early Treatment Diabetic Retinopathy Study Research Group: ETDRS report no. 10:
grading diabetic retinopathy from stereoscopic color fundus photographs: an extension of
the modified Airlie House classification. Ophthalmology 98:786806, 1991

11. Early Treatment Diabetic Retinopathy Study Research Group: ETDRS report no. 12:
fundus photographic risk factors for progression of diabetic retinopathy. Ophthalmology
98:823833, 1991

12. Mogensen CE: Management of diabetic renal involvement and disease. Lancet ii:867
870, 1988

13. Shima K, Endo J, Oimomi M, Oshima I, Omori Y, Katayama Y, Kanazawa Y, Kawai

T, Kawamori R, Kanno T, Kiyose H, Nakashima K, Nagamine Y, Baba S, Hoshino T:
Laboratory difference in HbA1c measurement in Japan: an interim report of the Committee
on an Interlaboratory Standardization of HbA1c Determination, the Japan Diabetes Society.
J Jpn Diabetes Soc 37:233243, 1994

14. Lee ET: Statistical Methods for Survival Data Analysis. Belmont, CA, Lifetime
Learning Publications, 1981

15. Kishikawa H, Hashimoto Y, Hashiguchi Y, Sasahara T, Nishiyama T, Yano T,

Yamaguchi K, Shichiri M: Insulin treatment regimes for type II (non-insulin dependent)
diabetes mellitus as judged by residual B-cell function. Diabetes Nutr Metab 4:305312,
1991
16. U.K. Prospective Diabetes Study Group: Intensive blood-glucose control with
sulphonylurea or insulin compared with conventional treatment and risk of complications in
patients with type 2 diabetes (UKPDS 33). Lancet 352:837853, 1998

17. U.K. Prospective Diabetes Study Group: Effect of intensive blood-glucose control with
metformin on complications in overweight patients with type 2 diabetes (UKPDS 34).
Lancet 352:854865, 1998

18. U.K. Prospective Diabetes Study Group: Tight blood pressure control and risk of
macrovascular and microvascular complications in type 2 diabetes (UKPDS 38). BMJ
317:703713, 1998

19. The Diabetes Control and Complications Trial Research Group: The relationship of
glycemic exposure (HbA1c) to the risk of development and progression of retinopathy in the
Diabetes Control and Complications Trial. Diabetes 44:968983, 1995

20. Shigeta Y, Haneda M, Kikkawa R: Early detection and progression of nephropathy in

patients with non-insulin-dependent diabetes. In Prevention and Treatment of NIDDM.
Goto Y, Lee TH, Kaneko T, Shigeta Y, Eds. London, Smith-Gordon, 1992, p. 289291

21. Kuzuya K, Akanuma Y, Akazawa Y, Uehata T: Prevalence of chronic complications in

Japanese diabetic patients. Diabetes Res Clin Pract 24 (Suppl.):S159S164, 1994

22. Jensen-Urstad KJ, Reichard PG, Rosfors JS, Lindblad LEL, Jensen-Urstad MT: Early
atherosclerosis is retarded by improved long-term blood glucose control in patients with
IDDM. Diabetes 45:12531258, 1996

23. The DCCT Research Group: Epidemiology of severe hypoglycemia in the diabetes
control and complications trial. Am J Med 90:450459, 1991

24. Shichiri M, Kishikawa H: Insulin treatment of non-insulin-dependent diabetes mellitus:

strict glycemic control and its effect on insulin sensitivity. In Prevention and Treatment of
NIDDM. Goto Y, Lee TH, Kaneko T, Shigeta Y, Eds. London, Smith-Gordon, 1992, p.
459463

25. The Diabetes Control and Complications Trial Research Group: Weight gain associated
with intensive therapy in the Diabetes Control and Complications Trial. Diabetes Care
11:567573, 1988

From the Department of Metabolic Medicine, Kumamoto University School of Medicine,

Kumamoto, Japan.
Address correspondence and reprint requests to Motoaki Shichiri, MD, PhD, Department of
Metabolic Medicine, Kumamoto University School of Medicine, 1-1-1, Honjo, Kumamoto,
860-8556, Japan.

Received for publication 8 July 1999 and accepted in revised form 19 November 1999.

Abbreviations: CIT, conventional insulin injection therapy; CV, coefficient of variation;

DCCT, Diabetes Control and Complications Trial; ETDRS, Early Treatment of Diabetic
Retinopathy Study; FBG, fasting blood glucose; MAGE, mean amplitude of glycemic
excursions; MBG, mean blood glucose; MIT, multiple insulin injection therapy; UKPDS,
U.K. Prospective Diabetes Study.

A table elsewhere in this issue shows conventional and Systme International (SI) units and
conversion factors for many substances.

This article is based on a presentation at a symposium. The symposium and the publication
of this article were made possible by an unrestricted educational grant from Aventis
Pharma.