Beruflich Dokumente
Kultur Dokumente
in any endpoint related to diabetes and in This article addresses the pathophys-
D
iabetes is a highly prevalent
chronic disease. The Third diabetes-related death, a 14% decrease in iology of type 2 diabetes, goals of thera-
National Health and Nutrition all-cause mortality and myocardial py, misconceptions about insulin,
Examination Survey, conducted between infarction, a 43% decrease in amputation restoration of natural insulin patterns,
1988 and 1994, estimated the prevalence or death from peripheral vascular dis- and ways to incorporate basal insulin
of diagnosed and undiagnosed diabetes ease, and a 37% decreased risk for into a strategy that promotes compliance.
in people aged 20 years and older at microvascular complications, each of
15.6 million.1 Of these people, ~90–95% which was statistically significant.5 The Pathophysiology of Type 2 Diabetes
have type 2 diabetes, with a higher Japanese Kumamoto study6 also found Type 2 diabetes is characterized by
prevalence seen among Native that intensive glycemic control reduced hyperglycemia caused by defects in
Americans and Americans of African, the risk for retinopathy, nephropathy, and insulin secretion (impaired -cell func-
Mexican, and Japanese descent.2 The neuropathy in patients with type 2 dia- tion) and insulin action (insulin resist-
prevalence of diabetes rose from 4.9% betes. ance by the liver and muscle tissue).8–10
in 1990 to 6.9% in 1999, primarily Although sulfonylurea therapy has These defects occur early in the course
because of an increase in the prevalence been the mainstay of treatment for type 2 of the disease and are often present
of obesity. It has been postulated that, diabetes for >40 years, the UKPDS before diagnosis.
with the growing obesity problem, dia- reported that over a 6-year period, ~53% In a prospective study11 of Pima Indi-
betes will become an even more perva- of patients who were randomized to ans, a group at high risk for developing
sive threat.3 receive treatment with sulfonylureas diabetes, body composition, insulin
Type 2 diabetes produces or is a con- needed additional insulin therapy, rein- action, insulin secretion, and endogenous
tributor to considerable morbidity in the forcing the concept that hyperglycemia glucose output were measured over sev-
form of metabolic complications, vision in type 2 diabetes is progressive.7 Clini- eral years in subjects whose glucose tol-
disorders, neuropathy, kidney disease, cians should consider this when estab- erance went from normal to impaired to
peripheral vascular disease, ulcerations lishing a therapeutic regimen for patients diabetic. A two-step hyperinsulinemic,
and amputations, heart disease, stroke, with type 2 diabetes. euglycemic glucose clamp test assessed
digestive diseases, infection, oral com- insulin action. During the transition from
IN BRIEF normal to impaired glucose tolerance,
plications, and depression. The associat-
ed mortality rate has been estimated at there was a 27% decrease in the acute
The prevalence of type 2 diabetes has
5.5% annually. Moreover, the disease insulin secretory response (AIR), the
been increasing rapidly and with it
reduces life expectancy by 5–10 years.1 average incremental plasma insulin con-
has been resultant morbidity and mor-
Although there is no cure for dia- centration from the third to the fifth
tality. Strict glycemic control reduces
betes, two large controlled studies, the minute after the glucose bolus. Further-
the progression of diabetic microvas-
Diabetes Control and Complications Tri- more, during the transition from
cular disease; however, most patients
al (DCCT)4 and the U.K. Prospective impaired glucose tolerance to diabetes,
treated with sulfonylureas require
Diabetes Study (UKPDS)5 have pointed there was an additional 57% decrease in
additional insulin therapy. This article
to the importance of intensive blood glu- AIR.
addresses common clinician concerns
cose control in reducing its associated Another controlled study in patients
about prescribing insulin early in type
morbidity. In fact, the UKPDS, the with type 2 diabetes who were either
2 diabetes. It presents strategies for
largest and longest trial ever conducted untreated or attempting to achieve con-
incorporating basal insulin therapy
in patients with type 2 diabetes, found trol using diet or oral hypoglycemic
with glargine (Lantus) into a regimen
that for each 1% reduction in hemoglo- agents9 found that basal and mean 24-
that promotes compliance.
bin A1c (A1C), there was a 21% decrease hour glucose concentrations were signif-
icantly higher in the diabetic patients, immune-mediated -cell destruction that diabetes,16 the use of insulin partially
pointing to potentially impaired insulin is characteristic of type 1 diabetes. This reversed the postbinding defect in
secretion. During the hyperglycemic disease, latent autoimmune diabetes of peripheral insulin action, produced
clamp portion of this study, patients adulthood, is often treated in the same near-normal basal hepatic glucose out-
secreted ~70% less insulin than control way as type 2 diabetes because it does put, and enhanced insulin secretion,
subjects (Table 1). In nondiabetic indi- not require insulin initially.13 thereby maintaining lower glucose val-
viduals, a biphasic insulin response ues. In addition, the mean daily insulin
begins upon glucose stimulation, starting Therapeutic Objective requirement fell by 23% after ~2 weeks
with a rapid rise in insulin 1–3 minutes The American Diabetes Association of therapy, leveling off thereafter.
after the glucose level is raised (first (ADA) recommends that patients with The use of insulin has been associat-
phase), returning toward baseline 6–10 diabetes receive care from a medical ed with weight gain, which in turn has
minutes after glucose stimulation, and team. Working with patients and their been considered a major factor in insulin
rising gradually once again (second families, these teams develop self-man- resistance. In the UKPDS, patients in
phase).12 Among patients in this study, agement and problem-solving plans that intensive therapy gained more weight
however, the first-phase response after consider each patient’s cultural, social, than those in conventional therapy
meals (glycemic load) was either absent physical, and medical needs. ADA sup- groups; patients taking insulin gained ~4
or greatly diminished.9 As a result of ports the findings of the DCCT and the kg compared to 2.6 kg for those on
bolus (mealtime) and basal (between- UKPDS for intensive glycemic control; chlorpropamide (Diabinese) and 1.7 kg
meal) defects in insulin activity in type 2 Table 2 lists its recommendations for for those on glibenclamide (glyburide
diabetes, bolus and basal glucose levels nonpregnant people with diabetes.14 [Micronase]).17 Yet patients in the inten-
are increased, producing hyperglycemia. sive therapy groups also had fewer
In the early stages of type 2 diabetes, Insulin: Misconceptions and Reality microvascular complications, suggesting
blood glucose levels can often be con- Contrary to some beliefs, there is no that tight glycemic control may be more
trolled with changes in diet and physical evidence that doses of insulin used in important in therapeutic decision-mak-
activity along with sulfonylureas. Unfor- clinical practice exacerbate insulin ing. The use of metformin (Glucophage)
tunately, the -cell dysfunction that resistance. It has long been recognized as an adjunct to insulin therapy provides
leads to impaired insulin secretion is that the chronic hyperglycemia associ- effective glycemic control without sig-
progressive, and eventually patients will ated with type 2 diabetes (glucose toxi- nificant weight gain.18,19
require a treatment strategy that includes city) leads to impairment in insulin The incidence of heart disease and
insulin, either alone or with oral agents.7 secretion and a possible defect in glyco- ischemic heart mortality is up to four
It should be noted that some patients gen synthesis.15 In a study of intensive times higher in people with diabetes.
who develop diabetes in adulthood have insulin therapy in patients with type 2 Ischemic heart disease, other heart dis-
ease, and cerebrovascular disease
Table 1. Glucose Concentrations and Insulin Secretion in Control and account for 40, 15, and 10% of all deaths
Diabetic Subjects Under Three Conditions in this population, respectively (Figure
1).20 A population-based survey of
Condition Control Patients With P almost 3,000 people21 identified high
Subjects Diabetes prevalence rates for a constellation of
Fasting disorders known as syndrome X, meta-
Glucose (mg/dl) 95.2 ± 2.1 221.4 ± 19.4 <0.0001 bolic syndrome, and insulin resistance
Insulin secretion syndrome interchangeably—abdominal
(nmol/m2/24 hours) 71.7 ± 9.5 82.7 ± 11.5 NS obesity, type 2 diabetes, glucose intoler-
ance, hypertension, hypertriglyc-
24-Hour Study eridemia, and hypercholesterolemia—
Glucose (mg/dl) 109.7 ± 1.9 282.3 ± 25.6 <0.0001 that far exceeded the rates for each
Insulin secretion disorder alone or in pairs. For each of
(nmol/m2/24 hours) 220.5 ± 30.4 201.7 ± 19.7 NS these conditions, fasting and post-glu-
cose hyperinsulinemia was a common
Hyperglycemic Clamp Study
Glucose (mg/dl) 303.6 ± 4.5 299.8 ± 1.0 NS thread that suggested the presence of
Insulin secretion insulin resistance. When insulin-resistant
(nmol/m2/24 hours) 80.6 ± 11.7 24.1 ± 4.1 <0.001 subjects were compared to control sub-
Adapted with permission from Ref. 9.
jects, significantly lower HDL choles-
terol levels were also seen.
Table 2. ADA Guidelines for Glycemic Control for Patients With Diabetes betes and several cardiovascular risk fac-
tors, ADA14 recommends the following
Normal Goal Additional action in addition to lifestyle alterations: blood
suggested* pressure measurement at routine medical
Plasma values (mg/dl)† visits; the use of antihypertensive agents
Average preprandial glucose <110 90–130 <90/>150 in patients with hypertension; testing for
Average bedtime glucose <120 110–150 <110/>180 lipid disorders at least annually; lower-
ing LDL and triglycerides; increasing
Whole blood values (mg/dl)‡ HDL and using statins as first-line lipid
Average preprandial glucose <100 80–120 <80/>140 therapy; using fibrates in patients with
Average bedtime glucose <110 100–140 <100/>160 low HDL; and using aspirin therapy to
prevent cardiovascular events.
A1C <6 <7 >8 Clinicians often cite hypoglycemia
The values shown in this table are by necessity generalized to the entire population of individuals as an adverse effect that might preclude
with diabetes. Patients with comorbid diseases, the very young and older adults, and others with the use of insulin. Indeed, in the DCCT
unusual conditions or circumstances may warrant different treatment goals. These values are for non-
pregnant adults. study of type 1 diabetes,4 tighter control
*Values above/below these levels are not “goals” nor are they “acceptable” in most patients. They are produced a risk of severe hypoglycemia
an indication for a significant change in the treatment plan. “Additional action suggested” depends on three times higher than that of conven-
individual patient circumstances. Such actions may include enhanced diabetes self-management edu-
cation, comanagement with a diabetes team, referral to an endocrinologist, change in pharmacologi- tional therapy (Figure 2). This must be
cal therapy, initiation of or increase in self-monitoring of blood glucose, or more frequent contact viewed, however, in the context of sub-
with the patient. A1C is referenced to a nondiabetic range of 4.0–6.0% (mean 5.0%, SD 5%). stantially reduced microvascular and
†Values calibrated to plasma glucose.
‡Measurement of capillary blood glucose. neurological complications. Further-
Reprinted with permission from Ref. 14. more, the rates of severe hypoglycemia
are quite low in type 2 diabetes. In the
Because of the connection among years; the absolute reduction in mortali- Kumamoto study of type 2 diabetes,6
hyperinsulinemia, insulin resistance, and ty was 11%.23 average A1C results were 7.1 and 9.4%
cardiovascular risk factors, the UKPDS17 Other studies24–26 have reported for tight and conventional groups,
compared cardiovascular events among improvement or neutral effects on other respectively. However, only mild hypo-
patients randomized to conventional cardiovascular risk factors—total choles- glycemic reactions occurred and at simi-
lifestyle and dietary management and terol, LDL cholesterol, HDL cholesterol, lar rates in both groups.
those on a tight glycemic control regi- triglycerides, or hypertension—with The UKPDS17 found that the rate of
men with sulfonylureas, metformin (in insulin, even among obese patients. major hypoglycemic episodes (defined
overweight patients), or insulin. No Considering the comorbidity of dia- as an episode in which help from another
adverse effects on cardiovascular out-
comes were seen with any of the treat- 4% 5%
ments, including insulin.
The Diabetes Mellitus Insulin Glu- 13%
cose Infusion in Acute Myocardial 40%
Infarction study22,23 assessed the effect
of acute insulin-glucose infusion fol- 65%
13%
lowed by long-term intensive (multi- 15%
dose) insulin treatment in diabetic
patients who have had an acute 10%
myocardial infarction. Among patients
who had received an acute infusion,
there was a significant decrease in glu- All other
Diabetes
cose. After 1 year, there was a signifi- Malignant neoplasms Ischemic heart disease
cant reduction in mortality in the group Pneumonia/influenza Other heart disease
who received intensive insulin treat-
Cerebrovascular disease
ment, particularly in patients who had a
low cardiovascular risk profile and
were insulin naive.22 These effects per- Figure 1. Cardiovascular and cerebrovascular diseases account for 65% of all
sisted after a mean follow-up of 3.4 deaths in patients with diabetes. Adapted from Ref. 20.
person or medical intervention was nec- concentration that result from the inges- mimic physiological basal insulin secre-
essary) was higher for patients taking tion of food. In people with diabetes, tion, with varying results. An insulin that
insulin (2.3%) than for patients undergo- however, bolus and basal glucose levels mimics basal secretion should be slowly
ing any other intensive therapy or con- are increased; thus, strategies for insulin and evenly absorbed with no peak, have
ventional treatment (<1%). The rate of replacement must focus on mimicking consistent bioavailability, and have a
any hypoglycemic episode (including the phases of insulin secretion. Figure 3 long half-life that permits once-daily
episodes that the patient was able to treat shows available insulins by onset, peak, administration.30,31 This has not been the
unaided) was 36.5% with insulin treat- and usual effective duration.27 case with most basal insulin products.
ment, compared to 11, 17.7, and 1.2% NPH and lente have early peaks with
for chlorpropamide, glibenclamide, and Prandial (Bolus) Insulin rapid waning of action; this may con-
diet, respectively. In a secondary analy- Prandial forms of insulin mimic the nor- tribute to both nighttime hypoglycemia
sis of obese patients on intensive mal first-phase response. These regular and the “dawn phenomenon,” a pre-
glycemic control, the rate of major hypo- insulin or rapid-acting insulin analogs breakfast rise in plasma glucose. There
glycemic events per year for intensive are administered 30–60 minutes (regu- is also variability in the absorption of
insulin treatment was 2.5% versus <1% lar) or 10–15 minutes (lispro [Humalog] NPH and lente.29,32 Data on ultralente
for other treatments. or aspart [Novolog]) before food con- vary; in one study of type 1 diabetes,30
Despite the increased risk of mild sumption. They must also mimic the the onset of action of human ultralente
hypoglycemia, aggressive therapy that second-phase response, requiring a more was 2–4 hours, and there was a broad,
combines patient education and self- prolonged duration of action to control variable peak 6–12 hours after injection.
management with a form of exogenous prolonged glucose elevation after the The authors concluded that this product
insulin that closely mimics normal meal.28 Although prandial forms offer did not provide a constant basal insulin
insulin secretion can help to reduce the flexibility in that they can be injected concentration.
morbidity and mortality associated with just before a meal, most patients also Basal insulin in type 1 diabetes. The
type 2 diabetes. require daily basal insulin injections.29 pharmacokinetics and pharmacodynam-
ics of the insulin analog glargine were
Restoring Natural Insulin Patterns Basal Insulin compared to those of NPH, ultralente,
Nondiabetic pancreases self-regulate the Both intermediate (NPH and lente) and and continuous subcutaneous insulin
amount of insulin secreted, acting in long-acting (ultralente and glargine infusion (CSII) of lispro in 20 type 1 dia-
response to changes in blood glucose [Lantus]) insulins have been used to betic patients using an isoglycemic 24-
hour clamp.33 Glargine was absorbed
120 slowly and produced no pronounced
peaks over a 24-hour period. Its onset of
100 action was ~1.5 hours, compared with
Rate of Severe Hypoglycemia
Table 3. Relative Risk of Hypoglycemic Episodes per Patient-Year affect compliance. The regimen itself is
also a factor; if it is difficult, costly, or
Hypoglycemia Glargine NPH (%) P Relative Risk With has many side effects, compliance may
(%) Glargine (%) diminish.46–49
All confirmed 13.9 17.7 <0.02 21 In the treatment of type 2 diabetes
Nocturnal confirmed 4.0 6.9 <0.001 42 with insulin, reluctance to inject oneself
and fear of weight gain or hypoglycemia
Adapted with permission from Ref. 40. may hinder compliance.35 Clinicians
need to explain to their patients that type
Compliance ance. In the management of diabetes, 2 diabetes is progressive and that insulin
Patients with diabetes play an integral this is more than a relationship between will probably have to be used at some
role in any treatment strategy. Lifestyle the patient and a single provider—it point; therefore, clinicians may need to
modification; goal setting; self moni- includes an entire health care team. dispel myths associated with insulin use,
toring; preventing, detecting, and treat- Other factors also influence compli- allay patient fears, and assure patients
ing acute complications; and using ance. On the patient’s side, the belief that that insulin will likely improve symp-
medications correctly are all important the benefits of therapy are worth the con- toms, enhance quality of life, and pro-
components in achieving glycemic sequences, a readiness to change, memo- vide a sense of well-being.
control.45 This makes patient education ry, communication skills, literacy level, “Resistance to insulin” on the part of
crucial, particularly when it comes to knowledge, competence, confidence, clinicians may also be a significant
dispelling myths about insulin therapy. skills, and a good support system work provider-driven factor in compliance.50
The content areas that must be clearly together to influence the patient’s accept- Concerns regarding hypoglycemia,
established for patients are listed in ance of therapy. On the team’s side, patients’ fear of needles, cultural health
Table 4. communication skills, the quality of beliefs, and the time necessary to teach
The relationship between health care information and instructions, and a will- self-injection can all emerge as barriers
provider and patient is crucial to compli- ingness to identify and address barriers to insulin use.
Retest for diabetes MNT, physical activity, DSME, Insulin therapy, MNT, physical
in 1 year risk factor reduction, SMBG activity, DSME, risk reduction
No more than 12 weeks
Figure 4. An algorithm of treatment for patients with type 2 diabetes. Reprinted with permission from Ref. 44.
42
ments in the pharmacological reduction of blood Yki-Järvinen H, Dressler A, Ziemen M, Stephen N. Davis, MD, FRCP, is chief
glucose in patients with type 2 diabetes. Clinical HOE 901/3002 Study Group: Less nocturnal
Diabetes 19:153–159, 2001 hypoglycemia and better post-dinner glucose of the Division of Diabetes, Endocrinol-
32
control with bedtime insulin glargine compared ogy, and Metabolism at Vanderbilt Uni-
Atiea JA, Luzio S, Owens DR: The dawn with bedtime NPH insulin during insulin combi-
phenomenon and diabetes control in treated nation therapy in type 2 diabetes. Diabetes Care versity Medical School in Nashville,
NIDDM and IDDM patients. Diabetes Res Clin 23:1130–1136, 2000 Tenn. Ramachandiran Cooppan, MD, is
Pract 16:183–190, 1992
33
43
Benedetti MM, Hamburg E, Dressler A, Zie- an assistant clinical professor at Har-
Lepore M, Pampanelli S, Fanelli C, Porcel- man M, for the 3002 Study Group: Lower inci-
lati F, Bartocci L, DiVincenzo A, Cordoni C, dence of nocturnal hypoglycaemia in patients
vard Medical School in Boston, Mass.
Costa E, Brunetti P, Bolli G: Pharmacokinetics with type 2 diabetes treated with insulin glargine Mayer B. Davidson, MD, is director of
and pharmacodynamics of subcutaneous injection compared with NPH insulin, given as a combina-
of long-acting human insulin analog glargine, tion regimen with oral agents (Poster). Presented
the Clinical Trials Unit at Charles R.
NPH insulin, and ultralente human insulin and at the 38th annual meeting of the European Asso- Drew University in Los Angeles, Calif.
continuous subcutaneous infusion of insulin ciation for the Study of Diabetes. Budapest, Hun-
lispro. Diabetes 49:2142–2148, 2000 gary, September 1–5, 2002
Kathryn Mulcahy, RN, MSN, CDE, is
34
Rosenstock J, Park G, Zimmerman J, U.S. 44
program director of the INOVA Dia-
IAFP Family Practice Education Network:
Insulin Glargine (HOE 901) Type 1 Diabetes Type 2 diabetes: diagnosis and management betes Center in Fairfax, Va. Gary A.
Investigator Group: Basal insulin glargine (HOE strategies for primary care clinicians: consensus Manko, MD, is president of Clinical
901) versus NPH insulin in patients with type 1 recommendations from an expert panel. Ill Acad
diabetes on multiple daily insulin regimens. Dia- Fam Phys 1–12, 2002 Associates, PA, in Reisterstown, Md.
betes Care 23:1137–1142, 2000 Donald Nelinson, PhD, is executive
45
35
Mensing C, Boucher J, Cypress M, Weinger
Nathan DM: Initial management of K, Mulcahy K, Barta P, Hosey G, Kopher W, director of the Diabetes Consortium.
glycemia in type 2 diabetes mellitus. N Engl J Lasichak A, Lamb B, Mangan M, Norman J, Tan-
Med 347:1342–1349, 2002 ja J, Yauk L, Wisdom K, Adams C: National
The Diabetes Consortium, Inc., is a
36
Lindstrom TH, Arnqvist HJ, von Schenck standards for diabetes self-management educa- nonprofit, multidisciplinary collabora-
HH: Effect of conventional and intensified insulin tion. Diabetes Care 25 (Suppl. 1):S140–S147, tion dedicated to the development and
therapy on free-insulin profiles and glycemic 2002
control in NIDDM. Diabetes Care 15:27–34, 46
dissemination of professional and
Meichenbaum D, Turk DC: Facilitating
1992 Treatment Adherence. New York, Plenum Press, patient initiatives implementing optimal
37
Shank ML, Del Prato S, DeFronzo RA: 1987 diabetes care. For more information,
Bedtime insulin/daytime glipizide: effective ther- 47
Ewart CK: Social action theory for a public write to: P.O. Box 8262, Parsippany,
apy for sulfonylurea failures in NIDDM. Dia- health psychology. Am Psychol 46:931–946, 1991
betes 44:165–172, 1995 NJ, 07054-8262, or call 877-462-4356.
48
38
Williams GC, Quill TE, Deci EL, Ryan RM:
Yki-Järvinen H, Kauppila M, Kujansuu E, “The facts concerning the recent carnival of
Lahti J, Marjanen T, Niskanen L, Rajala S, Ryysy smoking in Connecticut” and elsewhere. Ann Note of disclosure: Dr. White has served
L, Salo S, Seppälä P, Tulokas T, Viikari J, Kar- Intern Med 115:59–63, 1991 on an advisory board for Aventis; has
jalainen J, Taskinen M-R: Comparison of insulin
regimens in patients with non-insulin-dependent 49
Williams MV, Parker RM, Baker DW, received honoraria for speaking engage-
diabetes mellitus. N Engl J Med 327:1426–1433, Parikh NS, Pitkin K, Coates WC, Nurss JR: Inad- ments from Aventis, Takeda, Omron,
1992 equate functional health literacy among patients
39
at two public health hospitals. JAMA Pfizer, and TheraSense; and has received
Rosenstock J, Schwartz S, Clark C, Park 274:1677–1682, 1995
GD, Donley DW, Edwards MB: Basal insulin research support from Aventis. Dr. Coop-
50
therapy in type 2 diabetes: 28-week comparison Guise BJ, Nelinson DS, Beyond the Pre- pan has received honoraria from Bristol-
of insulin glargine (HOE 901) and NPH insulin. scription: Promoting Patient Compliance, CME
Diabetes Care 24:631–636, 2001 Monograph–Center for Health Education. Octo- Myers Squibb, Pfizer, Aventis, Novartis,
40
ber 30, 1999, p.2–3 Takeda, Eli Lilly, GlaxoSmithKline, and
Rosenstock J, Riddle M: Treatment to target
in type 2 diabetes: consistent risk reduction of Wyeth-Ayerst. Ms. Mulcahy has served
hypoglycaemia with basal insulin glargine as ACKNOWLEDGMENT on advisory boards and received hono-
compared with NPH insulin in insulin-naive
patients on oral agents (Poster). Presented at the raria from Aventis and Novo Nordisk. Dr.
38th annual meeting of the European Association The research for this article was sup- Manko has received honoraria from
for the Study of Diabetes. Budapest, Hungary,
September 1–5, 2002 ported by an unrestricted grant from Pharmacia, Wyeth-Ayerst, and Bristol-
41
Fonseca V, Bell D, Mecca T: Less sympto- Aventis. Myers Squibb. All of these companies
matic hypoglycemia with insulin glargine com- manufacture or market insulin or other
pared to NPH in patients with type 2 diabetes
(Poster). Presented at the 37th annual meeting of John R. White, Jr., PA-C, PharmD, is a pharmaceutical products for the treat-
the European Association for the Study of Dia- professor at Washington State Universi- ment of type 2 diabetes or its complica-
betes. Glasgow, United Kingdom, September
9–13, 2001 ty College of Pharmacy in Spokane. tions.