Content uploaded by Anders Green
Author content
All content in this area was uploaded by Anders Green
Content may be subject to copyright.
OBSERVATIONS
Please, Let My
Doctor Be Obese
Associations between the
characteristics of general
practitioners and their patients
with type 2 diabetes
O
besity is an important cardiovascu-
lar risk factor in type 2 diabetes (1).
Physician characteristics such as
age and sex are related to counseling for
overweight (2). Other physician charac-
teristics may also be related. Patients in-
dicated greater confidence in nonobese
versus obese physicians. Whether this
translates into increased success in obe-
sity management is unknown (3). We
aimed to study associations between the
weight of general practitioners and their
type 2 diabetic patients.
A postal survey was performed
among 36 general practitioners partici-
pating in a shared-care diabetes project in
2000. It contained questions about the
general practitioners’ age, sex, weight,
height, smoking behavior, work experi-
ence, practice population size, and opin-
ion regarding how much influence they
have on a patient’s weight and smoking
cessation. The project’s target population
consisted of type 2 diabetic patients who
were exclusively treated in primary care.
Patients who were cotreated in secondary
care or who were terminally ill or had de-
mentia were excluded. Participating pa-
tients (n ⫽ 1,441) represented 87% of the
target population. Data on patient age,
sex, diabetes duration, BMI, and smoking
were collected by nurses. We performed a
regression analysis with the mean BMI of
patients as the dependent variable and the
above variables as predictors.
The survey response rate was 100%.
Most general practitioners were non-
smoking (94%) men (83%) with a mean
(⫾SD) age of 51.1 ⫾ 7.0 years and work
experience of 18.1 ⫾ 8.9 years. The gen-
eral practitioners’ mean BMI was 24.4 ⫾
3.5 kg/m
2
(BMI ⬍25 in 72%). For pa-
tients (per general practitioner), mean age
was 68.2 ⫾ 2.9 years, diabetes duration
7.2 ⫾ 1.3 years, and BMI 29.3 ⫾ 0.85
kg/m
2
;44⫾ 9% were men. The mean
BMI of patients showed the strongest cor-
relation with the BMI of general practitio-
ners: ⫺0.40 (partial correlation) and
unstandardized coefficient B of ⫺1.05
(95% CI ⫺0.197 to ⫺0.013). The opti-
mal model (P ⫽ 0.07) had a multiple cor-
relation (R) of 0.56, and explained
variance (R
2
) was 31% (adjusted 17%).
We found a negative correlation be-
tween the BMI of type 2 diabetic patients
and their general practitioners. Obese doc-
tors had lean patients. Our study is limited
by the cross-sectional design; associations
were found, not causal relations.
Hash et al. (3) showed that patients
indicated greater confidence in nonobese
physicians. However, we found no trans-
lation into increased success in obesity
management. On the contrary, patients of
nonobese general practitioners had a
higher BMI compared with patients of
obese general practitioners. A discernable
negative impact of patient weight on phy-
sician behavior was shown earlier (4).
Could it be that nonobese general practi-
tioners lack motivation to treat over-
weight patients? Is it time for general
practitioners to search our own hearts?
LIELITH J. UBINK-VELTMAAT,
MD
1,2
ROGER A.J.M. DAMOISEAUX,
PHD
3
ROEL O. RISCHEN,
MD
4
KLAAS H. GROENIER
1
From the
1
Department of Family Practice, Univer-
sity of Groningen, Groningen, the Netherlands; the
2
Isala Clinics, Weezenlanden, Diabetes Outpatient
Clinic, Zwolle, the Netherlands; the
3
General prac-
tice ’de Hof van Blom’, Hattem, the Netherlands; and
the
4
General practice ‘t Veen, Hattem, the Nether-
lands.
Address correspondence to Lielith J. Ubink-
Veltmaat, MD, Vechtdijk 7, 7721 AG Dalfsen, Neth-
erlands. E-mail: l.j.veltmaat@isala.nl.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Abraham WT: Preventing cardiovascular
events in patients with diabetes mellitus.
Am J Med 116 (Suppl. 5A):39S– 46S, 2004
2. Heywood A, Firman D, Sanson-Fisher R,
Mudge P, Ring I: Correlates of physician
counseling associated with obesity and
smoking. Prev Med 25:268 –276, 1996
3. Hash RB, Munna RK, Vogel RL, Bason JJ:
Does physician weight affect perception of
health advice? Prev Med 36:41– 44, 2003
4. Hebl MR, Xu J, Mason MF: Weighing the
care: patients’ perceptions of physician care
as a function of gender and weight. Int J Obes
Relat Metab Disord 27:269 –275, 2003
Prognostic Factors of
Sudden
Sensorineural
Hearing Loss in
Diabetic Patients
S
udden sensorineural hearing loss
(SSNHL) is defined as the sudden
onset of unilateral sensorineural
hearing loss of ⱖ30 dB over at least three
contiguous audiometric frequencies (1).
Diabetes is a risk factor of SSNHL, possi-
bly due to microangiopathy (2). Cur-
rently, the clinical studies of SSNHL
rarely focus on diabetic patients. The cor-
relations between biochemical data and
hearing outcomes in SSNHL are seldom
analyzed. We analyzed the characteristics
of SSNHL in diabetic patients and tried to
elucidate the possible prognostic factors.
Medical charts of diabetic patients ad-
mitted for SSNHL from 1984 to 2003
were retrospectively reviewed. A total of
67 patients (38 men and 29 women) with
a mean age of 60.1 ⫾ 11.9 years were
included. Their mean duration of diabe-
tes was 7.5 ⫾ 7.7 years and mean HbA
1c
9.9 ⫾ 2.9%. Measuring by initial and fol-
low-up pure-tone audiometries, im-
provement in hearing was divided into
favorable outcome (FO) (hearing re-
turned to at least the same level as the
opposite ear or with improvement ⱖ30
dB) or poor outcome (PO) (hearing im-
provement ⬍30 dB) groups.
All of the patients denied previous ep-
isodes of SSNHL. After treatment, 23 pa-
tients (34.3%) had a favorable hearing
outcome, while 44 patients (65.7%) had a
poor hearing outcome. The age and sex
distributions were not statistically differ-
ent in the FO and PO groups. Duration of
diabetes, medication type, and control
levels of diabetes, hypertension, or hyper-
lipidemia were also not correlated with
hearing outcome. Factors such as the type
of initial audiogram, the period of SSNHL
onset to therapy, and the presenting
symptoms (vertigo, tinnitus, aural full-
ness, etc.) had no significant impact on
the hearing prognosis.
SSNHL occurring in the summer had
better a hearing outcome than that occur-
ring in the nonsummer seasons (adjusted
LETTERS
2560 DIABETES CARE, VOLUME 27, NUMBER 10, OCTOBER 2004
for sex and age, odds ratio [OR] for the PO
group 0.093 [95% CI 0.02–0.428], P ⫽
0.002). The glucocorticoid regimen had a
dose-dependent effect on the outcome
(
2
for trend test 5.4159, P ⫽ 0.020). Us-
ing no steroid treatment as a reference
group, the adjusted OR for the PO group
was 0.762 (0.231–2.516, P ⫽ 0.656) and
0.355 (0.157–0.802, P ⫽ 0.013) for the
low- and high-steroid groups (pred-
nisolone 1 mg 䡠 kg
⫺1
䡠 day
⫺1
for at least 7
days), respectively. When we compared
the hearing outcomes using binary co-
variates, such as using or not using ste-
roids, the difference was not significant.
Vitamin B treatment had a significantly
adverse effect for hearing recovery (ad-
justed OR for the PO group was 3.676
[1.081–12.5], P ⫽ 0.037). The adjusted
OR for the PO group for every gram per
liter increment of serum albumin was
0.659 (0.471–0.922, P ⫽ 0.015). The in-
crement in serum lactate dehydrogenase
was borderline significantly correlated
with poor hearing outcome (P ⫽ 0.050),
but the effect was not significant after ad-
justment for sex and age. In multivariate
logistic regression, only the increment in
albumin remained independently associ-
ated with the hearing outcome.
In conclusion, our data reveal that on-
set in summer and higher serum albumin
concentrations are favorable prognostic
factors of SSNHL in diabetic patients. A
vitamin B regimen may lead to a poor
hearing outcome. We suggest a high-dose
glucocorticoid treatment for SSNHL in di-
abetic patients. Further prospective stud-
ies are needed to confirm the true effects
of these prognostic factors.
S
HUEN-FU WENG,
MD
1
YUH-SHYANG CHEN,
MD
2
TIEN-CHEN LIU,
MD, PHD
2
CHUAN-JEN HSU,
MD, PHD
2
FEN-YU TSENG,
MD, MPH
1
From the
1
Department of Internal Medicine, Na-
tional Taiwan University Hospital, National Taiwan
University College of Medicine, Taipei, Taiwan; and
the
2
Department of Otolaryngology, National Tai-
wan University Hospital, National Taiwan Univer-
sity College of Medicine, Taipei, Taiwan.
Address correspondence to Dr. Fen-Yu Tseng,
MD, MPH, Department of Internal Medicine, Na-
tional Taiwan University Hospital, No. 7, Chung-
Shan South Road, Taipei, Taiwan. E-mail:
fytseng@ha.mc.ntu.edu.tw.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Gate GA: Sudden sensorineural hearing
loss. In The Ear: Comprehensive Otology.
Canalis RF, Lambert PR, Eds. Philadel-
phia, Lippincott Williams & Wilkins,
2000, p. 523–536
2. Shikowitz MJ: Sudden sensorineural hear-
ing loss. Med Clin North Am 75:1239 –
1250, 1991
Gastric Bypass
Surgery in Severely
Obese Type 1
Diabetic Patients
O
besity is common in type 2 diabetes
but is rarely seen in autoimmune
type 1 diabetes (1). However, as
shown in the Diabetes Control and Com-
plications Trial cohort (2,3), a subset of
type 1 diabetic patients are overweight,
and it has been suggested that intensive
insulin therapy may unmask the central
obesity or metabolic syndrome in suscep-
tible individuals.
Gastric bypass is considered a gold
standard procedure in the surgical treat-
ment of severe obesity, with ⬎50% of ex-
cess weight loss occurring within 2 years
after the surgery. The weight reduction is
maintained over the following years in a
vast majority of patients (4). Numerous
reports have been published on the ben-
eficial effects of obesity surgery on glucose
control and metabolic disorders in im-
paired glucose tolerant and type 2 dia-
betic subjects, with an impressive 70 –
90% of diabetic patients remaining
euglycemic without diabetes medications
several years after the surgery (5,6). This
significant impact of gastric bypass sur-
gery on glucose control is thought to not
only result from significant weight loss
but also from the exclusion of hormonally
active foregut (7,8). It is worth noting that
in recent years, gastric bypass has become
an increasingly safe procedure (9).
We performed gastric bypass opera-
tions in two young women with autoim-
mune type 1 diabetes associated with
severe obesity. Both patients had been
obese for many years and were unsuccess-
ful in their previous multiple attempts at
losing weight. Having been informed
about the benefits and risks of the sur-
gery, they gave their consent to have bari-
atric surgery performed.
The first patient was a 23-year-old
woman with type 1 diabetes since age 15
years, which was poorly controlled
(HbA
1c
9.5%) and treated with a basal bo-
lus insulin regimen (daily insulin dose 68
IU). Her body weight was 113.5 kg,
height 171 cm, and BMI 38.8 kg/m
2
. She
was mildly hypertensive and had hyper-
cholesterolemia treated with a low-
cholesterol diet. The patient underwent
Roux-en-Y gastric bypass surgery, as de-
scribed before (9), in August 2000. The
surgery and postoperative period were
uneventful. Six months after the surgery,
her body weight was 84.3 kg (BMI 28.8
kg/m
2
; 25.7% reduction). After 1 year, it
was 77.7 kg (BMI 26.6 kg/m
2
; 31.5% re-
duction). However, 2 years after the sur-
gery, her weight increased to 85.0 kg (BMI
29.1 kg/m
2
) and has remained stable ever
since. Eventually, the patient lost 28.5 kg,
i.e., 25% of her initial body weight. Her glu-
cose control improved markedly (HbA
1c
5.7%), with daily insulin dose reduced to
45 IU, and her blood pressure and plasma
lipids returned to normal.
The second patient was a 28-year-old
woman who was diagnosed with type 1
diabetes when she was 5 years old. She
had been obese since childhood, and her
body weight, as in the first patient, had
been steadily increasing since the diagno-
sis of diabetes. Upon presentation, her
weight was 126 kg, height 165 cm, and
BMI 46.3 kg/m
2
. She had also been diag-
nosed with hypertension, hypertriglycer-
idemia, and, more recently, Sjo¨ gren’s
Syndrome. She was also treated with in-
tensive insulin therapy (120 IU/day);
however, her blood glucose control had
been extremely poor for ⬎10 years
(HbA
1c
during this period was between
10.4 and 11.8%). Her other medications
included enalapril, bisoprolol, and feno-
fibrate. She had Roux-en-Y gastric bypass
successfully performed in November
2002. Her surgery was complicated by bi-
lateral pneumonia. Ten months after the
operation, her weight was 84 kg (BMI
30.1 kg/m
2
; 33% reduction) and, similar
to the previous patient, increased slightly
18 months after the surgery to 89.5 kg
(BMI 32.9 kg/m
2
; 29% reduction). Her
metabolic control has improved substan-
tially (HbA
1c
7.3%), and her daily insulin
dose is now 70 IU. In addition, she no
longer requires any treatment for hyper-
tension or lipid disorders and her blood
pressure and plasma triglycerides values
are now normal.
Letters
DIABETES CARE, VOLUME 27, NUMBER 10, OCTOBER 2004 2561
Our report is, to the best of our
knowledge, the first one describing the
effects of obesity surgery in type 1 diabe-
tes. In our opinion, gastric bypass surgery,
which is being performed increasingly of-
ten (⬃100,000 operations in the U.S. an-
nually [10]) in obese individuals, also
with type 2 diabetes (4–8), is a feasible,
safe, and effective method of weight re-
duction in young type 1 diabetic patients
with severe obesity and comorbidities
leading to metabolic syndrome (e.g., hy-
pertension, hyperlipidemia) (11). In our
patients, surgery-induced weight loss was
also associated with a decrease in insulin
requirement per kilogram of body weight
(0.60 to 0.53 IU/kg in the first patient and
from 0.95 to 0.83 IU/kg in the second
patient). This observation may suggest
the presence of clinically significant in-
sulin resistance in severely obese type 1
diabetic subjects (12), which was subse-
quently reduced once weight loss occurred.
Importantly, neither of the patients had
any significant hypoglycemic episodes af-
ter the surgery, despite considerable re-
duction in HbA
1c
level and apparent
increase in insulin sensitivity.
In conclusion, gastric bypass surgery
not only leads to a significant and main-
tained weight loss in type 1 diabetic pa-
tients, but also results in remarkable
improvement in metabolic control (abso-
lute reduction in HbA
1c
of 3–4%) and con-
comitant disorders. Interestingly, the need
for constant intensive insulin therapy in
these patients had no detrimental influence
on weight loss as an effect of obesity surgery.
Both patients lost 50 –60% of their exces-
sive body weight during the follow-up pe-
riod, which is also the rate reported in
nondiabetic subjects (4,5,7).
L
ESZEK CZUPRYNIAK,
MD, PHD
1
JANUSZ STRZELCZYK,
MD, PHD
2
KATARZYNA CYPRYK,
MD, PHD
3
MACIEJ PAWLOWSKI,
MD
1
DARIUSZ SZYMANSKI,
MD, PHD
2
ANDRZEJ LEWINSKI,
MD, PHD
3
JERZY LOBA,
MD, PHD
1
From the
1
Department of Diabetology and Metabolic
Diseases, Medical University of Lodz, Lodz, Poland;
the
2
Department of General and Transplant Surgery,
Medical University of Lodz, Lodz, Poland; and the
3
De-
partment of Endocrinology and Isotope Therapy, Pol-
ish Mother’s Memorial Hospital Research Institute,
Medical University of Lodz, Lodz, Poland.
Address correspondence to Leszek Czupryniak,
MD, PHD, Department of Diabetology and Meta-
bolic Diseases, Barlicki University Hospital, No. 1,
Ul, Kopcinskiego 22, 90-153 Lodz, Poland. E-mail:
bigosik@poczta.onet.pl.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Greenfield JR, Samaras K, Campbell LV,
Chisholm DJ: Type 1 diabetes is not associ-
ated with increased central abdominal obe-
sity (Letter). Diabetes Care 26:2703, 2003
2. Sibley SD, Palmer JP, Hirsch IB, Brunzell JD:
Visceral obesity, hepatic lipase activity, and
dyslipidemia in type 1 diabetes. J Clin Endo-
crinol Metab 88:3379 –3384, 2003
3. Kabadi UM, Vora A, Kabadi M: Hyperin-
sulinemia and central adiposity: influence
of chronic insulin therapy in type 1 dia-
betes (Letter). Diabetes Care 23:1024 –
1025, 2000
4. Fobi MA, Lee H, Holness R, Cabinda D:
Gastric bypass operation for obesity.
World J Surg 22:925–935, 1998
5. Pories WJ, Swanson MS, MacDonald KG
Jr, Long SB, Morris P, Brown BM, Bakarat
HA, deRamon RA, Israel G, Dolezal JM,
Dohm GL: Who would have thought it?
An operation proves to be the most effec-
tive therapy for adult-onset diabetes mel-
litus. Ann Surg 222:339 –350, 1995
6. Smith SC, Edwards CB, Goodman GN:
Changes in diabetic management after
Roux-en-Y gastric bypass. Obes Surg 6:
345–348, 1996
7. Eisenberg D, Bell RL: The impact of bari-
atric surgery on severely obese patients
with diabetes. Diabetes Spectrum 16:240 –
245, 2003
8. Clements RH, Gonzalez QH, Long CI,
Wittert G, Laws HL: Hormonal changes
after Roux-en-Y gastric bypass for morbid
obesity and the control of type-II diabetes
mellitus. Am Surg 70:1–5, 2004
9. Strzelczyk J, Czupryniak L, Loba J, Wa-
siak J: The use of polypropylene mesh in
midline incision closure following gastric
by-pass surgery reduces the risk of post-
operative hernia. Langenbecks Arch Surg
387:294–297, 2002
10. Brody JE: Surgical alternatives for the
truly overwieght. New York Times,31De-
cember 2002: Sect. F, p. 7
11. Expert Panel on Detection, Evaluation,
and Treatment of High Blood Cholesterol
in Adults: Executive summary of the
Third Report of the National Cholesterol
Education Programme (NCAP) Expert
Panel on Detection, Evaluation, and
Treatment of High Blood Cholesterol in
Adults (Adult Treatment Panel III).
JAMA 285:2486 –2497, 2001
12. Greenfield JR, Samaras K, Chisholm DJ:
Insulin resistance, intra-abdominal fat,
cardiovascular risk factors, and andro-
gens in healthy young women with type 1
diabetes mellitus. J Clin Endocrinol Metab
87:1036–1040, 2002
Dysadipocytokinemia
in Werner Syndrome
and Its Recovery by
Treatment With
Pioglitazone
W
erner syndrome (WS) (Men-
delian Inheritance in Man no.
277700) is an autosomal reces-
sive disorder known for progeroid pheno-
types including graying and loss of hair,
juvenile cataracts, insulin-resistant diabe-
tes, skin atrophy, premature atheroscle-
rosis, and cancer (1). Mutations in WRN,
a RECQ family DNA/RNA helicase gene,
have been identified to cause this disease.
The mechanism for insulin resistance in
WS remains to be elucidated.
Adipocytes secrete a number of hor-
mones (or adipocytokines), such as tumor
necrosis factor-␣ (TNF-␣), leptin, adi-
ponectin, and resistin, thereby regulating
insulin sensitivity (2). WS patients typically
show the lipoatrophic skinny extremities
with an obese trunk (1). The accumulated
intra-abdominal visceral fat (3) suggests an
altered production of adipocytokines.
To investigate the role of adipocyto-
kines in the pathophysiology of WS, we
examined the serum levels of TNF-␣ and
adiponectin in WS. Sera sampled from 24
WS patients (14 men and 10 women; 16
with and 8 without diabetes) proven to be
homozygous for WRN mutations, and 40
age- and sex-matched normoglycemic
healthy volunteers were assayed after in-
formed consent was obtained. Age (43 ⫾
8.1 vs. 41.6 ⫾ 7.5 years) and BMI (19.4 ⫾
1.9 vs. 18.8 ⫾ 2.0 kg/m
2
) were similar for
diabetic and nondiabetic WS patients.
The serum level of TNF-␣, a mediator
of insulin resistance, was significantly el-
evated in WS regardless of having diabe-
tes (21.8 ⫾ 8.7 pg/ml, P ⬍ 0.0001 by
Mann-Whitney test) or not having diabe-
tes (14.0 ⫾ 3.2 pg/ml, P ⫽ 0.002) com-
pared with the healthy control group
(6.05 ⫾ 3.0 pg/ml). Adiponectin levels in
diabetic WS patients (3.1 ⫾ 2.9 g/ml)
was significantly lower than in nondia-
betic WS patients (11.6 ⫾ 9.2 g/ml, P ⫽
0.006) or control subjects (14.4 ⫾ 8.8
g/ml, P ⬍ 0.0001). The growing evi-
dence indicates insulin sensitizing as well
as antiatherogenic actions of adiponectin
and the association of decreased serum
adiponectin with insulin resistance, obe-
Letters
2562 DIABETES CARE, VOLUME 27, NUMBER 10, OCTOBER 2004
sity, and type 2 diabetes (2,4). Although
WS patients are usually not obese by the
definition of BMI, the visceral fat specifi-
cally accumulated by an unknown mech-
anism (3) might cause high TNF-␣ and
low adiponectin levels, characteristics
similar to morbid obesity.
We recently reported the successful
improvement of glycemic control and in-
sulin sensitivity by pioglitazone in dia-
betic WS patients (5). Therefore, we next
assessed adipocytokines before and after
16 weeks on pioglitazone (15 mg/day) in
three diabetic WS patients. The treatment
significantly elevated adiponectin levels
from 2.57 ⫾ 1.36 to 7.07 ⫾ 2.48 g/ml
(P ⫽ 0.03 by paired t test). TNF-␣ and
HbA
1c
levels showed a tendency to de-
cline from 16.1 ⫾ 4.75 to 3.53 ⫾ 0.58
pg/ml (P ⫽ 0.052) and from 7.7 ⫾ 0.6 to
6.4 ⫾ 0.5% (P ⫽ 0.17), respectively.
To our knowledge, this is the first
study to examine serum adipocytokine
levels in WS patients. Reduced insulin
sensitivity with increased visceral adipos-
ity is the hallmark of both WS and normal
aging. Because pioglitazone achieved im-
provement of glycemic control as well as
correction of adiponectin and TNF-␣ lev-
els, these cytokines are likely to be at least
in part responsible for insulin resistance
in WS. Adipocyte function may be a key
element linking WRN mutation and the
metabolic abnormalities observed in WS. It
is also of our interest to know whether pio-
glitazone and other thiazolidinediones can
prevent or delay the onset of diabetes in WS
by modulating adipocytokines. Our present
findings raise a possibility that pioglitazone
could extend the lifespan of WS patients by
improving metabolism and preventing
early cardiovascular death.
KOUTARO YOKOTE,
MD
1,2
KAZUO HARA,
MD
3
SEIJIRO MORI,
MD
1,2
TAKASHI KADOWAKI,
MD
3
YASUSHI SAITO,
MD
1,2
MAKOTO GOTO,
MD
4,5
From the
1
Division of Endocrinology and Metabolism,
Department of Internal Medicine, Chiba University
Hospital, Chiba City, Japan; the
2
Department of Clin-
ical Cell Biology, Chiba University Graduate School of
Medicine, Chiba City, Japan; the
3
Department of Met-
abolic Diseases, Graduate School of Medicine, Univer-
sity of Tokyo, Tokyo, Japan; the
4
Department of
Rheumatology, Tokyo Metropolitan Otsuka Hospital,
Tokyo, Japan; and the
5
Institute of Bioengineering,
Toin Yokohama University, Yokohama, Japan.
Address correspondence to Koutaro Yokote, MD,
Division of Endocrinology and Metabolism, Depart-
ment of Internal Medicine, Chiba University Hospi-
tal, 1-8-1 Inohana, Chuo-ku, Chiba 260-8670,
Japan. E-mail: kyokote-cib@umin.ac.jp.
© 2004 by the American Diabetes Association.
Acknowledgments— We thank Dr. R.W.
Miller (National Cancer Institute) for helpful
comments. This work was supported in part
by the grants from the Japanese Ministry of
Education, Science and Sports and the Minis-
try of Health, Labor and Welfare.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Goto M: Clinical characteristics of Werner
syndrome and other premature aging syn-
dromes: pattern of aging in progeroid syn-
dromes. In From Premature Gray Hair to
Helicase-Werner Syndrome: Implication for
Aging and Cancer. Goto M, Miller RW, Eds.
Basel, Karger, 2001, p. 27–39
2. Yamauchi T, Kamon J, Waki H, Terauchi
Y, Kubota N, Hara K, Mori Y, Ide T, Mu-
rakami K, Tsuboyama-Kasaoka N, Ezaki
O, Akanuma Y, Gavrilova O, Vinson C,
Reitman ML, Kagechika H, Shudo K,
Yoda M, Nakano Y, Tobe K, Nagai R,
Kimura S, Tomita M, Froguel P, Kadow-
aki T: The fat-derived hormone adiponec-
tin reverses insulin resistance associated
with both lipoatrophy and obesity. Nat
Med 7:941–946, 2001
3. Mori S, Murano S, Yokote K, Takemoto M,
Asaumi S, Take A, Saito Y: Enhanced intra-
abdominal visceral fat accumulation in pa-
tients with Werner’s syndrome. Int J Obes
Relat Metab Disord 25:292–295, 2001
4. Daimon M, Oizumi T, Saitoh T, Yamagu-
chi H, Hirata A, Ohnuma H, Igarashi M,
Eguchi H, Manaka H, Tominaga M: De-
creased serum levels of adiponectin are a
risk factor for the progression to type 2
diabetes in the Japanese population. Dia-
betes Care 26:2015–2020, 2003
5. Yokote K, Honjo S, Kobayashi K, Fujimoto
M, Kawamura H, Mori S, Saito Y: Metabolic
improvement and abdominal fat redistri-
bution in Werner syndrome by pioglita-
zone (Letter). 52:1582–1583, 2004
Effect of ␣-Linolenic
Acid–Containing
Linseed Oil on
Coagulation in Type
2 Diabetes
B
lood coagulation in diabetes is
known to be increased (1,2). Be-
cause levels of n-3 and n-6 polyun-
saturated fatty acids (PUFAs) influence
the parameters of blood coagulation, the
aim of this study was to determine the
effects of n-3 PUFA supplementation on
coagulation and fibrinolytic factors in
type 2 diabetic subjects. While it is not
clear what the appropriate intake ratio of
n-6 to n-3 PUFAs should be for diabetic
subjects, it is known that the dietary in-
take ratio of n-6 to n-3 PUFAs is roughly
4:1 in Japanese subjects (3).
Ten subjects (six women and four
men, average age 59.6 years) with type 2
diabetes participated in this study as in-
patients. Their average BMI and HbA
1c
values were 20.9 ⫾ 3.8 kg/m
2
and 10.8 ⫾
1.1%, respectively. Their daily energy in-
take during the course of the study was
1,490 ⫾ 166 kcal. After 2 weeks on the
control diet, our subjects were placed on a
diet in which 5 g linseed oil was added (in
salads, miso soup, etc., without heating)
in exchange for 5 g cooking oil. The ratio
of PUFAs to saturated fatty acids in the
subjects’ prestudy and study diets were
1.2 and 1.6, respectively, while the ratios
of n-6 to n-3 PUFAs in their prestudy and
study diets were 3.6 and 1.5, respectively.
Blood samples were collected before and
14 days after initiation of the study. Plas-
min ␣2-plasmin inhibitor complex (PPI)
level and plasminogen activator inhibi-
tor-1 (PAI-1) activity in plasma was mea-
sured using a latex photometric
immunoassay, while thrombin anti-
thrombin III complex (TAT) level was
measured using an enzyme-linked immu-
noassay. Differences in these parameters
obtained at the start and end of the study
were analyzed using a paired t test; values
were considered to be significant if the P
value was ⬍0.05. Values are expressed as
the mean ⫾ SD.
After 2 weeks on a linseed oil–
supplemented diet, PPI level, PAI-1 activ-
ity, and TAT level fell significantly
(0.72 ⫾ 0.19 vs. 0.47 ⫾ 0.14 g/ml, P ⫽
0.0009; 73.3 ⫾ 37.5 vs. 51.6 ⫾ 25.0 ng/
ml, P ⫽ 0.02; and 9.6 ⫾ 9.1 vs. 2.5 ⫾ 1.1
ng/ml, P ⫽ 0.04; respectively).
Boberg et al. (4) reported that PAI-1
activity was increased in type 2 diabetic
subjects after supplementation of their
diet with 10 g eicosapentaenoic acid.
Kelly et al. (5) reported that a diet con-
taining flaxseed oil (60% ␣-linolenic acid)
did not alter indexes of blood coagula-
tion, i.e., bleeding time, prothrombin
time, and partial prothrombin time. Chan
et al. (6) showed that altering the dietary
n-6–to–n-3 PUFA ratio had no effect on
Letters
DIABETES CARE, VOLUME 27, NUMBER 10, OCTOBER 2004 2563
bleeding time. Our results, on the other
hand, showed that a diet-induced reduc-
tion in the n-6 –to–n-3 PUFA ratio af-
fected parameters of blood coagulation
and fibrinolysis. Finally, Freese et al. (7)
reported that supplemental ␣-linolenic
acid from vegetable oil and eicosapen-
taenoic and docosahexaenic acids from a
marine source had similar effects on he-
mostatic factors. In conclusion, our re-
sults showed that PPI level, PAI-1 activity,
and TAT level were significantly reduced
in type 2 diabetic subjects that had their
n-6–to–n-3 PUFA ratio lowered by di-
etary means.
N
ORIAKI TOHGI,
MD
From the Department of Endocrinology and Metab-
olism, Kurume University School of Medicine, Ku-
rume, Japan.
Address correspondence to Noriaki Tohgi, MD,
Midori Hospital 359-1 Yuu-chou Kuga-gun,
Yamaguchi 740-1404, Japan. E-mail: tohgi@
nanwakai.jp.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Szirtes M: Platelet aggregation in diabetes
mellitus. Adv Cardiol 4:179 –186, 1970
2. Jones DB, Haitas B, Bown EG, Carter RD,
Barker K, Jelfs R, Turner C, Mann JI, Pres-
cott RJ: Platelet aggregation in noninsu-
lin-dependent diabetes is associated with
platelet fatty acids. Diabet Med 3:52–55,
1986
3. Hirahara F: J Lipid Nutr 4:73– 82, 1995
4. Boberg M, Pollare T, Siegbahn A, Vessby
B: Supplementation with n-3 fatty acids
reduces triglycerides but increases PAI-1
in non-insulin-dependent diabetes melli-
tus. Eur J Clin I 22:645– 650, 1992
5. Kelly DS, Nelson JG, Love JE, Branch LB,
Taylor PC, Schmidt PC, Mackey BE, Ia-
cono JM: Dietary alpha-linolenic acid al-
ters tissue fatty acid composition, but not
blood lipids, lipoproteins or coagulation
status in humans. Lipids 28:533–537,
1993
6. Chan JK, Mcdonald BE, Gerrard JM,
Bruce VM, Weaver BJ, Holub BJ: Effect of
dietary alpha-linolenic acid and its ratio to
linoleic acid on platelet and plasma fatty
acids and thrombogenesis. Lipids 28:811–
817, 1993
7. Freese R, Mutanen M: Alpha-linolenic
acid and marine long-chain n-3 fatty acids
differ only slightly in their effects on he-
mostatic factors in healthy subjects. Am J
Clin Nutr 66:591–598, 1997
COMMENTS AND
RESPONSES
Use of Arterial
Transfer Functions
for the Derivation of
Central Aortic
Waveform
Characteristics in
Subjects With Type 2
Diabetes and
Cardiovascular
Disease
Response to Hope et al.
W
e read with interest the recent ar-
ticle by Hope et al. (1) concern-
ing the noninvasive estimation of
central aortic pressure waveforms in sub-
jects with type 2 diabetes. We entirely agree
with their view that aortic rather than bra-
chial artery pressure is likely to be of
greater prognostic value, as are indexes
such as aortic augmentation index. We
would also agree that the data presented
in the article clearly indicate that there
was a substantial difference between inva-
sively measured and derived aortic sys-
tolic pressure. However, the authors’
conclusions seem overstated and may be
misinterpreted by nonspecialist readers.
The transfer function used by Hope et al.
is actually their own and not that which is
used in the “commercial devices” that
they refer to in their introduction. Indeed,
they offer no data to suggest that other
transfer functions are unreliable in sub-
jects with diabetes. Therefore, all that
Hope et al. can actually conclude is that
their own generalized transfer function is
unreliable in subjects with diabetes. Thus,
other investigators may be better off using
“commercial devices” rather than the
seemingly discredited transfer function of
Hope et al.
I
AN B. WILKINSON,
DM, MA, MRCP
CARMEL MCENIERY,
BSC, PHD
From the Clinical Pharmacology Unit, Adden-
brooke’s Hospital, Cambridge, U.K.
Address correspondence to Dr. Ian B. Wilkinson,
Addenbrooke’s Hospital, Hills Road, CB2 2QQ
Cambridge, U.K. E-mail: ibw20@cam.ac.uk.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Hope SA, Meredith IT, Tay DB, Cameron
JD: Use of arterial transfer functions for
the derivation of central aortic waveform
characteristics in subjects with type 2 di-
abetes and cardiovascular disease. Diabe-
tes Care 27:746 –751, 2004
Use of Arterial
Transfer Functions
for the Derivation of
Central Aortic
Waveform
Characteristics in
Subjects With Type 2
Diabetes and
Cardiovascular
Disease
Response to Hope et al.
I
n a recent editorial, Mather and Lewanc-
zuk (1) describe the potential value of a
generalized transfer function to synthe-
size the ascending aortic pressure wave
from the radial pressure waveform, as re-
corded indirectly by applanation tonome-
try. The Food and Drug Administration
(FDA) had previously accepted validity un-
der different conditions of a generalized
transfer function used in a commercial de-
vice: “The SphygmoCor SCOR-Px can cal-
culate the calibrated ascending aortic
pressure waveform using the radial artery
pressure waveform recorded noninvasively
from a radial artery tonometer and a bra-
chial cuff measurement” (2). On the basis of
an accompanying article by Hope et al. (3)
in Diabetes Care, editorialists expressed con-
cern regarding whether such an approach
was applicable in diabetic subjects.
There are serious flaws in the report
of Hope et al. (3) Their transfer function is
different from that accepted by the FDA
(2) and had been determined from use of
conventional fluid-filled manometer sys-
tems of an unknown frequency response.
From their original dataset (4), they had
opined that different transfer functions
were necessary to characterize the vascu-
Letters
2564 DIABETES CARE, VOLUME 27, NUMBER 10, OCTOBER 2004
lar properties of men and women and,
subsequently from the same dataset (5),
that their transfer function was unreliable
for calculating the aortic augmentation
index and other features of the synthe-
sized aortic pressure wave. In their article,
Hope et al. developed a “diabetes-
specific” transfer function and showed
that this could accurately generate the
aortic from the radial pressure waveform.
The problem with such an approach is
that the same data used to validate were
also used to generate the transfer function.
Other investigators (6) have found
the approach used by Hope et al. to be
problematic, but these issues are not
taken into consideration. Other problems
are also apparent. The “representative”
(sic) synthesized aortic and measured
aortic pressure values in their Fig. 1B
show no evidence of the substantial (8
mmHg) diastolic pressure offset noted in
Table 2. The method used for calibration
(assuming mean and diastolic pressure
are equal [4,7]) cannot give a diastolic
pressure offset of 8 mmHg. In the system
approved by the FDA, the difference was
⬍1 mmHg (8). There are serious incon-
sistencies with other data published by
the same authors (9); the time to inflec-
tion (144 –163 ms) (their Table 2) is far
higher than that described for the carotid
artery (mean 81 ms), while the aortic aug-
mentation index is much lower (13.1–
21.3%) rather than higher than in the
carotid (34 ⫾ 13%).
Hope et al. (3) concluded that the use
of their nondiabetic transfer function was
inaccurate in generating aortic systolic
and diastolic pressures in diabetic sub-
jects. However, when the nondiabetic
transfer function was used to calculate
pulse pressure, as in the system approved
by the FDA (2) (without the unexplained
diastolic pressure offset), it was actually
more accurate in diabetic (mean error 2
mmHg) than in nondiabetic (mean error
6 mmHg in Table 2) subjects.
These points cast doubt over previous
publications from this group (4 – 6). Their
transfer function differs markedly from
that described by others (6) and those ac-
cepted by the FDA at frequencies above 3
Hz, which determines features of the
pressure waveforms such as the aortic
augmentation index (7).
The validity of the FDA-approved
system is supported by studies of changes
in heart rate and rhythm (8) and in arterial
pressure during Valsalva maneuver (7)
and with vasodilator challenge (8). It is
also endorsed by modeling studies that
have simulated changes in vascular prop-
erties in diabetic subjects (10). We believe
that editorialists and readers of Diabetes
Care can be reassured that no “diabetes-
specific” (3) or “gender-specific” (4)
transfer function is required to generate
aortic pressure from the radial pressure
waveform. An appropriately validated
and approved generalized transfer func-
tion, however, is required.
ALBERTO P. AVOLIO,
PHD
1
JOHN R. COCKCROFT,
BSC, MB
2
MICHAEL F. O’ROURKE,
MD, DSC
3
From the
1
Graduate School of Biomedical Engineer-
ing, University of New South Wales, Sydney, Aus-
tralia; the
2
Wales Heart Research Unit, College of
Medicine, University Hospital, University of Wales,
Cardiff, U.K.; and the
3
St. Vincent’s Clinic, Univer-
sity of New South Wales, Sydney, Australia.
Address correspondence to Dr. Michael F.
O’Rourke, Suite 810, St. Vincent’s Clinic, 438 Vic-
toria St., Darlinghurst, New South Wales 2010, Aus-
tralia. E-mail: m.orourke@unsw.edu.au.
A.P.A. has received honoraria from, is a consul-
tant for, and is a member of the advisory panel of
Atcor Medical. M.F.O. is a founder and director of
Atcor Medical and has received consulting fees from
Atcor Medical, Philips, Pfizer, Lilly, and Servier.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Mather K, Lewanczuk R: Measurement of
arterial stiffness in diabetes: a cautionary
tale (Editorial). Diabetes Care 831– 833,
2004
2. Food and Drug Administration: 510K
Summary for SphygmoCor Px Regulation
No. 21CFR870,1110 (K012487). Rockville,
MD, Food and Drug Administration, De-
partment of Health and Human Services,
2002
3. Hope SA, Tay DB, Meredith IT, Cameron
JD: Use of arterial transfer functions for
the derivation of central aortic waveform
characteristics in subjects with type 2 di-
abetes and cardiovascular disease. Diabe-
tes Care 27:746 –751, 2004
4. Hope SA, Tay DB, Meredith IT, Cameron
JD: Comparison of generalized and gender-
specific transfer functions for the derivation
of aortic waveforms. Am J Physiol Heart Circ
Physiol 283:H1150 –H1156, 2002
5. Hope SA, Tay DB, Meredith IT, Cameron
JD: Use of arterial transfer functions for
the derivation of aortic waveform charac-
teristics. J Hypertens 21:1299 –1305, 2003
6. O’Rourke MF, Kim M, Adji A, Nichols WW,
Avolio A: Use of arterial transfer function for
the derivation of aortic waveform character-
istics. J Hypertens 22:431– 432, 2004
7. Chen CH, Nevo E, Fetics B, Pak PH, Yin
FC, Maughan WL, Kass DA: Estimation
of central aortic pressure waveform by
mathematical transformation of radial
tonometry pressure: validation of gener-
alized transfer function. Circulation 95:
1827–1836, 1997
8. Pauca A, O’Rourke M, Kon N: Prospective
evaluation of a method for estimating as-
cending aortic pressure from the radial ar-
tery pressure waveform. Hypertension 38:
932–937, 2001
9. Gatzka CD, Cameron JD, Dart AM, Berry
KL, Kingwell BA, Dewar EM, Reid CM,
Jennings GL: Correction of carotid aug-
mentation index for heart rate in elderly
essential hypertensives: ANBP2 Investiga-
tors: Australian Comparative Outcome
Trial of Angiotensin-Converting Enzyme
Inhibitor- and Diuretic-Based Treatment
of Hypertension in the Elderly. Am J Hy-
pertens 14:573–577, 2001
10. Karamanoglu M, Gallagher DE, Avolio
AP, O’Rourke MF: Pressure wave propa-
gation in a multibranched model of the
human upper limb. Am J Physiol 269:
H1363–H1369, 1995
Use of Arterial
Transfer Functions
for the Derivation of
Central Aortic
Waveform
Characteristics in
Subjects With Type 2
Diabetes and
Cardiovascular
Disease
Response to Wilkinson and
McEniery and Avolio, Cockcroft,
and O’Rourke
W
e refer to the letters of Wilkinson
and McEniery (1) and Avolio,
Cockcroft, and O’Rourke (2). In
our article, we make it abundantly clear
that all results, including the transfer
function utilized, were generated from
data obtained in our laboratory (3). As
the above correspondents note, we have
previously reported at length in peer-
reviewed literature how this is imple-
mented by our group (4–9).
There is no published data to support
the contention that any transfer function
performs adequately in subjects with diabe-
Letters
DIABETES CARE, VOLUME 27, NUMBER 10, OCTOBER 2004 2565
tes (10 –13). This issue has not been ad-
dressed in the literature by enthusiasts of
the technique, yet it is of crucial importance
if the technique is to be used in clinical
practice.
Wilkinson and McEniery (1) make
the entirely unsupported suggestion that,
on the basis of our results, “investigators
may be better off using ‘commercial de-
vices.’” We note that both groups of corre-
spondents are prolific users of a particular
commercial device and therefore presum-
ably consider themselves “better off” (14–
24). We would caution them and others
against this leap of faith. While they may not
consider our data to support the use of ar-
terial transfer functions in general, there are
no data from any source suggesting that any
transfer function is able to perform better or
that our data might not reflect the maxi-
mum achievable accuracy of a “generalized”
arterial transfer function. We therefore
strongly urge all potential users, especially
the “nonspecialist readers” of such concern
to Wilkinson and McEniery, to carefully
evaluate the proven accuracy and validity of
such techniques in the specific population
and application of their interest (25).
Avolio, Cockcroft, and O’Rourke (2)
opine that “an appropriately validated and
approved generalized transfer function,
however, is required.” The implication in
this opinion, with which we agree, is that no
such entity exists. Whether our results are
indicative of the fundamental accuracy of a
generalized arterial transfer function can
only be disputed when appropriate and
comparable data are available concerning
other proposed transfer function tech-
niques, commercially available or not. Un-
fortunately, such data can only be provided
by those with access to and specific knowl-
edge of the individual implementation, and
we would strongly request that those
groups, which presumably include the cur-
rent correspondents, provide such data. Us-
ers could then move away from the “black-
box” approach with some confidence
regarding the likely accuracy for their appli-
cation, whether it be simple central pulse
pressure or more sophisticated waveform
analysis. Until such data are available, we
cannot share Avolio, Cockcroft, and
O’Rourke’s optimistic belief (no evidence
supplied) that “editorialists and readers of
Diabetes Care can be reassured that no ‘dia-
betes-specific’ or ‘gender-specific’ transfer
function, however, is required.”
As might be predicted, enthusiasts of
the technique do not like our findings.
However, the mature scientific response
is to disprove a hypothesis rather than as-
sign discredit (2). We therefore suggest
that such enthusiasts provide robust data
in support of oft-stated beliefs; in the ab-
sence of credible data we argue against an
“ignorance is bliss” approach.
SARAH A. HOPE,
MBCHB
1
DAVID B. TAY,
BE, PHD
2
IAN T. MEREDITH,
MBBS, PHD
1
JAMES D. CAMERON,
MBBS, MD
1
From the
1
Cardiovascular Research Centre, Monash
Medical Centre, Monash University, Melbourne,
Australia; and
2
Department of Electronic Engineer-
ing, La Trobe University, Melbourne, Australia.
Address correspondence to Sarah A. Hope,
MBChB, Cardiovascular Research Centre, Monash
Medical Centre, 246 Clayton Rd., Clayton, Victoria
3168, Australia. E-mail: sarah.hope@med.monash.
edu.au.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Wilkinson IB, McEniery CM: Use of arte-
rial transfer functions for the derivation of
central aortic waveform characteristics in
subjects with type 2 diabetes and cardio-
vascular disease (Letter). Diabetes Care 27:
2564, 2004
2. Avolio AP, Cockcroft JR, O’Rourke MF:
Use of arterial transfer functions for the
derivation of central aortic waveform
characteristics in subjects with type 2 di-
abetes and cardiovascular disease (Letter).
Diabetes Care 27:2564 –2565, 2004
3. Hope SA, Meredith IT, Tay DB, Cameron
JD: Use of arterial transfer functions for
the derivation of central aortic waveform
characteristics in subjects with type 2 di-
abetes and cardiovascular disease. Diabe-
tes Care 27:746 –751, 2004
4. Hope SA, Meredith IT, Cameron JD:
Reliability of transfer functions in deter-
mining central pulse pressure and aug-
mentation index. J Am Coll Cardiol 40:
1196, 2002 [author reply 40:1196–1197,
2002]
5. Hope SA, Tay DB, Meredith IT, Cameron
JD: Comparison of generalized and gen-
der-specific transfer functions for the der-
ivation of aortic waveforms. Am J Physiol
Heart Circ Physiol 283:H1150–H1156,
2002
6. Hope SA, Tay DB, Meredith IT, Cameron
JD: Use of arterial transfer functions for
the derivation of aortic waveform charac-
teristics (Letter). J Hypertens 21:2197–
2199, 2003
7. Hope SA, Tay DB, Meredith IT, Cameron
JD: Use of arterial transfer functions for
the derivation of aortic waveform charac-
teristics. J Hypertens 21:1299 –1305, 2003
8. Hope SA, Meredith IT, Cameron JD: Ef-
fect of non-invasive calibration of radial
waveforms on error in transfer function-
derived central aortic waveform charac-
teristics. Clin Sci (Lond) 107:205–211,
2004
9. Hope SA, Tay DB, Meredith IT, Cameron
JD: Use of arterial transfer functions for
the derivation of aortic waveform charac-
teristics (Letter). J Hypertens 22:433– 434,
2004
10. Lehmann ED: Cited “validation” refer-
ences for the SphygmoCor device. Kidney
Int 59:1597–1598, 2001
11. Lehmann ED: Citation of “validation” ref-
erences for sphygmocor-based estimates
of central aortic blood pressure. J Clin En-
docrinol Metab 86:1844 –1845, 2001
12. Lehmann ED: Where is the evidence that
radial artery tonometry can be used to ac-
curately and noninvasively predict central
aortic blood pressure in patients with di-
abetes? Diabetes Care 23:869 – 871, 2000
13. Lehmann ED: Regarding the accuracy of
generalized transfer functions for estimat-
ing central aortic blood pressure when
calibrated non-invasively. J Hypertens 18:
347–349, 2000 [author reply 18:349 –
350, 2000]
14. Wilkinson IB, Fuchs SA, Jansen IM, Spratt
JC, Murray GD, Cockcroft JR, Webb DJ:
Reproducibility of pulse wave velocity
and augmentation index measured by
pulse wave analysis. J Hypertens 16:2079–
2084, 1998
15. Wilkinson IB, MacCallum H, Rooijmans
DF, Murray GD, Cockcroft JR, McKnight
JA, Webb DJ: Increased augmentation in-
dex and systolic stress in type 1 diabetes
mellitus. QJM 93:441– 448, 2000
16. Wilkinson IB, MacCallum H, Flint L,
Cockcroft JR, Newby DE, Webb DJ: The
influence of heart rate on augmentation
index and central arterial pressure in hu-
mans. J Physiol 525:263–270, 2000
17. Wilkinson IB, Franklin SS, Hall IR, Tyrrell
S, Cockcroft JR: Pressure amplification
explains why pulse pressure is unrelated
to risk in young subjects. Hypertension 38:
1461–1466, 2001
18. Wilkinson IB, MacCallum H, Hupperetz
PC, van Thoor CJ, Cockcroft JR, Webb
DJ: Changes in the derived central pres-
sure waveform and pulse pressure in
response to angiotensin II and noradren-
aline in man. J Physiol 530:541–550, 2001
19. Wilkinson IB, Prasad K, Hall IR, Thomas
A, MacCallum H, Webb DJ, Frenneaux
MP, Cockcroft JR: Increased central pulse
pressure and augmentation index in sub-
jects with hypercholesterolemia. J Am Coll
Cardiol 39:1005–1011, 2002
20. Wilkinson IB, Mohammad NH, Tyrrell S,
Hall IR, Webb DJ, Paul VE, Levy T, Cock-
croft JR: Heart rate dependency of pulse
pressure amplification and arterial stiff-
Letters
2566 DIABETES CARE, VOLUME 27, NUMBER 10, OCTOBER 2004
ness. Am J Hypertens 15:24 –30, 2002
21. Wilkinson IB, Hall IR, MacCallum H,
Mackenzie IS, McEniery CM, van der
Arend BJ, Shu YE, MacKay LS, Webb DJ,
Cockcroft JR: Pulse-wave analysis: clinical
evaluation of a noninvasive, widely appli-
cable method for assessing endothelial
function. Arterioscler Thromb Vasc Biol 22:
147–152, 2002
22. Weber T, Auer J, O’Rourke MF, Kvas E,
Lassnig E, Berent R, Eber B: Arterial stiff-
ness, wave reflections, and the risk of cor-
onary artery disease. Circulation 109:184–
189, 2004
23. Vlachopoulos C, Alexopoulos N, Pana-
giotakos D, O’Rourke MF, Stefanadis C:
Cigar smoking has an acute detrimental
effect on arterial stiffness. Am J Hypertens
17:299–303, 2004
24. Vlachopoulos C, Hirata K, O’Rourke MF:
Effect of caffeine on aortic elastic proper-
ties and wave reflection. J Hypertens 21:
563–570, 2003
25. Cloud GC, Rajkumar C, Kooner J, Cooke
J, Bulpitt CJ: Estimation of central aortic
pressure by SphygmoCor requires intra-
arterial peripheral pressures. Clin Sci
(Lond) 105:219–225, 2003
Use of Arterial
Transfer Functions
for the Derivation of
Central Aortic
Waveform
Characteristics in
Subjects With Type 2
Diabetes and
Cardiovascular
Disease
Response to Avolio, Cockcroft, and
O’Rourke
A
volio, Cockcroft, and O’Rourke (1)
raise some valid concerns regarding
the approach taken by Hope et al.
(2) in generating and validating an arterial
transfer function in subjects with diabe-
tes. The rigorous evaluation and vali-
dation process, which undoubtedly
supported the Food and Drug Adminis-
tration approval of the SphygmoCor
SCOR-Px device, could be used as a
model for parallel evaluations in specific
populations of interest. Despite these au-
thors’ assurances of universal applicabil-
ity, however, given the profound effects of
diabetes on all dimensions of the vascular
tree, it does seem reasonable to specifi-
cally test whether generalized transfer
functions are valid in subjects with diabe-
tes. The manuscript by Hope et al. was of
value in that it raises questions about the
validity of this tool in this setting. We are
in full agreement with Avolio, Cockcroft,
and O’Rourke that appropriate validation
of the generalized transfer function in
subjects with diabetes is required before
the use of this tool in research or clinical
applications can be advocated.
K
IEREN MATHER,
MD, FRCPC
1
RICHARD LEWANCZUK,
MD, PHD, FRCPC
2
From the
1
Division of Endocrinology & Metabo-
lism, Indiana University School of Medicine,
Indianapolis, Indiana; and the
2
Division of En-
docrinology & Metabolism, University of Alberta,
Alberta, Canada.
Address correspondence to Kieren Mather, MD,
FRCPC, Indiana University School of Medicine, Di-
vision of Endocrinology & Metabolism, 975 West
Walnut St., IB424, Indianapolis, IN 46202. E-mail:
kmather@iupui.edu.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Avolio AP, Cockcroft JR, O’Rourke MF:
Use of arterial transfer functions for the
derivation of central aortic waveform
characteristics in subjects with type 2 di-
abetes and cardiovascular disease (Letter).
Diabetes Care 27:2564 –2565, 2004
2. Hope SA, Meredith IT, Tay DB, Cameron
JD: Use of arterial transfer functions for
the derivation of central aortic waveform
characteristics in subjects with type 2 di-
abetes and cardiovascular disease. Diabe-
tes Care 27:746 –751, 2004
Insulin Detemir
Offers Improved
Glycemic Control
Compared With NPH
Insulin in People
With Type 1
Diabetes
Response to Home et al.
T
he article by Home et al. (1) in the
May issue of Diabetes Care prompt-
ed us to respond and comment on
the current practice of reporting on clini-
cal trials when insulin analogs are con-
cerned. While a cure for type 1 diabetes
has not yet been achieved, comprehen-
sive, multidisciplinary treatment strate-
gies have lead the way to offer patients
and their families a near-normal life and
affected children a near-normal life ex-
pectancy. Multiple injections, insulin
pump therapy, and frequent blood glu-
cose measurements have provided the
basis for better quality of life and good
metabolic control in people with type 1
diabetes. However, the value of insulin
analogs is still in question (2). There is no
doubt that industry has invested vast re-
sources to develop new insulin types for
safety and efficacy reasons. However, at
present, there is no doubt that insulin an-
alogs are more expensive, and in view of
the rising tide of financial problems in
healthcare systems globally, this in itself
can potentially pose an additional threat
to the free availability and affordability of
insulin treatments for patients with type 1
diabetes worldwide. The Diabetes Con-
trol and Complications Trial (DCCT) re-
search group showed the importance of
strict metabolic control for the delay
and/or prevention of diabetes complica-
tions (3,4). The indiscriminate use of
terms such as “conventional” and “inten-
sified” insulin treatment has been aban-
doned, and “simplified” therapies should
not be recommended (4). Pre- or even
postprandial administration of rapid-
acting insulin analogs, especially in very
young children, has been reported by
some authors (5) to be safe and even ad-
vantageous. Multiple injection regimens
allow greater freedom in daily routines
and are therefore clearly the standard of
insulin-replacement therapy. However,
reports on the advantages of insulin
analog therapy have very often been
published only in abstract form and as
supplements to company-organized
meetings (1).
In their original article, Home et al.
included an example of another mislead-
ing way of reporting on insulin analog
treatment and its alleged advantages; their
investigations were setup mainly to study
safety and suitability issues. In the title,
and in some key passages of the article, it
is suggested that the insulin analog was
superior in terms of achieving glycemic
control and reducing hypoglycemic epi-
sodes when compared with NPH insulin.
The data from their article, however, do
not support this over-optimistic view and
clearly show that HbA
1c
levels for each of
Letters
DIABETES CARE, VOLUME 27, NUMBER 10, OCTOBER 2004 2567
the detemir groups were not different
from those of the NPH group. In addition,
and probably most importantly, the
reduction of hypoglycemic episodes so
enthusiastically reported was seen ex-
clusively for “minor events,” specifically
during the night, while “major” hypogly-
cemic events were not reduced or eventu-
ally even greater in their IDet
morn ⫹ bed
group when compared with their NPH
group. While we do not suggest that
Home et al. have failed to carry out a use-
ful and carefully executed study, we very
much regret to see that the interpretation
of their data is far from careful and not at
all balanced. Since co-authors have indus-
try affiliations and their interpretations
may naturally reflect company interests.
However, such reports as the article by
Home et al. influence clinical decision
making in daily practice and are based on
individual beliefs and personal interests
rather than on solid data when it comes to
conclusions and decision making. A
much more responsible attitude and more
careful interpretation of data is clearly
warranted and should guide clinical sci-
entists when interpreting their data and
writing articles for reputable peer-
reviewed journals. As such, wording and
biased phrasing in scientific papers is of-
ten more powerful than the actual data
and scientific work. The scientific com-
munity should therefore behave respon-
sibly when writing the results from
clinical trials in order to not tarnish its
reputation and most importantly to not
lead the public and ultimately the patient
to untimely and probably incorrect con-
clusions.
WIELAND KIESS,
MD
KLEMENS RAILE,
MD
ANGELA GALLER,
MD
THOMAS KAPELLEN,
MD
From the Hospital for Children and Adolescents,
Medical Faculty, University of Leipzig, Leipzig, Ger-
many.
Address correspondence to Prof. Dr. Wieland
Kiess, Hospital for Children and Adolescents, Uni-
versity of Leipzig, Oststr. 21-25, D-04317 Leipzig,
Germany. E-mail: kiw@medizin.uni-leipzig.de.
The authors of the original article did not wish to
respond to this letter.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Home P, Bartley P, Russel-Jones D,
Hanaire-Broutin H, Heeg J-E, Abrams P,
Landin-Olsson M, Hylleberg B, Lang H,
Draeger E, the Study to Evaluate the Ad-
ministration of Detemir Insulin Efficacy,
Safety and Suitability study group: Insulin
detemir offers improved glycemic control
compared with NPH insulin in people
with type 1 diabetes. Diabetes Care 27:
1081–1087, 2004
2. Siebenhofer A, Plank J, Berghold A,
Narath M, Gfrerer R, Pieber T: Short act-
ing insulin analogs versus regular human
insulin in patients with diabetes melli-
tus. Cochrane Database Syst Rev 2:
CD003287, 2004
3. The Diabetes Control and Complications
Trial research group: the effect of inten-
sive treatment of diabetes on the devel-
opment and progression of long-term
complications in insulin-dependent dia-
betes mellitus. N Engl J Med 329:977–986,
1993
4. Atkinson MA, Eisenbarth GS: Type 1 dia-
betes: new perspectives on disease patho-
genesis and treatment. Lancet 358:221–
229, 2001
5. Danne T, Aman J, Schober E, Deiss D,
Jacobsen JL, Friberg HH, Jensen LH, the
ANA 1200 study group: A comparison of
postprandial and preprandial administra-
tion of insulin aspart in children and ad-
olescents with type 1 diabetes. Diabetes
Care 26:2359 –2364, 2003
Global Prevalence of
Diabetes: Estimates
for the Year 2000
and Projections for
2030
Response to Wild et al.
G
lobal diabetes prevalence estimates
for adults in 2000, which were de-
rived from population-based data
using oral glucose tolerance tests, were
recently reported by Wild et al. (1). Be-
cause there are few OGTT-based data in
the European region, estimates from a re-
gional study carried out in the Nether-
lands were applied to 13 other countries,
including Germany (2). For Germany, a
prevalence of 4.1% was estimated, which
corresponds to 2.6 million people with
diabetes in 2000 (1).
In the German National Health Inter-
view and Examination Survey in 1998,
prevalence of known diabetes (self-
reported) was assessed in a representative
sample (aged 18–79 years) (3). Further-
more, from 1999 to 2001, glucose toler-
ance tests were carried out in the
population-based KORA Survey 2000
(Augsburg, Germany) among 1,353 sub-
jects aged 55–74 years (4). Both provided
higher age- and sex-specific prevalence
estimates than the Dutch study (2). Thus,
⬃5% of the adult German population had
known diabetes in 1998. In addition, at
least in the age-group ⬎55 years, one-half
of the total cases were undiagnosed (4).
The total diabetes prevalence (diagnosed
and undetected cases) in the 55- to 74-
year age-group was 16.6% in the KORA
Survey (4). As the one-for-one ratio for
known/undiagnosed case subjects is valid
for all age-groups, the total diabetes prev-
alence in the adult German population in
2000 was ⬃10%, corresponding to 6.3
million people. Germany should be listed
among the 10 countries with the highest
estimated number of people with diabetes
in the world (1).
Thus, there is a greater diversity of
diabetes prevalence in European coun-
tries than suggested by Wild et al. There
are also differences in known risk factors
for type 2 diabetes at the population level
in European countries, e.g., the preva-
lence of obesity in Germany (20%) was
almost twofold higher than in the Neth-
erlands (5). Furthermore, the percentage
of persons who did not partake in phys-
ical activity during their leisure time
was higher in Germany (22.6%) than in
the Netherlands (14.4%) and other Eu-
ropean countries, which may partly ex-
plain the wide variation of diabetes
prevalence (6).
WOLFGANG RATHMANN,
MD, MSPH
GUIDO GIANI,
PHD
From the Department of Biometrics and Epidemiol-
ogy, German Diabetes Research Institute, Heinrich
Heine University, Du¨ sseldorf, Germany.
Address correspondence to Dr. Wolfgang Rath-
mann, MD, MSPH, German Diabetes Research
Institute, Auf’m Hennekamp 65, D-40225 Du¨ ssel-
dorf, Germany. E-mail: rathmann@ddfi.uni-
duesseldorf.de.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Wild S, Roglic G, Green A, Sicree R, King
H: Global prevalence of diabetes: esti-
mates for the year 2000 and projections
for 2030. Diabetes Care 27:1047–1053,
2004
2. Mooy JM, Grootenhuis PA, de Vries H,
Valkenburg HA, Bouter LM, Kostense PJ,
Heine RJ: Prevalence and determinants of
Letters
2568 DIABETES CARE, VOLUME 27, NUMBER 10, OCTOBER 2004
glucose intolerance in a Dutch Caucasian
population: the Hoorn Study. Diabetes
Care 18:1270 –1273, 1995
3. Thefeld W: Prevalence of diabetes melli-
tus in the adult population in Germany
[in German]. Gesundheitswesen 61 (Suppl.
2):S85–S89, 1999
4. Rathmann W, Haastert B, Icks A, Lo¨wel
H, Meisinger C, Holle R, Giani G: High
prevalence of undiagnosed diabetes mel-
litus in southern Germany: target popula-
tions for efficient screening: the KORA
Survey 2000. Diabetologia 46:182–189,
2003
5. Seidell JC: Obesity in Europe: scaling an
epidemic. Int J Obes Relat Metab Disord 19
(Suppl. 3):S1–S4, 1995
6. Varo JJ, Martinez-Gonzalez MA, Irala-Es-
tevez J, Kearney J, Gibney M, Martinez JA:
Distribution and determinants of seden-
tary lifestyles in the European Union. Int J
Epidemiol 32:138 –146, 2003
Global Prevalence of
Diabetes: Estimates
for the Year 2000
and Projections for
2030
Response to Rathman and Giani
R
athmann and Giani (1) point out
that there is a greater diversity of
diabetes prevalence in Europe than
suggested by our recent article (2) on the
global prevalence of diabetes. We agree
with this point. Unfortunately, the data
from their study (3) in Augsburg, Ger-
many, had not been published when the
data for our article were assembled in
2002. As described in the article, we plan
to update the estimates of diabetes preva-
lence and will include the Augsburg data
in the next revision.
The global burden of disease study
(4) only included studies describing the
prevalence of diabetes defined using
oral glucose tolerance tests. Prevalence
studies based on self-reported diabetes
were excluded because, as Rathmann and
Giani found, ⬃50% of diabetes in Euro-
pean populations is undiagnosed. (This
proportion varies with age, sex, and
ethnicity).
We had similar concerns as Rath-
mann and Giani about applying diabetes
prevalence rates from the Netherlands to
other European countries, given the
higher levels of physical activity and
lower prevalence of obesity in the Neth-
erlands compared with many other coun-
tries. For example, total diabetes
prevalence, as derived from the study in
the Netherlands, was lower in middle-
aged people even when compared with
self-reported diabetes prevalence from
the Health Survey for England (5). A re-
cent article (6) on the prevalence of
known diabetes in eight European coun-
tries was published using data from sen-
tinel general practices. The results suggest
that the estimated prevalence of 2.7% for
both diagnosed and undiagnosed diabe-
tes for both sexes at all ages presented in
our article for the Netherlands may be an
underestimate, as the estimate for diag-
nosed diabetes for the Netherlands was
⬃2.6%. In Belgium, the prevalence of di-
agnosed diabetes at all ages was higher at
⬃3.3% than the estimate of 3.1% for both
diagnosed and undiagnosed diabetes
given in the global prevalence article.
The estimates of diabetes prevalence
presented in the global burden article (2)
are likely to represent conservative esti-
mates for most regions as a consequence
of the inclusion criteria and the need to
extrapolate the limited available data,
which may not reflect current patterns of
diabetes prevalence. Even these conserva-
tive estimates have major public health
implications. We hope that their publica-
tion will provide an incentive for better
data collection on diabetes prevalence
around the world, ideally as part of dia-
betes prevention programs.
SARAH H. WILD,
MBBCHIR, PHD
1
GOJKA ROGLIC,
MD
2
ANDERS GREEN,
MD, PHD, DRMEDSCI
3
RICHARD SICREE,
MBBS, MPH
4
HILARY KING,
MD, DSC
2
From the
1
Department of Public Health Sciences,
University of Edinburgh, Edinburgh, Scotland; the
2
Department of Non-Communicable Diseases,
World Health Organization, Geneva, Switzerland;
the
3
Department of Epidemiology and Social Medi-
cine, University of Aarhus, Aarhus, Denmark; and
the
4
International Diabetes Institute, Caulfield, Aus-
tralia.
Address correspondence to Dr. Sarah Wild,
MBBCHIR, PhD, Department of Public Health Sci-
ences, University of Edinburgh, Teviot Place, Edin-
burgh EH8 9AG, Scotland. E-mail: sarah.wild@
ed.ac.uk.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Rathmann W, Giani G: Global prevalence
of diabetes: estimates for the year 2000
and projections for 2030 (Letter). Diabetes
Care 27:2568 –2569, 2004
2. Wild S, Roglic G, Green A, Sicree R, King
H: Global prevalence of diabetes: esti-
mates for the year 2000 and projections
for 2030. Diabetes Care 27:1047–1053,
2004
3. Rathmann W, Haastert B, Icks A, Lo¨wel
H, Meisinger C, Holle R, Giani G: High
prevalence of undiagnosed diabetes mel-
litus in southern Germany: target popula-
tions for efficient screening: the KORA
Survey 2000. Diabetologia 46:182–189,
2003
4. DECODE Study Group: Age- and sex-
specific prevalences of diabetes and im-
paired glucose regulation in 13 European
cohorts. Diabetes Care 26:61– 69, 2003
5. Erens B, Primatesta P (Eds.): Health Survey
for England 1998. London, The Stationery
Office, 2000
6. Fleming DM, Schellevis FG, Van Casteren
V: The prevalence of known diabetes in
eight European countries. Eur J Public
Health 14:10 –14, 2004
Biological Variation
in HbA
1c
Predicts
Risk of Retinopathy
and Nephropathy in
Type 1 Diabetes
Response to McCarter et al.
W
e read the article by McCarter et
al. (1) with interest. Technically,
all nonanalytical variation, irre-
spective of its source, is biological varia-
tion. Thus, mean blood glucose (MBG)-
associated changes are included in
biological variation. It must also be
stressed that all population regression
equations have confidence limits that
need to be taken into account when com-
paring values from individuals to the pop-
ulation study mean.
However, such points do not take
away from the message of McCarter et al.
(1) that non-MBG–related biological vari-
ation may be an important prognostic in-
dicator. The real question is how health
care professionals are to identify this vari-
ation in routine clinical practice. MBG has
many problems such as a large variation,
which is common when many indepen-
Letters
DIABETES CARE, VOLUME 27, NUMBER 10, OCTOBER 2004 2569
dent analytes are measured (2), bias due
to calibration issues (3), or the time taken
for separation (4). Most importantly, it is
rarely used in routine clinical practice. In
addition, HbA
1c
also has its problems (5).
Accordingly, the calculation of the hemo-
globin glycation index is problematic in
routine clinical practice. Furthermore,
calculated indexes will suffer from the
propagation of error, contributing to mis-
classification and inaccurate prediction of
complications (6).
We previously recommended the use
of a rolling mean to reduce the effect of
analytical and biological variation (7).
The associated SD in stable patients
would reflect the total variation for
HbA
1c
. Since the majority of the total vari-
ation is nonanalytical, use of the SD
would easily identify those patients in
routine clinical practice with large non-
MBG–related biological variation. As well
as being easier to perform in routine clin-
ical practice, it would also be a more valid
way of identifying within-patient HbA
1c
variability. In addition, the use of a rolling
mean and its associated SD makes the de-
tection of critical changes in HbA
1c
levels
easier and more objective (5). Accord-
ingly, we recommend the use of a rolling
mean and its associated SD for the inves-
tigation of non-MBG–related biological
variation.
PATRICK J. TWOMEY,
MD, MRCPATH
1
ADIE VILJOEN
2
TIMOTHY M. REYNOLDS
3
ANTHONY S. WIERZBICKI
4
From the
1
Department of Clinical Biochemistry, The
Ipswich Hospital, Ipswich, U.K.; the
2
Department of
Clinical Biochemistry, Addenbrookes Hospital,
Cambridge, U.K.; the
3
Department of Clinical
Chemistry, Queen’s Hospital, Burton-on-Trent,
U.K.; and the
4
Department of Chemical Pathology,
Guy’s and St. Thomas’ Hospital, London, U.K.
Address correspondence to Patrick J. Twomey,
MD, MRCPath, The Ipswich Hospital, Department
of Clinical Biochemistry, Heath Road, Ipswich,
Suffolk IP4 5PD, U.K. E-mail: pattwomey@
doctors.org.uk.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. McCarter RJ, Hempe JM, Gomez R,
Chalew SA: Biological variation in HbA
1c
predicts risk of retinopathy and nephrop-
athy in type 1 diabetes. Diabetes Care 27:
1259–1264, 2004
2. Reynolds TM, Twomey P, Wierzbicki AS:
Accuracy of cardiovascular risk estima-
tion for primary prevention in patients
without diabetes. J Cardiovasc Risk 9:183–
190, 2002
3. Twomey PJ: Plasma glucose measurement
with the Yellow Springs Glucose 2300
STAT and the Olympus AU640. J Clin
Pathol 57:752–754, 2004
4. Twomey PJ: An audit of the potential ef-
fect of ex vivo glucose metabolism in pri-
mary care. Chem Pathol. In press
5. Twomey PJ, Reynolds TM, Wierzbicki AS:
Issues to consider when attempting to
achieve the American Diabetes Associa-
tion clinical quality requirement for hae-
moglobin A1c. Curr Med Res Opin 19:
719–723, 2003
6. Challand GS: Lies, damned lies and algo-
rithms. Ann Clin Biochem 37:741–750,
2000
7. Twomey PJ, Reynolds TM, Wierzbicki AS:
The casino of life: markets, mathematics
and medicine. J Clin Pathol 57:243–244,
2004
Biological Variation
in HbA
1c
Predicts
Risk of Retinopathy
and Nephropathy in
Type 1 Diabetes
Response to Twomey et al.
W
e appreciate the comments of
Twomey et.al. (1) in regard to
our recent article (2) about bio-
logical variation in HbA
1c
and its relation-
ship to microvascular complications.
They point out potential obstacles to the
general use of the hemoglobin glycation
index (HGI) by clinical practitioners. The
best approach to calculating the HGI of
individual patients from populations be-
sides the Diabetes Control and Complica-
tions Trial (2) and our original patient
data (3) still needs to be established for
the clinician.
We point out, however, that a moving
average of an individual patient’s HbA
1c
will not provide the same information as
the HGI. Our analyses indicate that
HbA
1c
carries two important components
of clinically relevant information: 1)an
estimate of the patient’s preceding mean
blood glucose (MBG) and 2) patient-
specific factors besides MBG that influ-
ence glycation of hemoglobin. Both of
these components are independently as-
sociated with the risk of development of
microvascular complications (2). Cur-
rently, clinical therapy is only directed at
altering the first component. Calculation
of the HGI allows us to separate the two
components of risk from the patient’s
HbA
1c
measurements. In this fashion, in-
dividual patients who have consistently
high or low hemoglobin glycation status-
can be identified. However, calculating a
patient’s HGI requires knowledge of the
patient’s preceding MBG, the population
relationship between HbA
1c
and MBG,
and the patient’s HbA
1c
.
We foresee the future development of
databases that will assist clinicians in cal-
culating an HGI necessary to assess the
hemoglobin glycation status of their pa-
tients. These computational sources will
need to be referenced to population data
that are specific for the methods of deter-
mining both the MBG and HbA
1c
used for
that particular patient.
STUART A. CHALEW,
MD
1
ROBERT J. MCCARTER,
SCD
2
JAMES M. HEMPE,
PHD
3
From the
1
Department of Endocrinology, Children’s
Hospital, New Orleans, Louisiana; the
2
Department
of Bioinforatics and Statistics, Children’s National
Medical Center, Washington, D.C.; and the
3
De-
partment of Pediatrics, Louisiana State University
Health Sciences Center, New Orleans, Louisiana.
Address correspondence to Stuart A. Chalew,
MD, Children’s Hospital, Department of Endocri-
nology, 200 Henry Clay Ave., New Orleans, LA
70118. E-mail: schale@lsuhsc.edu.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Twomey PJ, Viljoen A, Reynolds TM,
Wierzbicki AS: Biological variation in
HbA
1c
predicts risk of retinopathy and
nephropathy in type 1 diabetes (Letter).
Diabetes Care 27:2569 –2570, 2004
2. McCarter RJ, Hempe JM, Gomez R,
Chalew SA: Biological variation in HbA
1c
predicts risk of retinopathy and nephrop-
athy in type 1 diabetes. Diabetes Care 27:
1259–1264, 2004
3. Hempe J, Gomez R, McCarter R, Chalew
S: High and low hemoglobin glycation
phenotypes in type 1 diabetes: a challenge
for interpretation of glycemic control. J
Diabetes Complications 16:313–320, 2002
Letters
2570 DIABETES CARE, VOLUME 27, NUMBER 10, OCTOBER 2004
