Improvement of Temperature and Flow in Feet of Subjects with Diabetes With Use of a Transdermal Preparation of L-Arginine: A pilot study

Full text

OBSERVATIONS
Plasma Interleukin-
18 Concentrations
Are Elevated in Type
2 Diabetes
W
e read with interest the article by
Aso et al. (1) showing greater
plasma concentrations of inter-
leukin-18 (IL-18) in type 2 diabetic pa-
tients compared with matched control
subjects. While the authors found no sig-
nificant associations between fasting lev-
els of IL-18 and homeostasis model
assessment, a measure of insulin resis-
tance, they found associations between
IL-18 and C-reactive protein concentra-
tions. Moreover, carotid intima-media
thickness, a validated surrogate measure
of atherosclerosis, was greater in diabetic
patients with high IL-18 than in those
with normal IL-18. As a whole, these data
add to the mounting evidence that diabe-
tes may be regarded as a chronic low-
grade inflammatory state.
However, we disagree with the au-
thors’ conclusions that “the present study
demonstrated for the first time that
plasma IL-18 concentrations were signif-
icantly higher in type 2 diabetic patients
than in age-matched control subjects”
because we have reported similar find-
ings in Diabetes Care (2). In that study,
we demonstrated that 30 newly diag-
nosed, slightly overweight (BMI 26.9 ⫾
1.2 kg/m
2
, means ⫾SD) type 2 diabetic
patients without clinical or instrumental ev-
idence of micro- and macrovascular com-
plications presented higher circulating
concentrations of IL-18 compared with
nondiabetic subjects matched for sex, age,
and body weight (205 ⫾39 vs. 120 ⫾25
pg/ml, P⬍0.01). It is reassuring to see that
the fasting values of IL-18 in the Japanese
diabetic patients studied by Aso et al. (1)
were quite similar (203 ⫾153 pg/ml) to
those of our Caucasian diabetic patients and
that the relation between fasting plasma glu-
cose and IL-18 concentrations was present
in both studies (r⫽0.31, P⬍0.05; r⫽
0.24, P⬍0.02, respectively). These results
suggest that ethnicity does not to play a ma-
jor role in these associations.
IL-18 is a potent proinflammatory cy-
tokine reported to play a role in plaque
destabilization (3) and predict cardiovas-
cular death in patients with coronary ar-
tery disease (4). Although IL-18 is
produced mainly by monocyte/macro-
phages, a contribution from adipose tis-
sue has recently been suggested (5). In the
light of the evidence that IL-18 circulating
levels are higher in type 2 diabetic pa-
tients than matched nondiabetic subjects
and correlate with fasting glucose levels, it
does not seem hazardous to hypothesize
that IL-18 may play a role in acute coro-
nary syndromes through plaque destabi-
lization (6). This working hypothesis,
which deserves further investigation, also
finds support in the observation that
acute hyperglycemia may increase circu-
lation levels of IL-18 in both normal sub-
jects and patients with impaired glucose
tolerance (7).
KATHERINE ESPOSITO,
MD
RAFFAELE MARFELLA,
MD, PHD
DARIO GIUGLIANO,
MD, PHD
From the Department of Geriatrics and Metabolic
Diseases, Second University of Naples, Naples, Italy.
Address correspondence to Katherine Esposito,
MD, Dipartimento di Geriatria e Malattie del Me-
tabolismo, Policlinico Universitario, Piazza L. Mira-
glia, 80138 Napoli, Italy. E-mail: katherine.
esposito@unina2.it.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Aso Y, Okumura K-I, Takebayashi K,
Wakabayashi S, Inukai T: Relationships of
plasma interleukin-18 concentrations to
hyperhomocysteinemia and carotid inti-
ma-media wall thickness in patients with
type 2 diabetes. Diabetes Care 26:2622–
2627, 2003
2. Esposito K, Nappo F, Giugliano F, Di Palo
C, Ciotola M, Barbieri M, Paolisso G, Gi-
ugliano D: Cytokine milieu tends toward
inflammation in type 2 diabetes (Letter).
Diabetes Care 26:1647, 2003
3. Mallat Z, Corbaz A, Scoazec A, Besnard S,
Leseche G, Chvatchko Y, Tedgui A: Ex-
pression of interleukin-18 in human ath-
erosclerotic plaques and relation to
plaque instability. Circulation 104:1598–
1603, 2001
4. Blankenberg S, Tiret L, Bickel C, Peetz D,
Cambien F, Meyer J, Rupprecht HJ: Inter-
leukin 18 is a strong predictor of cardio-
vascular death in stable and unstable
angina. Circulation 106:24–30, 2002
5. Esposito K, Pontillo A, Ciotola M, Di Palo
C, Grella E, Nicoletti G, Giugliano D:
Weight loss reduces interleukin-18 levels
in obese women. J Clin Endocrinol Metab
87:3864–3866, 2002
6. Capes SE, Hunt D, Malberg K, Gerstein
HC: Stress hyperglycemia and increased
risk of death after myocardial infarction in
patients with and without diabetes: a sys-
temic overview. Lancet 355:773–778,
2000
7. Esposito K, Nappo F, Marfella R, Giugli-
ano G, Giugliano F, Ciotola M, Quagliaro
L, Ceriello A, Giugliano D: Inflammatory
cytokine concentrations are acutely in-
creased by hyperglycemia in humans: role
of oxidative stress. Circulation 106:2067–
2072, 2002
Similar A1C
Outcomes in Type 1
Diabetic Patients
Undergoing
Intensive Diabetes
Management With
Preprandial Rapid-
Acting Insulin and
Either CSII or
Glargine
T
he importance of achieving and
maintaining tight glycemic control
in patients with type 1 diabetes is
well known (1). Continuous insulin infu-
sion therapy (CSII) has been available for
many years, but only recently have re-
ports of efficacy with rapid-acting insulin
analogues been published. Similarly,
multiple daily injection (MDI) therapy us-
ing glargine insulin in conjunction with
premeal rapid-acting insulin is relatively
new (2).
In our diabetes center, all patients
with type 1 diabetes receive the same di-
abetes education, including instruction in
carbohydrate counting. All patients are
given the option of either CSII or MDI
therapy and are encouraged to use which-
ever treatment maintains their blood glu-
cose levels as close to normal as possible.
To assess our quality of care, we per-
formed a random chart audit of 150 pa-
tients. To be included in the analysis,
patients had to have type 1 diabetes and
be treated for at least 6 months with either
LETTERS
272 DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004

CSII using rapid-acting insulin (lispro or
aspart) or MDI with insulin glargine with
premeal rapid-acting insulin. Patients
who were pregnant, ⬍15 years of age, or
referred for treatment of severe recurrent
hypoglycemia were excluded.
There were 103 patients who met our
criteria; 58 were on CSII and 45 were on
MDI therapy. Glargine was given in the
morning in 11%, in the evening in 60%,
and twice a day in 29% of patients. Age,
duration of diabetes, and incidence of
complications were similar in both
groups. Duration of therapy was 16.0 ⫾
5.8 (means ⫾SD) (CSII) vs. 11.6 ⫾3.8
months (MDI) (P⬍0.0001). Most recent
A1C levels were the same in both
groups—6.79 ⫾1.07 (CSII) vs. 6.84 ⫾
0.90% (MDI) (P⫽0.82). One patient in
the CSII group and two patients in the
MDI group had an episode of severe hy-
poglycemia. One patient in the CSII
group had two episodes of diabetic
ketoacidosis.
Therefore, both CSII and MDI ther-
apy can be used to treat patients with type
1 diabetes to target. Prospective data are
needed to confirm the findings of this
cross-sectional analysis.
ANNE PETERS HARMEL,
MD
RUCHI MATHUR,
MD
From the Department of Endocrinology, University
of Southern California, Los Angeles, California.
Address correspondence to Dr. Anne Harmel,
Endocrinology, University of Southern California,
2905 Alma Ave., Manhattan Beach, CA 90266. E-
mail: momofmax@hotmail.com.
A.P.H. has received honoraria from and served on
advisory boards for NovoNordisk, Aventis, and Lilly
and has received research support from Minimed
and Aventis. R.M. is a member of the Aventis
speaker bureau.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. The Diabetes Control and Complications
Trial Research Group: The effect of inten-
sive treatment of diabetes on the develop-
ment and progression of long-term
complications in insulin dependent dia-
betes mellitus. N Engl J Med 329:977–986,
1993
2. DeWitt DE, Hirsch IB: Outpatient insulin
therapy in type 1 and type 2 diabetes mel-
litus: scientific review. JAMA 289:2254–
2264, 2003
Matrix
Metalloproteinase 2
May Be a Marker of
Microangiopathy in
Children and
Adolescents With
Type 1 Diabetes
P
atients with type 1 diabetes develop
microangiopathic complications
such as retinopathy, peripheral neu-
ropathy, and nephropathy (1), which are
responsible for morbidity in adulthood.
These complications usually have a pro-
longed asymptomatic phase, sometimes
starting in adolescence, characterized by
early subclinical functional and structural
abnormalities (2).
Since matrix metalloproteinases
(MMPs) represent a serum marker of vas-
cular disease (4), the aim of our study was
to detect MMP-2 and -9 levels and activity
in type 1 diabetic children and adolescents.
Twenty-five children and adolescents
(13 boys and 12 girls), median age 10.6
years (7.9⫺11.7), were longitudinally
evaluated at clinical diagnosis and during
a 5-year follow-up period.
Peripheral neuropathy, assessed by
peroneal motor nerve conduction veloc-
ity, developed in 12 patients (6 boys and 6
girls) 5.7 years (3.7⫺6.5) from disease di-
agnosis. Background diabetic retinopathy
(microaneurisms), assessed by fundus
photography, developed in three patients
(two boys and one girl) 6.6, 5.8, and 6.0
years from disease diagnosis, respectively.
For control subjects, we randomly
chose 19 nondiabetic subjects (9 boys and
10 girls), median age 12.0 years
(11.0⫺13.0), from among those report-
ing for their first hepatitis B virus vaccina-
tion and for the 5-year follow-up visit.
Informed consent was obtained from
all parents. The study was approved by
the local ethics committee. MMP-2 and -9
levels and activity were detected by ELISA
(Amersham, Pharmacia Biotech); GAD
antibody, IA-2 antigen, and insulin auto-
antibody levels were detected by radioim-
munoassay (CIS Bio International).
HbA
1c
levels were evaluated by high-
performance liquid chromatography
(BioRad). All samples were stored at
⫺80°C until analysis was performed.
No significant correlation was ob-
served among MMP results (both levels
and activity) and chronologic age, auto-
antibody, and HbA
1c
levels.
At baseline, MMP-2 levels were sig-
nificantly higher in type 1 diabetic pa-
tients and type 1 diabetic patients with
complications than in nondiabetic sub-
jects (1,100 [915–1,326], 1,742 [1,426–
1,908], and 907 ng/ml [735–970],
respectively), as was MMP-2 activity (31
[30–37], 152 [127–176], and 97% [88 –
101], respectively) (P⬍0.0001). No sig-
nificant differences were observed for
MMP-9 level and activity.
Patients who developed microangio-
pathic complications during the follow-up
period had significantly higher MMP-2
activity (P⬍0.001) and levels (P⫽0.009)
than patients without complications.
At 5-year follow-up, MMP-2 levels
were significantly higher in patients with
microangiopathic complications com-
pared with control subjects (1,782
[1,741–2,089] and 1,022 ng/ml [897–
1,125], respectively; P⬍0.0001), as was
MMP-2 activity (116 [104 –151] and 46%
[37–68], respectively; P⬍0.0005) and
compared with patients without compli-
cations (1,371 ng/ml [1,197–1,479] and
31% [30–34] for levels and activity, re-
spectively; P⬍0.0001).
MMP-9 levels were significantly
lower in patients with microangiopathic
complications (44 ng/ml [30–63]) com-
pared with control subjects (95 ng/ml
[58–126]; P⫽0.024) and patients with-
out complications (82 ng/ml [46–99];
P⫽0.0013), but no difference was found
between control subjects and patients
without complications.No difference was
observed for MMP-9 activity among the
three groups.
The three groups did not differ in
terms of percentage change of MMP-2 lev-
els and MMP-9 activity, but did differ in
terms of percentage change of MMP-2 ac-
tivity (P ⫽0.0036) and MMP-9 levels
(P⫽0.009).
Our results allow us to postulate that
MMP-2 may be a good index of severity
and stability of microangiopathy, and the
literature reports MMP-9 as a marker of
macroangiopathy (4).
The relationship between MMPs and
the presence of diabetic complications
needs to be elucidated; further studies are
necessary to clarify their possible involve-
ment in the onset or progression of dia-
betic complications.
Letters
DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004 273

GIUSEPPE DEROSA,
MD, PHD
1
MARIA ANTONIETTA AVANZINI,
BD, PHD
2
DIEGO GEROLDI,
MD
1
ROBERTO FOGARI,
MD
1
RENATA LORINI,
MD
3
ANNALISA DESILVESTRI,
BD
4
CARMINE TINELLI,
MD
4
GIORGIO RONDINI,
MD
5
GIUSEPPE D’ANNUNZIO,
MD
5
From the
1
Department of Internal Medicine and
Therapeutics, University of Pavia, Pavia, Italy; the
2
Research Laboratories for Pediatric Oncohematol-
ogy and Immunology, IRCCS Policlinico S. Matteo,
Pavia, Italy; the
3
Department of Pediatrics, Univer-
sity of Genova, G. Gaslini Institute, Genova, Italy;
the
4
Biometric Unit, IRCCS Policlinico S. Matteo,
Pavia, Italy; and the
5
Department of Pediatrics,
IRCCS Policlinico S. Matteo, Pavia, Italy.
Address correspondence to Giuseppe Derosa,
MD, PhD, Department of Internal Medicine and
Therapeutics, University of Pavia, P.le C. Golgi,
2-27100 PAVIA, Italy. E-mail: giuderosa@tin.it.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Law GR, McKinney PA, Staines A, Wil-
liams R, Kelly M, Alexander F, Gilman E,
Bodansky HJ: Clustering of childhood
IDDM: links with age and place of resi-
dence. Diabetes Care 20:753–756, 1997
2. Pomilio MP, Mohn A, Verrotti A, Chiarelli
F: Endothelial dysfunction in children
with type 1 diabetes mellitus. J Pediatr En-
docrinol Metab 15:343–361, 2002
3. Dollery CM, McEwan JR, Henney AM:
Matrix metalloproteinases and cardiovas-
cular disease. Circ Res 77:863–868, 1995
4. Maxwell PR, Timms PM, Chandran S, and
Gordon D: Peripheral blood level alter-
ations of TIMP-1, MMP-2 and MMP-9 in
patients with type 1 diabetes. Diabet Med
18:777–780, 2001
Plasma Adiponectin
and Pregnancy-
Induced Insulin
Resistance
L
indsay et al. (1) recently published
an article in Diabetes Care that claims
adiponectin is present in the cord
blood of the offspring of diabetic and non-
diabetic pregnant mothers and that its
level does not correlate with their birth
weight and skinfold thickness. Our data,
obtained from a case-control study of
nondiabetic women and women with ges-
tational diabetes mellitus (GDM), further
support the role of adiponectin in insulin
resistance.
We observed significantly decreased
plasma adiponectin levels (7.55 ⫾2.04
␮g/ml [means ⫾SD]) using radioimmu-
noassay (Linco Research, St. Charles,
MO) (intra-assay precision coefficient of
variation [CV] 3.86%, interassay CV
8.47%) in 30 women with GDM, all of
whom were treated with insulin (aged
28.12 ⫾2.71 years, gestational age
27.35 ⫾6.15 weeks), compared with 40
nondiabetic pregnant women tested with
oral glucose tolerance test (OGTT) (total
group 9.91 ⫾3.32, P⬍0.01, Mann-
Whitney, aged 26.91 ⫾2.65 years, gesta-
tional age 23.17 ⫾10.91 weeks, 15 in the
first, 12 in the second, and 13 in the third
trimester) and 30 age-matched nonpreg-
nant nondiabetic women (12.54 ⫾3.76,
P⬍0.01, aged 28.42 ⫾3.48 years). GDM
was diagnosed with a 75-g OGTT accord-
ing to the World Health Organization rec-
ommendations. In the nondiabetic
pregnant group, plasma adiponectin lev-
els were significantly lower in the second
(9.30 ⫾2.78) and third (8.06 ⫾2.44)
trimesters compared with women in the
first trimester (12.30 ⫾3.20, P⬍0.01).
After correction for gestational age and
BMI, the differences still remained signif-
icant. No difference was found in plasma
adiponectin levels before and after insulin
treatment in the GDM patients.
In the GDM group, plasma adiponec-
tin levels were in a significant negative
linear correlation with serum tumor ne-
crosis factor-␣(TNF-␣)(r⫽⫺0.65, P⬍
0.0001, Spearman’s rank correlation
test), leptin (r⫽⫺0.75, P⫽0.0004),
fasting C-peptide concentrations (r⫽
⫺0.83, P⬍0.0001), BMI (r⫽⫺0.67,
P⬍0.0001), and, as an indirect parame-
ter of insulin resistance, fasting C-
peptide/blood glucose ratio (r⫽⫺046,
P⫽0.0109). The same was found in the
total group of nondiabetic pregnant
women (TNF-␣:r⫽⫺0.56, P⫽0.0002;
leptin: r⫽⫺0.45, P⫽0.003; fasting C-
peptide: r⫽⫺0.70, P⬍0.0001; BMI:
r⫽⫺0.51, P⫽0.0007; C-peptide–to–
blood glucose ratio: r⫽⫺0.43, P⫽
0.0046). In the nonpregnant nondiabetic
control subjects, negative correlations
with serum leptin (r⫽⫺0.44, P⫽
0.0134), fasting C-peptide concentra-
tions (r⫽⫺0.46, P⫽0.01), and BMIs
(r⫽⫺0.57, P⫽0.0008) were found.
Maternal plasma adiponectin levels
correlated positively with the body weight
of the neonates, in both the GDM (new-
borns, n⫽30, 13 boys and 17 girls, ma-
ternal gestational age at delivery 38.22 ⫾
0.51 weeks, 14 Cesarean sections, body
weight 3,151 ⫾672 g, centile 55.87 ⫾
30.29%, r⫽0.4345 Spearman’s rank
correlation test, P⫽0.0164) and the non-
diabetic pregnant group (n⫽20, 9 boys
and 11 girls, gestational age at delivery
38.92 ⫾0.32 weeks, 6 Cesarean sections,
body weight 3,562 ⫾359 g, centile
63.46 ⫾15.80%, r⫽0.6124, P⫽
0.0041) after the correction for gesta-
tional age.
There is an association between adi-
ponectin concentrations and insulin sen-
sitivity. The protein is exclusively
produced in adipocytes. Its inverse rela-
tionship with the increasing BMI, fasting
C-peptide concentrations, and C-
peptide–to–blood glucose ratio of nondi-
abetic pregnant women and patients with
GDM underlines the importance of adi-
pose tissue and its secreted products,
such as adiponectin, in pregnancy-
induced insulin resistance. The negative
correlation between TNF-␣, leptin, and
adiponectin may suggest a negative regu-
latory role of these cytokines in the ex-
pression and secretion of adiponectin.
The peroxisome proliferator–activated re-
ceptor system in adipocytes is a known
regulator of fat cell differentiation and the
production of TNF-␣, leptin, and adi-
ponectin (2). On the other hand, TNF-␣
may inhibit peroxisome proliferator–
activated receptor function in adipocytes.
Through this mechanism, the TNF sys-
tem may have a suppressive effect on adi-
ponectin production during the course of
pregnancy (3).
Maternal insulin resistance and adi-
pocytokines may influence different an-
thropometric parameters (body weight,
body length, and head circumference) of
the neonates. TNF-␣and leptin were in a
negative correlation with these parame-
ters in neonates of mothers with GDM (4).
However, analyzing the relationship be-
tween these parameters and maternal
plasma adiponectin levels, the significant
positive linear correlation between adi-
ponectin, and the neonatal body weight,
both in GDM and nondiabetic pregnant
women, suggests a regulatory role of the
protein in neonatal development.
KA´ROLY CSEH,
MD, DSC
1
E
´VA BARANYI,
MD, PHD
2
ZSOLT MELCZER,
MD
3
Letters
274 DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004

EDIT KASZA
´S,
MD
1
E
´VA PALIK,
MD
4
GA´BOR WINKLER,
MD, DSC
5
From the from
1
1st Department of Internal Medi-
cine, Ka´rolyi Hospital, Budapest, Hungary; the
2
Na-
tional Health Center Diabetes Outpatient Unit,
Budapest, Hungary; the
3
2nd Department of Obstet-
rics and Gynecology, Budapest, Hungary; the
4
3rd
Department of Internal Medicine, Semmelweis Uni-
versity, Budapest, Hungary; and the
5
Department of
Internal Medicine II, St. John’s Hospital, Budapest,
Hungary.
Address correspondence to Ga´bor Winkler, MD,
DSc, Head of the Department of Internal Medicine
II, St. John’s Hospital, Budapest, Dio´sa´rok 1-3,
Hungary. E-mail: winkler.gabor@matavnet.hu.
© 2004 by the American Diabetes Association.
Acknowledgments—This work was sup-
ported by grants ETT 03/2000 and ETT 015/
2003 from the Ministry of Health.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Lindsay RS, Walker JD, Havel PJ, Hamil-
ton BA, Calder AA, Johnstone FD, on be-
half of the Scottish Multicentre Study of
Diabetes in Pregnancy: Adiponectin is
present in cord blood but is unrelated to
birth weight. Diabetes Care 26:2244–
2249, 2003
2. Chandran M, Phillips SA, Ciaraldi T,
Henry RR: Adiponectin: more than just
another fat cell hormone? Diabetes Care
26:2442–2450, 2003
3. Winkler G, Cseh K, Baranyi E
´, Melczer Z,
Speer G, Hajo´s P, Salamon F, Tury Z, Ko-
va´cs M, Vargha P, Kara´dy I: Tumor necro-
sis factor system in insulin resistance in
gestational diabetes. Diabetes Res Clin
Pract 56:93–99, 2002
4. Cseh K, Baranyi E
´, Melczer Z, Csa´ka´ny
GM, Speer G, Kova´cs M, Gero G, Kara´dy I,
Winkler G: The pathophysiological influ-
ence of leptin and tumor necrosis factor
system on maternal insulin resistance:
negative correlations with anthropomet-
ric parameters of neonates in gestational
diabetes. Gynecol Endocrinol 16:453–460,
2002
Vitiligo Associated
With Subcutaneous
Insulin Lispro
Infusion in Type 1
Diabetes
V
itiligo vulgaris, the loss of skin pig-
mentation, is known to occur with
increased frequency in patients
with type 1 diabetes and, based on a pre-
ponderance of circumstantial evidence
(1), presumed to be of autoimmune etiol-
ogy. For example, 20% of 39 patients
with vitiligo were found to have diabetes
in a Romanian community study (2), and
9% of 457 consecutive Italian patients
with diabetes had vitiligo in another study
(including 54% of the type 1 patients) (3).
However, the factors that can specifically
precipitate vitiligo in type 1 diabetes are
not known. Here, we present a case of
focal vitiligo vulgaris precipitated and ex-
acerbated by the subcutaneous infusion
of the human insulin analog, insulin
lispro.
A 32-year-old female with a 19-year
history of type 1 diabetes began continu-
ous subcutaneous insulin infusion (CSII)
therapy 3.5 years before presentation. She
had previously noted stable vitiligo vul-
garis of the elbows and knees for ⬃10
years. After initiating CSII therapy with
insulin lispro, she developed two sym-
metrical patches of depigmentation on
her abdomen ⬃6 cm in diameter sur-
rounding the insulin infusion sites bilat-
erally (Fig. 1). There was no known
antecedent inflammatory skin disease.
The antibody response to rapid-
acting human insulin analogs has been
shown to be similar in magnitude to that
triggered by human insulin (4,5). Most
cutaneous allergies to insulin, however,
manifest as IgE-mediated wheal and flare
responses (6). In this case, the focal vitil-
igo was apparently induced by insulin in-
fusion, raising questions about its
pathogenesis. Possible mechanisms in-
Figure 1—Vitiligo vulgaris on the abdominal skin of a young woman associated with the subcutaneous infusion of insulin lispro.
Letters
DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004 275

clude a postinflammatory, Koebner-type
response in which depigmentation occurs
in areas of mild injury or inflammation,
but no evidence of skin damage or inflam-
mation was present in the lesions. More
likely, a local allergic reaction to the con-
stituents of the insulin (or possibly the
infusion catheter) may have precipitated
an inflammatory response culminating in
depigmentation. Other scenarios include
molecular mimicry between the insulin
lispro molecule and various melanocyte
surface antigens, resulting in melanocyte
destruction.
This case represents the first report of
lispro insulin analog infusion as an etio-
logic factor in the development of focal
vitiligo in diabetes. Aside from the stan-
dard treatment options for vitiligo, other
options in this case include changing the
type of insulin used, changing the type of
infusion catheter used, and/or changing
the site of insulin infusion. The patient
was changed to insulin aspart and told to
place her infusion catheter into an entirely
new area of abdominal skin. Upon fol-
low-up 6 months later, however, the orig-
inal vitiligo lesions remained unchanged
and new lesions were forming around the
new infusion sites.
MARK R. BURGE,
MD
1
J. DAVID CAREY,
MD
2
From the
1
Department of Medicine, University of
New Mexico Health Science Center, Albuquerque,
New Mexico; and the
2
Albuquerque Dermatology
Associates, Albuquerque, New Mexico.
Address correspondence to Mark R. Burge, MD,
Associate Professor of Medicine, University of New
Mexico School of Medicine, Department of Medi-
cine/Endocrinology—5ACC, Albuquerque, NM
87131. E-mail: mburge@salud.unm.edu.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Kemp EH, Waterman EA, Weetman AP:
Autoimmune aspects of vitiligo. Autoim-
munity 34:65–77, 2001
2. Birlea S, Pop A, Haller M, Maier N, Das
PK: PP-31 A clinical and epidemiological
study on a small community with a prev-
alence of vitiligo (Abstract). Pigment Cell
Res 16:603, 2003
3. Romano G, Moretti G, Di Benedetto A,
Giofre C, Di Cesare E, Russo G, Califano
L, Cucinotta D: Skin lesions in diabetes
mellitus: prevalence and clinical correla-
tions. Diabetes Res Clin Pract 39:101–106,
1998
4. Fineberg NS, Fineberg SE, Anderson JH,
Birkett MA, Gibson RG, Hufferd S: Immu-
nologic effects of insulin lispro [Lys (B28),
Pro (B29) human insulin] in IDDM and
NIDDM patients previously treated with
insulin. Diabetes 45:1750–1754, 1996
5. Lindholm A, Jensen LB, Home PD, Raskin
P, Boehm BO, Rastam J: Immune re-
sponses to insulin aspart and biphasic in-
sulin aspart in people with type 1 and type
2 diabetes. Diabetes Care 25:876–882,
2002
6. Gonzalo MA, De Argila D, Revenga F,
Garcia JM, Diaz J, Morales F: Cutaneous
allergy to human (recombinant DNA) in-
sulin. Allergy 53:106–107, 1998
Aseptic Peritonitis
Revealed Through
Recurrent Catheter
Obstructions in Type
1 Diabetic Patients
Treated with
Continuous
Peritoneal Insulin
Infusion
R
eiterated catheter obstructions
thwart improved diabetes control
with continuous peritoneal insulin
infusion (CPII) from implantable pumps
(1). Occlusions, from either fibrin clots or
omental encapsulations, are promoted by
CPII and diabetes duration and insulin
instability (2,3). Pathological analysis of
encapsulation tissues disclosed, among
predominant collagen fibrosis, inflamma-
tory reactions, including lymphocytes
and amyloid-like deposits reacting to an-
ti-insulin antibodies, surrounded by his-
tiocytes or giant cells (2). However,
catheter obstructions were not related to
high plasma anti-insulin antibody levels
(4,5). Enhanced migration toward insulin
and the chemotactic peptide formyl-
methionyl-leucyl-phenylalanine of
monocyte-issued macrophages from
three patients with previous catheter en-
capsulations suggested that higher mac-
rophage chemotaxis might promote these
events (5). We report two unique obser-
vations of aseptic peritonitis with pre-
dominant macrophagic reactions that
occurred in patients using implantable
pumps with recurrent catheter obstruc-
tions, supporting this hypothesis.
Case 1 is a 54-year-old woman, type 1
diabetes duration 42 years. After 2 years
of CPII, she received an implantable
pump (model MIP 2007; MiniMed, Syl-
mar, CA) in July 2000. In December
2000, a fibrin clot occluding catheter tip
was removed using laparoscopy. A
shorter replacement catheter was im-
planted when obstruction recurred in
May 2001. Following surgery, CPII was
ineffective and ketosis required intrave-
nous insulin delivery. Computerized to-
mography scanning identified peritoneal
fluid accumulation and diffuse thickening
of mesenteric fat, suggesting possible neo-
plasic peritonitis. Laparoscopy revealed
diffuse peritoneal inflammation but no
cancer node. Neither bacterial infection
nor cancer cells were found in peritoneal
fluid, but a high content of fibrin, mono-
cytes, lymphocytes, and macrophages
were found. The catheter tip stuck to the
peritoneum and was surrounded by pre-
dominant macrophages among an inflam-
matory cell reaction. CPII became
effective again only after high doses of oral
prednisone (1 mg 䡠kg
⫺1
䡠day
⫺1
). Pred-
nisone (15 mg/day) remained necessary
to keep CPII effective, with each steroid
interruption resulting in recurrent hyper-
glycemia. No catheter obstruction re-
curred thereafter.
Case 2 is a 62-year-old man, type 1
diabetes duration 30 years, using CPII
since 1981 with previous implantable
pump catheter encapsulations from
1990. He received a new implantable
pump in December 2000. In June 2002,
the catheter encapsulation needed peel-
ing by laparoscopy. Removed tissue
showed a predominantly macrophagic in-
flammatory reaction, including some
lymphocytes, giant cells, and pseudo-
amyloid material among collagen fibrosis.
Catheter obstruction recurred in January
2003. Laparoscopy revealed diffuse peri-
toneal inflammation with whitish urticar-
ia-like plaques. Pathological analysis
identified granulomatous peritoneal le-
sions with histiocytes, fibrosis, and pseu-
do-amyloid material unlabeled by anti-
insulin antibodies. Similar histiocytic
reaction was found in collagen fibrosis
surrounding the catheter tip. Prednisone
(20 mg/day) was prescribed to treat peri-
toneal reaction until pump replacement
in July 2003 because an unexpected
pump failure precluded assessment of ste-
roid effect on CPII efficacy. CPII was ef-
fective with the new pump, and
prednisone could be stopped 2 weeks af-
ter surgery.
Letters
276 DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004

In our experience with 87 patients us-
ing an implantable pump since 1990,
such generalized peritoneal reactions
have not been seen previously or reported
elsewhere. In both cases, a predominant
macrophagic reaction was disclosed, as
previously described in encapsulation tis-
sues (2,5). Long-term CPII or diabetes
likely promoted these events in patients
who appear to be specifically reactive to
peritoneal infusion. Because pseudo-
amyloid material in Case 2 could not be
labeled by anti-insulin antibodies, contri-
bution of insulin in peritoneal macroph-
agic reaction cannot be argued. We
recommend that recurrent implantable
pump catheter obstructions should be ex-
plored by laparoscopy for peritoneal ex-
amination. Although steroid treatment
appeared to be effective on peritonitis and
restored CPII efficacy, continuation of
CPII in such cases must be debated.
ERIC RENARD,
MD, PHD
ISABELLE RAINGEARD,
MD
GUY COSTALAT,
MD
DOMINIQUE APOSTOL,
MD
DOMINIQUE LAUTON,
MD
FRANC¸OISE BOULET,
MD
JACQUES BRINGER,
MD
From the Endocrinology Department, Lapeyronie
Hospital, Montpellier, France.
Address correspondence to Eric Renard, MD,
PHD, Endocrinology Department, Lapeyronie Hos-
pital, F 34295 Montpellier cedex 5, France. E-mail:
e-renard@chu-montpellier.fr.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Broussolle C, Jeandidier N, Hanaire-
Broutin H, the Evadiac Study Group:
French multicentre experience with im-
plantable insulin pumps. Lancet 343:
514–515, 1994
2. Renard E, Baldet P, Picot MC, Jacques-
Apostol D, Lauton D, Costalat G, Bringer
J, Jaffiol C: Catheter complications with
implantable systems for peritoneal insulin
delivery: an analysis of frequency, predis-
posing factors, and obstructing materials.
Diabetes Care 18:300–306, 1995
3. Renard E, Bouteleau S, Jacques-Apostol
D, Lauton D, Boulet-Gibert F, Costalat G,
Bringer J, Jaffiol C: Insulin underdelivery
from implanted pumps using peritoneal
route: determinant role of insulin-pump
compatibility. Diabetes Care 19:812–817,
1996
4. Olsen CL, Chan E, Turner DS, Iravani M,
Nagy M, Selam JL, Wong ND, Waxman K,
Charles MA: Insulin antibody responses
after long-term intraperitoneal insulin ad-
ministration via implantable programma-
ble insulin delivery systems. Diabetes Care
17:169–176, 1994
5. Kessler L, Tritschler S, Bohbot A, Sigrist S,
Karsten V, Boivin S, Dufour P, Belcourt A,
Pinget M: Macrophage activation in type 1
diabetic patients with catheter obstruc-
tion during peritoneal insulin delivery
with an implantable pump. Diabetes Care
24:302–307, 2001
Comparative Study
of Prognostic Value
for Coronary Disease
Risk Between the
U.K. Prospective
Diabetes Study and
Framingham Models
A
ccording to epidemiological and
angiographic studies (1,2), diabetic
patients present a two to four times
greater risk for coronary artery disease
(CAD) than nondiabetic individuals.
The Framingham model (3) estimates
10-year CAD risk based on the traditional
risk factors, including age, sex, HDL and
LDL cholesterol, hypertension, and
smoking. In addition to these risk factors,
the U.K. Prospective Diabetes Study
(UKPDS) model, designed for people
with type 2 diabetes (4), incorporates
more specific variables, such as HbA
1c
,
age at diabetes diagnosis, and diabetes
duration.
The aim of this analysis was to com-
pare the accuracy of these models in the
prediction of 10-year risk for CAD in di-
abetic patients.
Clinical information related to the
above factors was retrieved from our dia-
betic outpatient database with 10-year
clinical follow-up. Of the 339 participants
(53% women and 47% men), 108 (32%)
presented with CAD. We did not observe
any statistically significant differences in
their demographic characteristics. The di-
agnosis of CAD was established by coro-
nary angiography. Diabetic patients
without a history of CAD were not con-
sidered to have CAD; however, if they had
presented symptoms of CAD, they would
undergo a treadmill test and myocardium
scanning with Th 201 (single photon
emission computerized tomography).
Receiver operating characteristic
curves were constructed for both models
to evaluate their accuracy in the predic-
tion of 10-year CAD risk (measured by
the area under the receiver operating
characteristic curve, range 0.5–1).
Areas under the curves were 0.61
(P⬍0.01) and 0.65 (P⬍0.01) for
UKPDS and Framingham, respectively.
The comparative analysis showed similar
sensitivity (56 vs. 55%) between these
models. A higher specificity in the Fram-
ingham model (65 vs. 56%) was noted.
We also noted higher positive and nega-
tive predictive values of Framingham, 43
and 75%, respectively, compared with 37
and 73% in the UKPDS.
According to the results of this analy-
sis, the Framingham model seems to be
more appropriate for the prediction of
CAD risk in diabetic patients. An expla-
nation for this could be the small contri-
bution of UKPDS variables (HbA
1c
, age at
diabetes diagnosis, and diabetes dura-
tion) to the 10-year risk for CAD.
Regarding the relation of HbA
1c
to the
development of CAD, data from the
UKPDS 23 (5) indicated that for each 1%
increment in HbA
1c
there was a 1.11-fold
increased risk of CAD, whereas for each
1-mmol/l increment in LDL concentra-
tion there was a 1.57-fold increased risk.
It should also be noted that an HbA
1c
in-
crement from 6.5 to 11% (6) just doubles
the risk of myocardial infarctions,
whereas an HbA
1c
increment of 1% mul-
tiplies the risk of microangiopathic inci-
dents by 10. Additionally, recent studies
(7,8) have confirmed the significant role
of traditional risk factors in the prediction
of CAD in contrast to the poor prognostic
value of blood glucose concentration. In
conclusion, comparing the above data,
the association of HbA
1c
with the risk for
CAD is considered rather weak.
Regarding the contribution of age at
diabetes diagnosis and diabetes duration
to the 10-year risk, it is well known that in
the early stages of the disease, when the
symptoms are not apparent, diabetes is
frequently underdiagnosed. Moreover, in
some cases, although diabetes complica-
tions (micro- and/or macrovascular) have
already presented, the existence of diabe-
tes is still ignored by the patient. How-
ever, it is difficult to determine the exact
age of diabetes onset and duration. There-
fore, the evaluation of the exact duration
of diabetes is potentially inaccurate,
which may result in underestimation of
CAD risk.
Letters
DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004 277

Future efforts should focus on the ul-
timate estimation and evaluation of the
prognostic value of both models using
randomized, prospective, comparative
studies.
IOANNIS D. PROTOPSALTIS,
MD, PHD
PANAYIOTIS A. KONSTANTINOPOULOS,
MD
ALEXANDROS V. KAMARATOS,
MD, PHD
ANDREAS I. MELIDONIS,
MD, PHD
From the Diabetes Center, Tzaneio General Hospital
of Piraeus, Piraeus, Greece.
Address correspondence to Dr. Ioannis D. Pro-
topsaltis, Tzaneio General Hospital of Piraeus, Dia-
betes Center, Zanni & Afentouli 1, Piraeus, 185 36,
Greece. E-mail: tzaniodiabetes@yahoo.com.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Melidonis A, Dimopoulos V, Lampidakis
E, Hatzisassavas J, Kouvaras G, Stefanidis,
Foussas S: Angiographic study of coro-
nary artery disease in diabetic patients in
comparison with nondiabetic patients.
Angiology 12:997–1006, 1999
2. Kannel WB, McGee DL: Diabetes and car-
diovascular disease: the Framingham
study. JAMA 241:2035–2038
3. Wilson PWF, D’Agostino RB, Levy D, Be-
langer AM, Silbershatz H, Kannel WB:
Prediction of coronary heart disease risk
using risk factor categories. Circulation
97:1837–1847, 1998
4. Stevens RJ, Kothari V, Adler AI, Stratton
IM, United Kingdom Prospective Diabe-
tes Study (UKPDS) Group: The UKPDS
risk engine: a model for the risk of coro-
nary heart disease in type II diabetes (UK-
PDS 56). Clin Sci (Lond) 101:671–679,
2001
5. Turner RC, Millns H, Neil HAW, Stratton
IM, Manley SE, Matthews DR, Holman
RR: Risk factors for coronary artery dis-
ease in non-insulin dependent diabetes
mellitus: United Kingdom Prospective Di-
abetes Study (UKPDS 23). BMJ 316:823–
828, 1998
6. Stratton IM, Adler AI, Neil HA, Mathews
DR, Manley SE, Cull CA, Hadden D,
Turner RC, Holman RR: Association of
glycaemia with macrovascular and micro-
vascular complications of type 2 diabetes
(UKPDS 35): prospective observational
study. BMJ 321:405–412, 2000
7. Stern MP, Fatehi P, Williams K, Haffner
SM: Predicting future cardiovascular dis-
ease: do we need the oral glucose toler-
ance test? Diabetes Care 25:1851–1856,
2002
8. Meigs JB, Nathan DM, D’Agostino RB,
Wilson PWF: Fasting and postchallenge
glycemia and cardiovascular risk: the Fra-
mingham Offspring Study. Diabetes Care
25:1845–1850, 2002
The Biological
Variation of Sex
Hormone–Binding
Globulin in Type 2
Diabetes
Implications for sex hormone–
binding globulin as a surrogate
marker of insulin resistance
Q
uantitative determination of insulin
resistance is technically demanding
and expensive. We have recently
shown (1) that insulin resistance deter-
mined using the homeostasis model as-
sessment of insulin resistance (HOMA-
IR) has a significantly greater biological
variability in individuals with type 2 dia-
betes than in healthy ones. A surrogate
marker of insulin resistance that was re-
producible, stable, and easily measured
would be invaluable for both research and
clinical practice, particularly for following
insulin-sensitizing therapy, such as met-
formin and the thiazolidinediones. A low
sex hormone–binding globulin (SHBG)
concentration reflects hyperinsulinemic
insulin resistance and has been proposed
as such a surrogate measure (2–4). This
study aimed to compare the biological
variation of SHBG and insulin resistance
in type 2 diabetes to determine the poten-
tial for SHBG as a surrogate marker of in-
sulin resistance in type 2 diabetes.
Subjects were initially recruited for a
study to assess the biological variation of
insulin resistance in individuals with type
2 diabetes (1). Postmenopausal Cauca-
sian subjects (n⫽12) with type 2 diabe-
tes (median age 62 years, range 50–73,
and median BMI 31.6 kg/m
2
, range 25.1–
35.7) and 11 age- and weight-matched,
healthy, postmenopausal Caucasian con-
trol subjects (median age 56 years, range
48–70, and median BMI 32.0 kg/m
2
,
range 26.6–44.4) participated. Fasting
blood samples were collected at 4-day in-
tervals on 10 consecutive occasions.
Plasma glucose was analyzed in singleton
within4hofcollection. Duplicate sam-
ples (i.e., two per visit) of stored serum
were randomized and then analyzed (in a
single continuous batch using a single
batch of reagents) for SHBG and insulin
(on a DPC Immulite 2000 analyzer; Euro/
DPC, Llanberis, U.K.). The coefficient of
variation for serum insulin and SHBG was
10.6% and 8.5%, respectively. The ana-
lytical sensitivity of the insulin assay was 2
␮U/ml, and there was no stated cross-
reactivity with proinsulin. All subjects
were asked to have an unrestricted diet
and instructed not to modify their eating
patterns during the sampling period. The
subjects were also advised to refrain from
excessive physical exercise and alcohol
before each fasting blood test.
The insulin resistance was calculated
using the HOMA-IR method (HOMA-
IR ⫽[insulin ⫻glucose]/22.5). Biovari-
ability data were analyzed by calculating
analytical, within-subject, and between-
subject variances (5,6). The critical differ-
ence (i.e., the smallest percentage change
unlikely to be due to biological variabil-
ity) between two consecutive SHBG sam-
ples in an individual subject with type 2
diabetes was calculated using the formula
2.77(CV
I
) (5), where CV
I
is the within-
subject biological coefficient of variation.
Figure 1 shows the mean and range of
HOMA-IR and SHBG for the individuals
in the two groups. In the type 2 diabetic
group, SHBG concentrations were lower
than those in the control subjects
(mean ⫾SD; 38.8 ⫾18.2 vs. 42.2 ⫾17.1
nmol/l, P⫽0.001), the insulin levels
were higher (13.1 ⫾5.4 vs. 9.42 ⫾3.4
␮IU/ml, P⫽0.0001), and the HOMA-IR
greater (4.33 ⫾2.3 vs. 2.11 ⫾0.79 units,
P⫽0.0001).
An inverse relationship was demon-
strated between SHBG concentration and
HOMA-IR in the group with type 2 dia-
betes (r⫽⫺0.32, P⫽0.001) and in con-
trol subjects (r⫽⫺0.28, P⫽0.003). The
intraindividual variance of SHBG rose lin-
early with increasing SHBG concentra-
tions (r⫽0.82, P⫽0.0001), and after
accounting for analytical variation, the
intraindividual variation of SHBG for the
group with type 2 diabetes was similar to
that seen in the control group (mean 2.35
vs. 2.44 nmol/l, P⫽0.93). In contrast,
the mean intraindividual variation of se-
rum insulin (mean 2.38 vs. 1.45 ␮U/ml,
P⫽0.016) and HOMA-IR (mean 1.05 vs.
0.15 units, P⫽0.001) was significantly
greater in the group with type 2 diabetes
than in the control subjects.
The critical difference between two
consecutive SHBG samples in an individ-
ual patient with type 2 diabetes was
14.5% at any initial level of SHBG, indi-
cating that a subsequent sample must rise
Letters
278 DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004

Figure 1—Means (range) of insulin resistance and SHBG (unadjusted for analytical variation) in control subjects and type 2 diabetic subjects.
Letters
DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004 279

or fall by ⬎14.5% to be considered signif-
icantly different from the first.
The subjects with type 2 diabetes
were hyperinsulinemic, insulin resistant,
and demonstrated lower SHBG levels
than control subjects. However, the more
variable fasting insulin/insulin resistance
in the subjects with type 2 diabetes was
not reflected by similarly more variable
SHBG readings compared with those of
the control subjects. This suggests that a
low SHBG concentration is a stable inte-
grated marker of insulin resistance and
therefore has the characteristics to be po-
tentially used as a surrogate measure of
insulin resistance, perhaps in monitoring
the response of an individual to insulin
sensitizers. However, although SHBG lev-
els differed significantly between those
with and without diabetes, the absolute
mean difference was small, indicating that
measurement of SHBG cannot be used as
a simple test for insulin resistance in dia-
betes. A much larger study is required to
investigate whether diagnostic cutoff val-
ues for low SHBG concentrations and in-
sulin resistance in type 2 diabetes can be
established. Without these parameters,
the utility of a low SHBG concentration as
areflection of insulin resistance in type 2
diabetes will be for the serial monitoring
of insulin resistance in individuals on
treatment after the presence of insulin re-
sistance has been established by conven-
tional means. The low variation of SHBG
compared with insulin resistance is likely
due to the inherent temporal volatility of
insulin and glucose levels as compared
with SHBG. In conclusion, in the evalua-
tion of serial measurements of SHBG con-
centration for an insulin-resistant
individual with type 2 diabetes, such as
before and after therapeutic intervention,
the critical difference value of 14.5% re-
ported here will identify whether any
change is beyond that of natural biologi-
cal variation and therefore a true
response.
VIJAY JAYAGOPAL,
MRCP
1
ERIC S. KILPATRICK,
MRCPATH
2
PAUL E. JENNINGS,
FRCP
3
STEVE HOLDING,
PHD
2
DAVID A. HEPBURN,
FRCP
1
STEPHEN L. ATKIN,
FRCP
1
From the
1
Department of Medicine, University of
Hull, Hull, U.K.; the
2
Department of Clinical Bio-
chemistry, Hull Royal Infirmary, Hull, U.K.; and the
3
Department of Medicine, York Hospital, York, U.K.
Address correspondence to Dr. V. Jayagopal, Mi-
chael White Centre for Diabetes and Endocrinology,
Brocklehurst Building, Hull Royal Infirmary, 220-
236 Anlaby Road, Hull, HU3 2RW, U.K. E-mail:
v.jayagopal@hull.ac.uk.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Jayagopal V, Kilpatrick ES, Jennings PE,
Hepburn DA, Atkin SL: Biological varia-
tion of homeostasis model assessment–
derived insulin resistance in type 2
diabetes. Diabetes Care 25:2022–2025,
2002
2. Nestler JE: Sex hormone-binding globu-
lin: a marker for hyperinsulinemia and/or
insulin resistance? (Editorial). J Clin Endo-
crinol Metab 76:273–274, 1993
3. Pugeat M, Crave JC, Tourniaire J, Forest
MG: Clinical utility of sex hormone-bind-
ing globulin measurement. Horm Res 45:
148–155, 1996
4. Jayagopal V, Kilpatrick ES, Jennings PE,
Hepburn DA, Atkin SL: The biological
variation of testosterone and sex hor-
mone-binding globulin (SHBG) in poly-
cystic ovarian syndrome: implications for
SHBG as a surrogate marker of insulin re-
sistance. J Clin Endocrinol Metab 88:1528–
1533, 2003
5. Fraser CG, Harris EK: Generation and ap-
plication of data on biological variation in
clinical chemistry. Crit Rev Clin Lab Sci 27:
409–437, 1989
6. Gowans EM, Fraser CG: Biological varia-
tion of serum and urine creatinine and
creatinine clearance: ramifications for in-
terpretation of results and patient care.
Ann Clin Biochem 25:259–263, 1988
Development of an
Assessment Tool for
Screening Children
for Glucose
Intolerance by Oral
Glucose Tolerance
Test
T
he American Diabetes Association
(ADA) has recommended screening
for type 2 diabetes by fasting plasma
glucose (FPG) in children who are over-
weight (BMI ⬎85th percentile) who have
two of the following risk factors: at-risk
ethnic minority origin, family history of
diabetes in a first- or second-degree rela-
tive, or insulin resistance (acanthosis nig-
ricans, polycystic ovarian syndrome,
hypertension, or dyslipidemia) (1).
The case for refining the criteria for
screening has been made previously (2).
In that study, the sensitivity of the criteria
was 24%, with a positive predictive value
of 3%, i.e., 40 children needed to be
tested to yield one abnormal result. In a
response to this, Rosenbloom (3) high-
lighted the need to test the ADA criteria in
high-risk populations to establish the
strength and risk level of different factors
that are influential in the development of
type 2 diabetes.
Our institution covers a population
where type 2 diabetes in childhood has
emerged (4). We describe our experience
of applying the ADA criteria and propose
a clinical assessment tool to refine the se-
lection of children for screening by the
oral glucose tolerance test (OGTT).
In the last 4 years, 66 children have
had OGTTs for suspected glucose intoler-
ance. The characteristics of this popula-
tion were mean age of 12.7 years (range
4.8–17.3), mean BMI standard deviation
score 3.0 (0.0–4.6), 71% female, 83%
ethnic minority origin (of whom 73%
were South Asian and 9% African Carib-
bean), 88% had acanthosis nigricans, and
67% had a first- or second-degree family
history of diabetes. Of these, 13 children
had abnormal glucose tolerance (4 diabe-
tes, 8 impaired glucose tolerance [IGT],
and 1 impaired fasting glycemia).
Applying the ADA criteria, 11 of the
13 children with abnormal results would
have qualified for screening, missing 1
child with diabetes and 1 with IGT.
Screening these 11 with FPG as per the
recommendations would have missed a
further 7 children, 1 with diabetes and 6
with IGT, as only 1 of the children with
IGT had impaired fasting glycemia. Over-
all, the sensitivity of the ADA criteria us-
ing FPG as a screening test in our
population was 31%. Use of the ADA cri-
teria to screen by OGTT would have iden-
tified 11 of the 13 abnormal results in this
cohort, giving a sensitivity of 85% for the
criteria and a specificity of 26%, with a
positive predictive value of 22%, i.e., five
children would need to be tested to yield
an abnormal result.
Using the clinical characteristics of
our cohort, we calculated the positive pre-
dictive value for each parameter singly
and in combination. We used this data to
weight each parameter and calculate a cu-
mulative risk score, dividing children into
low and high risk of abnormal glucose
tolerance when tested by OGTT (Fig. 1).
We then applied this risk score to our co-
Letters
280 DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004

hort: 36 children scored as high risk,
which included all 13 children with ab-
normal glucose tolerance. The sensitivity
of this scoring system in our cohort is
100%, specificity 57%, and positive pre-
dictive value 36%, i.e., three children
need to be tested to identify one abnormal
result. Prospective validation of this as-
sessment tool is now underway in our
clinic. We suggest that other units de-
velop risk assessment tools for their pop-
ulations so that screening at-risk children
for glucose intolerance becomes more fea-
sible and is not “a lost battle.”
SARAH EHTISHAM,
MBBCHIR
NICK SHAW,
MBCHB
JEREMY KIRK,
MD
TIMOTHY BARRETT,
PHD
From the Department of Diabetes and Endocrinol-
ogy, Birmingham Children’s Hospital, Birmingham,
U.K.
Address correspondence to Dr. Sarah Ehtisham,
Diabetes Homecare Unit, Birmingham Children’s
Hospital, Steelhouse Lane, Birmingham B4 6NH,
U.K. E-mail: s.ehtisham@bham.ac.uk.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. American Diabetes Association: Type 2
diabetes in children and adolescents
(Consensus Statement). Diabetes Care 23:
381–389, 2000
2. Fagot-Campagna A, Saaddine JB, Engel-
gau MM: Is testing children for type 2 di-
abetes a lost battle? (Letter). Diabetes Care
23:1442–1443, 2000
3. Rosenbloom AL: Is testing children for
type 2 diabetes a lost battle? Response to
Fagot-Campagna et al. (Letter) Diabetes
Care 23:1443, 2000
4. Ehtisham S, Barrett TG, Shaw NJ: Type 2
diabetes mellitus in UK children: an
emerging problem. Diabet Med 17:867–
871, 2000
5. Williams CL, Hayman LL, Daniels SR,
Robinson TN, Steinberger J, Paridon S,
Bazzarre T: Cardiovascular health in
childhood: a statement for health profes-
sionals from the Committee on Athero-
sclerosis, Hypertension and Obesity in
the Young (AHOY) of the Council on Car-
diovascular Disease in the Young, Ameri-
can Heart Association. Circulation 106:
143–160, 2002
Vinegar Improves
Insulin Sensitivity to
a High-Carbohydrate
Meal in Subjects
With Insulin
Resistance or Type 2
Diabetes
T
he number of Americans with type 2
diabetes is expected to increase by
50% in the next 25 years; hence, the
prevention of type 2 diabetes is an impor-
tant objective. Recent large-scale trials
(the Diabetes Prevention Program and
STOP-NIDDM) have demonstrated that
therapeutic agents used to improve insu-
lin sensitivity in diabetes, metformin and
acarbose, may also delay or prevent the
onset of type 2 diabetes in high-risk pop-
Figure 1—Assessment tool for the selection of children at risk of abnormal glucose tolerance. The assessment tool does not apply to children referred
with raised fasting blood glucose values or with suspected glucose intolerance posttransplantation or secondary to other diagnoses. Children are
scored on the clinical characteristics as shown. The scores for family history and insulin resistance are cumulative, giving a maximum score of 24.
*Definite diagnosis on scan or irregular menses and hirsutism; †systolic or diastolic blood pressure ⬎95th percentile for age (5). Scores: 0–11, low
risk; 12–24, high risk (these children should be screened by OGTT).
Letters
DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004 281

ulations. Interestingly, an early report
showed that vinegar attenuated the glu-
cose and insulin responses to a sucrose or
starch load (1). In the present report, we
assessed the effectiveness of vinegar in re-
ducing postprandial glycemia and insu-
linemia in subjects with varying degrees
of insulin sensitivity.
Our study included nondiabetic sub-
jects who were either insulin sensitive
(control subjects, n⫽8) or insulin resis-
tant (n⫽11) and 10 subjects with type 2
diabetes. Subjects provided written in-
formed consent and were not taking dia-
betes medications. Fasting subjects were
randomly assigned to consume the vine-
gar (20 g apple cider vinegar, 40 g water,
and 1 tsp saccharine) or placebo drink
and, after a 2-min delay, the test meal,
which was composed of a white bagel,
butter, and orange juice (87 g total carbo-
hydrates). The cross-over trial was con-
ducted 1 week later. Blood samples were
collected at fasting and 30 and 60 min
postmeal for glucose and insulin analyses.
Whole-body insulin sensitivity during the
60-min postmeal interval was estimated
using a composite score (2).
Fasting glucose concentrations were
elevated ⬃55% in subjects with diabetes
compared with the other subject groups
(P⬍0.01, Tukey’s post hoc test), and
fasting insulin concentrations were ele-
vated 95–115% in subjects with insulin
resistance or type 2 diabetes compared
with control subjects (P⬍0.01). Com-
pared with placebo, vinegar ingestion
raised whole-body insulin sensitivity dur-
ing the 60-min postmeal interval in insu-
lin-resistant subjects (34%, P⫽0.01,
paired ttest) and slightly improved this
parameter in subjects with type 2 diabetes
(19%, P⫽0.07). Postprandial fluxes in
insulin were significantly reduced by vin-
egar in control subjects, and postprandial
fluxes in both glucose and insulin were
significantly reduced in insulin-resistant
subjects (Fig. 1).
These data indicate that vinegar can
significantly improve postprandial insu-
lin sensitivity in insulin-resistant subjects.
Acetic acid has been shown to suppress
disaccharidase activity (3) and to raise
glucose-6-phosphate concentrations in
skeletal muscle (4); thus, vinegar may
possess physiological effects similar to
acarbose or metformin. Further investiga-
tions to examine the efficacy of vinegar as
an antidiabetic therapy are warranted.
CAROL S. JOHNSTON,
PHD
CINDY M. KIM,
MS
AMANDA J. BULLER,
MS
From the Department of Nutrition, Arizona State
University, Mesa, Arizona.
Address correspondence to Carol S. Johnston,
Department of Nutrition, Arizona State University,
East Campus, 7001 E. Williams Field Rd, Mesa, AZ
85212. E-mail: carol.johnston@asu.edu.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Ebihara K, Nakajima A: Effect of acetic
acid and vinegar on blood glucose and
insulin responses to orally administered
sucrose and starch. Agric Biol Chem 52:
1311–1312, 1988
2. Matsuda M, DeFronzo RA: Insulin sensi-
tivity indices obtained from oral glucose
tolerance testing. Diabetes Care 22:1462–
1470, 1999
3. Ogawa N, Satsu H, Watanabe H, Fukaya
M, Tsukamoto Y, Miyamoto Y, Shimizu
M: Acetic acid suppresses the increase in
disaccharidase activity that occurs during
culture of Caco-2 cells. J Nutr 130:507–
513, 2000
4. Fushimi T, Tayama K, Fukaya M, Kitako-
shi K, Nakai N, Tsukamoto Y, Sato Y: Ace-
tic acid feeding enhances glycogen
repletion in liver and skeletal muscle of
rats. J Nutr 131:1973–1977, 2001
Sleep Disturbance
and Onset of Type 2
Diabetes
S
leep disturbance, which is often ob-
served among patients with diabetes
(1), is possibly caused by impaired
glucose metabolism or physical and psy-
chological discomfort due to the disorder.
In addition, a recent prospective study of
women has indicated an interesting asso-
ciation between sleep patterns and later-
onset type 2 diabetes, with a greater
incidence among both short-term (⬍6h)
and long-term (⬎8 h) sleepers (2). Dis-
turbance in sleep quality may also affect
the later onset of overt diagnosis of type 2
diabetes. We investigated the association
between sleep disturbance and the subse-
quent onset of type 2 diabetes in a group
of Japanese male employees.
We analyzed the database of an 8-year
prospective study of male employees of an
electrical company in Japan (3). We fol-
lowed 2,649 male employees with no
medical history of diabetes or other
chronic illnesses at baseline for 8 years
from 1984 to 1992. Data from 2,265
(86%) male respondents, who were thor-
oughly followed, were analyzed. All sub-
jects received a medical checkup once a
year during the follow-up to identify
those with type 2 diabetes according to
World Health Organization criteria (4). A
mailed questionnaire was used to assess
sleep disturbance in the previous month
at baseline. Two single-item questions
Figure 1—Effects of vinegar (䡺) and placebo (}) on plasma glucose (A–C) and insulin (D–F)
responses after a standard meal in control subjects, insulin-resistant subjects, and subjects with
type 2 diabetes. Values are means ⫾SE. The Pvalues represent a significant effect of treatment
(multivariate ANOVA repeated-measures test).
Letters
282 DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004

were asked concerning difficulty initiat-
ing sleep (“Did you have trouble falling
asleep?”) and difficulty maintaining sleep
(“Did you often wake up in the middle of
the night?”). The subjects were classified
into one of two categories: low for those
who indicated “seldom”or “sometimes”
and high for those who indicated “often”
or “almost everyday”in response to the
questions.
During the 18,006 person-year ob-
servation, 38 incidents of type 2 diabetes
were identified (an incidence rate of 1.68
per 1,000 person-years). Those who ex-
perienced a high frequency of difficulty
initiating sleep had a significantly higher
age-adjusted hazard ratio (2.98, 95% CI
1.36–6.53) for type 2 diabetes compared
with those who experienced low-
frequency difficulty initiating sleep. A
similar hazard ratio was observed for dif-
ficulty maintaining sleep (2.23, 1.08–
4.61). These hazard ratios were almost
identical and were statistically significant
after controlling for other factors relevant
to type 2 diabetes (i.e., age, education,
occupation, shift work, BMI, leisure-time
physical activity, smoking, alcohol con-
sumption, and family history of diabetes).
Our analysis demonstrated that those
who had sleep disturbances showed a
two- to threefold higher risk of later onset
type 2 diabetes. The association was inde-
pendent of known risk factors for type 2
diabetes and was not attributable to treat-
ment for sleep disturbance. Sleep distur-
bance at baseline was unlikely due to
complications or disability from the treat-
ment of diabetes because we excluded
subjects with known diabetes at baseline.
A possible explanation is that an in-
creased sympathetic nervous activity as-
sociated with sleep disturbance (5) causes
glucose intolerance (6) and increases the
risk of type 2 diabetes. Physicians may
need to pay more attention to patients
with sleep disturbance because it may in-
dicate a higher risk for type 2 diabetes.
NORITO KAWAKAMI,
MD
1
NAOYOSHI TAKATSUKA,
MD
2
HIROYUKI SHIMIZU,
MD
2
From the
1
Department of Hygiene and Preventive
Medicine, Okayama University Graduate School of
Medicine and Dentistry, Okayama, Japan; and the
2
Department of Epidemiology and Preventive Med-
icine, Gifu University School of Medicine, Gifu,
Japan.
Address correspondence to Prof. Norito
Kawakami, Hygiene and Preventive Medicine,
Okayama University Graduate School of Medicine
and Dentistry, 2-5-1 Shikata-cho, Okayama 700-
8558, Japan. E-mail: norito@md.okayama-u.ac.jp.
© 2004 by the American Diabetes Association.
Acknowledgments—This study was sup-
ported in part by a grant-in-aid from the Jap-
anese Ministry of Education, Science, and
Culture.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Resnick HE, Redline S, Shahar E, Gilpin
A, Newman A, Walter R, Ewy GA,
Howard BV, Punjabi NM: Diabetes and
sleep disturbances: findings from the
Sleep Heart Health Study. Diabetes Care
26:702–709, 2003
2. Ayas NT, White DP, Al-Delaimy WK,
Manson JE, Stampfer MJ, Speizer FE, Pa-
tel S, Hu FB: A prospective study of self-
reported sleep duration and incident
diabetes in women. Diabetes Care
26:380–384, 2003
3. Kawakami N, Takatsuka N, Shimizu H,
Ishibashi H: Depressive symptoms and
occurrence of type 2 diabetes among Jap-
anese men. Diabetes Care 22:1071–1076,
1999
4. World Health Organization: WHO Expert
Committee on Diabetes Mellitus, Second Re-
port. Geneva, World Health Org., 1981
(Tech. Rep. Ser., no. 646)
5. Bonnet MH, Arand DL: Heart rate vari-
ability in insomniacs and matched normal
sleepers. Psychosom Med 60:610–615,
1998
6. Gonzalez-Ortiz M, Martinez-Abundis E,
Balcazar-Munoz BR, Pascoe-Gonzalez S:
Effect of sleep deprivation on insulin sen-
sitivity and cortisol concentration in
healthy subjects. Diabetes Nutr Metab 13:
80–83, 2000
Non-HDL Cholesterol
Contributes to the
“Hypertriglyceridemic
Waist”as a
Cardiovascular Risk
Factor
The Hoorn Study
L
emieux et al. (1) described the “hy-
pertriglyceridemic waist”as a
marker of the atherogenic metabolic
triad (hyperinsulinemia, hyperapoli-
poprotein B, and small, dense LDL) in
men. In 287 men, those with a waist cir-
cumference ⱖ90 cm and with triglyceride
levels ⱖ2 mmol/l had an odds ratio of 3.6
(95% CI 1.17–10.93) for having angio-
graphically diagnosed coronary artery
disease compared with those with smaller
waists and lower triglyceride concentra-
tions. We recently reported (2) the pre-
dictive value of non-HDL cholesterol and
triglyceride concentrations for 10-year
cardiovascular disease incidence in the
Hoorn Study, a population-based cohort
study of glucose tolerance. In people with
abnormal glucose metabolism, high tri-
glyceride concentration was associated
with the risk of cardiovascular disease,
particularly in people with high non-HDL
cholesterol, but not in those with normal
glucose metabolism (3).
We used the Hoorn Study data to pro-
spectively investigate whether the risk as-
sociated with the hypertriglyceridemic
waist differs between subjects with nor-
mal and abnormal glucose metabolism.
Additionally, we studied whether non-
HDL cholesterol adds to the predictive
power of the hypertriglyceridemic waist
in predicting cardiovascular disease. The
Hoorn Study is a cohort study among
2,484 subjects in the Netherlands that
started in 1989. Cardiovascular disease
was defined as first new cardiovascular
fatal or nonfatal event (3). Because the
cutoff points for waist girth in men used
by Lemieux et al. (1) are not applicable for
women, we used waist ⱖ94 cm in men
and ⱖ80 cm in women, according to the
European Group of Insulin Resistance
definition (4). The results show that in
particular the combination of a large waist
and a high triglyceride level (ⱖ2 mmol/l)
was associated with cardiovascular dis-
ease in subjects with both normal and ab-
normal glucose metabolism (hazard ratio
1.82 [95% CI 1.27–2.62] and 2.68
[1.89–3.81], respectively). These find-
ings concur with those of Lemieux et al.
(1). In addition, we observed that after
stratification for non-HDL cholesterol,
defined as the difference between total
cholesterol and HDL cholesterol concen-
tration, the risk associated with the hyper-
triglyceridemic waist was further
increased by 50% in the presence of high
non-HDL cholesterol concentrations
(above the median, i.e., 5.2 mmol/l for
men and 5.3 mmol/l for women) (Fig. 1).
The hazard ratio for subjects with a com-
bination of large waist, high triglycerides,
and high non-HDL cholesterol concentra-
tions was 2.94 (2.06–4.19).
Non-HDL cholesterol is closely
linked to visceral obesity (5). Non-HDL
cholesterol includes all cholesterol in po-
Letters
DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004 283

tentially atherogenic triglyceride-rich li-
poproteins and may be a better predictor
for the “bad”triglycerides, which are as-
sociated with increased risk of cardiovas-
cular disease (6).
In conclusion, in the Hoorn Study,
non-HDL cholesterol contributes consid-
erably to the risk associated with the hy-
pertriglyceridemic waist. Further studies
are clearly required to evaluate the clinical
relevance of monitoring these particular
variables for the assessment of cardiovas-
cular risk.
GRIE¨TBOS,
MSC
1
JACQUELINE M. DEKKER,
PHD
1
ROBERT J. HEINE,
PHD
1,2
From the
1
Institute for Research in Extramural Med-
icine, VU University Medical Center, Amsterdam,
the Netherlands; and the
2
Department of Endocri-
nology, Diabetes Center, VU University Medical
Center, Amsterdam, the Netherlands.
Address correspondence to G. Bos, MSc, Institute
for Research in Extramural Medicine, VU University
Medical Center Van der Boechorststraat 7, 1081 BT
Amsterdam, Netherlands. E-mail: g.bos.emgo@med.
vu.nl.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Lemieux I, Pascot A, Couillard C,
Lamarche B, Tchernof A, Almeras N,
Bergeron J, Gaudet D, Tremblay G,
Prud’homme D, Nadeau A, Despres JP:
Hypertriglyceridemic waist: a marker of
the atherogenic metabolic triad (hyperin-
sulinemia; hyperapolipoprotein B; small,
dense LDL) in men? Circulation 102:179–
184, 2000
2. Mooy JM, Grootenhuis PA, de Vries H,
Valkenburg HA, Bouter LM, Kostense PJ,
Heine RJ: Prevalence and determinants of
glucose intolerance in a Dutch Caucasian
population: the Hoorn Study. Diabetes
Care 18:1270–1273, 1995
3. Bos G, Dekker JM, Nijpels G, De Vegt F,
Diamant M, Stehouwer CD, Bouter LM,
Heine RJ: A combination of high concen-
trations of serum triglyceride and non-
high-density-lipoprotein-cholesterol is a
risk factor for cardiovascular disease in
subjects with abnormal glucose metabo-
lism: the Hoorn Study. Diabetologia 46:
910–916, 2003
4. Balkau B, Charles MA: Comment on the
provisional report from the WHO consul-
tation: European Group for the Study of
Insulin Resistance (EGIR). Diabet Med 16:
442–443, 1999
5. Bittner V, Hardison R, Kelsey SF, Weiner
BH, Jacobs AK, Sopko G: Non-high-
density lipoprotein cholesterol levels pre-
dict five-year outcome in the Bypass
Angioplasty Revascularization Investiga-
tion (BARI). Circulation 106:2537–2542,
2002
6. Brewer HB Jr: Hypertriglyceridemia:
changes in the plasma lipoproteins asso-
ciated with an increased risk of cardiovas-
cular disease. Am J Cardiol 83:3F–12F,
1999
Improvement of
Temperature and
Flow in Feet of
Subjects with
Diabetes With Use
of a Transdermal
Preparation
of L-Arginine
A pilot study
C
irculatory impairment and its se-
qulae have long been known to be
major complications of diabetes. It
has been shown that in diabetes, the func-
tionality of the endothelial nitric oxide
(NO)/nitric oxide synthase (eNOS) sys-
tem is impaired (1–3). NO is generated in
the endothelium through the oxidation of
the amino acid L-arginine by the enzyme
eNOS. NO causes vascular smooth mus-
cle to relax, resulting in increased blood
flow. In addition to being a substrate of
eNOS, L-arginine facilitates the dimeriza-
tion of two identical subunits, forming a
homodimer. The enzyme is only active in
the dimeric form. Under proper condi-
tions, dimerization occurs rapidly, on a
timescale of minutes. Once formed, the
dimer is stable (4).
Subjects with diabetes have abnor-
mally low levels of L-arginine (5) and ele-
vated levels of the eNOS inhibitor
asymmetric dimethylarginine (ADME)
(6) in their plasma. Though the value of
increasing L-arginine levels in cases of im-
paired circulation is now recognized,
practical schemes for therapeutic use of
L-arginine have been elusive. In this pilot
study, we sought to determine whether
supplying L-arginine transdermally
would improve vascular function of the
feet in patients with diabetes as indicated
by flow and temperature.
The study was designed as a double-
blind, vehicle-controlled, two-period,
crossover protocol with washout periods
of 1 week. Sixteen subjects were enrolled,
and 13 completed the study (aged 56 ⫾8
years). After analyzing the data, it was
clear that the effect of L-arginine persisted
throughout the washout periods (Tables
1 and 2). Because of this, except for the
initial exposure of L-arginine virgin feet,
the analysis was altered to determine the
effect from cumulative exposure to
L-arginine throughout the protocol. Flow
was measured at the metatarsal and Achil-
les area using a Doppler flow meter (7),
and temperature was measured at the
metatarsal and big toe areas using an in-
frared thermometer. The active cream
was a water-based moisturizing vehicle
containing 12.5% L-arginine hydrochlo-
ride in a hostile biophysical environment
comprised of high concentrations of cho-
line chloride, sodium chloride, and mag-
nesium chloride. The vehicle control was
identical except that the L-arginine was
omitted.
At the first visit, after baseline mea-
surements were made each subject
rubbed active cream (4 mg L-arginine/
cm
2
) into one foot and vehicle into the
other. After 30 min, measurements were
made again. A 1-week washout period
followed. Patients returned after the
Figure 1—Hazard ratios
(HRs) of cardiovascular dis-
ease for combined categories
of triglycerides and waist
circumference, stratified for
non-HDL cholesterol. 䡺,
waist ⬍94/80 cm; f, waist
ⱖ94/80 cm.
Letters
284 DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004

washout period and flow and tempera-
ture measurements were made. They
were then randomly given either active or
placebo cream and told to rub it into their
feet in the morning and evening every day
for 2 weeks. At the end of 2 weeks, sub-
jects returned and again measurements
were made. A second 1-week washout
period followed that third visit. At the end
of that period subjects returned and mea-
surements were made. They were given
the crossover product and told again to
rub it into their feet morning and evening
for 2 weeks. The subjects returned for fi-
nal flow and temperature measurements
at the end of that period.
At the first visit, flow was increased at
the Achilles area in the foot with active
cream from 8.1 ⫾3.3 to 11.5 ⫾5.5 AU
(P⫽0.05) 30 min after application. In the
foot that received placebo cream, flow
failed to increase (8.1 ⫾1.4 vs. 8.3 ⫾2.2
AU). Furthermore, at the last visit the
temperature at the metatarsal area had
risen from the initial value of 82.0 ⫾2.3
to 86.9 ⫾2.4°F(P⬍0.0001), and the
temperature of the big toe had risen from
the initial visit value of 74.4 ⫾4.2 to
82.4 ⫾4.8°F(P⬍0.0001). And at the
last visit the flow at the metatarsal area
had risen from 8.7 ⫾4.3 to 11.6 ⫾5.5
AU (P⬍0.0001), and flow at the Achilles
area had risen from 8.4 ⫾2.5 to 11.4 ⫾
5.5 AU (P⫽0.02). While the failure of
the L-arginine effect to wash out removed
the opportunity for placebo control, the
improvement in temperature and flow
were substantial and highly statistically
significant. Though puzzling, one expla-
nation of the persistence of the L-arginine
effect is that the local tissue concentration
of L-arginine becomes high enough to
cause inactive monomers of eNOS to form
active dimers.
We conclude that in the patients we
studied with diabetes, treatment of their
feet with a transdermal preparation of
L-arginine improved both flow and tem-
perature, and this effect was surprisingly
long lasting. Such improvement of com-
promised local blood flow should be
beneficial and could reduce the complica-
tions of the disease.
ERIC T. FOSSEL,
PHD
From Strategic Science and Technologies, Welles-
ley, Massachusetts.
Address correspondence to Eric T. Fossel, Strate-
gic Science and Technologies, 892 Worcester Rd.,
Wellesley, MA 02481. E-mail: efossel@strategicsci.
com.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Cooke JP, Dzau V: Nitric oxide synthase:
role in the genesis of vascular disease.
Annu Rev Med 28:489–501, 1997
2. Calles-Escandon J, Cipolla M: Diabetes
and endothelial dysfunction: a clinical
prospective. Endocr Rev 22:36–52, 2001
3. Kin KY, Ito A, Asagami T, Tsai PSS, Adi-
moolan S, Kimoto M, Hideaki T, Reaven
GM, Cooke JP: Impaired nitric oxide syn-
thase pathway in diabetes mellitus: role of
asymmetric dimethylarginine and dim-
ethylarginine dimethyoaminohydrolase.
Circulation 106:987–992, 2002
4. Panda K, Rosenfeld RF, Ghosh S, Meade
AL, Getzoff ED, Stuehr DJ: Distinct dimer
interaction and regulation in nitric oxide
synthase types I, II, and III. J Biol Chem
277:31020–31030, 2002
5. Pieper GM, Siebeneich W, Dondlinger
LA: Short-term oral administration of
L-arginine reverses defective endotheli-
um-dependent relaxation and cGMP gen-
eration in diabetes. Eur J Pharmacol 317:
317–320, 1996
6. Abbasi F, Asagmi T, Cooke JP, Lamendola
C, McLaughlin T, Reaven GM, Stueh-
linger M, Tsao PS: Plasma concentrations
of asymmetric dimethylarginine are in-
creased in patients with type 2 diabetes
mellitus. Am J Cardiol 88:1201–1203,
2001
7. Vongsavan N, Matthews B: Some aspects
of the use of laser doppler flow meters for
recording tissue blood flow. Exper Physiol
78:1–14, 1993
Increased
Prevalence of
Enteroviral RNA in
Blood Spots From
Newborn Children
Who Later
Developed Type 1
Diabetes
A population-based case-control
study
V
irus infections during fetal life may
lead to persistent infections due to
unresponsiveness of the immature
immune system and by different mecha-
nisms inducing autoimmunity in the
␤-cell (1). In a previous study, we found
group-reacting antibodies to enterovirus
more frequently increased in serum at de-
livery in a cohort of mothers whose chil-
dren later developed diabetes than in
control subjects (2). It has been argued
that mothers of later diabetic children
might have increased immune responses
to certain viruses (3). Therefore, detection
of the virus nucleic acid would be impor-
tant to confirm.
Enteroviral infections typically occur
as epidemics with a short viremic phase;
therefore, they are detectable for only a
short time in peripheral blood (4). Con-
sequently, large series of blood or serum
samples would be necessary. Since 1973,
blood spots routinely taken on days 2–4
Table 1—Effect of transdermal L-arginine cream on temperature
Metatarsal (°F) Pvs. visit 1 Big toe (°F) Pvs. visit 1
Visit 1 82.0 ⫾2.3 74.4 ⫾4.2
Visit 2 84.1 ⫾3.4 0.004 77.7 ⫾5.3 0.01
Visit 3 87.0 ⫾2.4 ⬍0.0001 83.6 ⫾4.9 ⬍0.0001
Visit 4 86.1 ⫾2.4 ⬍0.0001 80.6 ⫾5.4 ⬍0.0001
Visit 5 86.9 ⫾2.4 ⬍0.0001 82.4 ⫾4.8 ⬍0.0001
Data are means ⫾SD.
Table 2—Effect of transdermal L-arginine cream on flow
Metatarsal (AU) Pvs. visit 1 Achilles (AU) Pvs. visit 1
Visit 1 8.7 ⫾4.3 8.4 ⫾2.5
Visit 2 10.8 ⫾5.9 NS 8.5 ⫾3.9 NS
Visit 3 10.8 ⫾4.8 0.05 9.2 ⫾3.9 NS
Visit 4 11.6 ⫾8.3 NS 10.0 ⫾4.2 0.06
Visit 5 11.6 ⫾5.5 ⬍0.0001 11.4 ⫾5.5 0.02
Data are means ⫾SD.
Letters
DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004 285

of life for analysis of inherited metabolic
diseases in all newborns in Sweden are
stored in a biobank. From this biobank,
we collected blood spots from 600 chil-
dren in the Swedish childhood diabetes
register who were born during the years
1986–1995 and who had diabetes onset
before 1996. For each case, a control sam-
ple was included from a child born on the
same date and not found in the Swedish
childhood diabetes register, and the con-
trol sample was stored adjacent to the case
filter.
Nested enterovirus PCR was per-
formed essentially according to Puig et al.
(5). To exclude the possibility of false-
positivity due to contamination, we ex-
cluded all case-control pairs with double
positivity and each step of RNA extraction
and analysis included positive and nega-
tive controls. All case-control pairs were
analyzed in the same run. As references,
representatives of the most common con-
genital viral infections were chosen for
analysis: CMV (6) and parvo B19 virus
(7). A total of 542 pairs of samples were
valid and analyzed for enterovirus RNA.
Twenty-seven diabetic cases were
RNA positive, as compared with 14 con-
trol subjects (odds ratio 1.98 [95% CI
1.04⫺3.77], two-tailed P⫽0.04). Due to
limited material, typing of enteroviral
RNA by sequencing was not possible . For
CMV and parvo B19 viruses, no differ-
ences were shown between cases and ref-
erences in samples of 208 and 180 pairs,
respectively.
The findings support the hypothesis
that early enteroviral infections may play
a role in the pathogenesis of type 1 diabe-
tes. In our previous study of maternal
sera, the etiological fraction of enteroviral
infection in pregnancy (4) was calculated
at 27%. Thus, early fetal or neonatal in-
fections may explain a fraction of child-
hood diabetes cases.
GISELA G. DAHLQUIST,
MD, PHD
1
JENNY FORSBERG,
MS
1
LARS HAGENFELDT,
MD, PHD
2
†
JENS BOMAN,
MD
3
PER JUTO,
MD, PHD
3
From the
1
Department of Clinical Sciences–
Paediatrics, UmeåUniversity, Umeå, Sweden; the
2
Centre of Inherited Metabolic Diseases, Huddinge
University Hospital, Karolinska Institute, Stock-
holm, Sweden; and the
3
Department of Virology,
UmeåUniversity, Umeå, Sweden.
Address correspondence to Professor Gisela
Dahlquist, UmeåUniversity, Department of Clinical
Sciences–Paediatrics, SE-901 87 Umeå, Sweden. E-
mail: gisela.dahlquist@pediatri.umu.se.
†Deceased, summer 2003.
© 2004 by the American Diabetes Association.
Acknowledgements—This study was sup-
ported by grants from the Swedish Research
Council (Medicine) (project no. 07531), the
Swedish Diabetes Association, and the Va¨ster-
botten County Council.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Jun HS, Yoon JW: The role of viruses in
type I diabetes: two distinct cellular and
molecular pathogenic mechanisms of vi-
rus-induced diabetes in animals. Diabeto-
logia 44:271–285, 2001
2. Dahlquist GG, Ivarsson S, Lindberg B,
Forsgren M: Maternal enteroviral infec-
tion during pregnancy as a risk factor for
childhood IDDM: a population-based
case-control study. Diabetes 44:408–413,
1995
3. Graves PM, Norris JM, Pallansch MA,
Gerling IC, Rewers M: The role of entero-
viral infections in the development of
IDDM. Diabetes 46:161–168, 1997
4. Cherry JP: Enteroviruses: polioviruses,
coxsackieviruses, echoviruses and entero-
viruses. In Textbook of Pediatric Infectious
Diseases. Vol. 2, 3rd ed. Feigin RD, Cherry
JD, Eds. Philadelphia, PA, WB Saunders,
1992, p. 1705–1753
5. Puig M, Jofre J, Lucena F, Alland A,
Wadell G, Girones R: Detection of adeno-
viruses and enteroviruses in polluted wa-
ters by nested PCR amplification. Appl
Environment Microbiol 60:2963–2970,
1994
6. Brytting M, Sundqvist VA, Stålhandske P,
Linde A, Wahren B: Cytomegalovirus
DNA detection of an immediate early pro-
tein gene with nested primer oligonucle-
otides. J Virol Methods 32:127–138, 1991
7. Lundqvist A, Tolfvenstam T, Bostic J,
So¨derlund M, Broliden K: Clinical and
laboratory findings in immunocompetent
patients with persistent parvovirus B 19
DNA in bone marrow. Scand J Infect Dis
31:11–16, 1999
Physician Attitudes
Toward Foot Care
Education and Foot
Examination and
Their Correlation
With Patient Practice
F
oot complications are one of the
most serious causes of morbidity,
disability, poor quality of life, and
resource use among diabetic people (1).
The adoption of preventive strategies to
reduce the rate of foot problems thus rep-
resents an important priority. In fact, a
strategy that includes prevention, patient
and staff education, multidisciplinary
treatment of foot complications, and close
monitoring has been demonstrated to be
very effective in reducing amputation rate
(2).
In the context of a nationwide out-
comes research program (the QuED
project), we investigated several aspects
related to foot care in 3,564 patients with
type 2 diabetes enrolled by 125 diabetes
outpatient clinics and 103 general
practitioners.
Details on study design have been re-
ported elsewhere (3). Briefly, all patients
with type 2 diabetes were considered eligi-
ble, irrespective of age, duration of diabetes,
and treatment. Foot complications in-
cluded ulcers, gangrene, nontraumatic am-
putations, and claudicatio intermittens.
Patients filled out a questionnaire in-
vestigating whether they had received in-
formation about foot care, how often they
had had their feet examined in the last
year, and how often they usually checked
their feet. Analyses were adjusted for pa-
tient case mix and physician-level cluster-
ing using multivariate multilevel logistic
regression models (4).
The prevalence of lower limb compli-
cations was 6.8%. Seventy-two percent of
the patients declared that they had re-
ceived foot education, but only 49% re-
ported that they had had their feet
examined in the last year. Patients with
ⱕ5 years of school education (odds ratio
[OR] 1.3, 95% CI 1.1–1.6) and those with
low income (ⱕ$12,000) (1.2, 1.0–1.4)
were more likely not to receive foot edu-
cation. The presence of foot complica-
tions, peripheral vascular disease,
cardiac-cerebrovascular disease, and dia-
betic neuropathy were not independently
associated with a greater chance of receiv-
ing foot education. Foot examination was
more likely to be performed in low-
income patients (1.3, 1.1–1.6) and in
those with foot complications (1.5, 1.1–
2.1) but not in those with diabetic neu-
ropathy, peripheral vascular disease, or
cardiac-cerebrovascular disease. Foot ex-
amination tended to be performed less
frequently by general practitioners and
other specialists in diabetes outpatient
clinics as opposed to diabetologists, even
though the statistical significance was
reached only for the comparison between
Letters
286 DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004

general practitioners and diabetologists
(0.6, 0.4–1.0).
Overall, 33% of the patients declared
that they never checked their feet. Pa-
tients who had received foot education
(OR ⫽2.5, 95% CI 2.0–3.0) and those
who had had their feet examined by their
physician (1.7, 1.4–2.0) were more likely
to check their feet regularly. Similarly, pa-
tients with foot complications (2.2, 1.5–
3.2), but not those with peripheral
vascular disease, cardiac-cerebrovascular
disease, or diabetic neuropathy, were
more likely to check their feet.
In conclusion, the attention to foot
complications is generally poor, and a
substantial proportion of type 2 diabetic
patients is not offered foot education and
examination, even in those subgroups
showing a significant increase in the risk
of foot complications. Even in the pres-
ence of foot complications or major risk
factors, one-quarter of the patients did
not pay any attention to foot care. Those
patients who had received foot education
and had had their feet examined were sig-
nificantly more likely to regularly check
their feet. This finding underlines the cru-
cial role of physicians in orienting patient
practices.
GIORGIA DEBERARDIS,
MSC (CHEM PHARM)
1
FABIO PELLEGRINI,
MS
1
MONICA FRANCIOSI,
MSC (BIOL)
1
MAURIZIO BELFIGLIO,
MD
1
BARBARA DINARDO,
HSDIP
1
SHELDON GREENFIELD,
MD
2
SHERRIE H. KAPLAN,
PHD, MPH
2
MARIA C.E. ROSSI,
(CHEM PHARM)
1
MICHELE SACCO
MD
1
GIANNI TOGNONI
MD
1
MIRIAM VALENTINI
MD
1
ANTONIO NICOLUCCI
MD
1
THE QUED STUDY GROUP
From the
1
Department of Clinical Pharmacology
and Epidemiology, Istituto di Ricerche Farmaco-
logiche Mario Negri, Consorzio Mario Negri Sud,
S. Maria Imbaro, Italy; and the
2
Center for Health
Policy Research, University of California Irvine, Ir-
vine, California.
Address correspondence to Antonio Nicolucci,
MD, Department of Clinical Pharmacology and Ep-
idemiology, Consorzio Mario Negri Sud, Via Nazio-
nale, 66030 S. Maria Imbaro, Italy. E-mail:
nicolucci@negrisud.it.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Humphrey AR, Dowse GK, Thoma K,
Zimmet PZ: Diabetes and nontraumatic
lower extremity amputations: incidence,
risk factors, and prevention: a 12-year fol-
low-up study in Nauru. Diabetes Care 19:
710–714, 1996
2. Larsson J, Apelqvist J, Agardh CD, Sten-
strom A: Decreasing incidence of major
amputation in diabetic patients: a conse-
quence of a multidisciplinary foot care
team approach? Diabet Med 12:770–776,
1995
3. Belfiglio M, De Berardis G, Franciosi M,
Cavaliere D, Di Nardo B, Greenfield S,
Kaplan SH, Pellegrini F, Sacco M, Tog-
noni G, Valentini M, Nicolucci A, Caimi
V, Capani F, Corsi A, Della Vedova R,
Massi Benedetti M, Nicolucci A, Taboga
C, Tombesi M, Vespasiani G: The rela-
tionship between physicians’self-re-
ported target fasting blood glucose levels
and metabolic control in type 2 diabetes.
Diabetes Care 24:423–429, 2001
4. Snijders TAB, Bosker RJ. Multilevel Analy-
sis: An Introduction to Basic and Advanced
Multilevel Modeling. London, SAGE Publi-
cations, 1999
Hepatocyte Growth
Factor in the
Vitreous Fluid of
Patients With
Proliferative
Diabetic Retinopathy
Its relationship with vascular
endothelial growth factor and
retinopathy activity
T
he role of hepatocyte growth factor
(HGF) in the etiopathogenesis of
proliferative diabetic retinopathy
(PDR) remains to be elucidated. Others
and we (1–5) have found high intravitre-
ous concentrations of HGF in patients
with PDR. However, in these previous re-
ports, the main confounding factors that
could lead to misinterpretation of the re-
sults (vitreous hemorraghe, intravitreal
protein concentration, and plasma HGF
levels) have not been fully considered. In
the present study, we consider all these
confounding factors in order to evaluate
the vitreous levels of HGF in patients with
PDR and to investigate its relationship
with vascular endothelial growth factor
(VEGF) and retinopathy activity.
A total of 28 diabetic patients with
PDR, in whom a vitrectomy was per-
formed, were included in the study.
Thirty nondiabetic patients with other
conditions requiring vitrectomy but in
whom the retina was not directly affected
by neovascularization served as a control
group. Patients in whom intravitreous he-
moglobin was detectable by spectopho-
tometry were excluded. HGF and VEGF
were determined by enzyme-linked im-
munosorbent assay (R&D Systems,
Abingdon, U.K.). The results are ex-
pressed as the median and range.
Vitreal levels of both VEGF and HGF
were higher in diabetic patients with PDR
than in the control group (1.34 [0.16–
6.2] vs. 0.009 ng/ml [0.009–0.003] and
19.38 [0.4–80] vs. 6.04 mg/ml [1.8 –
17.34], respectively, P⬍0.0001). These
differences remained highly significant af-
ter adjusting for serum levels (P⬍
0.0001). To explore the influence of the
breakdown of the blood-retinal barrier
and, in consequence, the increased serum
diffusion that occurs in PDR patients, the
levels of both HGF and VEGF were nor-
malized for total vitreal protein concen-
tration. After correcting for total vitreous
protein concentration, the ratio of VEGF-
to-vitreal proteins remained significantly
higher in diabetic patients with PDR than
in the control group (0.34 [0.01–2.3] vs.
0.01 ng/mg [0.003–0.03], respectively,
P⬍0.0001). However, the ratio of HGF-
to-vitreal proteins was lower in diabetic
patients than in nondiabetic control sub-
jects (5.03 [2–20] vs. 7.52 ng/mg [1.9–
26], P⫽0.02). The lower intravitreous
levels of HGF obtained after correcting for
intravitreal proteins in patients with PDR
in comparison with nondiabetic control
subjects suggest that serum diffusion
largely explains the differences detected
in the intravitreous HGF levels between
these groups.
The vitreous concentrations of VEGF
were higher in patients with active PDR
than in patients with quiescent PDR (1.89
[0.2–6.2] vs. 0.78 ng/ml [0.1–1.7], re-
spectively, P⫽0.004). By contrast, vitre-
ous HGF was not related to PDR activity
(active vs. quiescent, 17.1 [7.3–46.6] vs.
23 ng/ml [0.4–80], respectively, P⫽
NS). Previous studies have found higher
HGF concentrations in patients with ac-
tive PDR than in those with quiescent
PDR (1,2). However, after carefully con-
sidering the main confounding factors
that could lead to a misinterpretation of
the results, we did not observe any rela-
tionship between PDR activity and intra-
vitreous HGF concentrations.
Letters
DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004 287

Finally, we did not find a relationship
between intravitreous levels of HGF and
VEGF. This was true in absolute terms
and after normalizing for vitreal proteins.
The different response of HGF and VEGF
to hyperglycemia (6) and hypoxia (7)
could explain the lack of relationship de-
tected between these growth factors in the
vitreous fluid of patients with PDR.
RAFAEL SIMO
´,
MD
1
ALBERT LECUBE,
PHD
1
JOSE´GARCı´A-ARUMı´,
MD
2
ESTHER CARRASCO,
PHD
1
CRISTINA HERNA´ NDEZ,
MD
1
From the
1
Department of Endocrinology, Hospital
General Universitari Vall d’Hebron, Barcelona, Spain;
and the
2
Department of Ophthalmology, Hospital
General Universitari Vall d’Hebron, Barcelona, Spain.
Address correspondence to Rafael Simo´, MD, Di-
abetes Unit, Endocrinology Division, Hospital Uni-
vesitari Vall d’Hebron, Pg. Vall d’Hebron 119-129,
08035 Barcelona, Spain. E-mail: rsimo@
hg.vhebron.es.
© 2004 by the American Diabetes Association.
Acknowledgments—This work was sup-
ported by grants from the Ministerio de Cien-
cia y Tecnologı´a (PM99-0136), Instituto
Carlos III (G03/212, C03/08), and Novonor-
disk Pharma (01/0,066)
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Katsura Y, Okano T, Noritake M, Kosano
H, Nishigori H, Kado S, Matsuoka T: He-
patocyte growth factor in vitreous fluid of
patients with proliferative diabetic reti-
nopathy and other retinal disorders. Dia-
betes Care 21:1759–1763, 1998
2. Nishimura M, Ikeda T, Ushiyama M,
Nanbu A, Kinoshita S, Yoshimura M: In-
creased vitreous concentrations of human
hepatocyte growth factor in proliferative
diabetic retinopathy. J Clin Endocrinol
Metab 84:659–662, 1999
3. Canto´nA, Burgos B, Herna´ndez C, Mateo C,
Segura RM, Mesa J, Simo´R: Hepatocyte
growth factor in vitreous and serum from
patients with proliferative diabetic retinop-
athy. Br J Ophthalmol 84:732–735, 2000
4. Umeda N, Ozaki H, Hayashi H, Kondo H,
Uchida H, Oshima K: Non-paralleled in-
crease of hepatocyte growth factor and
vascular endothelial growth factor in the
eyes with angiogenic and nonangiogenic
fibroproliferation. Ophthalmic Res 34:43–
47, 2002
5. Mitamura Y, Takeuchi S, Matsuda A, Ta-
gawa Y, Mizue Y, Nishihira J: Hepatocyte
growth factor levels in the vitreous of pa-
tients with proliferative vitreoretinopa-
thy. Am J Ophthalmol 129:678–680, 2000
6. Taniyama Y, Morishita R, Hiraoka K, Aoki
M, Nakagami H, Yamasaki K, Matsumoto
K, Nakamura T, Kaneda Y, Ogihara T:
Therapeutic angiogenesis induced by hu-
man hepatocyte growth factor gene in rat
diabetic hind limb ischemia model: mo-
lecular mechanisms of delayed angiogen-
esis in diabetes. Circulation 104:2344–
2350, 2001
7. Hayashi S, Morishita R, Nakamura S,
Yamamoto K, Moriguchi A, Nagano T, Taiji
M, Noguchi H, Matsumoto K, Nakamura T,
Higaki J, Ogihara T: Potential role of hepa-
tocyte growth factor, a novel angiogenic
growth factor, in peripheral arterial disease:
downregulation of HGF in response to hyp-
oxia in vascular cells. Circulation 100
(Suppl. 19):II301–II308, 1999
Autoimmune
Hypoglycemia in a
Type 2 Diabetic
Patient With Anti-
Insulin and Insulin
Receptor Antibodies
T
here are two types of autoimmune
hypoglycemia, one due to autoanti-
bodies acting against the insulin
receptor and the other due to autoanti-
bodies acting against insulin itself in indi-
viduals who have or have never received
exogenous insulin, respectively (1). Both
types are rare and can produce fasting and
postprandial reactive hypoglycemia.
A 72-year-old woman with frequent
severe hypoglycemia was admitted to the
emergency room, presenting with loss of
consciousness. Three weeks before her
admission, she was diagnosed with diabe-
tes and received insulin at a local hospital.
In the emergency room, her blood glu-
cose level was 40 mg/dl. She had been in
good general health, except for hyperten-
sion for 30 years and postmenopausal os-
teoporosis 10 years before admission. She
has no evidence of other diseases associ-
ated with altered immunity.
Three weeks ago, her biceps tendon
ruptured when she slipped and fell, and
during treatment, her blood glucose lev-
els were ⬎400 mg/dl. She was treated
with insulin. However, she stopped insu-
lin treatment because of frequent hypo-
glycemic events. Although she had had
intravenous glucose injections, she had
frequent hypoglycemic attacks, such as
disorientation, loss of consciousness, pal-
pitation, and diaphoresis. Her blood glu-
cose levels had been ⬍40 mg/dl on every
hypoglycemic event, especially during
fasting hypoglycemia.
Physical examination revealed nor-
mal vital signs except for a chronically ill
appearance. Her HbA
1c
was 6.3% (range
3–6%), plasma glucose 40 mg/dl, insulin
103.7 ␮U/ml, C-peptide 4.1 ng/ml, GAD
autoantibody levels 0.01 units/ml (nor-
mal range 0–1.45 units/ml; RSR, Cardiff,
U.K. ), and insulinoma-associated protein
2 autoantibody 0.01 units/ml (normal
range 0–1.1 units/ml; RSR). Her thyroid,
liver, and adrenal function studies were
normal. She was anemic, with a hemoglo-
bin level of 9.2 g/dl. Her creatinine level
was 0.8 mg/dl. Tests for anti-nuclear an-
tibody, anti-DNA antibody, anti-smooth
antibody, and anti-microsomal antibody
were all negative. Insulin antibody levels
were 58.5% (nonspecific binding, normal
range ⬍7%, measured by radioimmuno-
assay [Cobra 5010; Biosource Europe,
Nivelles, Belgium]), and she was positive
for insulin receptor antibodies (measured
by radioreceptor assay [LKB 1261; BML,
Tokyo, Japan]).
She was prescribed glucocorticoids
and glucose tablets. Treatment with pred-
nisone and glucose tablets was accompa-
nied by the resolution of hypoglycemic
episodes within 48 h. Six months later,
her insulin antibody level was 67%, and
she was negative for insulin receptor
antibodies.
Most cases of insulin autoimmune hy-
poglycemia described in Asian races (2,3)
have shown a strong correlation with cer-
tain HLA systems, suggesting the exis-
tence of a predisposing genetic
component. This subject’s HLA typing re-
sult was HLA-DRB1*. Autoimmune hy-
poglycemia is associated with certain HLA
systems, such as DR4 and DQw3, and es-
pecially DRB1*0406/DQA1*0302/
DQB1*0302. There have been 190 cases
of insulin autoimmune syndrome re-
ported over the past 20 years. It is note-
worthy that HLA-DRB1*0406 is quite
prevalent in Japanese patients (2,3). Our
patient’s HLA typing is predisposed to au-
toimmune hypoglycemia.
Several types of insulin antibodies
have been reported and are most fre-
quently seen in patients who receive in-
sulin injections, but there have also been
reports of them in nondiabetic patients
with such autoimmune disease. Postpran-
dial hypoglycemia is more common with
Letters
288 DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004

this syndrome than fasting hypoglycemia
(1), and the course of this condition is
benign and self-limited, with remission
usually occurring within 1 year.
The insulin receptor antibody is asso-
ciated with the inhibition of insulin bind-
ing to insulin receptors, accelerated
receptor degradation, receptor down-
regulation, and extreme insulin resistance
and hyperglycemia (4). Insulin receptor
antibodies act as agonists or antagonists to
the insulin receptor. Insulin receptor an-
tibodies may also inhibit insulin binding,
thereby inhibiting insulin clearance and
elevating levels of plasma insulin. The
most important laboratory test in autoim-
mune hypoglycemia is a direct assay for
the presence of antibodies directed
against the insulin receptor or insulin.
Patients with this condition have low
circulating insulin, C-peptide levels, and
refractory hypoglycemia. Antibody titers
generally decrease over time and remis-
sion eventually occurs in most patients.
However, because of the severity of the
hypoglycemia, aggressive treatment is in-
dicated. High-dose glucocorticoids, plas-
mapheresis, and alkylating agents have
been tried with varying success (5).
CHONG HWA KIM,
MD
JIHYUN PARK,
MD, PHD
TAE SUN PARK,
MD, PHD
HONG SUN BAEK,
MD, PHD
From the Division of Endocrinology and Metabo-
lism, Department of Internal Medicine, Chonbuk
National University Medical School, Chonju, Korea.
Address correspondence to T.S Park, MD, PHD,
Division of Endocrinology and Metabolism, Depart-
ment of Internal Medicine, Chonbuk National Uni-
versity Medical School, Chonju, Korea 634-18,
Keum-Am Dong, Dukjin Gu, Chonju, Chonbuk,
561-712, Korea. E-mail: pts@moak.chonbuk.ac.kr.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Taylor SI, Barbetti F, Accili D, Roth J, Gor-
den P: Syndromes of autoimmunity and
hypoglycemia: autoantibodies directed
against insulin and its receptor. Endocrinol
Metab Clin North Am 18:123–143, 1989
2. Cavaco B, Uchigata Y, Porto T, Amparo-
Santos M, Sobrinho L, Leite V: Hypogly-
caemia due to insulin autoimmune
syndrome: report of two cases with char-
acterisation of HLA alleles and insulin au-
toantibodies. Eur J Endocrinol 145:311–
316, 2001
3. Uchigata Y, Hirata Y, Omori Y: Hypogly-
cemia due to insulin antibody (Letter).
Am J Med 94:556–557, 1993
4. Elias D, Cohen IR, Schechter Y, Spirer Z,
Golander A: Antibodies to insulin recep-
tor followed by anti-idiotype: antibodies
to insulin in child with hypoglycemia. Di-
abetes 36:348–354, 1987
6. Dozio N, Scavini M, Beretta A, Sarugeri E,
Sartori S, Belloni C, Dosio F, Savi A, Fazio
F, Sodoyez JC, Pozza G: Imaging of the
buffering effect of insulin antibodies in
the autoimmune hypoglycemia syn-
drome. J Clin Endocrinol Metab 83:643–
648, 1998
COMMENTS AND
RESPONSES
Phenotypic
Heterogeneity and
Associations of Two
Aldose Reductase
Gene Polymorphisms
With Nephropathy
and Retinopathy in
Type 2 Diabetes
Response to Wang et al.
W
ang et al. (1) recently examined
aldose reductase as a susceptibil-
ity gene for diabetic nephropa-
thy among type 2 diabetic Chinese in
Hong Kong. Although there was a small
increase in the frequencies of the risk al-
leles of the (CA)
n
dinucleotide repeat and
C-106T polymorphisms, analysis of the
genotype distribution failed to detect any
significant association between these
polymorphisms and diabetic nephropa-
thy (1). This negative result persisted de-
spite confining the statistical analyses to
control subjects who were normoalbu-
minuric with at least 5 years of known
diabetes duration and case subjects with
both diabetic nephropathy and concomi-
tant diabetic retinopathy.
By and large, this study does not con-
firm earlier findings, which had impli-
cated aldose reductase as a genetic risk
factor for diabetic nephropathy among
Caucasians with type 1 diabetes (2). Al-
though the two studies were done on pa-
tients with different types of diabetes,
drawn from separate human populations,
which may arguably provide a basis for
the discordant findings, a distinct possi-
bility relates to the differential definitions
of diabetic nephropathy. In this Boston
study, diabetic nephropathy was defined
on the basis of persistent proteinuria, i.e.,
ⱖ1⫹on Multistix or albumin-to-
creatinine ratio (ACR) ⱖ300 mg/g, or
end-stage renal disease due to diabetic ne-
phropathy (2). In contrast, the definition
used in the current Hong Kong study was
less stringent and included patients with
microalbuminuria (albumin excretion
rate ⱖ20 ␮g/min or ACR ⱖ3.5 g/mmol)
(1). This latter criterion poses some con-
cern in terms of misclassification because
regression of micro- to normoalbumin-
uria is likely to be a common phenome-
non, as recently demonstrated in type 1
diabetic patients (3). In genetic epidemi-
ological studies, such misclassification
can diminish the power of a study to de-
tect an association. Therefore, the study
by Wang et al. does not negate the hy-
pothesis that aldose reductase could be a
susceptibility gene for advanced diabetic
nephropathy in type 2 diabetes. This pos-
sibility might be addressed by reanalyzing
their data using a stricter definition based
on advanced diabetic nephropathy.
DANIEL P.K. NG,
PHD
KEE-SENG CHIA,
FAMS, MD, FFOM
DAVID KOH,
MBBS, MSC, PHD
From the Department of Community, Occupational
and Family Medicine, National University of Singa-
pore, Singapore; and the Centre for Molecular Epi-
demiology, National University of Singapore,
Singapore.
Address correspondence to Daniel P K Ng, PhD,
Department of Community, Occupational and Fam-
ily Medicine (MD3), Faculty of Medicine, National
University of Singapore, 16 Medical Dr., Singapore,
117597 Singapore. E-mail: cofnpkd@nus.edu.sg.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Wang Y, Ng MC, Lee SC, So WY, Tong
PC, Cockram CS, Critchley JA, Chan JC:
Phenotypic heterogeneity and associa-
tions of two aldose reductase gene poly-
morphisms with nephropathy and
retinopathy in type 2 diabetes. Diabetes
Care 26:2410–2415, 2003
2. Moczulski DK, Scott L, Antonellis A, Ro-
gus JJ, Rich SS, Warram JH, Krolewski AS:
Aldose reductase gene polymorphisms
and susceptibility to diabetic nephropa-
thy in type 1 diabetes mellitus. Diabet Med
17:111–118, 2000
3. Perkins BA, Ficociello LH, Silva KH,
Finkelstein DM, Warram JH, Krolewski
Letters
DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004 289

AS: Regression of microalbuminuria in
type 1 diabetes. N Engl J Med 348:2285–
2293, 2003
Phenotypic
Heterogeneity and
Associations of Two
Aldose Reductase
Gene Polymorphisms
With Nephropathy
and Retinopathy in
Type 2 Diabetes
Response to Ng et al.
W
e thank Ng et al. (1) for their re-
sponse to our recent article on
the associations of two aldose re-
ductase gene polymorphisms, a (CA)n
microsatellite at the 5⬘region and a pro-
moter C/T polymorphism with nephrop-
athy and retinopathy in type 2 diabetes
(2).
However, we hold the view that Ng et
al. have misinterpreted our data and in-
advertenly commented that our “nega-
tive”results were related to our less
stringent definitions of nephropathy,
when in fact, we have provided clear evi-
dence to show that both the z-2 and T
allele of the aldose reductase gene poly-
morphisms were risk factors for diabetic
nephropathy in Chinese type 2 diabetic
patients.
In the consecutive cohort analysis in-
volving all 738 type 2 diabetic patients,
those with the T allele had higher albu-
minuria than noncarriers (30.2 vs. 21.9
␮g/min) (2). This difference remained
significant after adjustment for age, dura-
tion of disease, blood pressure, and
HbA
1c
.
We then excluded patients with a
short duration of disease (⬍5 years) and
used a case-control study design to fur-
ther test the hypothesis. We defined case
subjects as diabetic patients with both di-
abetic retinopathy and nephropathy,
whereas patients who had no complica-
tions were selected as control subjects.
Using this design, we found that both z-2
(odds ratio 2.64) and T alleles (odds ratio
2.48) were independent risk factors for
the coexistence of diabetic nephropathy
and retinopathy. The other predictors
were age, blood pressure, HbA
1c
, triglyc-
eride, and male sex (2).
Hence, contrary to the comments
made by Ng et al. that we failed to confirm
results from previous studies on aldose
reductase as risk genotypes for diabetic
nephropathy in type 1 diabetes, our study
has indeed provided corroborative evi-
dence to support these findings.
JULIANA CHAN,
MD, FRCP
YING WANG,
PHD
From the Department of Medicine and Therapeu-
tics, the Chinese University of Hong Kong, the
Prince of Wales Hospital, Shatin, Hong Kong.
Address correspondence to Juliana Chan, MD,
FRCP, The Chinese University of Hong Kong, De-
partment of Medicine & Therapeutics, Prince of
Wales Hospital, Shatin, N.T., Hong Kong. E-mail:
jchan@cuhk.edu.hk.
© 2004 by the American Diabetes Association.
●●●●●●●●●●●●●●●●●●●●●●●
References
1. Ng DPK, Chia K-S, Koh D: Phenotypic
heterogeneity and associations of two al-
dose reductase gene polymorphisms with
nephropathy and retinopathy in type 2
diabetes (Letter). Diabetes Care 27:289–
290, 2004
2. Wang Y, Ng MCY, Lee SC, So WY, Tong
CY, Cockram CS, Critchley JAJH, Chan
JCN: Phenotypic heterogeneity and asso-
ciations of two aldose reductase gene
polymorphisms with nephropathy and
retinopathy in type 2 diabetes. Diabetes
Care 26:2410–2415, 2003
Letters
290 DIABETES CARE,VOLUME 27, NUMBER 1, JANUARY 2004