IgA concentrations of all groups of diabetic patients were significantly higher than those of the corre- sponding subgroups of 943 control subjects, except for type.
Clinical Chemistry 1064-1067
42:7
(1996)
High serum IgA concentrations in patients with diabetes mellitus: agewise distribution and relation to chronic complications SANTIAGO
ANA
In this
M.
RODRIGUEZSEGADE,*
study
in,
FELIX CELSA
CAMIfA,
we investigated
the agewise distributions of in 1251 type I and 2224 type 2 diabetic patients, and the association between serum IgA concentration and diabetic complications (retinopathy, neuropathy, nephropathy, macroangiopathy, and hyperten-
e
common
Department
MARIA
LORENZO,
JESIITS
Lojo
among
patients
who had been under more
recent
treatment
for 10
patients.
Singh and Kulig [3] published tacit corroboration of high serum IgA concentrations among diabetics: The mean IgA concentration in their group of 149 diabetic patients was 3.5 ± 1.6 gIL, -70% higher than in 54 healthy subjects (2.1 ± 0.8 gIL) or 149 nondiabetic patients (2.3 ± 0.9 gIL), and the difference was maintained regardless of whether the diabetic patients had kidney disease (3.5 ± 1.6 g/L, n = 50), liver disease (3.7 ± 2.1 g/L, n = 15), both (3.6 ± 1.7 g/L, n = 65), or neither (3.4 ± 1.6 g/L, n = 84). To confirm the association between diabetes and high serum IgA concentration, and bearing in mind the statistically significant differences among the serum IgA concentrations of healthy subjects of different age groups [4], in this study we determined the agewise distribution of circulating IgA concentration in broad samples of control subjects and diabetic patients of both types. We also studied whether high IgA concentrations were associated with microvascular and (or) macrovascular diabetic
diabetes subtypes #{149} retinopathy #{149} neuropathy hypertension #{149} macrovascular disease
complications.
In an earlier study [1], we found that immunoglobulin A (IgA) has a marked effect on fructosamine determinations: 80% of a group of nondiabetics with abnormally high IgA concentrations but no hepatic disease had abnormally high fructosamine concentrations. Subsequently [2], we found that some 30% of a group of 169 diabetic patients had high serum IgA concentrations, the same prevalence holding among both insulin-depen-
Biochemistry,
CARNERO, SANTIAGO
years or more than among
sion). The IgA concentrations of all groups of diabetic patients were significantly higher than those of the corresponding subgroups of 943 control subjects, except for type 1 patients >60 years of age. High IgA concentrations were found in 23.1% of the whole diabetic group. The prevalence of high IgA was significantly greater in males than in females among type 1 patients (24.4% vs 18%). In conclusion, an increase in circulating IgA concentrations is a generalized phenomenon among diabetic patients; IgA concentrations above the reference range are more common among male than female diabetics; and diabetic complications are associated with a significant increase in serum IgA concentration. mfs: #{149} nephropathy
and
dent (type 1) and non-insulin-dependent (type 2) diabetics; and in both subgroups, high serum IgA concentration was more
serum IgA concentrations
INDEXING
AuxIuA1o1t
QUINTEIRO,
Materials and Methods SUBJECTS
The diabetic
outpatient
clinics at the Galician
General
Hospital,
Santiago de Compostela, Spain, are attended by almost all and pediatric type 1 diabetic patients in the district and by type 2 patients requiring treatment with insulin or oral diabetic medications. The sample studied in this work
of Biochemistry, Faculty of Medicine, and Division of Clinical Hospital Xeral de Galicia, University of Santiago de Compostela,
Santiago de Compostela,Spain. “Addrew correspondence to this author at: Clinical Biochemistry, Hospital Xeral de Galicia, RlGaleras s/n, 15706 Santiago de Compostela, Spain. Fax
adult most anticom-
prised 1251 type 1 patients and 2224 type 2 diabetic patients who attended our clinics at least three times between January 1, 1992, and December
7-34-81-57 01 02. Received September 15, 1995; accepted February 21, 1996.
determined
1064
31, 1994. Among
in all these patients.
other quantities,
IgA was
Clinical Cbemistiy
Type
1 and
type
2 diabetes
mellitus
were
diagnosed
in
42, No. 7, 1996
was >140
1065
mmHg
(160 mmHg
for patients
accordance with the criteria of the US National Diabetes Data Group [5]. For this study, patient age was defined as age at the time of the patient’s latest analytical determination. Reference ranges for serum IgA were established with sera from 943 healthy subjects who volunteered themselves or were volunteered by their parents; these subjects were recruited
(c) their
diastolic
patients
>60 years of age); blood
among hospital staff and their children, children attending playschools, and residents of old people’s homes belonging to the same geographical area as the diabetic patients. The good health of all of these subjects was confirmed by medical examination at the time of blood sampling. The approval of the Hospital Ethics Committee was obtained before initiation of the study.
values showed that none of the distrithe median was used as a measure of location. Reference ranges for the following age groups were established as the interval between the 2.5% and 97.5% quantiles of the corresponding control sample: 1-10 years (61 subjects), 11-20 years (147), 21-30 years (220), 31-40 years (182), 41-50 years (132), 5 1-60 years (94), and >60 years (107). Patients were deemed to have high serum IgA concentrations when their serum IgA concentration exceeded the upper end of
ANALYTICAL
PROCEDURES
Venous blood samples were taken from an antecubital vein. Serum IgA was determined in an Array Protein System Analyzer (Beckman Instruments, Fullerton, CA) by using appropriate antisera and controls. Twenty-four-hour urine was tested for proteinuria with Labstix dipsticks (Ames, Elkhart, IN); the total protein concentration in urine specimens for which dipstick testing showed any trace of protein was quantified by the trichloroacetic acid method [6].
in accordance
CRITERIA
FOR
ASSESSMENT
OF CLINICAL
this
with
was >90 mmHg pressures
definition
(95 mmHg
for
used for classification
the
were
means
of
two
determinations. STATI STICAL Since
butions
the
was
reference
follows,
gaussian,
range
we
percentile
ANALYSIS
and kurtosis
skew
for
their
will also use
interval
when
age
“range”
applied
group. to
to data
For refer
brevity
in what
the 2.5-97.5
to
for diabetic
groups.
The computer package SPSS version 6.0 for Windows 3.1 (SPSS, Cary, NC) was used to process the data. The significance of differences between groups was estimated by Wilcoxon’s test; P 60 years of age), or
by Pearson’s
of differences
between
percentages
was estimated
test.
COMPLICATIONS
Retinopathy. Patients were considered to have retinopathy if they exhibitedretinalmicroaneurysms, soft exudates, small intraretinal hemorrhages, venous bleeding,neovascularization, or retinal traction or detachment. The presence of retinopathywas determined by an experienced observer by direct ophthalmoscopy through pupils dilated with 5 g/L tropicamide. Neuropathy.Vibration thresholds at both medial malleoli and both great toes were determined by using a biothesiometer (Biomedical Instrument Co., Newbury, OH); neuropathy was considered to be present if the thresholds were >2 above the mean of the control subjects of the same age group [7]. Neuropathy was also diagnosed when there were symptoms compatible with sensorimotor polyneuropathy, autonomic neuropathy or mononeuropathy, absent ankle reflexes, or impairment of light touch or pin-prick sensation in the feet. Nephropathy. Patients were considered nephropathic if they exhibited proteinuria of 1 g/L (or 0.3 g/L on more than one occasion) in the absence of infection and of any evident cause of renal disease other than diabetes. Macrovascular disease. Macrovascular disease was considered to be present if there was a history of myocardial infarction, angina, stroke, intermittent claudication, vascular surgery, or amputation for atherosclerotic disease, or if one or more foot pulses were absent. Hypertension. Patients were defined as hypertensive if (a) they were on antihypertensive drugs, (b) their systolic blood pressure
Results SERUM
IGA CONCENTRATIONS
Table 1 lists the serum IgA concentrations (median and reference range; see Material and Methods) in each age group of the 943 control subjects, together with the median and range in each age group of the 1251 type I patients and the 2224 type 2 patients. Fig. 1 shows the frequency distributions of IgA concentrations among type 1 and type 2 diabetic patients. The observed reference ranges are similar to those reported by other authors who used an analytical method similar to ours [4]. The overall median among the control subjects was 1.46 g/L (range 0.22-3.89 g/L); in no age group was there a statistically significant difference between male and female controls. The overall median among type 1 patients was 2.01 g/L (range 0.29-7.01 g/L), and the overall median among type 2 patients was 2.67 g/L (range 0.67-9.86 g/L). These values differed significantly from the median for control subjects (P