RESEARCH ARTICLE
SA JOURNAL OF DIABETES & VASCULAR DISEASE
152
VOLUME 7 NUMBER 4 • NOVEMBER 2010
and more precisely DR15 (2) was statistically significantly different
between the children (3.53%) and the controls (15%), since the
H0 assumption was rejected (H0: P1
=
P2), with
ε
=
2.70 being
higher than 1.96. This was in contrast to the juvenile and LADA
groups, where the frequencies were respectively (4.44 and 10%)
with
ε
calculated at 1.87 and 1.05, respectively. This was lower
than 1.96, indicating there was no significant difference between
the frequency of the presumed marker of protection in the juvenile
and LADA groups, compared to the controls.
Comparison of the marker frequencies between the
diabetics groups
Comparison of the frequency of the HLA markers for T1D between
the three diabetic groups showed they were not statistically
significantly different. The
ε
values were calculated as follows;
ε
DR3
=
0.24;
ε
DR4
=
1.59;
ε
DQ2
=
0.22 and
ε
DQ8 (3)
=
2.23 for
the children, and
ε
DR3
=
0.39;
ε
DR4
=
0.39;
ε
DQ2
=
0.99 and
ε
DQ8 (3)
=
0.1 for the juveniles, compared to the LADA. The LADA
introduced the same markers of predisposition as type 1 diabetes.
By contrast, the values of the frequencies of the protective marker
-DR15 (2) were 3.53% in the children and 4.44% in the juveniles,
compared with the LADA (10%).
The
ε
value calculated for these markers was lower than 1.96
in each of the three diabetic groups, therefore (H0: P1
=
P2) was
retained. The frequency of these markers was slightly higher
in the children compared to the juveniles. However, this was
not statistically significant [
ε
DQ8 (3)
=
1.98 is higher than 1.96,
therefore H0: P1
=
P2 was rejected]. We also observed that the
marker DR15 (2), which was negatively associated with the disease,
was slightly higher for the juveniles (4.44%) compared to the
children (3.53%). This difference was not significant, according to
the reduced standard deviation test (
ε
=
0.26
<
1.96).
Discussion
T1D is the end result of an interaction between environmental
and host susceptibility factors, and variation in either or both of
these factors may account for the widely varying incidence of the
disease. The MHC contributes 50% of the inherited risk for T1D.
4
An additional 17 genes with variable but small effects, namely
IDDM1 to IDDM17, located on different chromosome have been
described.
17
The MHC which encodes the HLAs is located on
6p21.3.
18,19
The other genes are found on chromosomes 2, 11, 10,
14, 15 and 18, and four additional regions are on chromosome 6q.
Of these, the strongest candidates for disease susceptibility, other
than MHC, are on the insulin (Ins) locus on 11p15 and the cytotoxic
T-lymphocyte-associated antigen 4 (CTLA4) locus in 6q27.
20,21
The observations from our study are consistent with two previous
reports regarding T1D susceptibility, in that the frequencies of the
various markers of susceptibility, such as HLA-DR3, -DR4, -DQ2
and -DQ8 (3) differed significantly for children, juvenile and LADA
groups from those of the Algerian controls (O.R
=
5.54, 4.36, 5.09)
(OR
=
5.1, 2.66, 3.08) (OR
=
5.02, 3.09, 4.64) and (OR
=
7.67, 3.66,
3.79), respectively. The present study is the first on subjects from
the central area of Algeria and yielded similar results, not only with
regard to a positive disease association for all the same HLA class
II markers, but we also identified significantly lower HLA antigen
frequencies in each of the three diabetic groups: -DR5 (OR
=
0.03;
0.0; 0.0), -DR15 (OR = 0.21; 0.26; 0.63) and -DQ6 (OR
=
0.27;
0.31; 0.29), indicating these antigens have a protective role in T1D
disease.
The comparison of the frequencies and the reduced standard
deviation
ε
(H0: Pl
=
P2) enabled us to confirm that the LADA are
a population of patients which should be classified in the category
of type 1 diabetes, as there was no significant difference between
the frequencies. The results obtained suggest that the juveniles and
LADAs acquired type 1 diabetes later than the children because of
the expression of at least one of the protective markers, i.e. DR15
(2), -DQ6. Furthermore, the results indicate that the frequency of
this marker increases with increasing age of appearance of type 1
diabetes.
The HLA-DR antigen frequencies of our study population were
comparable with those previously reported.
22
In this earlier study,
type 1 diabetics (
n
=
39) and controls (
n
=
52) originating from the
west Algerian region of Tlemcen were compared and the result
showed no association of HLA-DR3 with the disease (
p
=
1.00, OR
=
0.95). The frequency of HLA-DR4 and the DR3/DR4 combination
of antigens however were associated with susceptibility to develop
type 1 diabetes in this Tlemcen population (DR4, OR
=
2.1; DR3/
DR4, OR
=
1.3). This previous study highlighted a characteristic of
the Tlemcen population with a history of consanguine marriages.
The population in our study were more heterogeneous than the
Tlemcen population, which may explain the different findings.
Another previously reported study
23,24
of a population in western
Algeria (Oran), which comprised 50 unrelated T1D patients and 46
controls. Both DRB1*0301-DQA1*0501-DQB1*0201 (DR3-DQ2)
and DRB1*04-DQA1*0301-DQB1*0302 (DR4-DQ8) haplotypes
were found at increased frequencies among the patients compared
to the controls (45 vs 13%, RR
=
5.5,
P
c
<
10
-5
and 37 vs 4%, RR
=
12.9,
P
c
<
10
-4
, respectively). This observation suggests that the
DRB1*0405 allele which encoded an Asp57-negative beta-chain
may contribute to T1D susceptibility in a similar way to the Asp
57-negative DQ beta-chain.
Our results comply with those reported by the other groups,
which showed that 90 to 95% of the type 1 diabetics were HLA-
DR3 and/or -DR4 positive, compared with 40 to 45% in the general
population, with relative risks, respectively, of 3.4 and 6.4, and 40
for the subjects positive for HLA-DR3/DR4. By contrast, a strong
protection was conferred by the haplotype DR15(2)-DQ6.
17
Previous
work reported by the Center for Blood Transfusion of Tunis,
including 43 patients who had developed diabetes type 1 and
were compared with a group of controls (
n
=
101), showed that
the genotype heterozygote HLA-DR3/DR4 was strongly associated
with type 1 diabetes and that the antigen DR15-DQ6 proved to be
protective in the diabetic Tunisian population.
25
Research on various groups including a Caucasian population
also revealed that the heterozygous genotype DR3/DR4 was strongly
associated with type 1 diabetes. On the other hand, the antigen
DR15(2)-DQ6 proved to be protective in the Caucasian population.
The HLA-DR15-DQ6 was negatively associated with the disease in
all populations.
26-30
Our results show that HLA-DR9 (OR
=
1.1; 1.04;
0.28,
p
>
0.05) was neutral. The common occurrence of -DR9 in the
Asian T1D patients, and in particular HLA-DR3/-DR9, was perhaps
because of the high frequency of HLA-DR9 in the general Asian
population. The predisposing HLA-DR3 in Algerian and Caucasian
populations is relatively rare in Asian populations.
31
Worldwide studies have therefore clearly shown that HLA class
II loci have an important role as susceptibility determinants for T1D.
Recently, other genes within the MHC have been suggested, as
class I alleles seem to modify the risk of expression, influencing the
