Table 3.
Basic data and 2 × 2 comparisons to detect the strongest human leukocyte antigen association of the risk factors a
|
A*02 |
B*51 |
BD patients |
Controls |
|||
| + |
+ |
72 |
29 |
|||
| + |
– |
76 |
112 |
|||
| – |
+ |
45 |
11 |
|||
| – |
– |
73 |
161 |
|||
|
A*24 |
B*51 |
|
|
|||
| + |
+ |
26 |
6 |
|||
| + |
– |
36 |
30 |
|||
| – |
+ |
91 |
34 |
|||
| – |
– |
119 |
237 |
|||
|
B*57 |
B*51 |
|
|
|||
| + |
+ |
8 |
0 |
|||
| + |
– |
24 |
11 |
|||
| – |
+ |
111 |
40 |
|||
| – |
– |
131 |
262 |
|||
|
Factor 1-B*51 |
A*02 |
|
A*24 |
|
B*57 |
|
| |
P
c
|
OR |
P
c
|
OR |
P
c
|
OR |
| Test 1 |
>0.05 |
0.62 |
>0.05 |
1.54 |
>0.05 |
6.17 |
| Test 2 |
>0.05 |
1.49 |
0.008 |
2.38 |
<0.0006 |
4.25 |
| Test 3 |
<0.0006 |
5.40 |
<0.0006 |
8.10 |
0.0012 |
33.9 |
| Test 4 | <0.0006 | 0.17 | >0.05 | 0.45 | >0.05 | 0.77 |
aBD, Behçet’s disease; HLA, human leukocyte antigen; odds ratio; Pc, corrected value of P. Test 1: ++ vs. –+ was performed to determine whether the HLA proposed factor was associated in B*51 positives. Test 2: +- vs. –– was designed to determine whether the HLA proposed factor was associated in B*51 negatives. Test 3: ++ vs. –– was conducted to determine the combined association between the HLA proposed factor and HLA-B*51. Test 4: +- vs. –+ was done to determine whether the associations between the HLA proposed factor and HLA-B*51 were different. Pc was calculated from the different 2 × 2 tables using Fisher’s exact test and corrected by multiplying by 6 (see [23]). Each comparison included B*51 and other HLA risk factors associated with the disease or suggestive of an association with the disease.