P. Kulmala et al., Genetic markers, humoral autoimmunity, and prediction of type 1 diabetes in siblings of affected children, DIABETES, 49(1), 2000, pp. 48-58
The relationships between genetic markers and disease-associated autoantibo
dies were studied in an unselected population of 701 siblings of children w
ith type 1 diabetes, and the predictive characteristics of these markers ov
er a period of 9 years mere determined. Increased prevalences of all the an
tibodies were closely associated with HLA identity to the index case, the D
R4 and DQB1*0302 alleles, and the DR3/4 phenotype and the DQB1*02/0302 geno
type. Antibodies to GAD (GADA) were also associated with the DR3 and DQB1*0
2 alleles. Siblings carrying the protective DR2 and DQB1*0602-3 alleles wer
e characterized by lower frequencies of islet cell antibodies (ICA), antibo
dies to IA-2 (IA-2A), and GADA. Higher levels of ICA were related to IFLA i
dentity, the DR4 and DQB1*0302 alleles, and the susceptible DQB1 genotypes,
while no significant differences mere observed in the levels of IA-2A, GAD
A, or insulin autoantibodies among siblings with different HLA risk markers
. The DR2 or DQB1*0602-3 alleles mere not related to the levels of any anti
body specificity. A combination of the genetic markers and autoantibodies i
ncreased the positive predictive values of all autoantibodies substantially
, which may have clinical implications when evaluating the risk of developi
ng type 1 diabetes at the individual level or when recruiting high-risk ind
ividuals for intervention trials. However, because such combinations also r
esulted in reduced sensitivity, autoantibodies alone rather than in combina
tion with genetic markers are recommended as the first-line screening in si
blings. Finally, not all siblings with a broad humoral autoimmune response
or high-risk genetic markers present with type 1 diabetes, while some with
a low genetic risk and weak initial signs of humoral autoimmunity may progr
ess to disease.