Genetic, autoimmune, and clinical characteristics of childhood- and adult-onset type 1 diabetes

OBJECTIVE - To assess whether there are any differences in genetic, autoimm une, or clinical features between type 1 diabetes presenting in childhood a nd that diagnosed later. RESEARCH DESIGN AND METHODS - We studied 352 individuals (252 children and adolescents <20 years of age and 100 adults greater than or equal to 20 yea rs of age) manifesting clinical signs of type 1 diabetes over a period of 7 .5 years at a university hospital in northern Finland with a primary catchm ent area population of similar to 300,000. The patients were analyzed for s usceptible and protective HLA-DQB1 alleles (*02, *0302, *0301, *0602, *0603 , and *0604), islet cell antibodies (ICA), insulin autoantibodies, and anti bodies to GAD and IA-2 (IA-2A). Their clinical symptoms and signs were reco rded at diagnosis. RESULTS - The adult patients carried the high-risk DQB1*02/0302 genotype le ss frequently than the children and more often had protective genotypes. Th ey also had a decreased frequency of all 4 single autoantibody specificitie s and of multiple (greater than or equal to 3) autoantibodies. The proporti on of patients testing negative for all autoantibodies was lower among the children than among the adults. IA-2A were associated with the DQB1*0302/x genotype in both the children and adults, and the same held true for ICA am ong the adults. The adults were characterized by a higher proportion of mal es, a longer duration of symptoms, and a lower frequency of infections duri ng the preceding 3 months. In addition, they had a higher relative body wei ght on admission and milder signs of metabolic decompensation (higher pH, b ase excess, and bicarbonate concentrations) and a lower glycated hemoglobin level at diagnosis than the children. CONCLUSIONS - Clinical manifestation of type 1 diabetes before the age of 2 0 years is associated with a strong HLA-defined genetic disease susceptibil ity an intensive humoral immune response to various beta-cell antigens, a h igher frequency of preceding infections, and a shorter duration of symptoms and more severe metabolic decompensation at diagnosis. Taken together, the se observations suggest that the age at clinical onset of type 1 diabetes i s determined by the intensity of the beta-cell-destructive process, which i s modulated by both generic and environmental factors.