The link between hyperglycemia and the complications of diabetes is unknown. It is still discussed whether oxidative stress precedes or merely reflects diabetic complications. To search for a familial predisposition to oxidative stress, we investigated indexes of glucose and lipid metabolism, markers of plasma and cell lipid oxidation, a marker of oxidant-induced protein damage, and the effects of oxygen radicals on erythrocytes (or red blood cells [RBCs]) of patients with type 1 diabetes and their relatives. RESEARCH DESIGN AND METHODS: We recruited 30 type 1 diabetic subjects (10 without diabetic complications, 10 with retinopathy, and 10 with nephropathy), 36 nondiabetic siblings, 37 nondiabetic parents of type 1 diabetic subjects, and 3 control groups of healthy subjects without a family history of diabetes. Levels of blood creatinine, glucose, HbA(1c), cholesterol, triglycerides, lipoprotein(a) (Lp[a]), fibrinogen, malondialdehyde (MDA), and advanced oxidation protein products were determined. The RBC response to oxidative stress (3-h incubation at 37 degrees C with or without a radical generating system) was evaluated by measuring RBC glutathione (GSH), RBC-MDA, and hemolysis. RESULTS: Diabetic patients had higher levels of blood glucose (P < 0.001), HbA(1c) (P < 0.001), Lp(a) (P < 0.01), and fibrinogen (P < 0.05) than control subjects. Siblings of diabetic patients had higher Lp(a) levels (P < 0.001). Parents had higher levels of plasma glucose (P < 0.05) and Lp(a) (P < 0.01). Plasma and RBC-MDA were significantly elevated in diabetic subjects and relatives compared with control subjects. Basal RBC-GSH was lower in diabetic subjects (P < 0.01). In diabetic subjects, incubations of cells caused a decrease in RBC-GSH of a lesser degree than that in control subjects, but they caused a significant increase in hemolysis. Among relatives, hemolysis was increased both at baseline and after incubation. Plasma MDA levels were associated with blood glucose, creatinine, and fibrinogen levels (multiple r = 0.5, P < 0.001), and basal RBC-MDA levels were associated with plasma Lp(a), fibrinogen, and plasma MDA levels (r = 0.6, P < 0.001). Basal RBC-GSH content correlated with serum glucose and RBC-MDA production (r = 0.3, P < 0.01). CONCLUSIONS: Our study is the first to present evidence that markers of lipoprotein metabolism (Lp[a]), oxidative stress (plasma and RBC-MDA), and cellular fragility (hemolysis) are abnormal in nondiabetic relatives of type 1 diabetic subjects, thereby supporting the view that familial elements of diabetes even precede the onset of diabetes. It seems reasonable that the same biological markers considered major predictors of cardiovascular disease can also trace familial susceptibility to type 1 diabetes, just as they have been associated with the development of type 2 diabetes.