Näntö-Salonen K et al. (2008) Nasal insulin to prevent type 1 diabetes in children with HLA genotypes and autoantibodies conferring increased risk of disease: a double-blind, randomised controlled trial. Lancet 372: 1746–1755

Prophylactic administration of insulin reduces the incidence of type 1 diabetes in mouse models. However, a study conducted in Finland has found that nasal insulin does not prevent diabetes in children genetically susceptible to the disease. This study was prematurely terminated in its 10th year, because the treatment lacked efficacy.

Among 116,720 children who were screened at birth, Näntö-Salonen et al. identified 224 with HLA-DQB1 alleles associated with high risk for type 1 diabetes and multiple diabetes-associated autoantibodies. These children (aged >1 year) were randomly assigned either nasal insulin or placebo. Of the 83 insulin-treated children who completed the study, 42 developed type 1 diabetes, compared with 38 of 85 placebo-treated children. Annual rates of progression to diabetes were 16.8% and 15.3%, respectively. The study also included 40 siblings of index participants, among whom similar outcomes were observed. Across all children included in the randomization, the hazard ratio for insulin treatment versus placebo was 0.98 (95% CI 0.67–1.43). However, in a subgroup of 68 children who presented with three or four different autoantibodies, the corresponding hazard ratio was 1.50 (95% CI 0.90–2.40), which raises the worrying possibility that nasal insulin could accelerate the development of diabetes in such children.

Although the insulin intervention was ineffective, the authors noted that study participants benefited from early HLA screening because diabetes was diagnosed promptly, which facilitated the achievement of metabolic control and prevented ketoacidosis.