Ill health in early life has a significant negative impact on school grades, grade repetition, educational level, and labor market outcomes. However, less is known about qualitative socio-economic consequences of a health shock in childhood or adolescence. We investigate the relationship between onset of type 1 diabetes up to age 15 and the probability of choosing and completing a health-oriented path at upper secondary and university level of education. We analyze the Swedish Childhood Diabetes Register, the National Educational Register, and other population registers in Sweden for 2756 people with type 1 diabetes and 10,020 matched population controls. Educational decisions are modeled as unsorted series of binary choices to assess the choice of educational field as a potential mechanism linking early life health to adult outcomes. The analyses reject the hypothesis of no systematic differences in choice of educational field between people with and without type 1 diabetes at both levels. The results are robust to selection on ability proxies and across sensitivity analysis. We conclude that the observed pro health-oriented educational choices among people with type 1 diabetes in our data are consistent with disease onset in childhood and adolescence having qualitative impact on life-course choices.