Objective To evaluate the role of grandparents' history of diabetes on defining family history of diabetes in order to identify adolescents at high risk of diabetes.

Study design and Setting We evaluated 1276 population-based adolescents, aged 13-year-old, from Portugal. Data were collected by self-reported questionnaires and a clinical evaluation was performed, including a fasting blood sample.

Results The prevalence of impaired fasting glucose (IFG)/diabetes was 4.1% using American Diabetes Association (ADA) criteria and 0.94% using the WHO criteria. Using data only from Parental History, 103 adolescents (8%) were identified has having a positive family history of diabetes, while combining both parental and grandparental history (Total Family History), 468 adolescents was additionally identified as having a positive history, performing a total of 571 adolescents (45%). After adjusting for sex and parents education, the OR for having fasting plasma glucose ≥75th percentile considering only Parental History was 0.91 (95% CI 0.57 to 1.47) and combining data from parents and grandparents the OR was 1.17 (95% CI 0.83 to 1.65). Parental History's sensitivity to identify IFG/diabetes by the ADA criteria was 2.6%, while total family history's sensitivity (including grandparents' data) was 62%.

Conclusion Although there's no significant association between IFG and a positive family history, combining parental with grandparental history (total family history) lead to a 5.5-fold increase in the identification of adolescents with a positive family history. So, on adolescent screening, it seems better to use parental plus grandparental information instead of just relying on parental data.