Adolescent Growth Spurt and Type 2 Diabetes Risk in an American Indian Population

Background: Some studies suggest that earlier pubertal development is associated with obesity and metabolic disorders in adulthood, but whether accelerated adolescent linear growth spurt influences type 2 diabetes (T2D) in populations at high risk of diabetes is unclear. Aim: To examine associations of age at initiation of the adolescent growth spurt (i.e., take-off) (ATO) , velocity at take-off (VTO) , and age at peak velocity (APV) with T2D incidence. Methods: Height measurements obtained from a longitudinal study of diabetes in American Indians were used to estimate biological growth parameters using the Preece-Baines growth model I, a parametric model of the human growth curve. Diabetes incidence was analyzed among individuals who had ≥ 4 height examinations at ages 5-17 years and who remained nondiabetic at first examination after age 18. Participants were followed until diabetes developed or last examination. Associations of growth parameters with diabetes incidence were analyzed by Poisson regression. Results: Among 649 originally nondiabetic participants, 135 females and 82 males developed diabetes in a median follow-up of years. With adjustment for age, birth date and maternal diabetes, younger ATO and APV were associated with T2D incidence in males (incidence rate ratio (IRR) =1.75 per year, 95% CI, 1.23, 2.50; and IRR=1.33, 95% CI, 1.01, 1.75, respectively) but not in females (IRR=0.93, 95% CI, 0.75, 1.16; and IRR=1.04, 95% CI, 0.86, 1.26, respectively) . Higher VTO was associated with T2D incidence in males (IRR=1.80 per cm/y, 95% CI, 1.15, 2.81) but not in females (IRR=1.15, 95% CI, 0.88, 1.50) . With further adjustment for adult and childhood BMI and fasting insulin, the associations were attenuated for APV and VTO but remained significant for ATO in males. Conclusion: Our findings suggest that an early and accelerated adolescent growth spurt is a risk factor for T2D, at least in males. These associations are only partially explained by measures of adiposity and insulinemia. Disclosure M.J.Ramirez-luzuriaga: n/a. S.Kobes: None. M.Sinha: None. W.C.Knowler: None. R.L.Hanson: None.