Strategy for early identification of prediabetes in lean populations: New insight from a prospective Chinese twin cohort of children and young adults

Aims: To identify certain subgroups in young and lean populations, who may be at a high risk of developing prediabetes/diabetes, which is not captured by current BMI-based screening algorithms. Methods: Incidence of prediabetes/diabetes was assessed using oral glucose tolerance tests among 1859 children and 1073 young adults from a prospective Chinese twin cohort. Results: Over a 6-year follow-up, 507 (27.3%) children and 293 (27.3%) adults developed prediabetes/diabetes. Of the 800 incidents, 737(92.1%) and 644(80.5%) were lean at baseline and follow-up, respectively. Baseline fasting glucose in the top tertile of the normal range was associated with an increased risk of prediabetes/diabetes: odds ratio, 1.85 (95% CI 1.32-2.59) and 3.29 (95% CI 2.10-5.17) among normal weight and underweight children, respectively, and 2.74 (95% CI 1.78-4.23) and 3.08 (95% CI 1.69-5.58) among normal weight and overweight/obese adults, respectively, compared with the low tertile of fasting glucose. Conclusions: We showed that majority incident cases of prediabetes/diabetes were not overweight/obese (at baseline), who would have been missed by traditional screening algorithm emphasizing overweight/obesity. Our findings revealed that an upper end of normal fasting glucose was a simple and robust predictor of future higher risk of prediabetes/diabetes in this young and lean population. (C) 2018 Elsevier B.V. All rights reserved.