OBJECTIVETo investigate the correlation between obesity in children and diabetes in adults from a cohort study, and further more to explore the necessity of preventing diabetes by controlling obesity in children.

METHODSIn 1987, 3 198 children and adolescents aged 6-18 were recruited from 6 elementary schools and 6 high schools located in 3 districts (Chaoyang, Haidian, and Xicheng) of Beijing using stratified cluster sampling design. The physical examination process included physical development test, blood pressure measurement, and questionnaire investigation. All children were invited to participate in the study, except for those who had history of congenital heart disease, chronic kidney disease, and limb disability. A total of 1,225 adults were enrolled in a prospective follow-up study from March 2010 to July 2012, anthropometric measures and blood sample were obtained. The obesity was defined by the following criteria: for children aged 6, the age-and the gender-specific 95th percentile of BMI from the US Centre for Disease Control and Prevention Growth charts 2000 as the baseline; for children age 7-18, recommendation from Working Group on Obesity in China (WGOC) as the standard; for adults, BMI≥28 kg/m(2) as the diagnosis standard. Diabetes was defined based on fasting plasma glucose(FPG) ≥7.0 mmol/L or 2 hours postprandial blood glucose (2 h PG) ≥11.1 mmol/L or glycosylated hemoglobin (HbA1c) ≥6.5% or current using blood glucose-lowering agents or current using insulin. Logistic regression was used to analyze the association obesity in children with diabetes in adults.

RESULTSThe prevalence of diabetes diagnosed by FPG and 2 h PG in adults who were obese children (16.2%, 18/111) was higher than those who were non-obese children (5.6%, 62/1,114)(χ(2)=18.76, P<0.001). The prevalence of diabetes diagnosed by HbA1c in adults who were obese children(18.1%,20/111) was higher than those who were non-obese children (6.9%, 77/1,114) (χ(2)=16.66, P<0.001). With multi-factor logistic regression analysis, we found that after controlling follow-up age, genders and lifestyle (smoking, alcohol consuming, dietary, and sleeping), in comparison with those non-obese from childhood to adulthood, those obese only in childhood or only in adulthood did not predict any risk of diabetes diagnosed by blood glucose in adults (OR(95%CI) were 1.90 (0.86-4.19), 1.71(0.50-5.79), respectively). Those obese both in childhood and in adulthood increased the risk of diabetes diagnosed by blood glucose in adults (OR(95%CI) was 4.50(2.22-9.14)). With multi-factor logistic regression analysis, we found that after controlling age, sex and lifestyle (smoking, alcohol consuming, dietary, and sleeping) in comparison with those non-obese from childhood to adulthood, those obese only in childhood or only in adulthood did not increase the risk of diabetes diagnosed by HbA1c in adults (OR(95%CI) were 1.42(0.71-2.86), 3.13(0.83-11.75), respectively). Those obese both in childhood and in adulthood increased the risk of diabetes diagnosed by HbA1c in adults (OR(95%CI) was 5.93(3.06- 11.49)).

CONCLUSIONObesity in children even sustained to adulthood was a risk factor for diabetes in adulthood. It is necessary to control obesity in children to prevent diabetes in adults.