OBJECTIVETo investigate the joint effect of birth weight and each of obesity measures (body mass index (BMI) and waist circumference (WC)) on abnormal glucose metabolism (including diabetes) at adulthood.

METHODSUsing the historical cohort study design and the convenience sampling method, 1 921 infants who were born in Beijing Union Medical College Hospital from June 1948 to December 1954 were selected to do the follow-up in 1995 and 2001 respectively. Through Beijing Household Registration and Management System, they were invited to participate in this study. A total of 972 subjects (627 were followed up in 1995 and 345 were followed up in 2001) with complete information on genders, age, birth weight, family history of diabetes, BMI, WC, fasting plasma glucose (FPG) and 2-hour plasma glucose (2 h PG) met the study inclusion criteria at the follow-up visits. In the data analysis, they were divided into low, normal, and high birth weight, respectively. The ANOVA and Chi-squared tests were used to compare the differences in their characteristics by birth weight group. In addition, multiple binary Logistic regression model was used to investigate the single effect of birth weight, BMI, and waist circumference on abnormal glucose metabolism at adulthood. Stratification analysis was used to investigate the joint effect of birth weight and each of obesity measures (BMI and WC) on abnormal glucose metabolism.

RESULTSThere were 972 subjects (males: 50.7%, mean age: (46.0±2.2) years) included in the final data analysis. The 2 h PG in low birth weight group was (7.6±3.2) mmol/L , which was higher than that in normal birth weight group (6.9±2.1) mmol/L and high birth weight group (6.4±1.3) mmol/L (F=3.88, P=0.021). After adjustment for genders, age, body length, gestation age, family history of diabetes, physical activity, smoking and alcohol consumption, and duration of follow-up, subjects with overweight and obesity at adulthood had 2.73 (95% confidence interval (CI) =2.06- 3.62) times risk to develop abnormal glucose metabolism when compared with norm weight ones. Likewise, subjects with central obesity were more likely to develop abnormal glucose metabolism than ones with normal waist (odds ratio (OR)=3.35, 95%CI=2.49-4.50). In addition, compared to subjects with normal birth weight and normal BMI at adulthood, ones with normal birth weight and overweight (including obesity) at adulthood were more likely to have abnormal glucose metabolism (OR= 2.60, 95%CI=1.94-3.49); subjects with low birth weight and overweight (including obesity) at adulthood had the highest risk for abnormal glucose metabolism (OR=4.70, 95% CI=1.84- 11.99). The attributable proportion of interaction between low birth weight and overweight (including obesity) at adulthood was 48.5%. In addition, compared to subjects with normal birth weight and normal WC at adulthood, one with normal birth weight and central obesity at adulthood were more likely to have abnormal glucose metabolism (OR=3.18, 95% CI=2.33- 4.32); subjects with low birth weight and central obesity at adulthood had the highest risk for abnormal glucose metabolism (OR=4.78, 95% CI=2.01- 11.38); subjects with high birth weight and central at adulthood also had high risk for abnormal glucose metabolism (OR=4.35, 95%CI=1.38- 13.65). We found that the attributable proportion of interaction between low birth weight and central obesity at adulthood was 38.5% , and was 28.3% for interaction between high weight and central obesity.

CONCLUSIONThere was strong interaction effect between birth weight and overweight (especially central obesity) at adulthood on abnormal glucose metabolism at adulthood. Effective measures should be adopted to prevent and control adult obesity in order to offset the adverse effect of birth weight on long-term health risk.