Blood glucose levels and the relationship of body mass index and circumference with blood glucose in China

Abstract

Objective: To describe and compare blood glucose levels in adults aged 18 years old and above in China and explore the relationship between BMI and waist circumference with blood glucose. Methods: China Chronic Disease and Risk Factor Surveillance were conducted in 298 counties/districts in China in 2018, covering 31 provinces (autonomous regions, municipalities). A multi-stage stratified cluster random sampling method selected permanent residents aged 18 years and above. Information on demographics, behavior-related risk factors, BMI, waist circumference, and blood glucose were collected through a face-to-face questionnaire, physical measurement, and laboratory examination. After complex weighting of data, they described the blood glucose levels of people with different characteristics and explored the relationship of BMI and waist circumference with blood glucose by multiple linear regression model analysis. Results: A total of 177 816 adults were included in the study. The average fasting blood glucose and average glycosylated hemoglobin were (5.73±1.46) mmol/L and (5.37±0.83) %, with people aged 60 years old and above group highest than that of other, with males higher than females (P<0.001); and urban was higher slightly than rural for the average of average glycosylated hemoglobin (P<0.001). The average fasting blood glucose and average glycosylated hemoglobin increased with increased BMI and waist circumference (P<0.001). Results from multiple linear regression model analysis showed that: 1) for each increase in BMI unit and waist circumference, the fasting glucose levels increased by 0.019 mmol/L and 0.008 mmol/L (all P<0.001) in those not diagnosed with diabetes, 2) by 0.021 mmol/L (P=0.163) and 0.014 mmol/L (P=0.004) in those newly detected as diabetes, and 3) by 0.028 mmol/L (P=0.088) and 0.023 mmol/L (P<0.001) in those self-reported as having been diagnosed as diabetes, respectively. However, glycosylated hemoglobin levels increased: 1) by 0.015% and 0.006% in those not diagnosed as diabetes (all P<0.001), 2) by 0.050% and 0.019% in those newly detected as diabetes (all P<0.001), and 3) by 0.033% and 0.019% in those self-reported as having been diagnosed as diabetes (all P<0.001), respectively. These associations with waist circumference were more robust than with BMI. Conclusions: Adults not diagnosed with diabetes with abnormal BMI or waist circumference are the key population for prevention and control. Measures improving the awareness rate of waist circumference should be taken to maintain average blood glucose in various groups.