Roles of weight change from early to middle adulthood and BMI in relation to cardiometabolic disease risks: findings from the China Kadoorie Biobank
Y Chen, A Iona, P Yao, Z ChenAbstract
Background/Introduction
BMI at different life stages and weight change in adulthood may reflect distinct aspects of adiposity exposure. Evidence comparing the contributions of early adulthood BMI, midlife BMI, and weight change from early to mid-adulthood to cardiometabolic disease risk remains limited.
Purpose
To compare the roles of weight change and BMI from early to middle adulthood in shaping risks of major cardiometabolic diseases (CMDs).
Methods
We analysed 461,046 adults (range: 30-79; mean: 52.0 years) in the China Kadoorie Biobank, excluding those with prior major chronic diseases at baseline. Weight change was defined as measured baseline weight minus recalled weight at age 25. BMIbaseline was calculated as baseline measured weight/height², and BMI25 as weight at age 25 divided by baseline height calibrated for age-related shrinkage. Outcomes were ascertained via linkage to health records. Cox regression models adjusted for sociodemographic, lifestyle factors and family history were used to examine associations of BMI25, BMIbaseline, and weight change with incident type 2 diabetes (T2D), ischaemic heart disease (IHD), ischaemic stroke (IS), and intracerebral haemorrhage (ICH). Sequential adjustment and joint analyses were conducted to disentangle independent and interrelated effects among adiposity measures.
Results
Mean BMI increased from 21.8 kg/m² at age 25 to 23.7 kg/m² at baseline, with an average weight gain of 4.4 kg (SD 8.8). During a median follow-up of 12.3 years (IQR 11.4-13.2), 18,109 T2D, 41,706 IHD, 42,357 IS, and 9,975 ICH incident cases were recorded. Higher BMI25 and BMIbaseline were each associated with higher risks of all four outcomes in models without mutual adjustment. Weight change was approximately log-linear for T2D and IS, whereas for IHD and ICH it showed a J/U-shaped association, with the lowest risks around 0 kg change. Each 10-kg weight gain was associated with higher risks of T2D (HR 1.92, 95%CI 1.88-1.96), IHD (1.23, 1.21-1.25), IS (1.17, 1.15-1.19), and ICH (1.14, 1.09-1.18). After further adjustment for BMI25, the associations with weight gain strengthened. Instead, after further adjustment for BMIbaseline, the associations with weight change remained strong for T2D and IHD but were attenuated to null for IS, and the association pattern reversed for ICH. In joint analyses stratified by BMIbaseline, weight gain was associated with a higher risk of T2D and IHD, weight loss with a higher ICH risk, and little variation was observed for IS across weight change categories.
Conclusion(s)
Weight change and BMI capture different dimensions of adiposity relevant to CMD risks. For T2D and IHD, weight change and BMIbaseline were both important determinants, whereas BMIbaseline appears dominant for IS and early-adulthood adiposity may be particularly important for ICH. Our findings highlighted heterogeneous life-course mechanisms linking adiposity to different CMDs.