Li, L;
(2011)
Joint Multivariate Response Modeling for Repeated BMI Measures and Single Measures of Adult Cardiovascular Disease Risk Factors.
In:
(Proceedings) 2011 Joint Statistical Meeting.
![[thumbnail of Li%20JSM2011.pdf]](https://discovery.ucl.ac.uk/style/images/fileicons/application_pdf.png) |
PDF
Li%20JSM2011.pdf Available under License : See the attached licence file. Download (425kB) |
Abstract
Studies of the association between developmental trajectories and adult health require methods relating developmental measures from different life stages to single measures of health outcomes. We present a joint multivariate response model to a longitudinal response variable, body mass index (BMI), and two single measure adult outcomes - systolic blood pressure (SBP) and high density lipoprotein cholesterol (HDL-C), to investigate the association between BMI trajectories and adult cardiovascular disease (CVD) risk factors. We adopt a linear spline model for repeated BMI measures to allow for distinct childhood and adult curves and separate models for SBP and HDL-C. The models are fitted simultaneously by assuming the joint distribution of random coefficients. The model is applied to the 1958 British Birth Cohort (n=17,000), whose BMI was recorded at six ages from 7 to 45y (16,820 with one or more measures) and SBP and HDL-C were measured at 45y. Results show that the rate of BMI gain in adulthood has a stronger association with SBP and HDL-C than the rate of BMI growth in childhood (p<0.05 for SBP in boys and for HDL-C in both genders). For SBP, the estimated correlation for the rate of adult BMI gain is 0.27 (95% CI: 0.23, 0.32) in males and 0.35 (0.31, 0.38) in females, compared to 0.22 (0.16, 0.27) and 0.18 (0.11, 0.25) respectively for the rate of childhood growth. For HDL-C the correlation is -0.43 (-0.48, -0.38) and -0.45 (-0.48, -0.41) for adult BMI gain compared to -0.14 (-0.17, -0.10) and -0.26 (-0.30, -0.22) for childhood growth. Furthermore, the rate of childhood growth is associated with adult outcomes, independent of adult BMI gain. Conclusions Joint multivariate response modeling is a useful approach for estimating the association between repeated exposure variables at different life stages and adult outcomes, therefore has important applications in the life-course epidemiology.
Archive Staff Only
 |
View Item |
