A Novel Synthetic Cohort Approach for the Study of the Developmental Origins of Cardiovascular Disease

• In Event: 1-005 - Data Integration

Abstract

The development of subclinical atherosclerosis and clinical cardiovascular and cerebrovascular disease (CVD) is an unfolding process across the lifespan. Increasingly, there is evidence that the origins of these processes begins in utero and spans the life course as clinical event rates increase. Most research on cardiovascular risk factors comes from studies examining the association between risk factor levels measured most often in adulthood, with incident disease over the short- to intermediate-term. Little is understood about the early life influences on long-term CVD health. Advances in the transdisciplinary developmental sciences could enable tracing of the evolution of risk and robustness beginning very early in life. Thus, the imperative for a life course approach is critical to understand the developmental origins of cardiovascular health and disease and we have the tools in hand to do so. No single study has collected detailed phenotypic data over individuals’ entire lifespans nor would this be feasible. In this paper we propose an innovative methodology to pool data from existing child and adult cohorts to create a “synthetic cohort” enabling the study of CV health across the lifespan. This approach is intuitively appealing, but methodologically complex. Synthetic cohorts offer the opportunity to understand CV health across the lifespan making efficient use of the observational cohort data already available. As an example of how this methodology can be implemented we developed and validated a synthetic cohort approach to examine CV risk factors from age 18-90 years of age. Among 41,387 participants from 7 cohort studies, we multiply imputed the participants’ CV risk profile and outcome using the available exam records based on a joint multi-level imputation model via the jomo package in R. To validate our imputed values, we removed selected observed data. We then imputed the CV risk profile of these deleted values and compared imputed and observed values. This synthetic cohort approach provides valid and unbiased estimates of CV factors across the lifespan. These modeling approaches lend themselves to lifespan analysis. We present a novel framework that fuses developmental and cardiovascular epidemiologic frameworks to “close the gap” in the early life origins of CVH/D by mining data from pediatric cohorts to model CVH trajectories in thousands of children from birth to age 12. Future studies using this synthetic cohort can provide novel insights into the origins of CV disease using cutting-edge analytic and developmental approaches.

Authors

• Rachel Flynn, Northwestern University
  Presenting Author

• Norrina B Allen, Northwestern University
  Presenting Author

• Hongyan Ning, Northwestern University

• Amy Krefman, Northwestern University

• Lauren Wakschlag, Northwestern University

• Lihui Zhao, Northwestern University

• Donald Lloyd-Jones, Northwestern University

• Juned Siddique, Northwestern University