Principal Component Scores in Predicting Cardiovascular Health

• Activity Number: 315
• Type: Contributed
• Date/Time: Tuesday, August 11, 2015 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Tamika Royal-Thomas* and Daniel McGee
• Companies: University of the West Indies and Florida State University
• Keywords: in utero ; longitudinal ; birth weight ; Linear Mixed Model ; principal component ; systolic blood pressure

Abstract

The Barker Hypothesis states that maternal and in utero attributes during pregnancy affects a child's cardiovascular health throughout life. Some published research have utilized birth weight as a proxy for in utero measurements. We present analyses of a unique dataset which consists of the mother's pregnancy measurement, offspring's in utero measurements and measurements of the offsprings from birth to 14 years. Measurements at birth and in utero are highly correlated which introduces multicollinearity into the model. An analysis of the relationship of these measurements will be presented using principal component analyses. In utero and birth component scores were developed. The results indicated that the in utero score predicted the future cardiovascular health of the child and that birth weight may be an intermediate predictor.