Estimating Onset Time from Longitudinal Data with Application to Estimating Gestational Age from Maternal Anthropometry During Pregnancy

• Activity Number: 430
• Type: Contributed
• Date/Time: Tuesday, August 2, 2016 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Ana Maria Ortega-Villa* and Paul Albert
• Companies: Eunice Kennedy Shriver National Institute of Child Health and Human Development and Eunice Kennedy Shriver National Institute of Child Health and Human Development
• Keywords: Time-zero ; Measurement Error ; Latent Variables ; Random Effects ; Linear Mixed Models

Abstract

Determining the date of conception is important for estimating gestational age for determining whether a fetus/mother is on track in their development/pregnancy. Various methods based on ultrasound have been proposed for dating a pregnancy in high resource countries. However, such techniques may not be available in under-resourced countries. We develop statistical methodology for estimating the date of conception using longitudinal assessment of multiple maternal anthropometry. We develop a multivariate linear mixed modeling framework where time-zero (conception time) is available on a training dataset and use the model for predict gestational age for women where time-zero is not available. We also consider the case where time-zero is unavailable on all subjects including those in the training set, and the case where time-zero is available, but with sizable measurement error. We illustrate this new methodology with data from the NICHD Fetal Growth Singleton Study.