Thomas Peyret (1), Samer Mouksassi (1) and Shasha Jumbe (2), representing the Healthy Birth, Growth and Development knowledge integration (HBGDki) community
(1) Certara Strategic Consulting, Montreal, QC, Canada, (2) Bill and Melinda Gates Foundation, Seattle, WA, USA
Background: Metabolic cost and metabolic consequences due to metabolic adjustments occurring during pregnancy and lactation to support fetal growth and milk synthesis are poorly understood.
Objectives: To develop a growth model of fetal weight based on fetus energy expenditure within a population approach modelling framework, in order to quantify required uniform flow of nutrients to the fetus and to the mammary gland, and the effect on maternal energy metabolism.
Methods: Daily energy deposition in a reference fetus was estimated by back calculating the necessary calories based on a published reference fetus mass growth equation[1]. Published energy densities of fat and fat free mass changes[2] and calculated fat and fat free masses based on literature data were used as energy sinks. Exponential, power, polynomial and Gompertz models were fitted to the cumulative energy deposition-gestational age (GA) curve and to the fraction of cumulative energy deposition in fat. Once the literature–based reference energy deposition curve was obtained, it was externally tested against ultrasound-based fetus and birth weight data from 1161 subjects [3] using a nonlinear mixed effects model estimating between-subject variability (BSV) on relevant parameters (FOCE-ELS engine in Phoenix® NLMETM v1.3,Certara, Princeton, NJ). Mother daily energy intake was estimated based on age, weight and height[2] and sex were also a priori included in the model as covariate.
Results: The piecewise equation for fat mass fraction consisted of three linear regressions for 0-25; 25-40 and > 40 weeks of GA. Gompertz equations obtained the best fitting performance for both cumulative energy deposition and its fat fraction. The energy-mass model predicted well the reference fetus weight-gestational age curve. The population model included BSV (<20%) on two parameters of the Gompertz model for cumulated energy deposition. Predicted individual fetal growth curve fitted well the trajectory of the observed fetus weight up to birth.
Conclusions: This study demonstrates the feasibility of using reverse engineering based on a closed model assumption of fetal caloric intake to predict fetus and birth weights.
Sponsored by the Bill & Melinda Gates Foundation, Healthy birth, growth and development initiative
References:
[1] Mikolajczyk et al. (2011) 377(9780):1855–1861.
[2] Hall et al. Lancet (2011) 378(9793):826-837.
[3] Soh et al. Int J Epidemiol (2014) 43(5):1401-1409.
Reference: PAGE 25 () Abstr 5811 [www.page-meeting.org/?abstract=5811]
Poster: Methodology - New Modelling Approaches