Premium
Proposing a new approach to measuring birth size asymmetry
Author(s) -
Betts Kim S.,
Kisely Steve,
Alati Rosa
Publication year - 2021
Publication title -
paediatric and perinatal epidemiology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.667
H-Index - 88
eISSN - 1365-3016
pISSN - 0269-5022
DOI - 10.1111/ppe.12684
Subject(s) - medicine , birth weight , asymmetry , etiology , pediatrics , multinomial distribution , sample size determination , obstetrics , pregnancy , demography , statistics , mathematics , genetics , physics , quantum mechanics , sociology , biology
Abstract Background Existing methods of measuring birth size asymmetry based on ratios of growth parameters are clinically useful but simplistic, and as such may have limited usefulness in studies of aetiology. Objectives We aimed to develop a novel method of measuring asymmetric fetal growth at birth and demonstrate its utility in characterising the perturbed growth associated with a number of prenatal exposures and neonatal outcomes. Methods Data were drawn from the Queensland (QLD) Perinatal Data Collection, which included all livebirths in the Australian state of QLD between July 2010 and December 2015, with analyses restricted to babies born between 32 and 42 weeks of gestation (n = 280 084). Novel measures of asymmetric birthweight, length, and head circumference were developed using a weighted average, representing “how far” an individual's given birth size measure deviated from the sample average and their other birth size measures. Associations among prenatal exposures and neonatal outcomes with the new asymmetry measures and traditional ratio measures (ie ponderal index, brain‐to‐body weight ratio, and birth length divided by head circumference) were then compared using log‐binomial and multinomial regressions. Results The new asymmetry measures clearly indicated that prenatal smoking was linked to a disproportionate decrease in all birth size measures and that low birthweight asymmetry and low birth head circumference asymmetry were specifically associated with neonatal respiratory distress and chromosomal abnormalities, respectively. When these same associations were tested using the traditional ratios, the estimates were weak, imprecise, and non‐specific. Conclusions We developed a new approach to measuring fetal growth asymmetry which provides complimentary insights against the existing ratios approach. Associations with the new asymmetry measures were more precise and easier to interpret than the associations obtained using the ratios, and may better reflect the underlying pathological processes, providing an advantage when investigating the aetiologies of perturbed fetal growth.