Trimester of maternal SARS-CoV-2 infection differentially predicts newborn size in the COVID-19 Mother Baby Outcomes (COMBO) cohort

Abstract

Background: In-utero exposures affect the developing fetus differentially and may manifest as growth and body size differences. We examined associations between maternal SARS-CoV-2 infection and newborn body size at a large medical center in New York City.

Methods: Between 3/22/2020 and 10/31/2020, birth outcomes for 544 newborns exposed to SARS-CoV-2 in utero and 3325 unexposed newborns were collected through retrospective chart review as part of the COVID-19 Mother Baby Outcomes (COMBO) Initiative. Data were also abstracted for 560 newborns born in February 2020 as a pre-pandemic comparison group. Growth outcomes at 6-10 weeks of life were available for a subset of 865 newborns followed in our hospital system.

Results: Newborns exposed to SARS-CoV-2 during the first trimester had significantly smaller birth weight, length, head circumference (HC) and earlier gestational age (GA) at birth than unexposed or second-to-third trimester exposed newborns. Mediation analysis found that GA mediated the relationship between first trimester SARS-CoV-2 exposure and body size at birth.

Conclusion: Maternal SARS-CoV-2 infection during the first trimester is associated with smaller body size at birth and this association is largely mediated by earlier GA. Size differences at birth can predict later adverse cardiometabolic and neurodevelopmental outcomes, warranting prospective analysis in this cohort.

Impact: Using data obtained through universal SARS-CoV-2 testing at a large, urban medical center, maternal SARS-CoV-2 infection during pregnancy was associated with newborn size. Previous literature shows that SARS-CoV-2 does not tend to cause severe respiratory disease in infants and that vertical transmission to the fetus is unlikely. However, maternal SARS-CoV-2 infection during pregnancy is associated with increased occurrence of pre-term delivery. This study contributes to the literature by showing that timing of maternal SARS-CoV-2 infection, specifically during the first trimester of pregnancy, predicts smaller newborn body size and that this relationship is mediated by earlier gestational age (GA) at birth.

Figure 1.. Relationship between trimester of pregnancy at maternal SARS-CoV-2 infection and infant anthropometric outcomes at birth.

(A) Graphical representation of the spread of the pandemic in New York City (NYC) during our study period and gestational timing of SARS-CoV-2 exposure in our cohort. New York City (NYC) pandemic case counts between 3/1/20 and 11/1/20 are shown (light grey), with maternal COVID-19 onset dates in the COMBO cohort overlaid (dark grey) as well as infant dates of birth in the COMBO cohort overlaid (purple). (B) GA at delivery in weeks in infants unexposed to SARS-CoV-2 in utero (blue) versus those exposed during third, second, and third trimester (red). (C) Birth weight in grams in infants unexposed to SARS-CoV-2 in utero (blue) versus those exposed during third, second, and third trimester (red). (D) Birth length in cm in infants unexposed to SARS-CoV-2 in utero (blue) versus those exposed during third, second, and third trimester (red). (E) Birth head circumference in cm in infants unexposed to SARS-CoV-2 in utero (blue) versus those exposed during third, second, and third trimester (red). *p<0.05, **p<0.01, ***p<0.001, on fully adjusted models.

Figure 2.. Mediation models examining gestational age as a mediator for effects of trimester of SARS-CoV-2 exposure on anthropometric outcomes at birth.

(A) A representation of the mediation model for birth weight is shown, with birth weight as the dependent variable, trimester of SARS-CoV-2 exposure as the independent variable, and GA at delivery as the mediating variable. (B) A representation of the mediation model for birth length is shown, with birth length as the dependent variable, trimester of SARS-CoV-2 exposure as the independent variable, and GA at delivery as the mediating variable. (C) A representation of the mediation model for birth head circumference is shown, with birth head circumference as the dependent variable, trimester of SARS-CoV-2 exposure as the independent variable, and GA at delivery as the mediating variable. For all mediation models, effect sizes are shown for the total effect of trimester of exposure on anthropometric outcomes without removing GA at delivery, the direct effect of trimester of exposure on anthropometric outcomes after removing GA at delivery, and the bootstrapped indirect effect of trimester of exposure on anthropometric outcomes through GA at delivery.