Non-linear associations of maternal pre-pregnancy body mass index with risk of stillbirth, infant, and neonatal mortality in over 28 million births in the USA: a retrospective cohort study Authors: Hannah V Thornton, Rosie P Cornish, Deborah A Lawlor Published: 2023 ABSTRACT SummaryBackgroundHigher maternal pre-pregnancy body mass index (BMI) has been associated with higher risk of stillbirth, infant and neonatal mortality. Studies exploring underweight have varied in their conclusions. Our aim was to examine the risk of stillbirth, infant and neonatal mortality across the BMI distribution and establish a likely healthy BMI range.MethodsIn this retrospective cohort study, we used publicly available datasets (covering 1st January 2014 to 31st December 2021) from the US National Center for Health Statistics National Vital Statistics System. All births were eligible; analyses included those with non-missing data. Fractional polynomial multivariable logistic regression was used to examine the associations of maternal pre-pregnant BMI with stillbirth (birth with no signs of life at ≥24 weeks), infant mortality (death of a live born baby aged <365 days) and neonatal mortality (death of a live born baby aged <28 days).FindingsThere were 77,896/28,310,154 (2.8 per 1000 births) stillbirths, 143,620/28,231,807 (5.1 per 1000 live births) infant deaths and 94,246/28,231,807 (3.3 per 1000 live births) neonatal deaths among complete cases. Mean (SD) BMI was 27.1 kg/m2 (6.7 kg/m2). We found non-linear associations between BMI and all three outcomes; risk was elevated at both low and high BMIs although, for stillbirth, the increased risk at low BMI was less marked than for infant/neonatal mortality. The lowest risk was at a BMI of 21 kg/m2 for infant and neonatal mortality and 19 kg/m2 for stillbirth.InterpretationPublic health messaging for preconception and postnatal care should focus on healthy weight to maximise maternal and child health, and not focus solely on maternal overweight or obesity.FundingEuropean Research Council, US National Institute of Health, UK Medical Research Council and National Institute for Health and Care Research.Keywords: Body mass index, Stillbirth, Infant mortality, Neonatal mortality Research in context Evidence before this study We carried out a Medline search (1st January 1946 to 14th December 2022) with no language restrictions, using used a combination of the following: body mass index; BMI; obesity; overweight; underweight; weight; low weight; low BMI; maternal; stillbirth; infant death; infant mortality; neonatal death; neonatal mortality; birth outcome. We identified five systematic reviews and four additional primary studies published after the most recent review; these consistently showed an association of increased risk of stillbirth, neonatal mortality and infant mortality in women with overweight or obesity whereas the results for underweight were inconsistent. Two of the reviews found evidence for non-linear associations, but again the results were inconsistent for low BMI. On the whole, the quality of studies included in the reviews was assessed as high. Added value of this study In this study of over 28 million births, we examine non-linear associations of maternal pre-pregnant BMI with stillbirth, infant and neonatal mortality using BMI as a continuous variable across the whole BMI distribution. This is the largest study to date (the largest systematic reviews pooled data on 10.8 million births for infant mortality, 3.8 million for neonatal mortality and 3.3 million for stillbirth), providing sufficient sample size to investigate the risk of these relatively rare outcomes across the whole maternal BMI range, in which we adjust for pre-specified confounders and include all women from a whole (US) population. We also compare results across three different definitions of stillbirth currently used in different regions of the world. Implications of all the available evidence The relationship between maternal pre-pregnancy BMI and risk of stillbirth infant and neonatal mortality is non-linear, with increased risk at both low and high BMIs. Public health messaging for preconception and postnatal care should focus on both maternal over and underweight. Introduction Stillbirth (birth with no signs of life from between 20 and 28 weeks of gestation, with definitions varying between populations and institutions and hence studies), infant mortality (death of a live born baby aged <365 days) and neonatal mortality (death of a live born baby aged <28 days, a subgroup of infant death) are devastating events with severe repercussions for families and healthcare services. Globally, an estimated 2.6 million third trimester (i.e., ≥28 weeks) stillbirths occurred in 2015 and, in 2021 over 3.7 million children died before the age of 1.1,2 Internationally, obesity has almost tripled since 1975 and the most recent WHO data shows that an estimated 25% of adults had obesity in 2020.3 The prevalence of underweight is slowly falling, yet an estimated 9.4% of all females had underweight in 2016.4 Whilst higher maternal body mass index (BMI) is consistently associated with a higher risk of stillbirth, neonatal mortality and infant mortality, there is less certainty over the associations of underweight with these outcomes1,5 There is an established higher risk of foetal growth restriction and low birth weight in underweight women, and a strong relationship between foetal growth restriction and stillbirth and infant mortality.6 One might therefore expect an increased risk of stillbirth, neonatal and infant mortality in women with underweight. Of the five previous systematic reviews5,7, 8, 9, 10 and four subsequent primary studies11, 12, 13, 14 on the relationship of pre-pregnancy maternal BMI on stillbirth, infant or neonatal mortality, all reported results using BMI categories, with three (including two of the reviews) not examining the association of underweight with these outcomes. These nine studies all reported analyses supporting an increased risk of all three outcomes among women with overweight or obesity, with less consistency regarding underweight. Two of the reviews used information from the BMI categories to estimate a relative risk for a unit increase in BMI and to examine non-linear associations, one with infant and neonatal mortality,10 one with all three outcomes.5 Both found evidence for non-linearities but, again, these were not consistent, with one indicating a decreasing risk of infant mortality with lower BMIs10 but the other finding a very slight increase below a BMI of around 20 kg/m2 but a decreasing risk of stillbirth with lower BMIs. The largest meta-analyses of stillbirth, infant mortality and neonatal mortality included 3.3 million, 10.8 million and 3.8 million births, respectively. Understanding the relationships between maternal underweight, as well as overweight and obesity, and stillbirth, neonatal and infant mortality might enable us to identify a “safe” target weight for women considering pregnancy; this has not yet been identified.5 This could help parents and healthcare providers make informed decisions about antenatal monitoring, delivery and infant care for mothers whose weight (whether under- or overweight) might confer increased risk of stillbirth and infant mortality. The aim of this study was to explore whether there were non-linear associations of maternal pre-pregnant BMI with stillbirth, neonatal and infant mortality in order to identify a “healthy” BMI for these outcomes. We add to existing literature by analysing the largest population to date, over 28 million recorded births in the United States (US) between 2014 and 2021, providing sufficient sample size to investigate the risk of these relatively rare outcomes across the whole maternal BMI range. Methods Dataset This study used publicly available US National Center for Health Statistics (NHCS) National Vital Statistics System datasets (covering the period 1st January 2014 to 31st December 2021, inclusive), which contain details of all births registered in the USA.15 For stillbirths we used the Birth and Fetal Death Data Files; for infant, including neonatal mortality, we used the Period/Cohort Linked Birth-Infant Death Datafiles. The Fetal Death files exclude all terminations.15 This timeframe was selected because maternal pre-pregnancy BMI was included in the foetal deaths data from 2014 onwards and, at the time of analysis, foetal and infant death data were available up to and including 2021. Each record in these datasets relates to a live birth or foetal death, rather than a pregnancy and there are no pregnancy or person-level identifiers in the dataset. To identify pregnancies, we matched multiple births occurring close in time in which birth and maternal characteristics were the same. Ethics These analyses used publicly available data containing no identifying information; no ethical approval was required for this study. We have adhered to the Vital Statistics Data User Agreement in the use of these data. The NHCS are bound by federal regulations to ensure data confidentiality. NCHS provides respondents with a description of the intended purpose(s) of the data they provide and informed consent is assumed if the respondent gives the requested information. Exposure Maternal pre-pregnancy height and weight were self-reported by the women at the time of birth via the following questions: What is your height? What was your pre-pregnancy weight? That is, your weight immediately before you became pregnant with this child? BMI was calculated from these. BMI was used in its continuous form in the primary analysis. For the descriptive and secondary analysis, we used eight World Health Organisation (WHO) BMI categories, including categories of underweight (severe underweight: <16 kg/m2, moderate underweight: 16–16.9 kg/m2, mild underweight: 17–18.49 kg/m2); normal: 18.5–24.9 kg/m2; overweight 25–29.9 kg/m2; obesity (class-I: 35–34.9 kg/m2, class-II: 35–39.9 kg/m2 and class-III: ≥40 kg/m2). Outcomes Gestational age at which a foetal death is classified as a stillbirth varies internationally. The WHO define stillbirth as a baby born with no signs of life at ≥28 weeks’ gestation. In many European countries a threshold of gestational age of ≥22 weeks is used; 24 weeks is used in the United Kingdom and 20 weeks in the US.16 To enable comparison, we report results using three different definitions. For the main results we used ≥24 weeks’ gestation, with results using ≥20 and ≥28 weeks also reported. WHO definitions were used for infant mortality (death of a live born baby aged <365 days) and neonatal mortality (death of a live born baby aged <28 days). Potential confounders We defined cofounders as any factors that are known or plausible causes of both the exposure (BMI) and outcomes (stillbirth and infant, including neonatal, mortality), as per best practice.17 The following variables were considered to be confounders, based on this definition: maternal age at birth (categorised as: <20, 20–24, 25–29, 30–34, 35–39, 40+ years); maternal education (