Seasonality Associated With Preterm Birth—Are We Borrowing From Our Children?

Preterm birth is a substantial global health issue with well-documented consequences for the infant, family, and society at large. Identified as the leading cause of death among children younger than 5 years, preterm delivery occurs in approximately 15 million births worldwide and remains one of the foremost issues in the field of obstetrics.¹ Preterm birth rates are increasing internationally.¹ In 2019, the most recent year with complete data available, the rate of preterm birth in the US was 10.23%, and the rate of births earlier than 28 weeks’ gestation was 0.66%.² Gestational age at delivery and the risk of neonatal morbidity and mortality are inversely associated. Neonates born earlier than 28 weeks’ gestation comprise the smallest proportion of births, but these infants experience disproportionately higher rates of prematurity-related complications, including death.³ For context, the infant mortality rate among those born earlier than 28 weeks’ gestation was 186 times higher than those born between 37 and 41 weeks’ gestation in the US in 2018.³ Infants who survive are at risk of developing a wide range of short- and long-term morbidities. Although the burden of preterm birth is clear, understanding the biological characteristics of human parturition remains elusive. To this end, modifiable risk factors are being explored to address this substantial public health issue.

Hviid and colleagues⁴ report data from a population-based retrospective cohort study in Denmark examining the association of birth season and pregnancy onset season with rates of preterm birth. Seasons were defined as winter (December, January, and February), spring (March, April, and May), summer (June, July, and August), and autumn (September, October, and November). Preterm births were categorized according to severity, with extremely preterm defined as live births occurring between 22 and 27 weeks’ gestation, very preterm as live births occurring between 28 and 31 weeks’ gestation, and moderately preterm as live births occurring between 32 and 36 weeks’ gestation. Because of the substantial burden of extreme prematurity, births between 22 and 27 weeks’ gestation were the primary study outcome. Sophisticated models for rates of preterm birth according to season were developed, with adjustment for socioeconomic and demographic factors, such as maternal health-related characteristics and smoking status.

There were 1 136 143 pregnancies between 1997 and 2016 that met the eligibility criteria for study inclusion. The overall rate of preterm birth was 3.4%, and the rate of extremely preterm birth for the cohort was 0.20%, with 2009 cases. The authors found that the cumulative incidence of extremely preterm birth was lowest in winter (0.16%; 95% CI, 0.14%-0.17%) and highest in autumn (0.20%, 95% CI, 0.18%-0.21%) followed by summer (0.18%; 95% CI, 0.17%-0.20%) and spring (0.17%; 95% CI, 0.16%-0.19%). Using winter as the referent season, adjusted hazard ratios (AHRs) were then calculated for season of birth and season of pregnancy onset using the strata of extremely preterm, very preterm, and moderately preterm birth. In the analysis of season of birth, after adjustment for maternal demographic and familial socioeconomic factors, AHRs for the risk of extremely preterm birth were 1.15 (95% CI, 1.02-1.31) for summer and 1.25 (95% CI, 1.10-1.42) for autumn. The number of extremely preterm births accounting for the increased risk in the spring, summer, and autumn was 56.1 (95% CI, 18.2-99.7), or 2.8% (95% CI, 0.9%-5.0%) of all extremely preterm births in the study. In the analysis of season of pregnancy onset, spring had the highest risk (AHR, 1.12; 95% CI, 0.99-1.27) of extremely preterm birth, and summer had the lowest risk (AHR, 0.87; 95% CI, 0.77-0.99) compared with winter. For the study cohort with information on maternal body mass index (calculated as weight in kilograms divided by height in meters squared) and smoking status during pregnancy (n = 689 680), similar HRs were found. In addition, analyses incorporating only extremely preterm births recorded as spontaneous (n = 811) or maternal preeclampsia (n = 32 543) did not substantively alter the findings. Statistical differences in seasonality and very preterm birth were similar. However, the analysis of moderately preterm birth did not identify an association with either season of birth or season of pregnancy onset.

Hviid and colleagues⁴ highlight the importance and challenges of examining factors associated with preterm birth. Their findings are particularly relevant given that the study used a fetuses-at-risk approach to assess nationwide data for more than 1 million births. That is, their analysis limited the confounding that can occur when a time-series approach is used, which does not account for seasonal variation in births. Furthermore, the authors emphasize that the analysis of live births accounted for the potential impact of stillbirths by using a competing-risk approach, and they performed adjusted analyses of various maternal and familial sociodemographic factors. Despite the lack of association with moderately preterm birth, the authors suggest that the findings for extremely preterm birth are important because of the substantial morbidity and mortality found in these vulnerable infants. In addition, the findings of reductions in extremely preterm births during winter may explain similar reductions observed during the COVID-19 lockdown period, which altered exposure to climate and increased hygiene practices.⁵

Before seasonality can be considered a definitive risk factor for preterm birth (extremely preterm or otherwise), several important caveats must be considered. First, there were substantial differences in the population studied compared with populations in other regions. The incidence of preterm birth in the Hviid et al⁴ study (3.4%) was significantly lower than the incidence reported in other countries.^1,2 For example, the rates of preterm birth in the US² are nearly 3-fold higher for both moderately preterm (<37 weeks’ gestation) and extremely preterm (<28 weeks’ gestation) birth compared with rates found in the study cohort. In addition, maternal factors in the study population, such as health-related characteristics and smoking status, were markedly different from populations in other regions. For example, the proportion of pregnant individuals with body mass index greater than 25 ranged from 21.3% to 24.5% in the study cohort compared with 55.9% in the US in 2019.² Race and ethnicity of participants present other obvious differences in Denmark compared with other regions.^1,2 Second, and importantly, other risk factors for preterm birth were not examined. The authors did not assess the association between previous preterm birth and the risk of recurrence, especially when considering the frequency, order, and severity of previous preterm births and the subsequent risk of preterm birth in a current pregnancy.⁶ Third, obstetric management strategies, such as cervical cerclage, were not reported, which could have had implications for rates of preterm birth, especially those with deliveries occurring in the second trimester between 22 weeks, 0 to 7 days’ gestation and 27 weeks, 6 to 7 days’ gestation. Fourth, the suggestion that reduced physical activity was associated with improvement in extremely preterm birth rates was not clearly supported by current evidence. Indeed, leisure activity may be a beneficial factor associated with reductions rather than increases in rates of preterm birth.^7,8

Despite these caveats, a provocative corollary of the findings from the Hviid et al⁴ study is the potential association of heat and climate exposure with preterm birth. Bekkar and colleagues⁹ recently conducted a systematic review of 68 studies totaling more than 32 million births and found that increases in environmental exposures, such as heat, which are exacerbated by climate change, were significantly associated with adverse pregnancy outcomes, including preterm birth. Although the mechanisms of preterm birth are complex, heterogenous, and multifactorial in origin, the current data paired with recent literature hints at the potential consequences of climate change for future health. It is often said that we do not inherit the earth from our ancestors. We borrow it from our children. Indeed, if these suggested findings are confirmed, the borrowed debt is being paid today. Health care expenditures are a substantial consequence of preterm birth when considering both direct and indirect costs. This issue was recently examined using data from the 2016 US birth cohort, with the total societal burden of preterm birth estimated at $25.2 billion, or $64 815 per preterm birth.¹⁰ Analyzed using another method to adjust for price increases, the total cost in 2016 would have been $32 billion.¹⁰ Medical care services for children born preterm comprised the largest category of societal costs by far, representing two-thirds of total costs, whereas indirect costs were $4.8 billion.¹⁰ Beyond financial costs, the impact of preterm birth cannot be overstated to the patients and families we serve. We look forward to future research on the consequences of season, climate, and modifiable factors to understand and prevent this global health crisis.

Published: February 3, 2022. doi:10.1001/jamanetworkopen.2021.45808

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2022 Nelson DB et al. JAMA Network Open.

Corresponding Author: David B. Nelson, MD, Department of Obstetrics and Gynecology, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390-9032 (davidb.nelson@utsouthwestern.edu).

Conflict of Interest Disclosures: None reported.