Winter, spring, summer or fall: temporal patterns in placenta-mediated pregnancy complications—an exploratory analysis

Placenta-mediated pregnancy complications include birth of small-for-gestational-age (SGA), intrauterine growth restriction (IUGR), gestational hypertension, pre-eclampsia, placental abruption and in some cases of preterm birth and intrauterine fetal death. Some of these complications are leading causes of maternal, fetal and neonatal morbidity as well as mortality in high-income countries and affected woman are at high risk of recurrence in their subsequent pregnancies [1].

The health and growth of the fetus is dependent on normal development of the placenta. This development begins early in pregnancy and undergoes adoption during pregnancy [2]. The vascularisation of placental villi involves the processes of vasculogenesis and angiogenesis, which form the maternal-fetal interface [2]. These processes depend on a tightly controlled balance between pro-angiogenic and anti-angiogenic pathways [1]. Furthermore, in a healthy pregnancy the maternal immune system adapts to protect the fetus and placenta. This means, the inflammatory and angiogenic systems are interdependent and tightly regulated across pregnancy. Therefore, disruption of either system could lead to a cascade of downstream events with negative impact on placental development [1].

The ethicology of placenta-mediated pregnancy complications is likely multifactorial but some of the known risk factors are maternal age > 35 years, smoking and maternal obesity [3‐9].

During a normal pregnancy, a systemic inflammatory response happens and the immune system adapts to this [10]. It has been suggested, that in preeclampsia this normal inflammatory response becomes excessive, causing a degree of decompensation in maternal system [10].

Furthermore, it is known that maternal infection (e.g. malaria, HIV) can result in immune activation and inflammation which dysregulates the tightly regulated processes of vasculogenesis and angiogenesis, contributing to poor function of placenta and thereby placenta-mediated complications [1]. Some studies even found an association between maternal influenza and preterm birth and foetal death [11]. This indicates that pre-existing/co-existing infections amplify the inflammatory response and thereby dispose to placenta-mediated complications during pregnancy.

Some studies have investigated seasonality in hypertension/preeclampsia during pregnancy. They found seasonality in hypertensive disorders during pregnancy with the highest incidence with estimated date of conception during autumn and summer months, but have not investigated other placenta-mediated complications [12‐14].

The aim of this study was to investigate if there is a seasonal variation in placenta-mediated pregnancy complications (small for gestational age, intrauterine growth restriction, preeclampsia, preterm birth and intrauterine death) in a large Danish cohort of singleton pregnancies.

Each pregnancy was assigned to the estimated date of conception by subtracting gestational age in days from the date of birth.

The risk factors and outcomes of interest were coded as binary (true/false) variables according to the following definitions (variable names used in figures and table in parentheses):

The number of pregnancies with each risk factor and outcome were aggregated by year and season: winter = Dec–Feb, spring = Mar–May, summer = Jun–Aug, and fall = Sep–Nov. Seasons were coded as dates referring to the first date of the season, e.g. ‘2007-12-01’ refers to the winter season starting 1 December 2007 and ending 29 February 2008.

We used SAS Software version 9.4 of the SAS System for Windows for data extraction and initial aggregation by month and year and R Statistical Software version 4.0.4 with the add-on packages dplyr version 1.0.4 and qicharts2 v. 0.7.1 for further data manipulation and visualization [16].

Table 1 describes the demographic of the included 550,897deliveries.

The number of deliveries showed a marked seasonal pattern with repeated peaks, i.e. more pregnancies were conceived during the fall season. There were no other certain signs of non-random variation over time (Fig. 1).

Figure 2 illustrate the proportion of mothers with risk factors over time.

Women with diabetes showed an increasing trend over time but there was no apparent seasonality. The proportion of pregnancies conceived by in-vitro fertilization showed an increasing trend during the years 2006–2011. From 2008 there was a marked sawtooth pattern with more pregnancies conceived in the spring and fall seasons.

The proportion of mother of age 40 or more shifted upwards around 2010. There was no apparent seasonality.

The proportion of obese mothers showed an increasing trend during the years 2008 to 2011. There was no apparent seasonality.

The proportion of mothers with previous cesarean delivery showed and increasing trend during the years 2006 to 2010. After 2012 there seemed to be a slightly decreasing trend. There was no apparent seasonality.

The proportion of primipara showed an increasing trend over the full time period. Also, there was a consistent seasonal pattern of fewer primipara pregnancies conceived in the summer seasons.

The proportion of smoking mothers showed a decreasing trend over the years with no apparent seasonality.

Outcomes are displayed in Figs. 3 and 4.

The proportion of pregnancies with hypertensive disorders trended upwards from 2006 to 2009. There was a consistent seasonal pattern with dips in the fall seasons. Preterm deliveries trended downwards until 2012. There was no apparent seasonality. Overall, the proportion of babies with SGA showed only random variation. However, moderate SGA showed a slight increasing trend while severe SGA trended downwards from 2010. There was no apparent seasonality. The proportion of stillbirths varied randomly over time and showed no apparent seasonality.

In this study, based on a large Danish cohort of singleton pregnancies, we found a seasonal pattern in hypertensive disorders during pregnancy with dips in pregnancies conceived in the fall season and highest risk by conception in the spring and summer season. We found no apparent seasonality in cases of preterm delivery, SGA and stillbirths.

Looking at the individual risk factors for placenta-mediated complicated over time, we found that the trends for pregnant women were in consistency with the general trends in Denmark: smoking showed a decreasing trend, while obesity and diabetes had an increasing trend [3, 21, 22].

The proportion of pregnancies conceived by in-vitro fertilization showed an increasing trend during the years 2007 to 2011. In 2007, the Danish legalisation for fertility treatment changed which, among others, made it possible for singles and homosexual women to receive treatment including donor eggs or sperm. From 2008 there was a marked sawtooth pattern in pregnancies conceived by in-vitro fertilization with more pregnancies conceived in the spring and fall seasons, which properly are explained by fertility clinics have summer- and Christmas vacation.

We found no seasonality of preterm birth. In the existing literature, different results are reported within this topic. A systemic review by Lee et al. describes a seasonality of preterm birth with highest risk by delivery in winter- and summer months [23]. The review included three studies from developing countries and three from developed countries. One of the studies from the developed countries did not find any seasonality. The two other studies from developed countries showed the rate of preterm birth was highest twice a year (summer and winter). Furthermore, Lee et al. made a London-based cohort study that found an increased risk of preterm birth during deliveries in the winter months [23].

A large Danish study concluded that the seasonality for extremely preterm birth before 28 completed weeks of gestation had highest risk at autumn and summer and lowest in winter. However, when all preterm births were investigated, the rate of preterm birth overall was not associated with season [14]. Our study differs from some of the international results. This can partly be explained by geographical, cultural and socio-economic differences between the populations studied in the review and our study. Consistent with the other Danish study, we also found no association with season when we investigated overall preterm births.

Other studies support our results of seasonality in hypertensive disorders during pregnancies. A Danish study by Thomsen et al. investigated seasonal variation in the risk of hypertensive disorders during pregnancy in a large cohort of nulliparous women. They found a higher risk for hypertensive disorders during pregnancy in women with the estimated date of conception in the summer months [24]. An Australian study by Verburg et al. investigated the seasonal variation of hypertensive disorders during pregnancy and found also the highest incidence of hypertensive disorders during pregnancy in pregnancies with conception in late spring and summer [25].

To make sense of seasonality in placenta-mediated pregnancy complications, it needs to be related to a plausible exposure at some gestational period or periods during pregnancy. This could be sunlight intensity, ambient temperature, infections, allergies or seasonality in exercise.

D-vitamin has been suggested as a potential contributor to the seasonal variation of hypertensive pregnancies during pregnancy [26]. The major natural source of the vitamin D is synthesis of cholecalciferol in presence of sunlight. Thereby the sunlight intensity (and thereby the ambient temperature) affects the status of d-vitamin.

In the Danish population, the level of D-vitamin is often low during the winter months and higher during summertime. This is supported by, among others, a study by Forman et al which found that low levels of vitamin D are associated with an increased risk of hypertension [27, 28]. On the other hand, a systemic review by O’Callaghan et al. from 2018 critically evaluated the current evidence for an association between vitamin D status and/or intake and risk of hypertensive disorders in pregnancy [29]. The study found that the current evidence base is weak. Furthermore, analysed observational studies showed a positive association between vitamin D deficiency and increased risk of preeclampsia but the results are hampered by suboptimal clinical phenotyping and large heterogenicity between the studies. Based on this, there are studies that indicate an association between vitamin D and hypertensive disorders during pregnancy but the association is not definitively described yet [29].

This study has both strengths and limitations. One of the strengths of this study is that we are able to investigate variations in trends in a large and diverse cohort including all singleton births in Denmark from 2007 to 2017. This is the largest cohort to date, by our knowledge, evaluating season variation in low birthweight, intrauterine growth restriction, preeclampsia, preterm birth and intrauterine death. This makes the study very adequate in the overall investigation of seasonal variation of placenta-mediated complications.

Another strength of the study is that it is based on data from registers with prospectively collected data where risks of recall bias are minimal. The registers have an extensive collection of variables and contains maternal, delivery and fetal characteristics.

A limitation of this study is that we did not have access to data about all potential risk factors for pregnancy complications and thus we were not able to adjust our results for, among others, ethnicity and socioeconomic factors. Furthermore, register-based studies are sensitive for changes in clinical practice, changes in definition as well as changes in registration.

Results from other studies have indicated that infection during pregnancy effects the placenta. During the Covid-19 pandemic several studies have examined placentas from woman with Covid-19 during pregnancy. A small study by Shanes et al. with 16 women included that placentas from pregnant women with covid-19 had increased prevalence of maternal vascular malperfusion [30]. This reflects abnormalities in oxygenation within the intervillous space associated with adverse perinatal outcome. Despite this, only one included pregnant woman with Covid-19 was hypertensive. This is supported by a study by Jaiswal et al. where 27 placentas from asymptomatic or mildly symptomatic SARS-CoV-2 positive pregnant women, with otherwise uncomplicated pregnancies, showed evidence of placental injury at a microscopic level [31]. This placental injury apparently does not lead to adverse pregnancy outcomes. Even though both studies found an increased maternal vascular malperfusion, only one woman included experienced a placenta-mediated complication. This shows that the pathogenesis between vascular malperfusion and hypertensive disorders during pregnancy has not been fully investigated. It could be considered to perform studies using artificial intelligence to investigate the individual risk profile of each pregnant woman to predict placenta-mediated complications.

In conclusion, we found a significant seasonal variation in the risk of the hypertensive disorders of pregnancy, including preeclampsia, with highest risk by conception in the spring and summer season. The risk was lowest by conception in the fall. This study found no seasonal variation in other placenta-mediated complications, including preterm delivery, small for gestational age, intrauterine growth restriction and intrauterine fetal death.