This matched nested case-control study included pregnant women and their newborns from the CHALLENGE study conducted at Shunyi Maternal and Children's Hospital of Beijing Children's Hospital between March 2018 and April 2021. Criteria for inclusion were pregnant women with (1) age ≥ 18 years, (2) gestational weeks ≤ 13 weeks, (3) having lived in the Shunyi area for at least 1 year without moving plan in recent 5 years, and (4) no diagnosed mental illness or intellectual disability. In the CHALLENGE study, all women were recruited at their first trimester of pregnancy and conducted 5 follow-up visits during their pregnancy and after their delivery. Eligible pregnant women were asked to sign the informed consent forms and completed baseline information. Five follow-up visits included three routine prenatal tests at 12–13 weeks, 24–28 weeks and 34–35 weeks, one postpartum checkup, and one postnatal checkup at 42 days after parturition. During these visits, routine tests including blood pressure, blood tests, and urine tests were conducted. Additional tests were performed at specific time points, such as group B streptococcus infection screening, ultrasound exam, nuchal translucency (NT) scan, and electrocardiogram (Figure 1). Along with each routine prenatal testing, online questionnaires were applied to collect baseline information, environmental exposures, sleep quality, depression scale, and dietary information. We constructed a database to integrate electronic medical records from hospital information systems and routine laboratory tests. The cohort had recruited 9,635 pregnant women and 8,183 babies as of April 2021. All data had been safely stored and deidentified. The chronological order of questionnaires, routine tests, and biological samples of CHALLENGE study are shown in Figure 1.

The pregnant women completed a baseline questionnaire at the time of recruitment, including demographic characteristics (age, ethnicity, educational level, address, institution of employment) of pregnant women and their spouses, health survey (including pre-pregnancy body weight and height, date of last menstrual period, basic information, development history, and physical examination, birth history, medication history for diseases) of woman. At 12 to 13 weeks of gestation, the questionnaire focused on life behaviors (such as smoking, alcohol drinking, tea drinking, nutrient supplements, physical activities, and use of electronic products), environmental exposures (work and residential environment), medication history during pregnancy and spouse's life behaviors. The above mentioned two questionnaires are shown in the Supplementary Material.

To explore the impact of COVID-19 pandemic on uninfected pregnant women, we conducted a nested case-control study building upon the CHANLLEGE study. Based on the estimated incidence rate of adverse pregnancy outcomes (0.4–0.9%) in northern China, the sample size was calculated as at least 562 cases and 562 controls to reach a power of 80% and a type I error of 5%. We divided the study objects into pandemic (case) and pre-pandemic (control) group from the CHALLENGE study, and none of them were infected with COVID-19. For the pandemic group, we chose pregnant women whose last menstrual period were from August 1, 2019 to November 30, 2019. 802 out of 863 participants of CHANLLEGE cohort were recruited in the pandemic group. These pregnant women delivered from April to September, 2020 during COVID-19 pandemic when strict measures of social isolation was executed in Beijing. As for the pre-pandemic group in matched period, we selected pregnant women who had their last menstruation between 1 August, 2018 and 30 November, 2018. They all delivered their babies before the COVID-19 pandemic. After matching the maternal age and delivery time with the pandemic group, 802 pregnant women out of 1,477 participants of CHANLLEGE study were included. The detailed selection criteria were shown in Figure 2.

All relevant diagnoses were obtained from the medical records, including pregnancy-related maternal diagnoses (anemia, vaginitis, infection), pregnancy-specific disorder (gestational diabetes mellitus, gestational hypertension, preeclampsia), pregnancy complications (placental and fetal membrane abnormalities, umbilical cord abnormalities), delivery complications (postpartum hemorrhage, laceration of cervix during labor, perineal laceration during delivery unspecified, shoulder dystocia), and adverse pregnancy outcomes (preterm birth, macrosomia, large for gestational age, small for gestational age, fetal death, low weight infants, congenital abnormalities). Based on corresponding diagnostic guidelines, the adverse pregnancy outcomes were defined as following. Preterm birth was defined as a birth before 37 weeks of gestation. Macrosomia was defined as a newborn whose birth weight was more than 4,000 g and low birth weight was defined as <2,500 g. Large for gestational age (LGA) referred to an infant whose birth weight was above the 90th percentile of the average weight for the same gestational age. Small for gestational age (SGA) referred to an infant whose birth weight fell below the 10th percentile of the average weight for the same gestational age. Stillbirth was defined as a baby born with no signs of life at a gestational age of 20 weeks or more. Congenital abnormalities were defined as structural or functional anomalies that present at birth.

Related environmental factors such as alcohol drinking, tea drinking, smoking, nutrient supplement, and frequency of using computers and watching TV were collected from the standardized questionnaires at early pregnancy (12–13 weeks).

All statistical analyses were conducted using IBM SPSS Statistics for Windows, Version 25.0 (IBM Corp., Armonk, NY, USA). Qualitative variables were expressed by frequency, and quantitative variables were expressed by mean (x) ± standard deviation (SD). We applied the X² test and Mann-Whitney U test to compare the basic characteristics and the pregnancy complications between two groups. The binary logistic regression model was used to investigate the relationship between the COVID-19 pandemic and pregnancy complications after controlling for pregnant women's and their spouse's demographic characteristics (age, ethnicity, education level, work), pre-pregnancy body mass index (BMI), number of pregnancies, bad obstetric history, and history of cesarean section. We performed X² test to analyze the factors associated with the adverse pregnancy outcomes that related to the COVID-19 pandemic, and logistic regression model was used for prediction analysis. The statistical significance level has been set to P = 0.05.

All study procedures were approved by the Institutional Review Board (IRB) at the Shunyi Maternal and Children's Hospital of Beijing Children's Hospital with the approval IRB number of 2021-01. Informed written consents were obtained from all participants at the initial hospital visits.

A total of 1,604 pregnant women (mean age, 32.48 ± 3.85 years) were enrolled in this study including age-matched 802 participants (mean age, 32.40 ± 3.82 years) in the pandemic group and 802 participants (mean age, 32.56 ± 3.88 years) in the pre-pandemic group, respectively (Figure 2). Among 1,604 women, 93.52% were Han ethnicity, 47.01% had a bachelor's degree or above, 66.96% of pregnant women received pregnancy education at their initial visit to the hospital, and their pre-pregnancy BMI was 22.61 ± 3.79 kg/m². The average age of the spouse was 33.65 ± 4.41 years old, in whom 93.81% were Han ethnicity, and 43.62% had a bachelor's degree or above. We found that the rates of stay-at-home were higher in the pandemic group (pregnant women: 16.3 vs. 13%; spouses: 17.5 vs. 11.5%), especially in the spouse group (P = 0.01), while the number of primiparas was significantly higher in the pre-pandemic group (Table 1).

In the primary analysis, we compared the diagnosed complications between two groups during pregnancy and at labor. During pregnancy, the proportion of anemia and vaginitis in the pandemic group was higher than the pre-pandemic group (46.63 vs. 37.16%, P < 0.001; 7.11 vs. 2.37%, P < 0.001) (Table 2). At labor, the rate of shoulder dystocia was higher in the pandemic group (3.87% vs. 0.75%, P < 0.001), but postpartum hemorrhage decreased (21.1 vs. 26.6%, P < 0.010) (Table 2). We further assessed the adverse pregnancy outcomes including preterm birth, macrosomia, large for gestational age, small for gestational age, fetal death, low weight infants, and congenital anomalies. We selected 1,604 pregnant women in total. Of them, we found 311 pregnant women who were diagnosed with at least one type of adverse pregnancy outcomes. The overall incidence of adverse pregnancy outcomes was higher in the pandemic group than the pre-pandemic group (21.45 vs. 17.33%, P = 0.037), such as preterm birth (5.86 vs. 3.24%, P = 0.012) (Table 2).

We applied several models to control for the confounding factors (Table 3). In the univariate model I, we observed anemia, vaginitis, shoulder dystocia, postpartum hemorrhage, and adverse pregnancy outcomes that were significantly associated with the pandemic. After adjustment for the demographic characteristics (age, ethnicity, education, and work) of pregnant women in model II, all the above pregnancy outcomes remained statistically significant. Apart from the demographic characteristics of pregnant women, we further controlled for the demographic characteristics of their spouses in model III, we observed a higher risk of all above pregnancy outcomes. In model IV, we additionally controlled for pre-pregnancy BMI, gravidity, history of abnormal pregnancy, history of cesarean section, all results remained significant (anemia: OR = 1.498, 95% CI = 1.216–1.846; vaginitis: OR = 3.294, 95% CI = 1.878–5.777; adverse pregnancy outcomes: OR = 1.301, 95% CI = 0.999–1.695; postpartum hemorrhage: OR = 0.779, 95% CI = 0.612–0.992; shoulder dystocia: OR = 5.117, 95% CI = 2.098–12.48; preterm birth: OR = 1.944, 95% CI = 1.144–3.302).

We analyzed the factors related to adverse pregnancy outcomes in 172 women in the pandemic group and 139 women in the pre-pandemic group. In a univariate analysis, the rate of stay-at-home was significantly higher in pregnant women (20.47 vs. 10.95%, P = 0.024) and their spouses (19.63 vs. 10.45%, P = 0.029) in the pandemic group, and higher risk of anemia (47.67 vs. 35.97%, P = 0.038) and vaginal candidiasis (5.20 vs. 0.72%, P = 0.055) in the pandemic group. However, compared with the pre-pandemic group, women in the pandemic group had a lower rate of exposure to smoking (14.91 vs. 26.67%, P = 0.027), tea drinking (4.39 vs. 15.83%, P = 0.004) in the first trimester. Moreover, women in the pandemic group had fewer exercises, such as reduced average walking time per week (4.65 vs. 5.69 h), and reduced proportion of participating in physical activities except walking (at least three times per week, at least 30 min each time) (12.5 vs. 18.33%), but it was not statistically significant (P = 0.22) (Table 4).

Furthermore, we conducted a multivariate analysis using a binary logistic regression model to predict the risk factors affecting adverse pregnancy outcomes during the pandemic. Compared with the pre-pandemic group, women who had adverse pregnancy outcomes in the pandemic group often had anemia (OR = 2.002, CI = 1.080~3.708, P = 0.027), less exposure to smoking (OR = 0.456 CI = 0.213~0.974, P = 0.043) and less tea drinking (OR = 0.193 CI = 0.057–0.653, P = 0.008) during pregnancy (Table 5).