Indicators of Cardiometabolic Function in Pregnancy and Long-Term Risk of COVID-19: Population-Based Cohort Study

Background: Pregnancy increases a woman’s susceptibility to severe COVID-19, especially those with metabolic dysfunction. It is unknown if markers of metabolic dysfunction commonly assessed around pregnancy are associated with COVID-19 illness after pregnancy. Aim: The aim of this study is to evaluate the indicators of metabolic dysfunction collected in pregnancy and the future risk of severe COVID-19 after pregnancy. Methods: This population-based cohort study was completed in all of Ontario, comprising 417,713 women aged 15-49 years with a hospital birth between April 2007 and March 2018. The main exposure was each 1-kg/m2 higher body mass index (BMI), 1-mmol/L higher glucose concentration at the 50-g glucose challenge test, and one-week earlier gestational week at delivery. The main outcome was severe COVID-19 illness or death, from the start of the pandemic period on March 1, 2020, till December 31, 2021. Results: The adjusted hazard ratio (aHR) of COVID-19 illness increased per 1-kg/m2 higher BMI (1.05, 95% CI 1.04-1.06), per 1-mmol/L higher serum glucose concentration (1.16, 95% CI 1.10-1.22), and for each one-week earlier gestational week at delivery (1.12, 95% CI 1.03-1.23). Relative to women with no dichotomized risk factors, the aHR for severe COVID-19 was 1.60 (95% CI 1.28-2.01) with one factor, 3.34 (95% CI 2.51-4.44) with two factors, and 4.52 (95% CI 2.11-9.67) with three factors. Conclusions: The number, and degree, of standard metabolic indicators measured around pregnancy predict the future risk of severe COVID-19 remotely after that pregnancy.


Introduction
Gestational diabetes mellitus (DM), higher body mass index (BMI) in pregnancy, and preterm delivery partly predict the onset of metabolic syndrome (MetSyn) in women years after pregnancy [1,2]. After the emergence of the SARS-COV-2 pandemic around March 2020, it became evident that pregnant women were prone to severe COVID-19 illness and adverse perinatal outcomes, such as preterm labor and preterm birth [3]. The risk of adversity was most pronounced in women with a high BMI or gestational DM within the index pregnancy [4]. What is not known, however, is whether the aforementioned factors, when measured in pregnancy, are associated with the onset of COVID-19 illness well after that pregnancy has ended. The availability of three standardized continuous measures, namely, pre-pregnancy BMI, glucose concentration at the time of gestational DM screening, and gestational week at delivery, enabled us to address this question.
This study evaluated the future risk of severe COVID-19 in relation to prior pregnancy BMI, serum glucose concentration, and gestational age at birth -both as continuous and dichotomized risk factors.
with pre-pregnancy DM and women not alive or eligible for OHIP on March 1, 2020 (the start of the SARS-CoV-2 pandemic) were excluded. If a woman had more than one eligible delivery during the study period, then her latest birth was considered. The last birth was considered up to March 31, 2018, to minimize the chance that a woman was pregnant, or had recently given birth, at the onset of the COVID-19 pandemic and because BMI data were only available up to that date.
Births and outpatient and inpatient encounters were captured in province-wide administrative datasets that were linked using unique encoded identifiers and analyzed at ICES (https://datadictionary.ices.on.ca/Applications/DataDictionary/Default.aspx), as described by Catov et al. and Li et at. [1,5] and in Appendix table. BMI was identified in the Better Outcomes Registry and Network (BORN) Information System and Niday Perinatal Databases with a 50-g GCT in the Ontario Laboratory Information System [6]. All SARS-CoV-2 vaccinations in Ontario are also captured at ICES (https://data.ontario.ca/dataset/covid-19-vaccine-data-in-ontario). More than 95% of pregnancies have an ultrasound enabling accurate pregnancy dating [7].
The main study outcome was severe COVID-19 illness arising from the start of the pandemic period of March 1, 2020 (i.e., time zero), to December 31, 2021 (the latest complete data). Severe COVID-19 illness was based on a positive SARS-CoV-2 PCR test within seven days preceding, or up to three days after, hospitalization or death [6].

Analyses
Time-to-event analyses started at time zero. Separate Cox proportional hazard models generated hazard ratios (HRs) and 95% CI for the respective relationships between the study outcomes and each 1-kg/m 2 incremental higher BMI; 1-mmol/L higher GCT glucose concentration, and one-week earlier gestational week at delivery, from ≥37 weeks (term birth) declining weekly down to ≤24 weeks. Next, each study outcome was assessed about having 0 (referent), 1, 2, or 3 dichotomized risk factors (i.e., BMI ≥ 30 kg/m 2 , positive 50g GCT ≥ 7.8 mmol/L, and/or preterm birth < 37 weeks).
HRs were adjusted for age, rural residence, and area income quintile at the index birth; a woman's age at time zero; chronic hypertension in the index pregnancy or up to two years before that pregnancy; and the time-varying SARS-CoV-2 first vaccination date. Censoring was based on death occurring prior to either outcome, loss of OHIP eligibility, or arrival at the end of the study period of December 31, 2021. A further analysis was censored at the start of any subsequent pregnancy during the SARS-CoV-2 pandemic.
Sample size estimations were not performed as the current study used a fixed population-based data sample.

Ethics approval
Datasets were linked using unique encoded identifiers and analyzed at ICES. The use of data in this project was authorized under section 45 of Ontario's Personal Health Information Protection Act, which does not require the approval of a research ethics board.

Results
A total of 417,713 women were included, followed by a median (IQR) of 1.8 (1.8, 1.8) years after time zero, totaling 756,471 person-years of follow-up. Their mean (SD) age at the index delivery was 31.1 (5.2) years, and 83.4% received at least one SARS-CoV-2 vaccination during the follow-up period ( Table 1). Chronic hypertension in the index pregnancy, or up to 2 years before 19,591 (4.7)

In the index delivery
Stillbirth

From the start of the COVID-19 pandemic on March 1, 2020 (time zero), to the end of t he study follow-up on December 31, 2021
Mean (   Relative to women with no risk factors at their conventional cut points, the aHR for severe COVID-19 was 1.60 (95% CI 1.28-2.01) with one factor, 3.34 (95% CI 2.51-4.44) with two factors, and 4.52 (95% CI 2.11-9.67) in women with three factors ( Table 2). There were 64,141 women (15.4%) who were pregnant during the SARS-CoV-2 pandemic, of whom 93 (1.5 per 1000) had severe COVID-19. Further censoring on pregnancy generated roughly the same aHRs ( Table 2).

Discussion
Elevated BMI and gestational DM are known risk factors for acquiring COVID-19 in pregnancy [4]. However, unlike prior studies, these two risk factors were handled here as continuous measures, and they were also assessed in relation to acquiring COVID-19 well beyond pregnancy. In a novel manner, gestational age at delivery was also assessed while avoiding any potential reverse causation introduced when COVID-19 and timing of birth were assessed in the same pregnancy because a severe infection may either precipitate or necessitate preterm birth [3].
The risk of severe COVID-19 was higher when each measure was analyzed continuously and even more so when combined together at conventional cut points. Recent data among middle-aged non-pregnant adults demonstrated a higher risk of COVID-19-related ICU admission, invasive mechanical ventilation, and mortality in the presence of the MetSyn [8]. While the current study could not formally assess all MetSyn criteria, maternal lipid profile, glucose intolerance in pregnancy, and pre-pregnancy BMI have been shown to predict the onset of MetSyn three months postpartum [9], as do a history of preterm delivery, gestational DM, and BMI years after pregnancy [1,2].
These study findings align with a body of work that considers in-pregnancy measures like BMI, glucose handling, and timing of delivery as a means to predict, and potentially modify, a woman's future health. To date, studies have largely focused on future cardiometabolic health [9][10][11], but not new-onset infection. The current study further suggests that metabolic measures may offer a future perspective on a woman's vulnerability to COVID-19. Even so, it remains to be determined whether these and other metabolic factors (e.g., blood pressure) influence her susceptibility to other types of viral and bacterial infections, or whether metabolic modification after birth can mitigate the onset of severe infectious illness.

Strengths and limitations
As a limitation, the current study adjusted for some confounders, such as rural residence and income level, but not race or ethnicity. Chronic hypertension was also accounted for, but blood pressure measures were not available. It is unlikely that caregiver burden can explain the relationship between preterm delivery and COVID-19 risk after pregnancy as parents of a preterm-born infant exhibit only slightly higher levels of stress than parents of term-born children [12]. Rather, women with the MetSyn are more likely to experience preterm delivery, especially provider-initiated preterm birth [13]. The study was conducted within a jurisdiction with relatively high vaccine uptake; so, the risk of severe COVID-19 may be even more pronounced in settings with lower vaccine use. While the first SARS-COV-2 vaccination was handled as a time-varying covariate, emerging SARS-CoV-2 variants were not differentiated.

Conclusions
The risk of severe COVID-19 remotely after pregnancy is higher in the presence of metabolic indicators standardly measured around the time of pregnancy.    BORN"), part of the Children's Hospital of Eastern Ontario. The interpretation and conclusions contained herein do not necessarily represent those of BORN Ontario. Data availability: The dataset from this study is held securely in the coded form at ICES. While data-sharing agreements prohibit ICES from making the dataset publicly available, access may be granted to those who meet pre-specified criteria for confidential access, available at www.ices.on.ca/DAS. The full dataset creation plan and underlying analytic code are available from the authors upon request, understanding that the computer programs may rely upon coding templates or macros that are unique to ICES and are therefore either inaccessible or may require modification.