Clinical evolution of COVID-19 during pregnancy at different altitudes: a population-based study

Introduction

Pneumonia in pregnant women, regardless of the etiology, is an important cause of morbidity and mortality. It is known that during pregnancy there are different anatomical and physiological changes such as diaphragm elevation due to the progressive increase in uterine dimensions, increased intra-abdominal pressure and intravascular volume, as well as decreased chest distension and immune response. As a consequence, the susceptibility to infection increases, in addition to the risk of developing complications and even respiratory failure (1-3).

In 2002, SARS-CoV infected 8,098 people with a case fatality rate (CFR) of 10.5% (4), while in 2012 MERS-CoV had a total of 2,519 confirmed cases and a CFR of 34.4% (5). On the other hand, in 2009, cases of influenza A H1N1 were estimated at 60.8 million, and approximately 18,500 deaths were reported (6-7). These diseases have been shown to be more severe in pregnant women.

We are currently facing the pandemic caused by the novel coronavirus, SARS-CoV-2, which has already exceeded 20 million cases worldwide (8). Cases and deaths are increasing rapidly and we still lack specific treatment for the disease, and although rates of hospitalization and admission to the intensive care unit (ICU) are variable, health systems are already saturated.

Recent studies suggest the possibility that being pregnant may be a risk factor for SARS-CoV-2 infection and its complications (9). Similarly, it was suggested that the infection rate of this virus might be lower at high-altitude (over 2,500m) 10). This study seeks to evaluate pregnancy and high-altitude as factors that could intervene in the risk of contracting SARS-CoV-2 infection, as well as the severity of its presentation.

Materials and methods

Results

Summary characteristics of the study population are presented in Table 1. Pregnant women represented 2.23% of the study population and were on average six years younger than non-pregnant women (28.2 versus 34.9 years, p<0.0001). Pregnant women had a 11.3% higher proportion of COVID-19 compared to non-pregnant women (p<0.0001), and a following 132% higher proportion of hospital admission (p<0.0001) and 48.6% higher proportion of ICU admission (p=0.0014). Also, pregnant women had a 45% higher proportion of COVID-19 pneumonia, 20.3% higher proportion of admission, 90.7% higher ICU admission and a 37.8% lower case-fatality rate (p<0.0001) (Crude prevalences).

Most of pregnant women who developed COVID-19 pneumonia (5.06%; 95% CI: 4.5-5.6), required hospitalization (84.08%; 95% CI: 80.0-87.4), a proportion that nearly quadrupled that of COVID-19 pregnant women admitted without pneumonia (24.26%; 95% CI: 22.8-25.7%). The case-fatality rate of COVID-19 pregnant women with pneumonia was ten-fold of that of pregnant women without pneumonia (11.67% [95% CI:8.8-15.3] versus 1.69% [95% CI:1.3-2.2], p<0.0001). Half of pregnant women lived at altitudes above 1185 meters (50.03%, 3724/7444) and 25.02% lived in altitudes above 2187 meters (1863/7774); while 52.4% were born in altitudes above 1567 meters (3903/7444) and 25% in altitudes above 2240 meters (1863/7444). When COVID-19-related symptoms appeared, pregnant women showed a tendency to consult a healthcare center earlier than non-pregnant women (p<0.0001). Although the median time until death of COVID-19 pregnant women was two days longer than non-pregnant patients, no significant differences were found.

Controlling for age and altitude of residence and origin, pregnancy was independently associated with a 15% higher probability of COVID-19 (aPR: 1.15; 47.7% versus 41.4%), a 116% higher probability of its following admission (aPR: 2.169; 22.5% versus 10.4%) and a 127% higher probability of ICU admission (aPR: 2.275; 14.5% versus 6.4%) compared to non-pregnant women without any risk factors. Also, pregnancy was independently associated with 84.2% higher probability of developing COVID-19 pneumonia (aPR: 1.842; 6.4% versus 3.5%) and a 163% higher probability of its following admission (aPR 2.639; 25.3% versus 9.6%). COVID-19 pregnant women seemed to have a higher probability of intubation and a higher case-fatality rate, and those who had pneumonia seemed to have a higher probability of admission and intubation and a lower case-fatality rate compared to non-pregnant women, however no significant differences were found (p=0.279, 0.48, 0.1772, 0.0758 and 0.4182 respectively) (Table 2).

In Figure 1–A, we show that the probability of COVID-19 increased with age for both pregnant and non-pregnant women. However, it was always higher in pregnant women, with those over 35 years found with a 50% or higher probability. The chance of developing pneumonia also increased with age for both groups and never exceeded the 13.7% in pregnant women and 7.5% in non-pregnant women (Figure 1–B). Hospitalization curves had inverse trends, diminishing for pregnant and increasing for non-pregnant patients. However, the 95% confidence bands for COVID-19 pneumonia overlapped in women older than 45 years (Figure 1 – C). The mean chance of intubation for pregnant women over 30 years old was higher compared to non-pregnant women in SARS-CoV-2 infection and over 25 years old in COVID-19 pneumonia, however no significant differences were found for all ages (p>0.05) (Figure 2 – A-B). The probability of ICU admission in all COVID-19 cases was higher in pregnant women older than 29 years old and women older than 25 years old in COVID-19 pneumonia compared to non-pregnant women (Figure 2 – C-D). Consistent with the adjusted prevalence ratio found, no significant differences were found for case-fatality rates (Figure 2 – E-F).

• Download figure
• Open in new tab

Image 1

– Mild and moderate COVID-19 and the subset of COVID-19 pneumonia along ages. Gray: Nonpregnant women. Orange: Pregnant women. Non-overlapping confidence bands represent statistically significant differences at the p<0.05 level.

• Download figure
• Open in new tab

Image 2

– Severe COVID-19 and the subset of COVID-19 pneumonia along ages. Gray: Non-pregnant women. Orange: Pregnant women. Non-overlapping confidence bands represent statistically significant differences at the p<0.05 level.

As for altitude, the probability trend of COVID-19 (all cases) registered a decreasing trend with pregnant women who resided below 1,750 meters at increased risk and COVID-19 pneumonia for those who resided below 2750 meters (Figure 3 – A-B). The chance of admission also decreased with increasing altitudes and was consistently higher in pregnant women (Figure 3 – C-D). Pregnant women had the same chance of intubation than no-pregnant across all altitudes (p>0.05), (Figure 4 – A-B). The mean probability of ICU admission decreased with altitude but was significantly higher in pregnant women with COVID-19 who lived above 750 meters and all altitudes above sea level in pneumonia cases, compared to non-pregnant women (Figure 4 – C-D). In contrast with the adjusted prevalence ratio, the case-fatality rate was found to be lower in pregnant women with COVID-19 who lived above 1000 meters and lower in all pregnant women with pneumonia (Figure 4 – E-F).

• Download figure
• Open in new tab

Image 3

– Mild and moderate COVID-19 and the subset of COVID-19 pneumonia along altitudes. Gray: Nonpregnant women. Orange: Pregnant women. Non-overlapping confidence bands represent statistically significant differences at the p<0.05 level.

• Download figure
• Open in new tab

Image 4

– Severe COVID-19 and the subset of COVID-19 pneumonia along altitudes. Gray: Non-pregnant women. Orange: Pregnant women. Non-overlapping confidence bands represent statistically significant differences at the p<0.05 level.

Discussion

In the Mexican population, pregnancy was associated with a higher probability of COVID-19 and its following hospitalization and ICU admission; and pneumonia and its following ICU admission compared to non-pregnant women of reproductive age.

Pregnant women displayed a seemingly lower mean probability of intubation until 25-30 years old and a significant lower risk of ICU admission until 25-30 years old than non-pregnant women. With increasing altitudes, pregnant women displayed both a lower probability of COVID-19 and its corresponding case-fatality rates. The data so far does not allow for compelling conclusions about intubation differences according to pregnancy status (Figure 1-4).

Pregnant women with influenza, SARS-CoV or MERS-CoV have shown more complications and lethality. During H1N1 Influenza, hospitalization and ICU admission rates were higher compared to the general population. Hospital admission was especially higher in pregnant women in the second half of pregnancy and 2-64% were severe cases (defined as ICU admission or maternal death). Among pregnant women admitted to the ICU, the maternal mortality rate was 0-33% (13). Pregnant women with SARS-CoV infection showed a higher probability of intubation and ICU admission compared to nonpregnant women, in addition to an increased risk of kidney failure and disseminated intravascular coagulopathy. The case fatality rate in pregnant patients with this infection was reported at 25% (14). MERS-CoV infection during pregnancy was only reported in few patients (n=12). Seven of them required admission to the ICU, 3 died, of which 2 contracted MERS-CoV during the third trimester and 1 during the second trimester (15-16).

A systematic review of 538 pregnant women with COVID-19 identified that, according to Wu and McGoogan’s (17) classification of disease severity, 86.1% had mild disease, 15.3% severe disease and 1.4% critical disease; while only 3.0% of cases were reported as ICU admissions (18). Although the Mexican dataset did not allow for a standard definition of COVID-19 pneumonia, some parallels could be drawn. Mild cases, defined as those who were treated as outpatients, accounted for 75.74% of all pregnancies with COVID-19 and 15.92% of all pregnancies with pneumonia without other risk factors recorded. While “severe” cases (in which the clinician identified non-reassuring signs and symptoms related or not to pregnancy) that were more likely to be admitted represented 24.26% of pregnant women with COVID-19 and 84.08% in the subset of pregnant women with COVID-19 pneumonia. The rate of pregnant women with COVID-19 “critical” cases (Those who required ICU admission) stood at 9.63%; however, pregnant patients with COVID-19 pneumonia had a higher rate of ICU admission (18.10%) followed by an overall COVID-19 1.69% case-fatality rate that increased by several orders of magnitude if pneumonia was also present (11.67%).

The largest case series of women of reproductive age (15-45 years) reported by the Centers of Disease Control and Prevention (19) found that among the 9% of pregnant women (8207/91412), 31.5% required hospitalization, 1.5% needed ICU support and 0.2% died. This represents up to 46.4% additional hospitalizations in comparison to the Mexican series. The difference in the percentages could be due to the lower proportion (2.23%) of pregnancies in women of reproductive age (15-49 years) in our study, in addition to characteristics inherent to their population or their health system.

The Angiotensin-Converting Enzyme 2 (ACE2) is a receptor on the cell surface to which binds SARS-CoV-2 for entering inside it. During pregnancy, ACE2 activity can increase up to twofold, being the placenta and the uterus the most important sources of ACE2 (20). This increased activity responds to the need of blood pressure regulation in pregnancy (21). Also, ACE2 expression is critical for both an adequate pregnancy progression and fetal development, as its deficiency relates to both impaired gestational weight and fetal growth restriction (22). Lower estrogen activity, a down regulator of ACE2 expression, may contribute in its increased activity seen in pregnancy. 17β-estradiol (E2) is physiologically found in women during pregnancy and has been shown to reduce ACE2 mRNA in differentiated normal human bronchial epithelial cells (NHBE), which does not necessarily mean a reduction of surface ACE2 protein (23). However, these reports could be important to understand why pregnant women did not have a higher risk of mortality from SARS-CoV-2. Therefore, the increased ACE2 pool could render pregnant women more susceptible to developing COVID-19 pneumonia. However, it remains unclear how the complexity of the immune system that renders pregnant women more susceptible to viral, bacterial, parasitic and autoimmune disorders interact with the inflammatory response associated with severe COVID-19 cases (24).

Regarding altitude, we found that in pregnant women altitude influenced the trend of COVID-19, development of pneumonia, admission, intubation, and maternal case-fatality rates (Figure 3-4). The main significant relationship was found for pregnant women with COVID-19 compared to their baseline probability (of non-pregnant women) as all saw falling rates that were higher that of non-pregnant women only below 1750 meters. The admission trend was always higher for pregnant women but tended to decay with higher altitudes (Figure 3). Finally, the last difference was seen at case-fatality rates, finding that pregnant women are at lower CFR above 750 meters (Figure 4 – E). This paper confirms that predictive models could be used to monitor populations at risk through probability plots (25). This approach has the advantage of performing a regression on the continuous spectrum of altitudes instead of assigning a static value for altitude ranges, which could introduce cluster effect and difficult the interpretation of results.

Arias-Reyes first observed less COVID-19 rates in high-altitude dwellers compared to populations living below 3,000 meters. The authors related this association with a possible lower ACE2 expression in inhabitants of high altitude (26). This observation was further observed by the same authors in patients living below 1000 meters (27). In Peru, researchers also found that COVID-19 rates and its subsequent mortality fall logarithmically with increasing altitudes (10). Of note, hypobaric hypoxia is classically described to start at 1500 meters due to a reduced alveolar diffusion pressure and reduced mass transport (28). As the hypobaric hypoxia found at high altitudes upregulates the hypoxia inducible factor 1 alpha (HIF1-a), ACE2 downregulation ensues (29). Under experimental conditions, TMPRSS2 is also downregulated via HIF-signaling (30). As these membrane enzymes represent the gate and the doorkeeper of viral cell entry (31), patients already residing at high altitudes would benefit with a reduced chance of SARS-CoV-2 infection. Although both SARS-CoV and SARS-CoV-2 share the same entry receptor (ACE2), historical observations under hypoxia in SARS-CoV were not feasible due to the geographical distribution of the disease that was not reported in high-altitude areas within the affected countries (Image 5).

• Download figure
• Open in new tab

Image 5

– Geographical distribution of SARS-CoV.

Apart from ACE2, erythropoietin (EPO) was also suggested as a putative factor protecting human population at high altitude from COVID-19 (32). Hypobaric hypoxia induces 2−4-fold increases in serum levels of EPO in permanent citizens at high altitude compared to lowlanders (33). However, independently from altitude, EPO levels rise during pregnancy, especially during the first trimester, to 2-4 times the concentration compared with nonpregnant control subjects (34). EPO is certainly the essential hormone for the stimulation of red blood cells production and reticulocytosis in the bone marrow (35-36). This aspect is central in the infectious scenario of COVID-19, as EPO may help restoring hemoglobin levels and improve oxygen delivery to the tissues. In addition, EPO also promotes adaptive cellular responses to hypoxic challenges and tissue-damaging insults in non-hematopoietic tissues (37). Non-erythroid effects of EPO include the activation of anti-apoptotic, anti-oxidative, anti-cytotoxic, and anti-inflammatory mechanisms (38) in several tissues, including the brain, lung, kidney, vascular bed, and heart (39-43). As such, EPO may play a key role against generalized tissue damage, multiple organ dysfunction and inflammation induced by SARS-CoV-2 in lowland pregnant women in general, but especially in pregnant women at high-altitude (44-45). In fact, increased levels of EPO with altitude could explain, partially, the altitude-related gradual decreasing tendency of pregnant women to acquire SARS-CoV-2, develop COVID-19 pneumonia, hospitalization, and maternal death observed in our work. For other respiratory viral infections, as influenza, high altitude rates are only 30% lesser, (46) but for COVID-19 in Peru they are 350% lesser, being the decrease in ECA-2 expression related to hypobaric hypoxia probably the reason (10).

Limitations of the study included that the dataset did not record the trimester of pregnancy, therefore we couldn’t determined the clinical reason for admission to regular hospitalization or ICU or whether the patient was admitted preventively and subsequently worsened their respiratory parameters requiring intubation. Because of this, it remains unclear how SARS-CoV-2 infection in the first trimester affects women and fetal stability across the course of pregnancy, despite that SARS-CoV-2 transplacental transmission has been reported previously (47). Also, pregnant women’s physiology varies across gestational age and might interact with the hypobaric hypoxia seen at high altitudes. In order to isolate the effects of pregnancy, we only included women with no other recorded risk factors in the analysis, so the burden of chronic diseases such as hypertension and obesity remains to be scrutinized.

Conclusion

Pregnancy was associated with a higher probability of SARS-CoV-2 infection, COVID-19 pneumonia and need for hospitalization, but not for case fatality rates. Furthermore, the still developing information on this new coronavirus limits the understanding of its possible effect during pregnancy. However, our results suggest that as the altitude increase, the risk of SARS-CoV-2 infection decreases. Despite, various aspects of this virus are yet unknown, ACE2 and EPO may play important role in this phenomenon. Moreover, our data suggests that knowledge of the mechanisms of physiological adjustment to hypoxia may help to better understand COVID-19. At the same time, this knowledge can lead to the development of treatment strategies specifically focused on the population of pregnant women.

Declaration of conflicts of interest

The authors declare that no financial, academic or institutional conflicts of interest that affected the conception, elaboration, execution and processing of this paper.

Funding

The paper has not received any kind of funding aid from public agencies, the commercial sector or non-profit entities.