The effects of biological sex and cardiovascular disease on COVID-19 mortality.

Cardiovascular disease (CVD) is a common comorbidity observed in patients with coronavirus disease 2019 (COVID-19), which is associated with increased severity and mortality. However, the effects of biological sex on CVD-associated mortality in patients with COVID-19 are poorly established, particularly among Hispanic/Latin Americans. We examined the association of preexisting CVD with COVID-19 mortality in hospitalized Latin American men and women. This multicenter study included Mexican patients hospitalized with a positive diagnosis of COVID-19. The main outcome was in-hospital mortality. Multivariable regression analyses were used to calculate the adjusted odds ratio with 95% confidence interval for mortality in women and men. Of 81,400 patients with a positive diagnosis for SARS-CoV-2 infection, 28,929 (35.54%) hospitalized patients were evaluated. Of these, 35.41% (10,243) were women. In-hospital death was higher in men than in women. In relation to CVD between the sexes, women had a higher incidence of CVD than men (4.69 vs. 3.93%, P = 0.0023). The adjusted logistic regression analyses showed that CVD was significantly associated with COVID-19 mortality in women but not men. We then stratified by sex according to age <52 and ≥52 yr old. Similar significant association was also found in prespecified analysis in women ≥52 yr old but not in men of similar age. We conclude that CVD's effect on mortality among patients hospitalized with COVID-19 is dependent on biological sex and age in this Latin American cohort. These results suggest that therapeutic strategies for Latin American women with CVD and COVID-19 should include particular attention to their cardiovascular health.NEW & NOTEWORTHY CVD's effect on COVID-19 mortality is dependent on biological sex and age. CVD in women but not men with COVID-19 is associated with significantly unfavorable outcomes.

INTRODUCTION

A high prevalence of cardiovascular diseases (CVD) including chronic heart disease and vascular diseases have been reported in patients with coronavirus disease 2019 (COVID-19) in several countries. Indeed, CVD is recognized as one of the most common comorbidities observed in both elderly and young patients with COVID-19 and is considered a risk factor for developing severe COVID-19. Recent quantitative meta-analyses that included 24,032,712 cases showed that patients with CVD had a significantly greater risk of COVID-19 severity and associated mortality (1). CVD is an age-related condition that is more commonly observed in men as compared with women. However, a growing body of evidence shows that women with CVD have worse outcomes and increased risk of mortality than men (2).

Early studies in China showed that COVID-19 was less prevalent in women as compared with men (3). Others have shown that women have lower proportion of hospitalizations, intensive care unit (ICU) admissions, and deaths than men (4). However, the report of the Global Health 50/50 data showed that the number of COVID-19 confirmed cases were similar between men and women (4, 5). In fact, meta-analyses of more than 3 million reported cases from government-sponsored websites in different countries collected between January 1, 2020 and June 1, 2020 revealed that the proportion of subjects segregated by sex with confirmed COVID-19 was similar between the sexes, but men had three times the risk of ICU admission and twice the risk of death (6). Consistent with these observations, studies evaluating risk factors for poor outcomes among hospitalized patients from China, United States, and European countries showed that women, as compared with men, were associated with lower COVID-19 severity and mortality (7). These findings suggest that biological sex may play a key role in the progression, severity, and mortality among patients with COVID-19. However, whether CVD affects COVID-19-associated mortality in men and women differently remains poorly understood as sex-disaggregated data and findings on sex-specific COVID-19 outcomes are not frequently reported.

Hispanic/Latin American women and men have been strongly affected by the current pandemic. Hispanic/Latin American patients were more likely than White patients to test positive for COVID-19 (8). This is clinically relevant as Hispanic/Latin American patients when compared with non-Hispanic White patients in the United States are ethnic groups who are more likely to have COVID-19-associated hospitalizations, ICU admissions, as well as in-hospital deaths (9). We hypothesized that among hospitalized, Latin American patients with COVID-19, preexisting CVD would be associated with higher mortality in men than in women. We examined 81,400 laboratory-confirmed severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2)-infected women and men from Mexico collected between February 27, 2020 and May 28, 2020 and reported on the relationship between biological sex, CVD, and age with mortality among hospitalized patients.

METHODS

Study Population and Design

This is a retrospective, observational, and multicentric study from the public database of the Epidemiological Surveillance System of Viral Respiratory Diseases of confirmed cases drawn from a larger database of suspected cases identified by the healthcare system in Mexico (10). This database is updated daily with all suspected cases of COVID-19 from all health institutions across Mexico.

This study is a multicenter cohort of 81,400 patients with positive diagnosis of COVID-19 in Mexico that was collected between February 27, 2020 and May 28, 2020. We studied the publicly available anonymized data only. A review by Institutional Review Board was not required according to the open data policy of the Ministry of Government of Mexico. We focused our analyses on hospitalized patients with confirmed SARS-CoV-2 and had at least two COVID-19 symptoms and at least one of the clinical conditions associated with COVID-19 disease. Our study cohort consisted of 28,929 hospitalized patients of both sexes. This study follows the reporting guidelines of the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Collaborative Initiative.

Definitions of Suspected and Confirmed COVID-19 Cases

The Ministry of Health in Mexico defines a suspected COVID-19 case as a person of any age who presented at least two of the following symptoms—cough, fever, or headache—that were accompanied by at least one of the following clinical conditions: dyspnea, arthralgia, myalgia, odynophagia, sore throat, rhinorrhea, conjunctivitis, and chest pain in the past 7 days. Confirmed COVID-19 case was defined as a patient who met the operational definition of suspected cases and was positive for SARS-CoV-2 by polymerase chain reaction (PCR) and certified by the Institute of Epidemiological Diagnosis and Reference (InDRE) in Mexico (11).

Data Collection and CVD Definition

Available information for all cases consisted of age, sex, current pregnancy, state of birthplace, place of residency, nationality, identification of individuals who speak indigenous languages, type of medical attention (hospitalization/ambulatory), admission date, symptom onset date, invasive mechanical ventilation (yes/no), admission to intensive care unit (ICU) (yes/no), pneumonia (yes/no), date of death, contact with confirmed COVID-19 cases (yes/no), and SARS-CoV-2 RT-PCR result. Comorbidity information was derived from a previous medical diagnosis and self-reported by the patient or responsible caregiver. The comorbidities included were hypertension, diabetes, obesity, CVD, chronic kidney disease (CKD), immunosuppression, asthma, chronic obstructive pulmonary disease (COPD), and smoking.

CVD was defined as all individuals who declared to have a previous medical diagnosis of a heart-related disease according to the pathological entities included and recognized by the Mexican clinical practice guidelines from the Mexican Society of Cardiology and the Ministry of Health in Mexico. These included coronary heart disease (CHD), cerebrovascular disease, rheumatic heart disease, congenital heart disease, deep vein thrombosis, and pulmonary embolism (12, 13).

Statistical Analyses

Data were analyzed and presented using descriptive statistics. Continuous variables were expressed as medians and interquartile ranges and categorical variables as frequencies and percentages. Comparisons between groups were tested using the χ2 test for categorical variables and Student’s t test for continuous variables. Multivariable logistic regression analyses, adjusting for potential confounders, were performed to determine factors associated with COVID-19 mortality in women and men. Stratified regression analysis according to sex was performed to assess the lethality of COVID-19 infection. Further stratified regression analysis by sex and age was performed to elucidate the association of CVD with COVID-19 mortality. Odds ratios (OR) and 95% confidence intervals (95% CI) were presented. All analyses were carried out using SAS v9.1 package (SAS Institute). A two-tailed P value of <0.05 was considered a statistically significant result.

RESULTS

Characteristics of Study Population

As of May 28, 2020, 254,794 patients were tested for SARS-CoV-2 in Mexico: 81,400 (31.90%) cases were confirmed with SARS-CoV-2 infection, and of these, 29,128 patients were hospitalized. In this study, we excluded 199 (0.68%) cases because of pregnancy or unknown pregnancy status. Therefore, the data analyses shown below were performed on 28,929 (35.54%) hospitalized patients of which 10,243 (35.41%) were women.

The characteristics of the study population separated by sex are presented in Table 1. Median (IQR) age by sex was 54 (44–65) yr for men and 56 (45–66) yr for women. Comparative analysis showed that ICU, intubation, and pneumonia were also more common in men as compared with women. Also, smoking was higher in men than in women. However, CVD, hypertension, diabetes, obesity, CKD, asthma, immunosuppression, and the presence of COPD were more common in women when compared with men. In addition, in-hospital death was higher in men than in women.

Table 1.

Characteristics of COVID-19 hospitalized study population by sex

	Men	Women (Nonpregnant)	P
n	18,686	10,243
Age, yr	54 (44–65)	56 (45–66)	<0.0001
Smoking, n (%)	2,068 (11.16)	441 (4.33)	<0.0001
Mortality, n (%)	5,419 (29.00)	2,631 (25.69)	<0.0001
ICU, n (%)	1,948 (10.44)	875 (8.55)	<0.0001
Intubation, n (%)	2,021 (10.83)	876 (8.56)	<0.0001
Pneumonia, n (%)	12,600 (67.43)	6,713 (65.54)	0.0011
CVD, n (%)	728 (3.93)	477 (4.69)	0.0023
Hypertension, n (%)	5,611 (30.29)	3,960 (38.91)	<0.0001
Diabetes, n (%)	5,246 (28.32)	3,561 (35.01)	<0.0001
Obesity, n (%)	4,067 (22.01)	2,908 (28.63)	<0.0001
Asthma, n (%)	321 (1.73)	366 (3.60)	<0.0001
COPD, n (%)	604 (3.26)	498 (4.89)	<0.0001
Immunosuppression, n (%)	406 (2.19)	339 (3.33)	<0.0001
CKD, n (%)	828 (4.47)	534 (5.25)	0.0031
Indigenous population, n (%)	321 (1.72)	164 (1.60)	0.472

COVID-19, coronavirus disease 2019; ICU, intensive care unit; COPD, chronic obstructive pulmonary disorder; CKD, chronic kidney disorder; CVD, cardiovascular disorder. n, number of subjects. χ2 test for categorical variables and Student’s t test for continuous variable were used.

COVID-19 Mortality

Logistic regression analyses were used to evaluate mortality of patients with COVID-19 stratified by sex that was adjusted by age (continuous), smoking, ICU, intubation, pneumonia, and comorbidities including diabetes, obesity, COPD, and immunosuppression (Table 2). Our data show that these factors were significantly associated with COVID-19 mortality in both men and women. However, CVD was significantly associated with COVID-19 mortality in women (OR = 1.248, 95% CI = 1.012–1.540, P = 0.0385) but not in men (OR = 1.050, 95% CI = 0.887–1.243, P = 0.5712).

Table 2.

Logistic regression of COVID-19 mortality stratified by sex

	Men			Women (Nonpregnant)
	OR	95% CI	P	OR	95% CI	P
CVD*	1.050	0.887–1.243	0.5712	1.248	1.012–1.540	0.0385

*Adjusted for age, smoking, intensive care unit, intubation, pneumonia, diabetes, obesity, chronic obstructive pulmonary disorder, and immunosuppression. COVID-19, coronavirus disease 2019; OR, odds ratio; CI, confidence interval; CVD, cardiovascular disorder.

Stratified Analyses According to Biological Sex and Age

We then grouped our study population according to the average reported postmenopausal age among Latin American women (52 yr), defined using the criteria of the Stages of Reproductive Aging Workshop (STRAW), absence of menses for at least 1 yr (14) and age-matched men. When stratified by sex according to age <52 and ≥52 yr old, we observed that in women ≥52 yr old, CVD was again significantly associated with COVID-19 mortality (OR = 1.254, 95% CI = 1.003–1.567, P = 0.0466) when adjusted for age (continuous), smoking, ICU, intubation, pneumonia, diabetes, obesity, COPD, and immunosuppression. In men of both age groups and women with age <52 yr, CVD had no effect on COVID-19 mortality (Table 3).

Table 3.

Logistic regression of COVID-19 mortality stratified by sex and age

	Men (<52 yr)			Men (≥52 yr)			Women (<52 yr) (Nonpregnant)			Women (≥52 yr) (Nonpregnant)
	OR	95% CI	P	OR	95% CI	P	OR	95% CI	P	OR	95% CI	P
CVD*	1.217	0.778–1.906	0.3899	1.053	0.879–1.262	0.5728	1.212	0.668–2.199	0.5275	1.254	1.003–1.567	0.0466

Adjusted for age, smoking, intensive care unit, intubation, pneumonia, diabetes, obesity, chronic obstructive pulmonary disorder, and immunosuppression. COVID-19, coronavirus disease 2019; OR, odds ratio; CI, confidence interval; CVD, cardiovascular disorder.

DISCUSSION

CVD is a leading cause of mortality worldwide. Premenopausal women seem to be protected from CVD while postmenopausal women have the same risk for CVD as do men (15). However, women appear to be at greater risk than men for developing adverse CVD outcomes (2). In this study, we evaluated the relationship between CVD, age, and mortality in Latin American women and men with COVID-19. Our results show that in a cohort of Mexican patients, both the incidence and in-hospital COVID-19-associated mortality were higher in men than in women. However, our present findings show that CVD is a COVID-19 risk factor for mortality among women but not men and especially in women older than 52 yr of age. To the best of our knowledge, our study is the first to report the effects of CVD on mortality from COVID-19 in a sufficiently large cohort with sufficient power to properly contrast mortality outcomes in Hispanic/Latin American women and men with COVID-19 and CVD. Our data show that the proportion of COVID-19 in women was 35.41%, which is slightly smaller than those reported for women in the general Mexican population with 42% according to reported studies (16–18). Hospitalized Latin American patients show similar proportions of women as shown herein (19–22). These findings are significant as sex-specific COVID-19 outcomes are not well established and remain poorly understood, in part because of the scarcity of sex-disaggregated reports.

The presence of CVD is associated with vascular and cardiac aging patterns that differ in men and women. Increasing left ventricular (LV) wall thickness and LV mass are associated with higher incidence of cardiovascular events (23). LV wall thickness has been observed to increase in women as they age but not in men (24). Also, postmenopausal women have stiffer arteries when compared with men (25). These findings suggest that aging is associated with cardiac and vascular maladaptation in women, which may contribute to worse outcomes in older women with CVD and COVID-19. Our study population showed a higher proportion of older as compared with younger women (60.72% were women <52 yr of age and 39.27% were women <52 yr of age). Age and postmenopausal status are associated with higher CVD rates (26, 27).

Sex is a biological variable that also may affect the response to SARS-CoV-2 infection and COVID-19 severity. Various groups have proposed that the increased severity and mortality in men versus women with COVID-19 may be due to expression and activity of the angiotensin-converting enzyme-2 (ACE2), the putative receptor that facilitates SARS-CoV-2 cellular internalization and infectivity (28–31). Early studies showed that ACE2 activity was higher in men than in women. Consistent with these observations, studies have shown that 1) ACE2 expression in type II pneumocytes was higher in men as compared with women, 2) circulating ACE2 levels were higher in older men, and 3) ACE2 levels are associated with the presence of CVD. However, there are reports showing that ACE2 levels in skeletal muscle are higher in women than in men (32) and that ACE2 levels are higher in postmenopausal women when compared with premenopausal women (33, 34).

Limitations

Some limitations exist in our study. First, the retrospective and cross-sectional nature of our study does not allow us to determine cause and effect. Second, CVD of patients was self-reported at the time of medical care and could lead to underreporting of cases. In addition, the instrument used to collect the information was an official standardized questionnaire used by the federal government’s Epidemiological Surveillance System. However, CVD subtypes and pharmacological treatment information were not included. As such it was not possible to carry out a group subanalyses to identify specific CVD etiologies related to COVID-19 mortality in women and men in our study population. Our results show a CVD incidence of ∼4.3% (4.69% in women vs. 3.93% in men) that was not very different from that reported in the 2020 National Health and Nutrition Survey report issued by the National Institute of Public Health of Mexico, where they reported a CVD prevalence of 1.7% (1.8% in women vs. 1.6% in men) in adults older than 20 yr of age (35). However, it is important to note that CVD prevalence in Mexico is significantly lower when compared with other populations in the United States and the United Kingdom (36, 37). Furthermore, there are data showing reduced CVD-associated mortality in Hispanic/Latin Americans in the United States when compared with Whites (38). Moreover, there is evidence that CVD prevalence is lower among Hispanic/Latin American women as compared with men (39). Third, laboratory measures specific to CVD, as well as the hormonal profile, were not available in this study. This limited analyses of potential impact of specific types of CVD and menopausal status that may affect mortality in women with COVID-19. Indeed, it may have underestimated the risk of mortality in women, as there is evidence that women older than 50 yr of age who were on estradiol replacement therapy showed greater than 50% reduction in the risk of death by SARS-CoV-2 infection (40).

Our findings in this large population underscore the importance of studying CVD in women with COVID-19. Our data in women provide the rationale for developing prospective studies in women with CVD that have acute COVID-19 as well as in those with postacute sequelae of SARS-CoV-2 infection, an array of symptoms that persist for several weeks, months, and potentially years after an infection has been diagnosed. Also, given the high COVID-19 case fatality rates among Hispanic/Latin American subjects in the United States and Mexico (41, 42), further studies in these populations are critically needed to confirm our results.

Conclusions

In this study, hospitalized Latin American men had higher rates of mortality than women, whereas the incidence of CVD was higher in women than in men. Furthermore, our findings show that CVD’s effect on mortality is dependent on sex and age. Older women with CVD are associated with unfavorable outcomes if infected with SARS-CoV-2. Thus, therapeutic strategies in women, particularly older women, with CVD and COVID-19 should include attention to their cardiovascular health.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

M.E.H-H., R.Y.L.Z., P.P.-P., E.T.-R., and J.R.R. conceived and designed research; M.E.H.-H., R.Y.L.Z., and J.R.R. analyzed data; M.E.H.-H., R.Y.L.Z., E.T.-R., and J.R.R. interpreted results of experiments; R.Y.L.Z. prepared figures; M.E.H.-H., R.Y.L.Z., P.P.-P., E.T.-R., and J.R.R. drafted manuscript; M.E.H.-H., R.Y.L.Z., P.P.-P., E.T.-R., and J.R.R. edited and revised manuscript; M.E.H.-H., R.Y.L.Z., P.P.-P., E.T.-R., and J.R.R. approved final version of manuscript.