Introduction

The coronavirus disease 2019 (COVID-19) pandemic, caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), presented a major global public health challenge. In Mexico, the burden associated with the disease was high, and by the end of August 2023, more than 7.6 million confirmed cases were registered along with more than 330 thousand deaths among patients with confirmed SARS-CoV-2 infection1.

With the progression of the pandemic, the disease transitioned from an emergency phase, marked by surges in cases and overwhelmed healthcare systems, to an endemic phase where the virus persists at a baseline level within populations2. As the pandemic evolved, a comprehensive understanding of the risk factors for severe outcomes among COVID-19 patients, such as the development of pneumonia, has become increasingly critical3.

Pneumonia is a common and severe complication of COVID-19, often requiring hospitalization and intensive medical care4. Accurate risk assessment for pneumonia development in COVID-19 cases is relevant for clinical decision-making and resource allocation and provides valuable insights into the changing dynamics of the disease5.

When compared to women, men have an increased risk for severe outcomes in COVID-196. Published data comprehensively evaluating its effect across critical phases of the pandemic, as well as in the post-pandemic era, remain limited7,8,9,10,11,12. Moreover, and to the best of our knowledge, there are not published studies evaluating the effect of the patients’ sex on the risk of COVID-19 related pneumonia encompassing all the emergency phases of the pandemic together with the early endemic phase of the diseases.

This study aimed to evaluate the effect of patients’ sex on the risk of developing COVID-19-related pneumonia during the pandemic and the early months of the endemic phase of the disease. Overall, this study aims to contribute to the understanding of how sex influences the risk of pneumonia development in COVID-19 patients across different phases of the pandemic, thereby informing clinical practices and enhancing our knowledge of the evolving disease dynamics.

Materials and methods

Study design

We conducted a retrospective cohort study based on data derived from the epidemiological surveillance of COVID-19 throughout Mexico. The analyzed dataset, an integral component of the System for the Epidemiological Surveillance of Respiratory Viral Diseases (SISVER, the Spanish acronym), compiles all suspected and confirmed COVID-19 cases reported by both public and private healthcare facilities, with mandatory reporting requirements13. The data collection period extends from the onset of the pandemic in February 2020 to its conclusion in May 2023, encompassing the transition into the endemic phase of the disease (May 2023–August 2023)14,15,16,17.

Study population

Individuals aged 18 years or older at the onset of respiratory symptoms, later confirmed as cases of SARS-CoV-2 infection through reverse transcription-polymerase chain reaction (RT-PCR) or serologic assays, were eligible. Patients with missing final case classification (confirmed or discarded COVID-19), as well as those lacking clinical or epidemiological data of interest (e.g., comorbid conditions status), were excluded (\(\varvec{n}\)= 88,902). All patients meeting the eligibility criteria were included.

We utilized data from genomic surveillance of SARS-CoV-2 to categorize the included individuals based on the epidemiological week of patient-reported symptoms onset and the predominant variant during that period. The categorization was as follows: Wild type, weeks 08/2020 to 39/2020; B.1.1.519, weeks 40/2020 to 20/2021; B.1.617.2 (Delta), weeks 21/2021 to 48/2021; B.1.1.529 (Omicron), weeks 49/2021 to 41/2022; BQ1, weeks 42/2022 to 04/2023; and XBB.1.5, weeks 05/2023 to 35/2023. In Mexico, the end of the pandemic emergency was officially declared on May 9, 2023, corresponding to epidemiological week 1918.

Main outcome and data collection

The primary binary outcome of this study was pneumonia (no/yes), which was defined in accordance with established normative guidelines19. These guidelines resulted from a national consensus of institutions that are part of the Mexican National Healthcare System and were coordinated by the Ministry of Health. Pneumonia was diagnosed based on the presence of clinical symptoms (fever, cough, dyspnea, and rapid breathing)19 in individuals with laboratory-confirmed symptomatic SARS-CoV-2 infection. Although the guidelines were updated in August 2021, the criteria for pneumonia diagnosis were retrospectively applied to participants with earlier symptoms to ensure a consistent and uniform approach. Data related to the primary outcome, as well as other clinical and epidemiological variables, were sourced from the audited surveillance system. Medical records served from the patients as the primary data source for this system.

Statistical analysis

Summary statistics were computed, and the significance level was set at 5%. We used Risk Ratios (RR) and 95% confidence intervals (CI), estimated using log-binominal generalized linear regression models, to evaluate the effect of patient sex on the risk of being diagnosed with COVID-19-related pneumonia. The multiple model was adjusted for relevant clinical predictors, namely age, predominant SARS-CoV-2 genotype at illness onset, and comorbidities (present/absent) included in the evaluated dataset (obesity, type 2 diabetes mellitus, arterial hypertension, chronic obstructive pulmonary disease, asthma, chronic kidney disease, and immunosuppression due to any cause).

Additionally, eta-squared (η2) was used to evaluate the sample size effect on the obtained coefficients as published previously20. The analytical procedure was conducted using the statistical package Stata® MP 15 (College Station, TX, USA).

Ethical considerations

Since we analyzed publicly available data solely for academic purposes, ethics approval from a research committee was waived. The waiver for ethics approval, as well as the waiver for obtaining written informed consent to participate, was provided by the Committee of Ethics in Health Research 601 of the Mexican Institute of Social Security. However, we adhered to relevant data protection regulations and guidelines to ensure responsible and ethical data use.

Results

Data from approximately 1.6 million (n = 1,640,256) adult participants with laboratory confirmed COVID-19 were analyzed. Most of the evaluated patients (53.7%) were women, and the mean age (± standard deviation) was 43.6 ± 15.7 years. As presented in Table 1, individuals diagnosed with pneumonia were more likely to be men and exhibit any of the chronic comorbidities examined.

Table 1 Characteristics of the study sample for selected variables, Mexico 2020–2023.
Full size table

The overall risk of developing pneumonia was 9.3% (95% CI 9.2–9.4%), with sex-stratified estimates of 7.0% (95% CI 6.9–7.1%) for women and 12.0% (95% CI 11.9–12.1%) for men. This sex-related difference was statistically significant (p < 0.001). In the analysis stratified by the dominant SARS-CoV-2 genotypes, the pneumonia risks were as follows: wild-type, 21.9% (95% CI 21.7–22.0%); B.1.1.519, 12.7% (95% CI 12.6–12.8%); B.1.617.2 (Delta), 16.0% (95% CI 13.4–19.0%); B.1.1.529 (Omicron), 2.5% (95% CI 2.4–2.6%); BQ1, 2.2% (95% CI 2.2–2.3%); and XBB.1.5, 2.0% (95% CI 1.9–2.1%). As show in Fig. 1, when compared to women, the risk of pneumonia was consistently higher in men during all phases of the pandemic emergency and during the endemic phase when the XBB.1.5 subvariant was dominant (p < 0.001 in all intervals).

Fig. 1
figure 1

Sex-stratified risk of developing pneumonia according to the dominant SARS-CoV-2 variant at the date of symptom onset, Mexico 2020–2023.

Full size image

In the multiple analysis (Table 2), it was observed that when compared to women, men exhibited a 3.3% increased risk of developing pneumonia (RR = 1.033, 95% CI 1.032–1.034). This estimate was adjusted for the dominant SARS-CoV-2 variant and other patient characteristics associated with more severe COVID-19. Among these characteristics, a personal history of chronic kidney disease was found to confer the highest increase in the risk of pneumonia (RR = 1.166, 95% CI 1.162–1.170).

Table 2 Factors associated with the pneumonia risk in COVID-19, Mexico 2020–2023.
Full size table

The η2 values for all the coefficients presented in Table 2 were found to be below 0.043, signifying a limited individual contribution of each variable to the overall variance in pneumonia risk20. The highest η2 estimate was associated with the dominant SARS-CoV-2 genotype (0.058). In contrast, the influence of a personal history of arterial hypertension had the lowest estimate (0.001), underlining its comparatively modest impact on pneumonia risk. The η2 for age was intermediate at 0.004, signifying a moderate contribution of age to the variance in pneumonia risk.

Discussion

The presented results suggest sex-based disparities in pneumonia risk along the entire emergency of the COVID-19 pandemic and the first months of the endemic phase of the disease. Throughout the analyzed period, men were at a higher risk of developing pneumonia (RR = 1.033, 95% CI 1.032–1.034) compared to women. The observed difference may have implications for public health strategies and clinical management, highlighting the need for tailored approaches to address the specific risk profiles of both sexes.

Similar findings have been previously reported in the literature. However, these studies have predominantly focused on analyzing specific periods within the emergency phase of the COVID-19 pandemic21,22,23 or its subsequent endemic phase24, rather than providing a comprehensive evaluation across all phases of the disease.

Sex differences in COVID-19 outcomes seem to be influenced by a complex interplay of genetic, hormonal, immunological, and other factors25. One relevant element is hormonal variation. Estrogens, which are predominantly present in women, have been linked to an array of immunoprotective effects26. These hormones can stimulate both innate and adaptive immune responses, augmenting the production of antiviral interferons and cytokines27. This could provide women with a more robust defense mechanism against viral infections, including SARS-CoV-2. Conversely, men, who have lower estrogen levels but higher levels of testosterone, might experience a comparatively subdued immune response, potentially reducing the strength of their defense against the virus8.

Another crucial determinant of sex disparities is the distinct immune responses exhibited by men and women. Published data have indicated that women tend to launch more vigorous immune responses to infections28. This heightened immune reactivity may offer a protective advantage by recognizing and eliminating pathogens more efficiently. However, it also raises the potential for an overly aggressive immune response, which can lead to immune-related pathology, such as the cytokine storm observed in severe COVID-19 cases29.

Genetic factors also play a role in shaping sex disparities in COVID-19 outcomes. The X chromosome carries immune-related genes, encompassing those responsible for encoding toll-like receptors and cytokines30. This genetic landscape may endow women with a more extensive genetic repertoire, facilitating a diverse and adaptable immune response. Furthermore, genetic variants on the X chromosome could potentially contribute to the observed variability in COVID-19 outcomes among individuals31.

The angiotensin-converting enzyme 2 (ACE2) receptor serves as the portal for SARS-CoV-2’s entry into human cells32. Published studies suggest that the expression levels of ACE2 may differ between men and women. While high ACE2 expression could potentially facilitate viral entry, low ACE2 levels might serve as a protective factor33,34. These sex specific ACE2 expression patterns could contribute to differences in viral infectivity and the severity of the ensuing disease.

In our analysis, the overall pneumonia risk was 9.3%, with a range (95% CI) from 9.2 to 9.4%. However, it is important to note that in Mexico, no population-based SARS-CoV-2 screening was conducted, and the analyzed patients were those who exhibited symptoms and underwent laboratory testing. Therefore, the computed risk may overestimate the population-based risk of pneumonia during and after the emergency phase of the COVID-19 pandemic. For instance, the risk observed in our study during the wild type emergency phase was 21.8% (\(\varvec{n}\) = 87,749/402,923), which is higher than the risk documented in the Chinese population early during the same period (16.7%)35.

The increased risk of pneumonia in men during the pandemic, including the current endemic phase, highlights the necessity for a targeted approach. Public health campaigns and educational resources should be designed to raise awareness of this heightened risk among men and individuals with comorbid conditions, encouraging them to seek healthcare promptly if they exhibit COVID-19 symptoms36.

We must highlight the importance of vaccination and booster programs as crucial preventive measures against severe COVID-19 outcomes, including pneumonia37. Policies promoting high vaccination coverage and timely booster shots for all individuals, regardless of patients’ sex, remain essential.

Lastly, the finding that chronic kidney disease is a significant risk factor for pneumonia in COVID-19 patients highlights the necessity for integrated healthcare management38. Strengthening chronic disease management programs, particularly for individuals with kidney disease, could potentially reduce the risk of severe COVID-19 complications, including pneumonia.

This provides data on the differential impact of COVID-19 on pneumonia risk based on sex. However, further research can expand our understanding in areas such as gaining a deeper insight into the biological mechanisms associated with the risk of pneumonia in women and men, the long-term effects of the infection, the impact of future genomic variants of SARS-CoV-2 on pneumonia risk, sex-specific preventative and treatment strategies tailored towards the higher risk population, among others.

The strengths of this study include the analysis of a database containing data from all public and private healthcare institutions across Mexico, which enhances the potential generalizability of our results to the entire country. In addition, we analyzed patients from all the emergency phases of the pandemic, as well as its early endemic transmission, providing a broad background for risk quantification.

We would also like to highlight that our study aligns with several Social Development Goals (SDGs). Understanding the differential pneumonia risk in men and women could contribute to SDG 3 (Good Health and Well-being) by informing targeted healthcare interventions. Our results emphasize the importance of data-driven healthcare planning and policy development, aligning with SDG 17 (Partnerships for the Goals). The presented results underscore the need for collaborative efforts between healthcare professionals, policymakers, and researchers to address health disparities and improve health outcomes for all populations.

The limitations of this study should also be acknowledged. First, our analysis depended on an epidemiological surveillance system that facilitated the examination of a substantial sample size but, at the same time, limited our access to certain data. Specifically, data regarding the COVID-19 vaccination status was unavailable, which hindered our ability to evaluate its impact on the pneumonia risk. It is important to note that in Mexico, the COVID-19 vaccination campaign commenced in late December 2020, primarily targeting first-line healthcare personnel39. The combined effects of vaccination efforts, increasing rates of individuals with protective levels of SARS-CoV-2 antibodies, and the implementation of non-pharmacological interventions were factors that contributed to the observed reduction in pneumonia risk.

Second, the audited publicly available government dataset does not indicate whether the patient was confirmed by RT-PCR or serological assay. However, according to normative standards13, we can be certain that all the analyzed individuals were confirmed through one of these analytical procedures. From a public health and clinical perspective, serological assays generally have good accuracy in diagnosing COVID-1940. Third, since we analyzed an already integrated and fully de-identified dataset, we were unable to perform quality control measures. Fourth, we assumed that each participant was infected by a specific COVID-19 variant based on the epidemiological week of patient-reported symptoms, which may be a source of bias.

Finally, the audited database does not specify whether patients received medical attention in private or public healthcare settings. While lower mortality rates associated with COVID-19 have been documented in private settings compared to public ones41, our analysis focused on pneumonia risk as an intermediate outcome. Discrepancies between these two healthcare environments may arise from variations in resource availability and access to timely interventions, potentially limiting the generalizability of the study findings and serving as a source of bias.

Conclusions

Our research emphasizes the role of patient’s sex in determining the risk of pneumonia among patients diagnosed with COVID-19 during both the peak of the pandemic and its early transition to an endemic state. This effect was independent of age, the dominant SARS-CoV-2 genotype, and preexisting medical conditions. Our findings suggest that patient sex may be a relevant factor to consider in healthcare planning and policy development. Given the observed differences in pneumonia risk between men and women, future strategies might benefit from focusing on targeted interventions for populations that may be at higher risk.