key: cord-0766402-mg4djnpj
authors: Torres-Ibarra, Leticia; Basto-Abreu, Ana; Carnalla, Martha; Torres-Alvarez, Rossana; Reyes-Sanchez, Francisco; Hernández-Ávila, Juan E; Palacio-Mejia, Lina S; Alpuche-Aranda, Celia; Shamah-Levy, Teresa; Rivera, Juan A; Barrientos-Gutierrez, Tonatiuh
title: SARS-CoV-2 infection fatality rate after the first epidemic wave in Mexico
date: 2022-02-14
journal: Int J Epidemiol
DOI: 10.1093/ije/dyac015
sha: 4c49005880e6e2c014b6963601f61ca490ce75c5
doc_id: 766402
cord_uid: mg4djnpj

BACKGROUND: Estimates of SARS-CoV-2 infection fatality rates (IFRs) in developing countries remain poorly characterized. Mexico has one of the highest reported COVID-19 case-fatality rates worldwide, although available estimates do not consider serologic assessment of prior exposure nor all SARS-CoV-2-related deaths. We aimed to estimate sex- and age-specific IFRs for SARS-CoV-2 in Mexico. METHODS: The total number of people in Mexico with evidence of prior SARS-CoV-2 infection was derived from National Survey of Health and Nutrition-COVID-19 (ENSANUT 2020 Covid-19)—a nationally representative serosurvey conducted from August to November 2020. COVID-19 mortality data matched to ENSANUT’s dates were retrieved from the death-certificate registry, which captures the majority of COVID-19 deaths in Mexico, and from the national surveillance system, which covers the subset of COVID-19 deaths that were identified by the health system and were confirmed through a positive polymerase chain reaction test. We analysed differences in IFRs by urbanization and region. RESULTS: The national SARS-CoV-2 IFR was 0.47% (95% CI 0.44, 0.50) using death certificates and 0.30% (95% CI 0.28, 0.33) using surveillance-based deaths. The IFR increased with age, being close to zero at age <30 years, but increasing to 1% at ages 50–59 years in men and 60–69 years in women, and being the highest at ≥80 years for men (5.88%) and women (6.23%). Across Mexico's nine regions, Mexico City (0.99%) had the highest and the Peninsula (0.26%) the lowest certificate-based IFRs. Metropolitan areas had higher certificate-based IFR (0.63%) than rural areas (0.17%). CONCLUSION: After the first wave of the COVID-19 pandemic, the overall IFR in Mexico was comparable with those of European countries. The IFR in Mexico increased with age and was higher in men than in women. The variations in IFRs across regions and places of residence within the country suggest that structural factors related to population characteristics, pandemic containment and healthcare capabilities could have influenced lethality at the local level.

Throughout the pandemic, the estimation of deaths attributable to SARS-CoV-2 infection has remained controversial. Studies that analysed lethality during the first pandemic wave (up to August 2020) reported a case-fatality rate (CFR) ranging from 2% to 3% across different countries. 1,2 These estimates were under-ascertained both in the numerator (due to delays or failures in the registry of deaths) and in the denominator, which only considered cases of COVID-19 as registered by surveillance systems. Problems with the denominator were evidenced as soon as seroprevalence estimates became available. 3, 4 These estimates evidenced the large differences between CFRs and infection fatality rates (IFRs), driven by asymptomatic cases and the intensity of testing, that lead to case underreporting.

In March 2021, Mexico ranked second in COVID-19 CFR in the world, with 9.1 deaths per 100 COVID-19 cases, compared to 2.2% worldwide. 5 However, this reported CFR overestimates the IFR because Mexico has one of the lowest testing rates globally, undercounting the number of cases exposed to SARS-CoV-2. The surveillance system in Mexico tests 100% of severe cases and a varying proportion of cases with mild to moderate symptoms, with no coverage of asymptomatic cases or cases that did not seek or receive medical attention. The surveillance system also underestimates COVID-19 deaths, as it covers only 62% of deaths as registered in death certificates 6 ; however, considering that the extent of undercounting of cases (denominator) is greater compared with the undercounting of deaths (numerator), the CFR overestimates the IFR.

In this study, we aimed to estimate sex-and age-specific IFR for SARS-CoV-2 in Mexico. Based on serologic data from the 2020 National Survey of Health and Nutrition-COVID-19 (ENSANUT 2020 Covid- 19) , we estimated the cumulative number of people who were exposed to the SARS-CoV-2 virus in Mexico and who survived, being surveyed in the ENSANUT data collection period (August to Key Messages • The COVID-19 case-fatality rate was estimated at 9.1% early in Mexico's epidemic. This estimate was based on sentinel surveillance data and largely limited to symptomatic COVID-19 cases that were tested, thus biasing the fatality estimate upwards.

• During the study period, 145 975 people died because of COVID-19 according to death certificates; of those, 94 217 were registered in the surveillance system, which covers cases that were captured by the health system and had a positive polymerase chain reaction test.

• The infection fatality rate in Mexico was 0.47% using certificate-based deaths up to November 2020 and 0.30% using surveillance-based deaths, comparable to infection fatality rates (IFRs) observed in other countries.

• Within the country, we observed large heterogeneity of IFRs, being as high as 0.99% in Mexico City or as low as 0.26% in the Peninsula.

• The heterogeneity of infection fatality rates across regions and places of residence within the country suggests that SARS-CoV-2 lethality could have been influenced by structural factors, such as population density, hospital saturation or quality of care.

November 2020). Then, we used two sources of data to incorporate all COVID-19 deaths accumulated up to one month after the mid-point of ENSANUT's data collection by region: (i) the number of COVID-19 deaths as reported by the national surveillance system, and (ii) COVID-19 deaths as captured by death certificates in the Epidemiological and Statistical Death Registry. Finally, we analysed the IFR by urbanization level and region of the country.

To estimate the total number of people infected with SARS-CoV-2 in Mexico between the onset of the pandemic and late 2020, we obtained seroprevalence data from ENSANUT 2020 Covid-19-a survey that is representative at the national and regional levels (nine regions), and is also representative for both urban and rural areas. 7, 8 Briefly, fieldwork teams visited households and randomly selected one person from each of six potential age groups (1-4, 5-9, 10-19, 20-34, 35-49 and 50 years) to provide a blood sample. A total of 9640 serum samples (51% participation rate) for SARS-CoV-2 antibody determination were obtained between August and November 2020 (Supplementary Table S1 , available as Supplementary data at IJE online, presents all regions surveyed and collection dates). The ENSANUT 2020 Covid-19 survey did not include virologic testing for prevalent SARS-CoV-2 infection.

Seroprevalence (N-protein) was corrected by an internal validation study, which established that Roche's Elecsys assay had a 92.02% sensitivity (95% CI 88.57, 94.5) and a 99.52% specificity (95% CI 97.35, 99.92). 8 Estimates of seroprevalence were calculated using sampling weights, whereas the absolute number of people with evidence of recent or prior infection was estimated by multiplying the age-and sex-specific adjusted seroprevalence by the total size of the corresponding population group, according to the 2020 census. 9 The same procedure was used to calculate the number of seropositive people by region and urbanization strata.

Considering the challenges to generalizability conferred by the 51% participation rate, in a previous study, we conducted a sensitivity analysis of the magnitude of possible selection bias in ENSANUT 2020 Covid-19. 8 After adjustment, the national seroprevalence changed from 24.9% (95% CI 22.2, 26.7) to 23.8% (95% CI 21.3, 25.4). More information on this adjustment can be found elsewhere. 8 Because the 95% CIs overlapped, 8 we used the original estimate in this paper, without adjusting.

We used two overlapping sources of information for COVID-19 deaths: deaths registered by death certificates, which cover most COVID-19 deaths in Mexico; and deaths captured by the surveillance system, which includes the subset of COVID-19 deaths that were captured by the health system and had a positive reverse transcription polymerase chain reaction (RT-PCR) test.

Deaths in the surveillance system The Surveillance System for Viral Respiratory Disease (SISVER, by its acronym in Spanish) is the national epidemiological surveillance system for monitoring viral respiratory infections, including SARS-CoV-2, and it is managed by the Mexican Ministry of Health. It includes individuals who sought medical attention and were tested for SARS-CoV-2 in the public sector. SISVER captures 100% of hospitalized patients, but a varying proportion of outpatients (minimum 10%), depending on the resources of each monitoring healthcare facility. We included all deaths confirmed by a RT-PCR test. We used a dataset updated on 6 January 2021, to account for delays in reporting.

We used the Epidemiological and Statistical Deaths Registry (SEED, by its acronym in Spanish) for the year 2020, updated on 22 February 2021, as issued by the General Direction for Health Information of the Health Ministry. SEED is an information subsystem based on death certificates that contains information of all certified deaths by cause, place of residence, date and other sociodemographic characteristics. We selected all death records with a basic cause of death coded as U07.1 (COVID-19, virus identified) or U07.2 (COVID-19, virus not identified: clinically epidemiologically diagnosed COVID-19, probable and suspected COVID-19), according to the 10th revision of the International Classification of Disease, updated in 2020. 10 COVID-19 deaths reported by death certificates are a subset of all-cause excess mortality which includes COVID and non-COVID related deaths. Until February 2021, COVID-19-specific mortality, as reported by death certificates, represented 70% of all excess deaths in Mexico. 6 By law, every death in Mexico needs to be certified before burial or cremation 11 , thus, >90% of deaths are certified. 12

We estimated IFRs by dividing the number of surveillance or death-certificate COVID-19 deaths by the estimated total number of seropositive people in Mexico during the same time period. To match survey dates by region, we used the cumulative number of deaths that were reported 1 month after the midpoint date of the survey period at each region (Supplementary Tables S2 and S3, available as Supplementary data at IJE online), as suggested by Levin et al. 4 IFRs were then stratified by sex, age, geographical region and urbanization. Sex and age in years were selfreported for living persons in the individual questionnaire of the ENSANUT 2020 Covid-19 or (for deceased persons) retrieved from the death-certificate database or the SISVER data set. To estimate the level of urbanization, we linked the individual data of place of residence of the deceased with the database of census results by locality. Information on the population size of each locality was obtained from the Unique Catalog of Keys for Localities and Geostatistical, State, and Municipal Areas updated to 2021 and published by the National Institute of Statistics and Geography. 13 Urbanization was classified according to the number of inhabitants in each locality: rural <2500, urban 2500-100 000 and metropolitan >100 000 inhabitants. We estimated region-specific sex-and agestandardized IFRs using Mexico City as the reference. Besides, with the intent of comparing IFRs by level of urbanization, we estimated sex-and age-standardized IFRs in the three categories of urbanization with the metropolitan area as the reference. We computed 95% CIs using simulation with 100 000 random replications from a normal distribution for adjusted seroprevalence and observed deaths for each sex and age stratum. Confidence intervals were obtained from the 2.5 and 97.5% percentiles of the simulated distribution as described in Supplementary Methods S1 (available as Supplementary data at IJE online).

It is challenging to define the time lags between the onset of infection, death and death reporting. For that reason, we conducted a sensitivity analysis using the cumulative deaths registered 1 month after the end of the fieldwork in each region for both sources of deaths. We obtained IFR estimates using this 1-month lag in each sex-and agespecific stratum.

From 27 February 2020-when the epidemic began in Mexico-to the collection date of the survey in each region, 94 217 cumulative surveillance-based deaths and 145 975 certificate-based deaths were registered. Across sex and age groups, the ratio of certificate-based to surveillance-based deaths ranged from 1.1 in men <10 years of age to 1.8 in men aged 20-29 years, whereas for women it varied from 1.3 for those <10 years old to 1.7 in women >80 years old. We also found heterogeneity in the ratio of deaths across the nine geographic areas, ranging from 1.3 in Central Pacific, North Pacific and Center-North to 2.0 in Mexico City. Table 1 presents the IFR calculated using the surveillance-and certificate-based deaths. Overall, we found 94 217 deaths with the surveillance system and 145 975 with the death-certificate system in the analysed period. All of the deaths recorded by the surveillance system were also reported in the death-certificate system, since surveillance-based deaths are a subset of certificate-based deaths. IFR was 0.30% (95% CI 0.28, 0.33) with surveillancebased deaths and 0.47% for certificate-based deaths (95% CI 0.44, 0.50). Surveillance-based IFRs were 0.39% and 0.22% for men and women, respectively, compared with 0.61% in men and 0.33% in women for certificatebased IFR. We observed that IFR increased with age. Surveillance-based IFR for people <20 years of age was <0.01%, increasing to 1.03% in men and 0.46% in women aged 50-59 years, compared with 1.61% in men and 0.68% in women of the same age for certificate-based IFR. The highest IFR was observed among people aged 80 years, in whom the surveillance-based IFR was 3.71% for men (95% CI 2.34, 8.76) and 3.73% for women (95% CI 2.18, 11.56) and 5.88% for men (95% CI 3.72, 13.88) and 6.23% for women (95% CI 3.63, 19.27) with the certificate-based IFR. After stratification by urbanization, metropolitan areas had the highest IFR due to COVID-19 ( Table 3 ). The estimate was 0.40% (95% CI 0.37, 0.44) for surveillancebased IFR and 0.63% (95% CI 0.58, 0.69) for certificatebased IFR. In contrast, rural areas had the lowest IFR due to COVID-19, with a surveillance-based IFR of 0.15% (95% CI 0.14, 0.18) and a certificate-based IFR of 0.17% (95% CI 0. 15, 0.20) . Similar results were observed after sex-and age-standardization.

The surveillance-and certificate-based IFR estimates, considering deaths that occurred up to 1 month after the last survey date (rather than the survey midpoint date), were higher than the estimates from the main analysis (Supplementary Tables S1 and S2, available 

We aimed to estimate the overall and sex-and age-specific IFRs for SARS-CoV-2-related illness in Mexico, using national estimates of seroprevalence and death counts as registered in the national surveillance system and in death certificates. Overall, the COVID-19 IFR was 0.30% when using deaths based on the surveillance system and 0.47% when using death certificates; IFRs were higher for men than for women and increased with age. Urban and metropolitan areas experienced higher IFRs than rural areas, and important regional differences were observed, with the highest IFR in Mexico City and the lowest in the Peninsula.

Our study is the first to provide estimates of the COVID-19 IFR for Mexico. Prior analyses calculated that CFRs in Mexico ranged from 9.4% in April 2020 14 up to 20.2% at a major social security institution by November 2020. 15 Unlike these studies, which were conducted with administrative data and based largely on symptomatic cases, we used nationally representative data showing that one-quarter of the population (31 million people) had been infected as compared with the 847 108 cases reported in official statistics by mid-October 2020. 16 There is worldwide interest in estimating COVID-19 IFRs and identifying sources of variation. Supplementary Figure S1 (available as Supplementary data at IJE online) shows IFRs using national serosurveys conducted in 2020 from Luxembourg, Hungary, England, Spain and Brazil, in comparison with Mexico. The IFR in Mexico was smaller (0.47%) than the European estimates. Meyerowitz-Katz and Merone reported a 0.60% COVID-19 IFR in a metaanalysis based on serosurveys predominantly from Europe conducted from February to June 2020. 17 Differences across sites could be due to variations in age distribution. European countries have an average life expectancy of 81 years, whereas countries like Mexico and Brazil have life expectancies of 75 years. Therefore, with the same number of infections and healthcare capability, we would expect a higher IFR in European countries in comparison to Mexico. 18 Levin et al. estimated that 90% of the variation in IFRs across regions was attributable to age. 4 In the Levin et al. study, IFRs ranged from 0.5% to 2.7% using representative samples from four European cities and regions and 11 states of the USA, all conducted before September 2020. 4 Differences in methods, population age structures and time frames limit comparability across estimates. 4, [19] [20] [21] Comparisons could also be affected by the epidemic dynamic experienced by each country. Many European countries experienced a sharp increase in COVID-19 cases during their first wave, which overloaded health services and could have led to a higher initial IFR. In contrast, Mexico experienced a more aggressive second wave from November 2020 to February 2021 that was not captured by our analysis.

To our knowledge, most IFR studies have used surveillance-based deaths. Only the Spanish study, by Pastor-Barriuso et al., included excess deaths from all causes to estimate the IFR as an upper bound of IFR that includes both COVID-19 and non-COVID-19 deaths. 22 Compared with our death-certificate estimate (comprising only COVID-19 deaths), a higher IFR for excess deaths in Mexico would be expected using the methods of Pastor-Barriuso et al. That increase in overall IFR would be driven by the extent of non-COVID-19 deaths in Mexico, where excess deaths from all causes are 30% higher than COVID-19 deaths from death certificates. 6 Notably, we found a higher IFR in Mexico than in Spain among adults aged 40-69 years (Supplementary Figure S2 , available as Supplementary data at IJE online). The higher prevalence of chronic diseases such as diabetes and obesity at younger ages in Mexico could partially explain that difference, considering that one in four COVID-19 deaths among young adults is attributable to chronic co-morbidities. 23 Still, both studies concur on the high IFR in elderly people and provide a more encompassing estimate of the burden of the pandemic by including deaths that were not registered in the epidemiological surveillance system.

Our findings suggest that the COVID-19 IFR was higher in more urbanized regions, which could be related to population density and challenges in healthcare delivery.

With the certificate-based IFR, Mexico City experienced four times the COVID-19 mortality than the Peninsula (0.99% vs 0.26%). Mexico City has more healthcare facilities than any other region, 24 yet, during the first pandemic wave, it experienced one of the highest bed occupancies in the country (>70%). 25 The population density in Mexico City is higher (6163 inhabitants/km 2 ) than in cities in the Peninsula (from 16.1 in Campeche to 75.6 inhabitants/km 2 in Chiapas). 26 Crowded hospitals could increase the IFR by compromising the quality of care, limiting admissions or lacking the necessary resources. [27] [28] [29] According to a previous report in Mexico, inadequate availability of intensive care beds was associated with an increase of 45% in fatal outcomes. 27 Other factors could have also contributed to the high IFR observed in Mexico City. Evidence suggests that exposure to particulate matter is associated with higher COVID-19 mortality, likely related to chronic pollution exposure. 30 Additionally, overcrowding could increase the severity of COVID-19 outcomes, as a result of exposure to a higher viral load. 31 These factors could influence the IFR and explain some of the variability across regions; yet, studies analysing the causes behind IFR heterogeneity are needed.

Some limitations should be considered when interpreting our results. Estimating IFRs with the death-certificate database allowed us to capture a larger proportion of deaths than using only surveillance-based deaths. Mortality information from SEED has a delay of 3-5 months, thus we obtained the data set in February of 2021, when at least 4 months had passed since the end of ENSANUT 2020 Covid-19. Misclassification due to lack of testing could have occurred, particularly at the beginning of the pandemic; however, in April 2020, the Health Ministry published guidelines to include suspected COVID-19 cases through ICD-10 codes, 32 which should have reduced misclassification. In our sensitivity analysis, we observed a 10% variation in the IFR when including deaths that occurred 1 month after the end of ENSANUT's fieldwork, compared with the midpoint. Variation was higher in regions where fieldwork took longer to complete, where the gap between the mid and final dates of fieldwork completion was wider. This finding shows the variability in IFRs that can be introduced depending on selected dates; yet the range explored should cover most deaths, since >90% of people in Mexico died <25 days after symptom onset. Some subgroups at high risk of SARS-CoV-2 infection are not considered in ENSANUT 2020 Covid-19 since the survey does not include institutionalized people, such as those incarcerated (221 000 as of July 2021) or living in nursing homes (22 600 in 2015) . 33, 34 These groups have experienced high rates of infection and death, yet their size relative to the population is small, which could produce a small underestimation of seroprevalence. However, these deaths are represented in the numerator, since being institutionalized increases the probability of requesting and getting a death certificate. Finally, acknowledging that IFR is not a fixed metric, our results should be interpreted in the context of the first COVID-19 wave.

Having reliable epidemiological estimates to guide public policies is a fundamental step towards pandemic control. IFRs in Mexico show a different probability of death when compared with CFRs. They also show that the reported probability of dying upon infection after the first wave of COVID-19 in the country was highly heterogeneous across regions and age groups. IFR information is particularly useful to prioritize groups for nonpharmacological interventions and vaccination. Death certificates provide a more encompassing picture of COVID-19 fatality rates in Mexico, yet the administrative pathway of death certificates is long, making sentinel surveillance estimates a better option to track rapid changes in lethality. Strengthening the timeliness of death certificates through improved electronic records and simplified processes could help the country to produce complete and rapid fatality estimates. As vaccines and new drugs roll out, covering the groups with a higher risk of death upon infection will help to reduce the burden of COVID-19 and could provide much-needed relief for the health system. A closer look at the reasons behind the heterogeneity of IFR observed in the country is needed, including disparities in health system capacity and utilization, differences in the registries and the large social and economic inequalities that exist across the country.

This analysis was conducted as part of the "System for the temporospatial analysis and quantitative visualization of the general situation of health problems, resources and costs, through linkage of big data from the health sector in Mexico" project, approved by the Institutional Review Board of the National Institute of Public Health of Mexico (No. CI-1706-2020). The ENSANUT 2020 Covid-19 study was approved by the Institutional Review Board of the National Institute of Public Health of Mexico No. CI-450-2021; written informed consent was obtained from the ENSANUT 2020 Covid-19 participants.

The de-identified data from the seroprevalence study ENSANUT 2020 Covid-19 are publicly available at https://ensanut.insp.mx/ encuestas/ensanutcontinua2020/descargas.php. The COVID-19 mortality data from the national surveillance system is publicly available at https://www.gob.mx/salud/documentos/datos-abiertos-152127. Death certificate data was obtained through a data use agreement with the General Direction of Information for Health (www.dgis.salud.gob.mx). Aggregated and unidentified COVID-19 mortality data from death certificates used for this analysis are available in Torres, Leticia (2022): COVID-19 mortality data from death certificates SEED. figshare. Dataset. https://doi.org/10.6084/ m9.figshare.19134959.v1

COVID-19 case-fatality rate and demographic and socioeconomic influencers: worldwide spatial regression analysis based on country-level data

Estimation of global case fatality rate of coronavirus disease 2019 (COVID-19) using metaanalyses: comparison between calendar date and days since the outbreak of the first confirmed case

Prevalence of SARS-CoV-2 infection in India: findings from the national serosurvey

Assessing the age specificity of infection fatality rates for COVID-19: systematic review, meta-analysis, and public policy implications

COVID-19)

Bolet ın Estad ıstico Sobre el Exceso de Mortalidad por Todas las Causas Durante la Emergencia por COVID-19. Nú mero 11. Semana 6

Metodolog ıa de la Encuesta Nacional de Salud y Nutrició n 2020 sobre Covid-19

Nationally representative SARS-CoV-2 antibody prevalence estimates after the first epidemic wave in Mexico

Censo de Població n y Vivienda 2020

World Health Organization. International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD-10)-WHO Version for

Decreto por el que se reforman los art ıculos 314, 348 y 419; y se adicionan los art ıculos 348 Bis, 348 Bis 1 y 348 Bis 2 de la Ley General de Salud

Demographic and Social Statistics: Civil Registration and Vital Satatistics

Catá logo Único de Claves de Áreas Geoestad ısticas Estatales, Municipales y Localidades

Impact of comorbidities in Mexican SARS-CoV-2-positive patients: a retrospective analysis in a national cohort

Hypertension, diabetes and obesity, major risk factors for death in patients with COVID-19 in Mexico

Informe T ecnico Diario COVID-19 M exico

A systematic review and metaanalysis of published research data on COVID-19 infection fatality rates

Life Expectancy at Birth, Total (Years)

The contribution of the age distribution of cases to COVID-19 case fatality across countries: a nine-country demographic study

Factors associated with COVID-19-related death using OpenSAFELY

Understanding and interpretation of case fatality rate of coronavirus disease 2019

ENE-COVID Study Group. Infection fatality risk for SARS-CoV-2 in community dwelling population of Spain: nationwide seroepidemiological study

Fraction of COVID-19 hospitalizations and deaths attributable to chronic diseases

Subsecretar ıa de Integració n y Desarrollo del Sector Salud. Informe sobre la Salud de los Mexicanos

Inhospital mortality from severe COVID-19 in a tertiary care center in Mexico City; causes of death, risk factors and the impact of hospital saturation

Gó mez Del Corral L, Alvis-Guzmá n N. Is Colombia an example of successful

Covid-19 pandemic? A critical analysis of the epidemiological data

Virus Spread Pushes Italian Hospitals Toward Breaking Point. Bloomberg

The role of air pollution (PM and NO2) in COVID-19 spread and lethality: a systematic review

Association of poor housing conditions with COVID-19 incidence and mortality across US counties

Correcto llenado del Certificado De Defunción Muertes por Virus SARS-CoV2 (COVID-19)

Cuaderno Mensual de Informació n Estad ıstica Penitenciaria Nacional

Censo de Alojamientos de Asistencia Social (CAAS)

We are thankful to the teams who participated in the fieldwork of the ENSANUT 2020 Covid-19 at the National Institute of Public Health and to all the personnel from the Institute for Epidemiological Diagnosis and Reference who conducted the analyses of the serum samples for SARS-CoV-2 antibody testing.

None declared.

Supplementary data are available at IJE online. All authors contributed to data interpretation, critically reviewed the first draft and approved the final version of the manuscript, and agreed to be accountable for the work. All authors had full access to all the data in the study and had final responsibility for the decision to submit for publication.

This work was supported by the Mexican Government through the National Health and Nutrition Survey provision to the National Institute of Public Health and by the Centers for Disease Control and Prevention of the United States of America (grant: NU50CK000493) through the CDC-Mexico Cooperative Agreement for Surveillance, Epidemiology, and Laboratory Capacity. The "System for the temporo-spatial analysis and quantitative visualization of the general situation of health problems, resources and costs, through linkage of big data from the health sector in Mexico" project was supported by the Consejo Nacional de Ciencia y Tecnolog ıa [FORDECYT-PRONACES/1303/20]. The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.