We analyze and quantify the ways the COVID-19 pandemic affected other causes of death in Brazil in 2020 and 2021. We decompose age-standardized mortality rate time series for 2010–2021 into three additive components: trend, seasonal, and remainder. Given the long-term trend and historical seasonal variation, we assume that most of the impact of the COVID-19 pandemic will be left in the remainder. We use a regression model to test this assumption. We decompose the contributions of COVID-19 deaths (direct effect) and those of other causes (indirect effects) to the annual change in life expectancy at birth (⁠ $e_{0}$ ⁠) from 2017 to 2021. The COVID-19 pandemic not only increased rates for other causes of death but also decreased rates for some causes. Broadly, the remainders mirror the COVID-19 pandemic waves. The direct effects of the pandemic reduced $e_{0}$ by 1.88 years in 2019–2020 and by 1.77 in 2020–2021. Indirect effects increased $e_{0}$ by 0.44 in 2019–2020 and had virtually no effect on $e_{0}$ in 2020–2021. Whether the trajectories of mortality rates and annual gains in $e_{0}$ will return to prepandemic levels and their interregional gradients depend on whether a nonnegligible number of patients who recovered from COVID-19 will suffer premature mortality.

Life expectancy at birth, COVID-19 mortality, Time-series decomposition, Life expectancy decomposition, Indirect effects of COVID-19 mortality

Introduction

The impact of the COVID-19 pandemic on human life was overwhelming. Brazil was one of the hardest-hit countries, experiencing approximately 702,000 cumulative COVID-19 deaths, the second-highest number reported (World Health Organization 2023). Between 2020 and 2021, the COVID-19 pandemic claimed approximately 638,000 lives in Brazil, accounting for 11.7% of the reported global COVID-19 deaths (Brasil 2023a; World Health Organization 2022a). Brazil's public health system, Sistema Único de Saúde – SUS, is universal, unified, and decentralized. However, in the absence of federal intervention, state and local government responses were inconsistent and poorly coordinated. As a result, COVID-19 deaths were unequally distributed across social and spatial boundaries, which was further exacerbated by socioeconomic disparities (Castro, Kim et al. 2021; Rocha et al. 2021). The COVID-19 vaccination campaign in Brazil began on January 18, 2021, and 434 million vaccine doses had been administered by April 2022 (Brasil 2023b).

The mortality burden of the COVID-19 pandemic was not limited to deaths directly attributable to COVID-19. It also included indirect effects on other causes of death (Castro et al. 2023). Indirect effects could manifest in different and contradictory ways. Specific comorbidities or conditions increase the risk of dying from COVID-19, such as cancer (Venkatesulu et al. 2021), cardiovascular and cerebrovascular diseases (Thakur et al. 2021), diabetes (Castro, Gurzenda, Macário et al. 2021; Dorjee et al. 2020), hypertension (Dorjee et al. 2020), mental and behavioral disorders (Fond et al. 2021), obesity (Yang et al. 2021), pregnancy (Zambrano et al. 2020), pulmonary diseases (Dorjee et al. 2020), and renal system diseases (Dorjee et al. 2020; Thakur et al. 2021). Therefore, competing risks (Chiang 1991; Yashin et al. 1986) between COVID-19 and these other causes of death might reduce their mortality, a phenomenon known as mortality displacement (Zeger et al. 1999). Conversely, post-COVID syndrome (known as long COVID) (Mantovani et al. 2022) has contributed to increased mortality from cancer, cardiovascular, and respiratory diseases (Uusküla et al. 2022). COVID-19 infection might lead to higher maternal mortality rates (Villar et al. 2021) and an increased burden of cardiovascular diseases (Xie et al. 2022). Additionally, the COVID-19 pandemic caused disruptions in the health system, health care access bias, and reduced demand for health services owing to fear of contracting COVID-19. These disruptions increased mortality rates from causes that are amenable to primary care, such as diabetes, leukemia, and maternal deaths (Bigoni et al. 2022; Dey and Davidson 2021; Griffin 2021; Lai et al. 2020). Finally, indirect effects on external causes are uncertain and difficult to predict. For example, COVID-19 lockdowns reduced traffic volume, the number of road traffic accidents, and the number of traffic fatalities. Nevertheless, some countries observed the opposite because of increased risky driving behaviors (Yasin et al. 2021).

Assessments of the impact of the COVID-19 pandemic on mortality include estimates of years of life lost, excess deaths, and changes in life expectancy at birth and other specific ages (e.g., 65 years). Some of these studies explored the direct (COVID-19 deaths) and indirect (non-COVID-19 deaths) mortality impacts of the COVID-19 pandemic according to age, sex, race and ethnicity, socioeconomic status, regional distribution, and mortality causes (Ackley et al. 2022; Andrasfay and Goldman 2021; Arias et al. 2021; Arias et al. 2022; Brant et al. 2020; Castro, Gurzenda, Turra et al. 2021; Castro et al. 2023; Chan et al. 2021; Cronin and Evans 2021; dos Santos et al. 2021; Guimarães et al. 2022; Iuliano et al. 2021; Jardim et al. 2022; Kelly et al. 2021; Kontopantelis et al. 2021; Lima et al. 2021; Marinho et al. 2020; Sanmarchi et al. 2021; Stokes et al. 2021; World Health Organization 2022b). Besides socioeconomic inequalities and unequal responses to the COVID-19 pandemic, the direct or indirect mortality impacts of COVID-19 might be explained by differences in health systems, COVID-19 testing, the ability to work from home, and shelter-in-place policies. Thus, detailed analyses by mortality causes are essential.

Which causes of death were most impacted by the COVID-19 pandemic in Brazil in 2020 and 2021? We perform three analyses to understand and quantify the different and paradoxical ways the COVID-19 pandemic indirectly affected other causes of death. First, we use time-series decomposition from 2010 to 2021 to describe and quantify changes in the age-standardized mortality rates of other causes of death and COVID-19. Second, we use a regression model to examine the association between the age-standardized mortality rates from other causes of death and COVID-19. Finally, we conduct life expectancy decompositions to measure the contribution of other causes of death and COVID-19 to the annual changes in life expectancy at birth from 2017 to 2021.

We consider the spatial heterogeneity of the COVID-19 pandemic in Brazil (Castro, Kim et al. 2021) and the distinct mortality pattern by age and sex. We categorize causes of death other than COVID-19 into 14 groups (Table A1; all tables and figures designated with an “A” appear in the online appendix), including 10 previously shown to be positively or negatively correlated with COVID-19, along with septicemia (Lastinger et al. 2023; Weiner-Lastinger et al. 2022), digestive system diseases (5% of total deaths in Brazil), ill-defined and unknown causes, and the remaining causes.

Our study does not elucidate every direct and indirect mechanism behind the COVID-19 pandemic. However, it assesses the pandemic's most significant mortality impacts, revealing how it has altered the historical patterns of other causes of death in Brazil.

Data and Methods

Population Data

We use midyear (July 1) population estimates and projections from the Instituto Brasileiro de Geografia e Estatística (Brazilian Institute of Geography and Statistics [IBGE]) from 2010 to 2022 for the five Brazilian regions (North, Northeast, Midwest, Southeast, and South) by five-year age groups (0–4 to 90+) and sex (male and female) (IBGE 2018). We apply cubic splines (Forsythe et al. 1977; Zeileis and Grothendieck 2005) to these midyear populations to estimate 144 midmonth populations (i.e., from January 15, 2010, to December 15, 2021) by region, age group, and sex. We compute the midmonth population totals for regions and sex by adding corresponding categories (Figure A1).

Mortality Data

We use death registers from the Mortality Information System (SIM) of the Brazilian Ministry of Health from January 1, 2010, to December 31, 2021, comprising 15,847,407 records (Brasil 2023a). We reclassify records indicating pregnancy, childbirth, and puerperium as the cause of death for males as unknown (n = 8). We exclude records with missing information on sex (n = 7,289) or age (n = 38,645) and those indicating COVID-19 as the cause of death but with a death date before 2020 (n = 6). Because the exclusion criteria overlap, the final database has 15,804,832 records. For each Brazilian region, we classify deaths by month, five-year age group, sex, and the 14 causes of death (Table A1).

COVID-19 Deaths

At the beginning of 2020, Brazilian doctors used the International Classification of Diseases (ICD-10) code U04.9—for severe acute respiratory syndrome (SARS)—to categorize most COVID-19 deaths owing to the absence of specific guidelines. In April 2020, the WHO published guidelines for the certification and classification of COVID-19 infection as a cause of death (World Health Organization 2020), including two new ICD-10 codes: U07.1 (COVID-19 virus identified) and U07.2 (COVID-19 virus not identified). In April 2020, the Brazilian Ministry of Health determined the use of the ICD-10 code B34.2 for COVID-19 deaths. Prior COVID-19 deaths categorized as U04.9, U07.1, or U07.2 were corrected to B34.2 (Brasil 2020, 2021a, 2021b). Nevertheless, the SIM database contains 9,278 records for 2021, with U04.9, U07.1, or U07.2 as the cause of death. Consequently, all records with the underlying cause codes B34.2, U04.9, U07.1, or U07.2 were reclassified as COVID-19 deaths.

Decomposition of Monthly Time Series of Mortality Rates

We calculate annualized mortality rates for 144 months (⁠ $t$ ⁠) from January 2010 to December 2021 by region (Brazil and its five regions), age group (0–4, 5–19, 20–64, and 65+), sex, and the 14 causes of death, including COVID-19 (⁠ $c$ ⁠). To ensure that the region-, age-, sex-, and cause-specific mortality rates are consistent within each region (i.e., they add to the region's midmonth crude death rate), we use the midmonth population for both sexes in each age group by region as the denominators. We use Brazil's total population age structure on July 1, 2021, to construct age-standardized and annualized time series of mortality rates by cause of death (⁠ $m_{t, c}^{s}$ ⁠).

We use seasonal-trend decomposition with Loess (STL) (Cleveland et al. 1990) to decompose each 144-month

m_{t, c}^{s}

time series into three additive components. The first component is the trend (⁠

T_{t, c}^{s}

⁠), which reflects long-term changes. The second is the seasonal component (⁠

S_{t, c}^{s}

⁠), which captures intraperiod fluctuations. The third component is the remainder (⁠

R_{t, c}^{s}

⁠), which echoes changes in the data not explained by the first two components:

m_{t, c}^{s} = T_{t, c}^{s} + S_{t, c}^{s} + R_{t, c}^{s} .

The STL method is deterministic and uses the local weighted regression (Loess) smoother technique (Cleveland and Devlin 1988; Cleveland et al. 1988) to estimate the trend and seasonal components. The remainder is calculated by subtracting the estimated trend and seasonal components from the original data. Loess is a nonparametric regression method suitable for capturing complex and nonlinear patterns. Because Loess is nonparametric, it does not assume any functional form of the model structure and does not provide statistical measures, such as significance tests.

Given the long-term trend and historical seasonal fluctuations (signal) in standardized mortality rates, we hypothesize that the COVID-19 pandemic mainly affected the remainder (noise). Besides the intrinsic advantages of Loess (e.g., capturing complex and nonlinear patterns), the STL method has two other main benefits (Cleveland et al. 1990). First, it is robust to outliers, which is essential to our hypothesis that most of the mortality impact of the COVID-19 pandemic will be reflected in the remainder. Second, it allows the seasonal component to change over time.

We set the trend window equal to the total number of observations (144 months), which helps ensure the trend is smooth and reflects only long-term variations in $m_{t, c}^{s}$ ⁠. Additionally, we assume that the season follows a yearly cycle of 12 months (Rau 2007). For COVID-19, we presume that the trend and seasonal components equal zero. Within each region, we compute the trend, seasonal, and remainder totals by age group, sex, and cause of death by adding their respective categories because the STL method and the mortality rates are additive. We conduct a sensitivity analysis to test our hypothesis that most of the mortality impact of the COVID-19 pandemic will be reflected in the remainder. Specifically, we compare the trend, seasonal, and remainder results for the 14 causes of death with (144 months) and without (the 120 months from January 2010 to December 2019) data from the two COVID-19 pandemic years.

Regression of Monthly Time Series of Remainders of Mortality Rates

We use a linear regression model for each of the 14 causes of death to statistically identify the remainders associated with the COVID-19 pandemic. Our model estimates the remainder of standardized mortality rates time series (⁠

R_{t, c}^{s, r}

⁠) for each cause of death (⁠

c

⁠), region (⁠

r

⁠), and month (⁠

t

⁠), as follows:

\begin{array}{l} R_{t, c}^{s, r} = μ_{c, r, 2010 - 19} + γ_{c, 2020} + δ_{c, 2021} \\ + \sum_{i = 1}^{2} \sum_{a = 1}^{2} β_{c, r, i, a} (R e g i o n_{r} \times S e x_{i} \times A g e_{a} \times m_{t, C O V I D - 19}^{s, r}) + ε_{t, c}^{s, r} . \end{array}

We use ordinary least-squares (OLS) to estimate region-specific fixed effects for the non-COVID-19 pandemic years (⁠ $μ_{c, r, 2010 - 19}$ ⁠) and fixed effects for the first (⁠ $γ_{c, 2020}$ ⁠) and second (⁠ $δ_{c, 2021}$ ⁠) COVID-19 pandemic years. We also estimate the marginal contributions (⁠ $β_{c, r, i, a}$ ⁠) of the interactions between the region, sex (⁠ $i$ ⁠, male or female), and age group (⁠ $a$ ⁠, 20–64 or 65+) given a one-unit change in the total COVID-19 standardized mortality rate (⁠ $m_{t, C O V I D - 19}^{s, r}$ ⁠) of that region and month. The model is robust to the additive properties of the age-, sex-, and cause-specific mortality rates within each region.

Decomposition of Changes in Life Expectancy at Birth by Causes of Death

We combine death and midyear population data to calculate five-year age group annual life tables from 2017 to 2021 by region and sex. We also estimate annual mortality rates by region, five-year age group, sex, and cause-of-death group (⁠ $m_{x, n}^{t, c}$ ⁠). To examine the contributions of mortality changes across age groups and causes of death to the total difference in life expectancy, we apply a continuous decomposition method proposed by Pollard (1982, 1988). Specifically, we compute age-, sex-, and cause-specific contributions to each region's total annual changes in life expectancy at birth (⁠ $e_{0}$ ⁠). We use discrete equations recommended by Murthy (2005) to approximate Pollard's continuous approach (see the online appendix). Because of errors inherent in discrete approximations of continuous methods, our decomposed totals do not perfectly match the observed period changes in life expectancy at birth (Murthy 2005). Nevertheless, our estimates show differences ranging from −1.6% to 0.9%, with an average difference of 0.4%.

Results

Monthly age-standardized mortality rates for all other causes of death were impacted by COVID-19 (Figure 1). Mortality rates for pregnancy, childbirth and puerperium, diabetes mellitus, and hypertension and hypertensive renal diseases increased following the pattern of the two major COVID-19 pandemic waves in 2020 and 2021. Mortality rates for mental and behavioral disorders, as well as transport accidents, reverted their trends of decline. The typical fluctuations in mortality rates for influenza, pneumonia, and chronic and other lower respiratory diseases were not observed. Additionally, mortality rates for heart disease and stroke fell sharply in 2020 but then increased in 2021. The patterns of these changes were not uniform across regions (Figure A7).

Decomposition of Monthly Time Series of Mortality Rates

We assess the effect of the COVID-19 pandemic on mortality rates from other causes by decomposing each standardized time series into trend, seasonal, and remainder components. The remainder in 2020 and 2021 reflects the two major COVID-19 pandemic waves, except for obesity and mental and behavioral disorders (Figure 2). In these two years, the trend and seasonal components do not change significantly (Figures A2 and A3).

The sensitivity analysis shows minor differences in the trend (Figure A4) and seasonal (Figure A5) components between 2010–2021 and 2010–2019. For obesity, mental and behavioral disorders, diabetes mellitus, hypertension and hypertensive renal diseases, transport accidents, and ill-defined and unknown causes, the trends from 2017 onward are slightly higher in the 2010–2021 series. Conversely, the trends for kidney and urinary system diseases, malignant neoplasms, and remaining causes from 2017 onward are slightly lower in the 2010–2021 series. In both cases, the 2010–2021 trends incorporate a small part of the increasing or decreasing mortality impacts of the COVID-19 pandemic. This finding implies that the 2010–2021 remainders of these causes of death (Figure A6) somewhat underestimate the increasing or decreasing mortality impacts of the COVID-19 pandemic. Nevertheless, these differences are minor, and the results are robust enough to support our hypothesis that most of the mortality impact of the COVID-19 pandemic is reflected in the remainder.

The COVID-19 pandemic had increasing and decreasing indirect effects on non-COVID-19 causes, particularly in 2020. However, the extent of change in the remainder varied by cause of death, as shown in Table 1 and Figure 2. Table 2 presents regional differences in the increases and decreases of the remainder for the five Brazilian regions (Figures A7–A10).

When examining the estimates for males and females (Figures A11–A14) across two major age groups, 20–64 and 65+ (Figures A15–A18), we notice a significant increase in the remainder for transport accidents and mental and behavioral disorders, particularly among men and individuals aged 20–64. Our results are consistent with findings from other countries showing increased transport accident deaths despite reduced traffic volume during the COVID-19 lockdowns. This increase is attributed to the rise in risky driving behaviors during the COVID-19 pandemic (Yasin et al. 2021).

Regarding mental and behavioral disorders, the increase initially contradicts the findings that these causes of death lead to mortality displacement during the COVID-19 pandemic (Fond et al. 2021). However, the uncertainty, anxiety, social isolation, and economic problems that arose during the COVID-19 pandemic impacted more vulnerable individuals, such as those with preexisting mental and behavioral disorders or low resilience (Sher 2020). It is unclear why the effect is negligible among women, who were more affected than men by depressive and anxiety disorders in 2020 owing to the social and economic impact of the COVID-19 pandemic (Santomauro et al. 2021).

Our results support the indirect impacts of stressors on health during disasters. Stressors include grief, fear for the safety of family and friends, and lack of access to medical care and medications. Some of these stressors persist over time, leading to long-term health effects (Ng et al. 2020; Raker et al. 2020a, 2020b).

Regression of Monthly Time Series of Remainders of Mortality Rates

We assess the impact of the COVID-19 pandemic on the remainder of other causes of death using a linear regression model for each of the 14 causes of death (Table 3). Our results show that the region-specific fixed effects in the years preceding the pandemic (⁠ $μ_{c, r, 2010 - 19}$ ⁠) are significant 35% of the time—predominantly for malignant neoplasms; influenza, pneumonia, and chronic and other lower respiratory diseases; and ill-defined and unknown causes. The fixed effects for the two COVID-19 pandemic years (⁠ $γ_{c, 2020}$ ⁠, $δ_{c, 2021}$ ⁠) are of higher magnitude and significant 78% of the time. Table 4 summarizes the cause-of-death groups with statistically significant coefficients, indicating the year for the fixed effects for the two COVID-19 pandemic years as well as the sex, region, and major age group for the interaction coefficients (⁠ $β_{c, r, i, a}$ ⁠). These results are primarily consistent with our initial assessments of increases and decreases in the remainder. For example, the remainder of septicemia decreased in 2020 and increased in 2021, and the remainder of mental and behavioral disorders increased for males aged 20–64.

Decomposition of Changes in Life Expectancy at Birth by Causes of Death

We analyze the impact of COVID-19 on life expectancy, considering COVID-19 deaths (direct effects) and other causes (indirect effects). This analysis helps us understand how much each category contributed to the annual change in life expectancy at birth. We compare the two years before the COVID-19 pandemic (2017–2018 and 2018–2019) with the first two years of the pandemic (2019–2020 and 2020–2021).

In 2020 and 2021, the COVID-19 pandemic reduced life expectancy at birth by 3.23 years, with 1.44 years lost in 2020 and 1.79 years lost in 2021 (Figure 3 and Table A4). On the one hand, our analysis shows that the direct effects of the COVID-19 pandemic were responsible for a reduction of 1.88 years in 2020 and 1.77 years in 2021. On the other hand, indirect effects increased life expectancy at birth by only 0.44 years in 2020 and had a minimal impact (of 0.02 years) in 2021. Regarding specific causes of death, our findings align with the overall trend for the remainder. Causes of death that showed substantial increases in the remainder contributed to reducing life expectancy at birth by 0.30 years in 2020 and 0.05 years in 2021. Conversely, causes of death with decreases in the remainder increased life expectancy at birth by 0.75 years in 2020 and 0.04 years in 2021.

Regarding the regions (Figure A19 and Table A5), the North experienced the largest decline in life expectancy at birth in 2020 (2.76 years), and the South saw the biggest decline in 2021 (3.32 years). In both cases, the impact was greater for males. For 2020 and 2021 combined, the Midwest had the highest reduction in life expectancy (4.30 years), whereas the Northeast had the lowest (2.41 years). The direct effects had the most significant impact on reducing life expectancy at birth in the Midwest (4.62 years) and the least impact in the Northeast (2.60 years). Conversely, indirect effects increased life expectancy in all regions. Table 5 shows the relevant changes in life expectancy at birth from indirect effects, indicating regions and specific causes of death. Additionally, major age groups contribute differently to changes in life expectancy (Table 6, Figure A20, and Table A6).

Discussion

This study examines and quantifies the different and paradoxical ways the COVID-19 pandemic indirectly impacted other causes of death in Brazil in 2020 and 2021. The results reveal that the consequences varied because of several factors, such as cause of death, age group, sex, and geographical region. Although the mortality rates from cause-of-death groups followed the pattern of the two major COVID-19 pandemic waves, the impact involved both increases and decreases. As a result, the direct effects of the COVID-19 pandemic in Brazil reduced life expectancy at birth by 1.88 years between 2019 and 2020 and by 1.77 years between 2020 and 2021. On the other hand, indirect effects increased life expectancy at birth by 0.44 between 2019 and 2020 and had little to no impact between 2020 and 2021.

This study has two major strengths. First, it thoroughly analyzes the impact of the COVID-19 pandemic on mortality in Brazil by cause of death, sex, and region. Second, the study is unique in utilizing a deterministic time-series decomposition to measure the effects of mortality shocks. The analyses use Brazil's most comprehensive dataset of mortality records, yielding reliable findings and providing valuable insights into the impact of COVID-19 on mortality in Brazil.

This study has some limitations. The first of these is primarily related to the population estimates and projections provided by IBGE. We use the 2018 IBGE revision, which will be updated with new data from Brazil's 2022 demographic census (IBGE 2023). The updated population estimates and projections might indicate that the Brazilian population was smaller than that from the 2018 IBGE revision, but it is unlikely to result in significant disparities in specific years. Therefore, errors are expected to be minor when analyzing time series of mortality rates or decomposing differences in life expectancy.

Second, death records might have issues with the completeness and classification of the cause of death. According to IBGE, deaths were underreported by approximately 1.5% between 2015 and 2019 and 1.4% in 2020, which varied by 0.8% to 3.4% across regions. However, underreporting was roughly equal across age groups and sexes (IBGE 2022a, 2022b). Also, the misclassification of COVID-19 deaths (Gill and DeJoseph 2020; Moghadas and Galvani 2021) could have led to an overestimation of specific causes of death, and this error might vary by sex and region. However, the Ministry of Health has provided guidelines for the proper ICD-10 categorization of COVID-19 deaths (Brasil 2020, 2021a, 2021b), minimizing this issue. Furthermore, the proportion of ill-defined and unknown deaths has remained at roughly 5% since 2019. Although underreporting and misclassification of death registrations are possible, data from the SIM have been widely used in studies of mortality in Brazil (Castro, Gurzenda, Turra et al. 2021; dos Santos et al. 2021; Hone et al. 2019; Kim and Castro 2020; Orellana and de Souza 2022; Rasella et al. 2013).

Some of our findings support earlier studies showing that the COVID-19 pandemic might have increased mortality from other causes of death (Arias et al. 2021, 2022; Brant et al. 2020; dos Santos et al. 2021; Guimarães et al. 2022; Jardim et al. 2022; Kontopantelis et al. 2021; Stokes et al. 2021). Such increases could be attributed to a rise in risky driving behaviors and increased transport accident deaths (Yasin et al. 2021). Additionally, individuals with preexisting mental and behavioral disorders might be more affected by anxiety and economic problems, increasing mortality rates (Sher 2020). Disaster-related stressors could indirectly affect health (Ng et al. 2020; Raker et al. 2020a, 2020b). Furthermore, the disruptions in the health care system caused by the COVID-19 pandemic might have contributed to increased mortality, especially from causes responsive to primary care (Bigoni et al. 2022; Dey and Davidson 2021; Griffin 2021; Lai et al. 2020; Nolte and McKee 2004).

Our results also corroborate studies revealing that the COVID-19 pandemic might have decreased mortality from other causes of death, such as heart disease and stroke (Brant et al. 2020; dos Santos et al. 2021; Jardim et al. 2022); influenza, pneumonia, and chronic and other lower respiratory diseases (Arias et al. 2021, 2022; dos Santos et al. 2021; Guimarães et al. 2022; Kontopantelis et al. 2021); malignant neoplasms (Arias et al. 2021, 2022; Jardim et al. 2022); and digestive, kidney, and urinary system diseases. These decreases in mortality confirm that specific medical conditions increase the risk of COVID-19 mortality (Castro, Gurzenda, Macário et al. 2021; Dorjee et al. 2020; Fond et al. 2021; Thakur et al. 2021; Venkatesulu et al. 2021; Yang et al. 2021; Zambrano et al. 2020). Therefore, our analysis suggests that mortality displacement occurred as a result of competing risks (Chiang 1991; Yashin et al. 1986) between deaths from COVID-19 and other causes (Zeger et al. 1999).

The misclassification of causes of death might explain some variations, particularly in the North or Northeast regions, which presented the largest increases or decreases in the time-series remainders. However, these areas exhibit striking inequalities, with some of Brazil's worst socioeconomic, infrastructure, and access indicators (IBGE 2021; Rocha et al. 2021). In this context, the restricted availability of resources, physicians, and intensive care unit beds is a limiting factor (Rocha et al. 2021) that contributed to the rapid spread of COVID-19 and its adverse mortality outcomes (Castro, Kim et al. 2021).

Regarding life expectancy at birth, the North and Northeast regions experienced the largest declines in 2020, whereas the Midwest and South regions showed the steepest drops in 2021. These findings support earlier research (Castro, Gurzenda, Turra et al. 2021) and reflect the regional spread of COVID-19. The South and Midwest regions experienced late virus transmission, leading to a more intensive decline in life expectancy at birth in 2021 (Castro, Kim et al. 2021). Most of the decrease in life expectancy at birth can be attributed to the direct effects of COVID-19. Therefore, as the mortality burden of the COVID-19 pandemic decreases, life expectancy at birth will progressively return to its prepandemic trajectory. However, patients who recovered from COVID-19 have a higher mortality risk than those who have not contracted the disease (Al-Aly et al. 2022; Uusküla et al. 2022; Xie et al. 2022). Some 36.9 million people have recovered from COVID-19 (World Health Organization 2023), but this number could be an underestimate because of limited testing and the lack of a self-reporting system. Therefore, a nonnegligible number could potentially suffer premature mortality, which might slow the pace of gains in life expectancy at birth and alter its interregional gradient.

Acknowledgments

This study was financed partly by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (Finance Code 001), which funds the Demography Program at the Federal University of Minas Gerais. Fernando Fernandes acknowledges support from the Brazil Office of the David Rockefeller Center for Latin American Studies, Harvard University. Cássio M. Turra acknowledges support from the Conselho Nacional de Desenvolvimento Científico e Tecnológico.

References

Ackley, C. A., Lundberg, D. J., Ma, L., Elo, I. T., Preston, S. H., & Stokes, A. C. (

2022

County-level estimates of excess mortality associated with COVID-19 in the United States

SSM–Population Health

101021

Notable Increases	Notable Decreases
Septicemia (2021)	Septicemia (2020)
Obesity	Digestive system
PGN, childbirth, and PPR	Kidney and urinary system diseases
Mental and behavioral disorders	Malignant neoplasms
Diabetes mellitus	Heart disease and stroke
HPT and hypertensive renal diseases	FLU, PNA, and chronic and other LR diseases
Transport accidents	Remaining causes
Ill-defined and unknown causes

Year/Region	Notable Increases	Notable Decreases
2020
North	Obesity; diabetes mellitus; HPT and hypertensive renal diseases; transport accidents; ill-defined and unknown causes; FLU, PNA, and chronic and other LR diseases	Malignant neoplasms, remaining causes
Northeast	Mental and behavioral disorders, diabetes mellitus, HPT and hypertensive renal diseases, transport accidents	Kidney and urinary system diseases; malignant neoplasms; FLU, PNA, and chronic and other LR diseases
Midwest	Transport accidents
Southeast	Mental and behavioral disorders, HPT and hypertensive renal diseases, transport accidents	Septicemia; digestive system diseases; kidney and urinary system diseases; malignant neoplasms; heart disease and stroke; FLU, PNA, and chronic and other LR diseases; remaining causes
South		Septicemia; heart disease and stroke; FLU, PNA, and chronic and other LR diseases; remaining causes
2021
North	Obesity; PGN, childbirth, and PPR	Malignant neoplasms
Northeast	Mental and behavioral disorders, HPT and hypertensive renal diseases	Kidney and urinary system diseases; malignant neoplasms; FLU, PNA, and chronic and other LR diseases
Midwest	Septicemia; PGN, childbirth, and PPR; transport accidents	Malignant neoplasms
Southeast	Mental and behavioral disorders, HPT and hypertensive renal diseases	Malignant neoplasms; FLU, PNA, and chronic and other LR diseases
South	PGN, childbirth, and PPR	FLU, PNA, and chronic and other LR diseases

	Septicemia	Obesity	PGN, Childbirth, and PPR	Mental and Behavioral Disorders	Digestive System Diseases	Kidney and Urinary System Diseases	Diabetes Mellitus	HPT and Hypertensive Renal Diseases	Transport Accidents	Malignant Neoplasms	Heart Disease and Stroke	FLU, PNA, and Chronic and Other LR Diseases	Ill-Defined and Unknown Causes	Remaining Causes
North × 2010–2019	–0.0167 (0.0180)	0.0106 (0.0066)	0.0006 (0.0047)	–0.0308^† (0.0160)	0.0906^** (0.0346)	0.0959^** (0.0277)	0.1297^** (0.0473)	–0.0318 (0.0307)	–0.0074 (0.0316)	0.3416^** (0.0629)	0.4766^** (0.1164)	0.2225 (0.1392)	–0.4134^** (0.0624)	0.3870^** (0.0871)
Northeast × 2010–2019	0.0341^† (0.0180)	–0.0085 (0.0066)	–0.0010 (0.0047)	–0.0428^** (0.0160)	0.0778^* (0.0346)	0.0924^** (0.0277)	–0.0100 (0.0473)	–0.1454^** (0.0307)	0.0050 (0.0316)	0.2397^** (0.0629)	0.4905^** (0.1164)	0.3821^** (0.1392)	–0.2468^** (0.0624)	0.2939^** (0.0871)
Midwest × 2010–2019	–0.0452^* (0.0180)	0.0076 (0.0066)	0.0009 (0.0047)	0.0065 (0.0160)	0.0143 (0.0346)	0.0540^† (0.0277)	0.0739 (0.0473)	–0.0602^† (0.0307)	–0.0521^† (0.0316)	0.2655^** (0.0629)	–0.0930 (0.1164)	0.5462^** (0.1392)	–0.1676^** (0.0624)	0.1739^* (0.0871)
Southeast × 2010–2019	0.0228 (0.0180)	–0.0125^† (0.0066)	–0.0006 (0.0047)	–0.0434^** (0.0160)	0.0080 (0.0346)	0.0930^** (0.0277)	–0.1438^** (0.0473)	–0.0684^* (0.0307)	–0.0179 (0.0316)	0.1628^** (0.0629)	0.1578 (0.1164)	0.3768^** (0.1392)	–0.2648^** (0.0624)	0.0689 (0.0871)
South × 2010–2019	0.0364^* (0.0180)	–0.0061 (0.0066)	0.0005 (0.0047)	–0.0366^* (0.0160)	–0.0469 (0.0346)	0.0227 (0.0277)	–0.0473 (0.0473)	–0.0556^† (0.0307)	–0.0558^† (0.0316)	0.1263^* (0.0629)	0.1222 (0.1164)	0.2762^* (0.1392)	–0.1658^** (0.0624)	0.2460^** (0.0871)
2020	–0.1407^** (0.0289)	–0.0044 (0.0106)	–0.0480^** (0.0076)	0.0705^** (0.0257)	–0.0629 (0.0556)	–0.2057^** (0.0446)	0.2372^** (0.0761)	0.4096^** (0.0495)	0.2216^** (0.0509)	–0.1297 (0.1013)	–1.0867^** (0.1873)	–1.1387^** (0.2240)	0.0324 (0.1005)	–0.4404^** (0.1401)
2021	0.3966^** (0.0363)	–0.0124 (0.0133)	–0.0303^** (0.0096)	0.0864^** (0.0322)	0.4646^** (0.0698)	0.0428 (0.0560)	–0.8329^** (0.0955)	–0.0999 (0.0621)	0.3128^** (0.0639)	–0.5113^** (0.1271)	0.5393^* (0.2351)	–2.5971^** (0.2811)	0.2087^† (0.1261)	0.4409^* (0.1758)
North × Male × 20–64 × C19SMR	0.0000 (0.0003)	0.0008^** (0.0001)	0.0001 (0.0001)	0.0006^* (0.0003)	–0.0011^* (0.0005)	0.0005 (0.0004)	0.0030^** (0.0008)	0.0017^** (0.0005)	0.0004 (0.0005)	–0.0011 (0.0010)	–0.0012 (0.0018)	0.0195^** (0.0022)	0.0059^** (0.0010)	–0.0080^** (0.0014)
Northeast × Male × 20–64 × C19SMR	–0.0004 (0.0005)	0.0006^** (0.0002)	0.0002 (0.0001)	0.0051^** (0.0004)	–0.0010 (0.0009)	–0.0002 (0.0007)	0.0051^** (0.0012)	0.0041^** (0.0008)	0.0008 (0.0008)	–0.0035^* (0.0017)	–0.0055^† (0.0031)	0.0112^** (0.0037)	0.0082^** (0.0016)	–0.0014 (0.0023)
Midwest × Male × 20–64 × C19SMR	0.0000 (0.0003)	0.0003^* (0.0001)	0.0001 (0.0001)	0.0009^** (0.0003)	0.0007 (0.0006)	0.0009^† (0.0005)	0.0023^** (0.0008)	0.0006 (0.0005)	0.0028^** (0.0005)	–0.0007 (0.0011)	0.0017 (0.0019)	0.0087^** (0.0023)	0.0016 (0.0010)	0.0023 (0.0015)
Southeast × Male × 20–64 × C19SMR	–0.0005 (0.0004)	0.0008^** (0.0001)	0.0001 (0.0001)	0.0024^** (0.0004)	–0.0007 (0.0008)	0.0007 (0.0006)	0.0039^** (0.0010)	0.0025^** (0.0007)	0.0011 (0.0007)	–0.0021 (0.0014)	–0.0053^* (0.0026)	0.0087^** (0.0031)	0.0080^** (0.0014)	0.0029 (0.0019)
South × Male × 20–64 × C19SMR	–0.0002 (0.0004)	0.0001 (0.0001)	0.0001 (0.0001)	0.0011^** (0.0003)	0.0017^* (0.0007)	0.0010^† (0.0006)	0.0029^** (0.0010)	0.0009 (0.0007)	0.0003 (0.0007)	–0.0009 (0.0014)	0.0018 (0.0025)	0.0067^* (0.0030)	0.0032^* (0.0013)	0.0027 (0.0019)
North × Female × 20–64 × C19SMR	–0.0001 (0.0003)	0.0007^** (0.0001)	0.0030^** (0.0001)	0.0000 (0.0003)	–0.0011^* (0.0005)	–0.0003 (0.0004)	0.0026^** (0.0008)	0.0011^* (0.0005)	–0.0005 (0.0005)	–0.0012 (0.0010)	–0.0014 (0.0018)	0.0126^** (0.0022)	0.0022^* (0.0010)	0.0000 (0.0014)
Northeast × Female × 20–64 × C19SMR	–0.0003 (0.0005)	0.0004^* (0.0002)	0.0027^** (0.0001)	0.0002 (0.0004)	–0.0021^* (0.0009)	–0.0004 (0.0007)	0.0045^** (0.0012)	0.0023^** (0.0008)	–0.0012 (0.0008)	–0.0030^† (0.0017)	–0.0015 (0.0031)	0.0093^* (0.0037)	0.0044^** (0.0016)	–0.0015 (0.0023)
Midwest × Female × 20–64 × C19SMR	0.0003 (0.0003)	0.0002^* (0.0001)	0.0028^** (0.0001)	0.0000 (0.0003)	–0.0007 (0.0006)	0.0006 (0.0005)	0.0015^† (0.0008)	0.0004 (0.0005)	0.0001 (0.0005)	–0.0005 (0.0011)	0.0016 (0.0019)	0.0053^* (0.0023)	0.0005 (0.0010)	0.0021 (0.0015)
Southeast × Female × 20–64 × C19SMR	–0.0007^† (0.0004)	0.0004^** (0.0001)	0.0020^** (0.0001)	0.0001 (0.0004)	–0.0016^* (0.0008)	0.0001 (0.0006)	0.0037^** (0.0010)	0.0016^* (0.0007)	–0.0007 (0.0007)	–0.0019 (0.0014)	–0.0026 (0.0026)	0.0080^** (0.0031)	0.0036^** (0.0014)	0.0014 (0.0019)
South × Female × 20–64 × C19SMR	–0.0009^* (0.0004)	0.0001 (0.0001)	0.0032^** (0.0001)	–0.0001 (0.0003)	–0.0008 (0.0007)	0.0002 (0.0006)	0.0032^** (0.0010)	0.0004 (0.0007)	–0.0005 (0.0007)	–0.0013 (0.0014)	0.0018 (0.0025)	0.0057^† (0.0030)	0.0009 (0.0013)	–0.0004 (0.0019)
North × Male × 65+ × C19SMR	–0.0001 (0.0003)	0.0003^** (0.0001)	0.0001 (0.0001)	0.0008^** (0.0003)	–0.0022^** (0.0005)	–0.0010^* (0.0004)	0.0064^** (0.0008)	0.0061^** (0.0005)	–0.0008^† (0.0005)	–0.0048^** (0.0010)	0.0028 (0.0018)	0.0223^** (0.0022)	0.0249^** (0.0010)	0.0008 (0.0014)
Northeast × Male × 65+ × C19SMR	–0.0016^** (0.0005)	0.0001 (0.0002)	0.0002 (0.0001)	0.0012^** (0.0004)	–0.0036^** (0.0009)	–0.0031^** (0.0007)	0.0081^** (0.0012)	0.0105^** (0.0008)	–0.0013 (0.0008)	–0.0068^** (0.0017)	–0.0016 (0.0031)	–0.0109^** (0.0037)	0.0119^** (0.0016)	–0.0026 (0.0023)
Midwest × Male × 65+ × C19SMR	0.0004 (0.0003)	0.0002^* (0.0001)	0.0001 (0.0001)	0.0001 (0.0003)	–0.0020^** (0.0006)	0.0007 (0.0005)	0.0021^** (0.0008)	0.0012^* (0.0005)	–0.0005 (0.0005)	–0.0035^** (0.0011)	0.0021 (0.0019)	–0.0029 (0.0023)	0.0021^* (0.0010)	–0.0030^* (0.0015)
Southeast × Male × 65+ × C19SMR	–0.0017^** (0.0004)	0.0002 (0.0001)	0.0001 (0.0001)	0.0002 (0.0004)	–0.0025^** (0.0008)	–0.0014^* (0.0006)	0.0049^** (0.0010)	0.0048^** (0.0007)	–0.0010 (0.0007)	–0.0038^** (0.0014)	–0.0077^** (0.0026)	–0.0107^** (0.0031)	0.0076^** (0.0014)	–0.0026 (0.0019)
South × Male × 65+ × C19SMR	–0.0009^* (0.0004)	0.0003^* (0.0001)	0.0001 (0.0001)	0.0000 (0.0003)	–0.0027^** (0.0007)	0.0004 (0.0006)	0.0032^** (0.0010)	0.0016^* (0.0007)	–0.0011 (0.0007)	–0.0008 (0.0014)	0.0029 (0.0025)	–0.0095^** (0.0030)	0.0049^** (0.0013)	–0.0024 (0.0019)
North × Female × 65+ × C19SMR	0.0001 (0.0003)	0.0008^** (0.0001)	0.0001 (0.0001)	0.0003 (0.0003)	–0.0020^** (0.0005)	–0.0012^** (0.0004)	0.0046^** (0.0008)	0.0054^** (0.0005)	–0.0008 (0.0005)	–0.0040^** (0.0010)	–0.0040^* (0.0018)	0.0122^** (0.0022)	0.0148^** (0.0010)	–0.0015 (0.0014)
Northeast × Female × 65+ × C19SMR	–0.0027^** (0.0005)	0.0003^† (0.0002)	0.0002 (0.0001)	0.0003 (0.0004)	–0.0042^** (0.0009)	–0.0046^** (0.0007)	0.0072^** (0.0012)	0.0098^** (0.0008)	–0.0014^† (0.0008)	–0.0078^** (0.0017)	–0.0068^* (0.0031)	–0.0432^** (0.0037)	0.0130^** (0.0016)	–0.0076^** (0.0023)
Midwest × Female × 65+ × C19SMR	0.0006^* (0.0003)	0.0002^† (0.0001)	0.0001 (0.0001)	–0.0004 (0.0003)	–0.0027^** (0.0006)	–0.0017^** (0.0005)	0.0020^* (0.0008)	0.0023^** (0.0005)	–0.0008 (0.0005)	–0.0009 (0.0011)	0.0009 (0.0019)	–0.0044^† (0.0023)	0.0038^** (0.0010)	–0.0042^** (0.0015)
Southeast × Female × 65+ × C19SMR	–0.0030^** (0.0004)	0.0003^* (0.0001)	0.0001 (0.0001)	0.0000 (0.0004)	–0.0036^** (0.0008)	–0.0036^** (0.0006)	0.0051^** (0.0010)	0.0062^** (0.0007)	–0.0010 (0.0007)	–0.0053^** (0.0014)	–0.0087^** (0.0026)	–0.0174^** (0.0031)	0.0104^** (0.0014)	–0.0046^* (0.0019)
South × Female × 65+ × C19SMR	–0.0013^** (0.0004)	0.0002 (0.0001)	0.0001 (0.0001)	–0.0002 (0.0003)	–0.0015^* (0.0007)	–0.0005 (0.0006)	0.0049^** (0.0010)	0.0026^** (0.0007)	–0.0012^† (0.0007)	–0.0008 (0.0014)	0.0039 (0.0025)	–0.0148^** (0.0030)	0.0057^** (0.0013)	–0.0036^† (0.0019)
Number of Observations	2,880	2,880	2,880	2,880	2,880	2,880	2,880	2,880	2,880	2,880	2,880	2,880	2,880	2,880

Year and Sex/Region	Positive: Increases the Remainder	Negative: Decreases the Remainder
2020 (⁠ $γ_{c, 2020}$ ⁠)	Mental and behavioral disorders, diabetes mellitus, HPT and hypertensive renal diseases, transport accidents	Septicemia; PGN, childbirth, and PPR; kidney and urinary system diseases; heart disease and stroke; FLU, PNA, and chronic and other LR diseases; remaining causes
2021 (⁠ $δ_{c, 2021}$ ⁠)	Septicemia, mental and behavioral disorders, digestive system diseases, transport accidents, heart disease and stroke, ill-defined and unknown causes, remaining causes	PGN, childbirth, and PPR; diabetes mellitus; malignant neoplasms; and FLU, PNA, and chronic and other LR diseases
Male (⁠ $β_{c, r, i, a}$ ⁠)
All regions	Mental and behavioral disorders (20–64); diabetes mellitus (20–64 and 65+); HPT and hypertensive renal diseases (65+); FLU, PNA, and chronic and other LR diseases (20–64); ill-defined and unknown causes (65+)	Digestive system diseases (65+)
North	Obesity (20–64 and 65+); mental and behavioral disorders (65+); HPT and hypertensive renal diseases (20–64); FLU, PNA, and chronic and other LR diseases (65+); ill-defined and unknown causes (20–64)	Digestive system diseases (20–64), kidney and urinary system diseases (65+), transport accidents (65+), malignant neoplasms (65+), remaining causes (20–64)
Northeast	Obesity (20–64), mental and behavioral disorders (65+), HPT and hypertensive renal diseases (20–64), ill-defined and unknown causes (20–64)	Septicemia (65+); kidney and urinary system diseases (65+); malignant neoplasms (20–64 and 65+); heart disease and stroke (20–64); FLU, PNA, and chronic and other LR diseases (65+)
Midwest	Obesity (20–64 and 65+), kidney and urinary system diseases (20–64), transport accidents (20–64)	Malignant neoplasms (65+), remaining causes (65+)
Southeast	Obesity (20–64), HPT and hypertensive renal diseases (20–64), ill-defined and unknown causes (20–64)	Septicemia (65+); kidney and urinary system diseases (65+); malignant neoplasms (65+); heart disease and stroke (20–64 and 65+); and FLU, PNA, and chronic and other LR diseases (65+)
South	Obesity (65+), digestive system diseases (20–64), kidney and urinary system diseases (20–64), ill-defined and unknown causes (20–64)	Septicemia (65+); FLU, PNA, and chronic and other LR diseases (65+)
Female (⁠ $β_{c, r, i, a}$ ⁠)
All Regions	PGN, childbirth, and PPR (20–64); diabetes mellitus (20–64 and 65+); HPT and hypertensive renal diseases (65+); FLU, PNA, and chronic and other LR diseases (20–64); ill-defined and unknown causes (65+)	Digestive system (65+)
North	Obesity (20–64 and 65+); HPT and hypertensive renal diseases (20–64); FLU, PNA, and chronic and other LR diseases (65+); ill-defined and unknown causes (20–64)	Digestive system diseases (20–64); kidney and urinary system diseases (65+); malignant neoplasms (65+); heart disease and stroke (65+)
Northeast	Obesity (20–64 and 65+); HPT and hypertensive renal diseases (20–64); ill-defined and unknown causes (20–64)	Septicemia (65+); digestive system diseases (20–64); kidney and urinary system diseases (65+); transport accidents (65+); malignant neoplasms (20–64 and 65+); heart disease and stroke (65+); FLU, PNA, and chronic and other LR diseases (65+); remaining causes (65+)
Midwest	Septicemia (65+), obesity (20–64 and 65+)	Kidney and urinary system diseases (65+); FLU, PNA, and chronic and other LR diseases (65+); remaining causes (65+)
Southeast	Obesity (20–64 and 65+), HPT and hypertensive renal diseases (20–64), ill-defined and unknown causes (20–64)	Septicemia (20–64 and 65+); digestive system diseases (20–64); kidney and urinary system diseases (65+); malignant neoplasms (65+); heart disease and stroke (65+); FLU, PNA, and chronic and other LR diseases (65+); remaining causes (65+)
South		Septicemia (20–64 and 65+); transport accidents (65+); FLU, PNA, and chronic and other LR diseases (65+); remaining causes (65+)

Year/Region	Decreases in $e_{0}$	Increases in $e_{0}$
2019–2020
All Regions	Diabetes mellitus, HPT and hypertensive renal diseases	Malignant neoplasms, heart disease and stroke
North	FLU, PNA, and chronic and other LR diseases; ill-defined and unknown causes	Remaining causes
Northeast	Mental and behavioral disorders (male), ill-defined and unknown causes, remaining causes	FLU, PNA, and chronic and other LR diseases
Midwest	Transport accidents (male)	FLU, PNA, and chronic and other LR diseases
Southeast	Transport accidents (male), ill-defined and unknown causes	FLU, PNA, and chronic and other LR diseases; remaining causes
South		FLU, PNA, and chronic and other LR diseases; remaining causes
2020–2021
All Regions		Malignant neoplasms
North		FLU, PNA, and chronic and other LR diseases
Northeast	Ill-defined and unknown causes	FLU, PNA, and chronic and other LR diseases
Midwest	Ill-defined and unknown causes	FLU, PNA, and chronic and other LR diseases
Southeast	Ill-defined and unknown causes	FLU, PNA, and chronic and other LR diseases
South	Heart disease and stroke, ill-defined and unknown causes

Year/Age Group	Decreases in $e_{0}$	Increases in $e_{0}$
2019–2020
20–64 and 65+	Diabetes mellitus; HPT and hypertensive renal diseases; FLU, PNA, and chronic and other LR diseases (North); ill-defined and unknown causes (except Midwest and South)	Malignant neoplasms, heart disease and stroke
20–64	Mental and behavioral disorders (Northeast), transport accidents (Midwest), remaining causes (Northeast and Midwest)	Remaining causes (North)
65+		FLU, PNA, and chronic and other LR diseases (except North); remaining causes (Southeast and South)
2020–2021
20–64 and 65+	Heart disease and stroke (South), ill-defined and unknown causes (except North)
20–64	Heart disease and stroke (Midwest), remaining causes (North and South)	Remaining causes (Northeast)
65+	Heart disease and stroke (Southeast)	FLU, PNA, and chronic and other LR diseases (except South)

Mortality by Cause of Death in Brazil: A Research Note on the Effects of the COVID-19 Pandemic and Contribution to Changes in Life Expectancy at Birth

Abstract

Introduction

Data and Methods

Population Data

Mortality Data

COVID-19 Deaths

Decomposition of Monthly Time Series of Mortality Rates

Regression of Monthly Time Series of Remainders of Mortality Rates

Decomposition of Changes in Life Expectancy at Birth by Causes of Death

Results

Decomposition of Monthly Time Series of Mortality Rates

Regression of Monthly Time Series of Remainders of Mortality Rates

Decomposition of Changes in Life Expectancy at Birth by Causes of Death

Discussion

Acknowledgments

References

Supplementary data

Data & Figures

Contents

Supplements

Supplementary data

References

Related

Citing articles via

Email alerts

Related Articles

Related Topics

This Feature Is Available To Subscribers Only