Skip to main content
Download PDF
Download ePub

Social epidemiology of urban COVID-19 inequalities in Latin America and Canada

International Journal for Equity in Health volume 23, Article number: 212 (2024) Cite this article

Abstract

Background

The COVID-19 pandemic has spread through pre-existing fault lines in societies, deepening structural barriers faced by precarious workers, low-income populations, and racialized communities in lower income sub-city units. Many studies have quantified the magnitude of inequalities in COVID-19 distribution within cities, but few have taken an international comparative approach to draw inferences on the ways urban epidemics are shaped by social determinants of health.

Methods

Guided by critical epidemiology, this study quantifies sub-city unit-level COVID-19 inequalities across eight of the largest metropolitan areas of Latin America and Canada. Leveraging new open-data sources, we use concentration indices to quantify income- and vulnerability-related inequalities in incidence, test positivity, and deaths over the first 125 weeks of the pandemic between January 2020 and May 2022.

Results

Our findings demonstrate that incidence, deaths, and test positivity are all less concentrated in low-income sub-city units than would be expected, with incidence ranging concentration in lower income neighbourhoods in Toronto (CI = -0.07) to concentration in higher income neighbourhoods in Mexico City (CI = 0.33). Drawing on relevant studies and evaluations of data reliability, we conclude that the best available public surveillance data for the largest cities in Latin America are likely not reliable measures of the true COVID-19 disease burden. We also identify recurring trends in the evolution of inequalities across most cities, concluding that higher income sub-city units were frequent early epicentres of COVID-19 transmission across the Latin America and Canada.

Conclusions

Just as critical epidemiology points to individuals biologically embodying the material and social conditions in which we live, it may be just as useful to think of cities reifying their material and social inequities in the form of sub-city unit-level infectious disease inequities. By shifting away from a typical vulnerability-based social determinants of health frame, policymakers could act to redress and reduce externalities stemming from sub-city unit-level income inequality through redistributive and equity-promoting policies to shift the centre of gravity of urban health inequalities before the next infectious disease epidemic occurs.

Background

The COVID-19 pandemic has spread through pre-existing fault lines in societies, exposing deeply ingrained health inequities in countries and cities throughout the world. While many studies have quantified the magnitude and evolution of inequalities in COVID-19 distribution within cities, few have taken an international comparative approach to draw inferences on the ways urban epidemics were shaped by social determinants of health. This study leverages new open-data resources to quantify sub-city unit-level inequalities in COVID-19 incidence, test positivity, and deaths in the eight largest metropolitan areas of the Americas outside the United States over the first 125 weeks of the pandemic (Tables S1-S2).

As one of the most urbanized and economically unequal regions of the world [1], the Americas provide a compelling opportunity to better understand the ways in which urban epidemics become inequitable. Since the World Health Organization (WHO) declared COVID-19 a pandemic in March 2020, the Americas has been one of the most severely affected regions in the world. By March 2021, the Pan American Health Organization (PAHO) estimated that the region accounted for 45% of the global cumulative cases and 48% of COVID-19 deaths despite representing less than 15% of the global population [2]. Seroprevalence surveys of the largest metropolitan areas between 2020 to 2022 estimated that the proportion of urban residents that had been exposed to COVID-19 ranged from 1.0% in Toronto, 11.0% in Santiago, 11.4% in Buenos Aires, 19.6% in Mexico City, 21.6% in Lima, 26.0% in São Paulo, and 26.5% in Bogotá [3,4,5,6].

Although urban areas are generally wealthier than rural areas, wide disparities in socioeconomic status (SES) and segregation into wealthy and poor sub-city units mean that cities contain a microcosm of social gradients – and inequities in health – present in the rest of the country [7,8,9]. The structural barriers faced by precarious workers, low-income populations, and racialized communities that are disproportionately located in lower income sub-city units have resulted in these populations bearing a disproportionate burden of COVID-19 [10,11,12]. The first two years of the pandemic have also shown that sub-city unit SES is frequently associated with severity of COVID-19 outcomes, with evidence of higher incidence and deaths in Belo Horizonte, Chicago, Philadelphia, New York, and Toronto, among other cities [10, 13,14,15,16].

There is now a preponderance of evidence demonstrating that the epidemic has disproportionately affected lower income sub-city units, and the Americas is no exception [17,18,19,20,21,22]. In Latin America, it is estimated that between 20–30% of the region’s population – approximately 113 million people – lives in an urban informal settlement, which have been identified as being at elevated risk to COVID-19 due to poverty, violence, and overcrowding [7, 23]. With informal work making up approximately 54% of all employment in the region – as well as more than 60% in countries like Colombia and Peru – many informal workers were impacted by the lack of employment protections and have had to continue to work in precarious conditions [24,25,26,27,28]. This has led to groups such as marginalized women and Indigenous populations being disproportionately affected by the pandemic, each of whom are more likely to live in lower income sub-city units [28,29,30]. Even in the higher income city of Toronto, sub-city units with lower income, higher proportions of racialized residents, and greater concentration of “essential workers” were found to be disproportionately affected [31, 32].

Here we present the first international comparative social epidemiology of sub-city unit-level SES-based inequalities in COVID-19 incidence, deaths, and test positivity in the eight largest metropolitan areas in Canada and Latin America. Leveraging new open-data sources, we quantify overall and longitudinal inequalities from January 1, 2020 to May 27, 2022. Guided by strategic positivism and ecosocial theory, we compare sub-city unit-level COVID-19 inequalities to probe the reliability of official surveillance data and explore similarities and differences in emergent longitudinal trends across cities.

Methods

Theoretical approaches

This analysis makes strategic use of positivist methodologies to probe complex social phenomena through a social epidemiology lens. This kind of strategic positivism is particularly justified where inequities and oppression have been justified by tools of positivism such as maps, models, data, and statistics [33]. Although we primarily conduct quantitative analyses, results are interpreted through the lens of critical epidemiology, which probes the complex, multi-pathway, cumulative mechanisms through which we biologically embody the material and social conditions in which we live [34,35,36,37]. This study aims to leverage the strengths of empirical quantitative epidemiological methods with an understanding that social arrangements of power, property, and the production and reproduction of life shape the distribution of disease.

Data collection

Case selection

We limit our study to Canada and Latin America due to the prevalence of large urban agglomerations, the presence of open-access surveillance data documenting the introduction and early dissemination of COVID-19, and to prevent duplication of analyses focused on the United States. Latin America is the most urbanized region in the world with over 80% of its inhabitants living in cities, including six cities with over 10 million inhabitants [38]. Like Latin America, over 70% of Canadians live in census metropolitan areas, and over a third live in the three largest cities of Toronto, Montréal, and Vancouver [39]. Similar to their respective countries, each of these cities also experience high levels of economic inequality with Gini indices as high as 0.61 in Rio de Janeiro and 0.57 in São Paulo (Table 1).To avoid duplicating work being conducted by several other research groups in the United States, we excluded New York and Los Angeles from the analysis [40,41,42,43]. We then selected the eight most populous metropolitan area, resulting in the inclusion of Bogotá, Buenos Aires, Lima, Mexico City, Rio de Janeiro, Santiago, São Paulo, and Toronto.

Table 1 Descriptive statistics for the eight study cities
Full size table

Outcomes

This study is made possible by open data resources that enable comparative and longitudinal analyses of COVID-19 inequities in urban areas, often collected and disseminated by municipal departments of health. Three outcome variables were collected to minimize the impacts of missing data and barriers to accessing COVID-19 tests and healthcare services. Our primary outcome is weekly incidence of COVID-19, aggregated to the smallest possible municipal subdivision (i.e., neighbourhoods, distritos, comunas, etc.; hereafter sub-city units). While incidence maximizes the amount of information available for analysis, inequitable access to COVID testing means that these data will always result in the under-estimation of SES-related inequalities [44].

To evaluate the impact of these barriers on incidence-based inequality calculations, we also collected data on deaths attributed to COVID-19 and test positivity rates at the same sub-city unit-level as COVID-19 incidence whenever possible. Test positivity rates reduce the impact of sub-city unit-level inequities in access to healthcare by considering the proportion of tests conducted in each sub-city unit that are positive for COVID-19. However, they cannot eliminate biases for marginalized populations that lacked access to testing, and this data was not available in every city. Similarly, the near-comprehensive identification of deaths occurring in urban areas in each of the sample countries should reduce biases emerging from lack of access to healthcare, but the relative infrequency of deaths reduces the amount of available data and is dependent on consistent identification of COVID-19 as a cause of death in each sub-city unit. Nevertheless, we hypothesize that inequalities will be higher for deaths and test positivity than incidence because of under ascertainment of true incidence.

We were able to identify COVID-19 incidence and death data for every city in our sample, but test positivity rates could not be identified at the sub-city unit level for Bogotá, Lima, Rio de Janeiro, São Paulo, and Toronto. Full information on data sources, date ranges available, and other information can be found in Table S1. Bogotá, Lima, Santiago, and São Paulo each recorded over 1.5 million cases over the study period, but the highest average weekly incidence (254.1 per 100,000) was recorded in Buenos Aires (Table 1). Lima recorded both the highest number of cumulative deaths (95,577) and the highest weekly death rate (8.0 per 100,000). By contrast, Toronto recorded the fewest cases (321,234) and deaths (4,214), while Mexico City recorded the lowest weekly incidence (35.4 per 100,000) and death rates (1.0 per 100,000).

Sub-city unit characteristics

To facilitate the calculation of bivariate health inequality, a measure of sub-city unit-level SES is needed. We chose median household income as our primary measure of SES due to this data being available in every sample city at the sub-city unit level and widespread international recognition as a valid measure of SES in urban contexts [45]. Income was only used to compare sub-city units within cities, so there was no need to convert to a common currency. Because household income may not capture diverse conditions of vulnerability, we also collected multidimensional poverty indices (MPIs) such as the Mexican Índice de Vulnerabilidad Social or the Brazilian Índice de Desenvolvimento Social wherever possible [46, 47]. Although each index is calculated using different metrics, the common underlying logic of measuring context-specific multidimensional poverty allows for meaningful sub-city unit comparisons. We identified household income for every city in our sample, but an MPI could not be identified for São Paulo at the sub-city unit level and only income poverty could be found for Lima at the sub-city level (Tables S2-S3).

Finally, sub-city unit population was collected for the closest available year to 2020 to calculate incidence and deaths on a per capita basis. For some cities, incidence or SES data could not be found for every sub-city unit in its greater metropolitan area. Notably, the data needed to calculate bivariate health inequalities could not be found for the metropolitan areas of Buenos Aires and Toronto, which led to the restriction of analysis to their urban cores. This resulted in our sample populations ranging from a low of 2,731,571 in Toronto to a high of 21,800,000 in Mexico City. Similarly, the smallest geographic subdivision for which data could be found resulted in a range of sub-city unit sizes from a low of 19,511 in Toronto to a high of 404,037 in Bogotá (Table 1).

Data analysis

Every city’s data was downloaded from the sources listed in Tables S1-S2 and evaluated for comparability and anomalies, after which data was aggregated from daily to weekly time intervals to mitigate the impacts of reduced testing and reporting over weekends and reduce volatility in daily reports of cases and deaths. Incidence and deaths were divided by sub-city unit population, which produced weekly population-adjusted COVID-19 incidence and death rates. Measures of health inequalities were then calculated starting on the week that each city registered at least 50 cumulative cases or 20 cumulative deaths. Absolute differences in sub-city unit health inequalities were quantified by determining the top and bottom quartile (25%) of sub-city units in each city based on income, and then calculating the population-weighted weekly average of incidence, deaths, and test positivity for these richest and poorest sub-city units.

Concentration indices (CIs) were used to quantify income- and MPI-related inequalities in COVID-19 incidence, test positivity, and deaths. These bivariate measures of inequality are commonly used in epidemiology because they are robust to differences in sub-city unit population size, measure the spectrum of inequality across the entire city’s population, and are widely used in health inequality research [48, 49]. CIs are based on concentration curves, which order the sample population along the x-axis in order of lowest to highest socioeconomic status and the cumulative distribution of the health outcome along the y-axis. Curves above the diagonal imply concentration among the poor, while curves below the diagonal imply concentration among the rich. The CI is equivalent to double the proportion of the area between the concentration curve and the line of equality, ranging between -1 if the curve is above (more common among the poor) and 1 if the curve is below (more common among the rich). As a measure of relative inequality, this measure explores the role of socioeconomic inequality as a determinant of health inequalities rather than focusing on a threshold measure such as poverty.

CIs were calculated for every city and week in which data was available, and for every combination of outcome (COVID-19 incidence, test positivity, and deaths) and SES measure (income and MPI). All data aggregation, processing, and calculations were run in Stata. We used the conindex command with sub-city unit-level population weights, and all outcomes were treated as continuous interval variables with a true zero lower limit and no adjustment for upper bounding [50, 51]. The resulting CIs were then plotted as scatter plots overlaid on absolute case counts for reference. CIs using cumulative sub-city unit-level incidence, test positivity, and test data over the entire study period were also calculated to generate cumulative concentration curves and sub-city unit-level maps using ESRI ArcGIS. We used the Jenks natural breaks classification method to create 11 divisions for incidence, deaths, and test positivity, or 10 divisions for income.

Patient and public involvement

Although patients and the public were not involved in the design and conduct of this descriptive study, we aim to widely disseminate our results to researchers and policymakers who can contribute to the work of better understanding and addressing the social determinants of health inequities among linked communities.

Results

Geographic case distribution and concentration curves

On visual analysis, the sub-city units with the highest average incidence rates are generally not concentrated in areas with the lowest household incomes for most cities (Fig. 1). In fact, there is a near complete discordance between the lowest income sub-city units and highest incidence sub-city units observed most clearly in Bogotá, Mexico City, and Lima, and to a lesser degree in Rio de Janeiro and São Paulo. By contrast, overall concordance is difficult to evaluate in Santiago and only Toronto displays the expected concordance between the lowest income and highest incidence sub-city units. These trends are generally mirrored in the distribution of sub-city unit-level death rates and test positivity rates, with several notable exceptions (Figures S1-S8). The exceptions to this concordance include differences in the distribution of test positivity and incidence rates in Mexico City (Figure S3); differences between incidence and deaths rates in Buenos Aires (Figure S2), Lima (Figure S4), Rio de Janeiro (Figure S5), and São Paulo (Figure S7); and differences between all three measures in Santiago (Figure S6).

Fig. 1
figure 1

Sub-city unit level income and COVID-19 incidence. Maps of average household income (high = green; low = red) and cumulative incidence (high = red; low = green) are presented for Bogotá, Buenos Aires, Lima, Mexico City, Rio de Janeiro, Santiago, São Paulo, and Toronto. Jenks natural breaks classification method is used to divide income into ten categories and incidence into eleven categories

Full size image

The general discordance between the lowest income and highest incidence sub-city units results in concentration curves that are either close to or below the line of equality in most cities, implying that cases are concentrated in higher-income sub-city units (Fig. 2). The only city that displays the expected association between the lowest income and highest incidence sub-city units is Toronto (CI = -0.07; Table S4). Overall, there is no significant income-related inequality measured in Buenos Aires, Santiago, and São Paulo, while we found a significant concentration among higher income sub-city units in Bogotá (CI = 0.11), Lima (CI = 0.24), Mexico City (CI = 0.33), and Rio de Janeiro (CI = 0.06).

Fig. 2
figure 2

Population adjusted concentration curves for cumulative sub-city unit-level COVID-19 incidence. Each city’s sample population is ordered along x-axis in order of increasing relative sub-city unit income percentile and the cumulative distribution of incidence is plotted along the y-axis. The red diagonal line represents a CI of 0 and 95% confidence intervals are shaded in gray

Full size image

There were some differences in income-based inequalities in death rates, which were generally more concentrated in lower income sub-city units than those based on incidence (Figures S9-S10; Table S4). Buenos Aires (CI = -0.04) and Santiago (CI = -0.07) joined Toronto (CI = -0.12) in displaying significant concentration among lower income sub-city units, Bogotá’s CI was not statistically significant, and São Paulo (CI = 0.03) displayed a small, but statistically significant concentration among higher income sub-city units (Table S4). Finally, higher test positivity rates are significantly concentrated in higher income sub-city units for Mexico City (CI = 0.05) and Buenos Aires (CI = 0.02), and among lower income sub-city units for Santiago (CI = -0.04), while test positivity rates could not be found at the sub-city unit level for other cities.

Evolution of inequalities over time

Looking beyond overall differences in COVID-19 inequalities between cities, a remarkably consistent pattern in the evolution of inequalities over time emerges in Fig. 3. In every sample city, the earliest cases are concentrated in higher income sub-city units, after which a rapid decrease in CI is observed, resulting in what is often the highest concentration in lower income sub-city units over the study period. After this rapid decrease in CI, a gradual increase is generally observed, eventually resulting in CIs that approach the early peaks of the first weeks of the epidemics. Bogotá, Lima, Santiago, São Paulo, and Toronto all experienced this generalized U-shaped evolution, while Buenos Aires only experienced a short-lived decrease in CI from May to September 2020, Mexico City’s CI remained far more stable than other cities, and Rio de Janeiro’s U-shaped evolution was punctuated by a notable dip in CI from October to December 2021. None of these trends were significantly altered by using MPI as an alternate measure of SES (Figure S11-S13).

Fig. 3
figure 3

Evolution of income-based CIs for sub-city unit level COVID-19 incidence and average incidence for richest and poorest quartile sub-city units over time. CIs are weighted by sub-city unit population and calculated starting in the week in which each city surpassed 50 cumulative cases. Positive CIs imply concentration among higher income sub-city units and negative CIs imply concentration among lower income sub-city units. Y axes for weekly incidence rates are rescaled for each city to more clearly visualize within-city trends

Full size image

These trends in relative inequalities are largely reproduced by comparing absolute differences between the richest and poorest quartile of sub-city units, with incidence in richer sub-city units outpacing that of poorer sub-city units whenever the CI rises above 0, and vice versa. However, what becomes clear is the scale of inequalities become much larger in periods of intense transmission, as compared to the relatively few reported cases in the early weeks of the epidemics. As seen in Table S5, the cities with the largest absolute differences in weekly incidence per 100,000 between these quartiles are Lima (166.6), Bogotá (77.6), and Mexico City (58.9).

Comparing the evolution of weekly cases, incidence rates, death rates, and test positivity in each city (Fig. 4) provides the richest picture of each city’s epidemic trends over the first 125 weeks of the pandemic. In Bogotá, both incidence and deaths produced a nearly identical early peak in positive CIs with a rapid decrease and gradual increase over time, although the low number of deaths between epidemic peaks between September and December 2021 do not allow for direct comparison. In Buenos Aires, deaths and test positivity largely follow incidence trends, with some discordance during the early dip in June 2020 and late rise in March 2022 observed with incidence. In Lima, there is a clear and growing divergence between relatively stable death rate CIs and increasing incidence rate CIs over the entirety of the study period. By contrast, in Mexico City, trends across the entirety of the study period are largely mirrored in incidence, deaths, and test positivity, but CIs in test positivity are almost 0.3 points lower than incidence and death CIs for the entire study period.

Fig. 4
figure 4

Evolution of incidence CIs, death rate CIs, test positivity CIs, and weekly reported cases over the sample period. All CIs are weighted by sub-city unit-level population and calculated starting in the week in which each city surpassed 50 cumulative cases or 20 cumulative deaths. These line plots are overlaid on gray bar graphs of weekly incidence

Full size image

In Rio de Janeiro, death rate CIs are more stable than incidence rate CIs, but the initial peak in March 2020 is closely mirrored in both outcomes, with divergences occurring in periods of lower incidence. In Santiago, test positivity CIs largely follow trends in incidence, with lower peaks in periods of lower incidence and after August 2021, while death rate CIs remain fairly stable throughout the study period. In São Paulo, death rate CIs and incidence rate CIs both display a very similar early peak and follow similar trends, but death rate CIs are generally higher than incidence CIs in periods of lower transmission. Lastly, in Toronto, lower death rates limit the number of weeks those CIs are available, but they generally follow incidence trends except for moderately higher CIs in May and November 2020.

Discussion

This comparative analysis of sub-city unit-level COVID-19 inequalities across the eight largest cities in Latin America and Canada produces two notable findings that require further exploration. First, incidence, deaths, and (to a lesser extent) test positivity are all far less concentrated in low-income sub-city units than would be expected. This could be due to one of several reasons, which could include our assumption that COVID-19 was concentrated in lower income sub-city units being incorrect, a problem with our analysis of imperfect real-world data, or systematic bias due to inequitable access to testing and healthcare.

Second, inequality trends over the first 125 weeks of the pandemic evolved in surprisingly similar ways across every city, with early cases highly concentrated in higher income sub-city units, after which cases spread quickly into lower income sub-city units and gradually reverted to higher income sub-city units by the end of the study period. These trends could be statistical artifacts due to early healthcare and testing access in higher income sub-city units, could point to higher income sub-city units as early COVID-19 epicentres across the American continent, or could be the result of a combination of these factors. We explore each of these possibilities in greater length and contextualize our findings using the best available evidence.

Officially reported incidence data are likely unreliable

Contrary to many of our findings, there is ample research supporting the expectation that COVID-19 cases and deaths would be concentrated in lower income sub-city units. Differences between reported cases and deaths at the sub-city unit level have been noted in several studies, including one finding that age-standardized mortality over the age of 60 was significantly higher in disadvantaged comunas in Buenos Aires [52] and another study finding a 50% higher risk of mortality in lower income sub-city units of São Paulo [20]. Separately, population-level serosurveys have found 20% higher seroprevalence in low SES households in Lima [53], as well as higher rates of statistically adjusted COVID-19 cases and deaths and six times higher seroprevalence in vulnerable municipalities of Santiago [6, 54]. These studies point to the potential for misreporting of cases at the sub-city unit level, and of the importance of conducting serosurveys to establish more reliable population-representative estimates of disease distribution. Even deaths may have been under-reported in official datasets, as deaths coded as being caused by other respiratory disease or due to cardiovascular disease may in fact have been proximally caused by COVID-19 [55].

Another potential confounding factor in our analysis is the modifiable areal unit problem, which could result in substantially different inequality outcome measures purely as a result of the size and composition of sub-city unit boundaries [56]. With average sub-city unit sizes in this analysis ranging from less than 20,000 to more than 400,000, we must acknowledge that these differences could affect overall inequality levels. However, these differences would not affect the longitudinal evolution of inequality trends and no direct association is apparent between sub-city unit size and CIs (Table S4). Furthermore, a sensitivity analysis changing the scale of sub-city units analyzed in Rio de Janeiro from 159 bairros (mean population = 39,103) to 10 áreas de planejamento (mean population = 625,645) resulted in a minor reduction in incidence- (0.06 to 0.03) and death-based (0.09 to 0.02) CIs but did little to change the evolution of inequality over time (Fig. S14).

In contrast, there are several indications that the barriers faced by lower income sub-city units in accessing healthcare and testing resources are systemically biasing case-based inequality outcomes. The most dramatic examples from this study include a 0.28 difference between incidence-based and test positivity-based CIs in Mexico City, and differences of 0.13, 0.07, and 0.07 between incidence-based and death-based CIs in Lima, Bogotá, and Santiago, respectively. Although test positivity data was only available in three cities, the number of reported weekly tests per capita was more than ten times higher in Santiago than Mexico City (Table 1), pointing to meaningful differences in access to testing between different cities. These inequities in access may have even grown over time, with divergence between test positivity and incidence increasing over time in Buenos Aires and Santiago. One study in Bogotá even suggests that the share of detected cases from June 2020 to March 2021 grew from 5 to 52% among the wealthiest residents, while the rate remained stable among the poorest residents [57].

Not only are the highest levels of discrepancy between different outcomes found among the cities with the highest concentration of incidence in higher income sub-city units, but the WHO’s P-Scores, which measure the ratio of the excess to the reported COVID-19 deaths in the corresponding country, are highly associated with increases in city-level CIs [58]. The WHO found Canada to have the most accurate data reporting of countries in our sample (P-Score = 3.8), and Toronto was the only city to display the expected concentration of both incidence and death rates in lower income sub-city units (Table S4), in spite of being the city with the lowest levels of income inequality (Table 1). In contrast, Mexico (P-Score = 41.7) and Peru (P-Score = 97.0) were found to have the least reliable data reporting, and Mexico City and Lima displayed the greatest concentration of incidence and death rates in wealthier sub-city units. Taken together, this evidence suggests that the best available public incidence data for the largest cities in Latin America are likely not reliable measures of the true COVID-19 disease burden.

Higher income sub-city units as initial epicentres

The unreliability of reported incidence data compels us to consider whether the high positive CIs observed in the first weeks of each city’s epidemic are entirely due to these biases, or if they truly point to the possibility that higher income sub-city units were initial epicentres in nearly every major city of the Americas. Inequalities in test positivity data do not reproduce the same initial peak in CI for the three cities with data, suggesting that incidence-based peaks may be accounted for by higher levels of testing in higher income sub-city units during the initial weeks of the epidemics. However, there are clear initial peaks in death-based CIs in Bogotá, Lima, Rio de Janeiro, and São Paulo, suggesting these initial peaks may truly indicate initial spread in higher income sub-city units.

It is important to note that residents living in lower income sub-city units would have been less likely to have access to COVID-19 testing, and therefore, it may have been more common to overlook COVID-19 as a cause of death in these sub-city units. Importantly, the likelihood that initial cases would have been imported by individuals more likely to be travelling internationally – and therefore more likely to be living in higher income sub-city units – lends plausibility to this apparent trend. In fact, there is evidence of this process documented in contemporaneous reports identifying early disease clusters in country clubs and high income sub-city units of Rio de Janeiro and São Paulo, high income sub-city units of Toronto, and travellers from affluent municipalities in Santiago [17, 32, 54].

Inverse equity hypothesis and implications

In many ways, the COVID-19 pandemic appears to be following a similar but accelerated inequality trajectory as the HIV/AIDS pandemic, which was observed to first affect relatively wealthy individuals until they quickly took up effective health interventions, which inverted its concentration to poor and marginalized populations [59]. One way of explaining these trends is through the inverse equity hypothesis, which posits that advantaged populations are more likely to take up new health interventions and programmes, resulting in an increase in inequities as new interventions are developed [60, 61].

In the case of COVID-19, wealthier individuals were more likely to be exposed to COVID-19 through international travel or close contact with those that had recently travelled, after which epidemics quickly spread to lower income sub-city units. This gradual shift in concentration to lower income areas can be explained by inequities in working conditions, overcrowding, underlying chronic disease, and other social determinants of health. However, concomitant changes in behaviours in higher income sub-city units are often overlooked. Individuals living in higher income sub-city units were the first to adopt social technologies that prevent the spread of COVID-19, such as remote work, reducing social contact, the use of masks, and the active use of rapid and molecular testing to know whether they had contracted COVID-19 and subsequently take appropriate precautions [54].

This integrated adoption of medical and social technologies, in combination with relative advantages in the social determinants of health, allowed higher income sub-city units across the American continent to mitigate the impacts of the COVID-19 pandemic. This also could have had the effect of systematically biasing official case records that misleadingly imply higher income sub-city units were more heavily impacted than disadvantaged urban areas. These dynamics call for further study to understand how the inverse equity hypothesis can be applied to rapidly evolving viral epidemics.

Although unreliable case data prevents direct city-to-city comparisons in this case, it seems possible that no matter which sub-city unit is first affected by the introduction of a novel infectious disease, cities have a context-dependent “centre of gravity” for inequities in disease distribution which is reached after a period of attenuation that corresponds in speed with the rate of disease transmission. Just as critical epidemiology points to individuals biologically embodying the material and social conditions in which we live, it may be just as useful to think of cities reifying their material and social inequities in the form of sub-city unit-level infectious disease inequities.

More concretely, we should also consider the social and policy implications of managing externalities related to international travel and access to scarce medical technologies. Risks related to infectious disease are often framed in terms of uncontrolled disease transmission in lower income sub-city units posing a threat to higher income sub-city units [62]. A more critical frame would consider the risks posed by frequent international travel in higher income sub-city units for vulnerable populations in lower income sub-city units, as well as higher income sub-city units’ privileged access to molecular and rapid testing resources [57]. Just as intense competition for scarce COVID-19 vaccines resulted in lack of access for many low- and middle-income countries [63], lower income sub-city units were left without access to the medical technologies that were needed to inform protective social technologies to prevent infectious disease transmission.

For policymakers and public health professionals, these results underscore the importance of proactively implementing equity-focused strategies to ensure that disadvantaged populations can access and benefit from health innovations as early as possible. Some strategies that have shown promise include strengthening primary healthcare systems in marginalized areas [64, 65], targeted early implementation of health innovations in disadvantaged areas [66], equity-driven monitoring and evaluation [67, 68], resilient licensing and procurement strategies [69,70,71], subsidies and financial support to for low-income households [72], and equity driven communications to promote uptake of medical innovations [73]. Even well-intentioned deployment of medical technologies can have unintended consequences and the best time to begin planning to reduce health inequities is well in advance of the start of a health emergency.

Strengths and limitations

Besides the challenges posed by inequities in healthcare access explored earlier in the discussion, this study is limited by the potential impacts of inequities in vaccine uptake and the lack of disaggregated analyses. Because the degree to which COVID-19 vaccines prevented disease transmission in each city is unknown, potential impacts on incidence as vaccines were rolled out in 2021 are difficult to predict, but inequities in access to COVID-19 vaccines would likely have disproportionately protected higher income sub-city units from hospitalization and death [74]. Comparing absolute levels of inequality between cities is complicated by the lack of data on greater metropolitan areas in Buenos Aires, Rio de Janeiro, São Paulo, and Toronto, and because differences in the degree of testing and place-based inequalities in testing varied greatly between cities.

In many cities, the large size of sub-city units conceals substantial heterogeneity in income and health outcomes that would become more apparent if data was available at a more granular level. Because of this limitation, these estimates likely represent an underestimation of the true scale of inequality, and this effect would be greater in cities with more pronounced sub-city unit segregation. The lack of age-, race-, and gender-stratified data is another limitation of this initial analysis, which could result in overlooking important information that would affect our understanding of these trajectories of inequity and should be explored in further study. Finally, this study of place-based health inequalities, primarily focuses on compositional rather than contextual factors as social determinants [75, 76]. Additional study to investigate space-based inequalities and sub-city unit contextual factors such as the built environment and social networks would add additional nuance to our understanding of these health inequalities.

Despite these limitations, this study is – to the best of our knowledge – the first international comparative evaluation of urban COVID-19 inequity dynamics. This comparative approach limits the detail and complexity with which analyses can adjust for context-specific idiosyncrasies but allows us to put forward more generalizable inferences across eight cities, 125 weeks, 608 sub-city units, over 280,000 deaths, and nearly 10 million COVID-19 cases.

Conclusions

This comparative study of sub-city unit-level COVID-19 incidence, test positivity, and deaths in the eight largest Latin American and Canadian cities produces evidence that official incidence data are likely unreliable due to significant under-ascertainment of disease burden in lower income sub-city units. However, remarkably similar trajectories in most cities at the start of the pandemic suggest that higher income sub-city units frequently experienced the first COVID-19 outbreaks, after which epidemics rapidly shifted to lower income sub-city units. With evidence that higher income sub-city units were frequently early epicentres of transmission and maintained privileged access to scarce testing resources, policymakers could benefit from a more critical approach than typical vulnerability-based social determinants of health frames. By doing so, redressing and reducing these externalities of income inequality could be accomplished through redistributive and equity-promoting policies to shift the centre of gravity of urban health inequalities before the next infectious disease epidemic occurs.

Data availability

All research data was obtained from open-access data portals which are detailed in Supplementary Material Tables S1-S2.

Abbreviations

CI:

Concentration index

MPI:

Multidimensional poverty index

PAHO:

Pan American Health Organization

SES:

Socioeconomic status

WHO:

World Health Organization

References

  1. United Nations, Department of Economic and Social Affairs. World Urbanization Prospects The 2018 Revision. New York: United Nations; 2019. Report No.: ST/ESA/SER.A/420.

    Book  Google Scholar 

  2. PAHO. Epidemiological Update: Coronavirus disease (COVID-19) - 11 March 2021. Washington, D.C: PAHO; 2021 Mar. Available from: https://www.paho.org/en/documents/epidemiological-update-coronavirus-disease-covid-19-11-march-2021.

  3. Public Health Ontario. COVID-19 Seroprevalence in Ontario: September 3 to October 30, 2020. 2020. Available from: https://www.publichealthontario.ca/-/media/documents/ncov/epi/2020/12/covid-19-epi-seroprevalence-in-ontario-oct-30.pdf?la=en.

  4. Levin AT, Owusu-Boaitey N, Pugh S, Fosdick BK, Zwi AB, Malani A, et al. Assessing the burden of COVID-19 in developing countries: systematic review, meta-analysis and public policy implications. BMJ Glob Health. 2022;7:e008477.

    Article  PubMed  Google Scholar 

  5. Saeed S, Drews SJ, Pambrun C, Yi Q-L, Osmond L, O’Brien SF. SARS-CoV-2 seroprevalence among blood donors after the first COVID-19 wave in Canada. Transfusion. 2021;61:862–72.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Vial P, González C, Icaza G, Ramirez-Santana M, Quezada-Gaete R, Núñez-Franz L, et al. Seroprevalence, spatial distribution, and social determinants of SARS-CoV-2 in three urban centers of Chile. BMC Infect Dis. 2022;22:99.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Corburn J, Vlahov D, Mberu B, Riley L, Caiaffa WT, Rashid SF, Slum Health: Arresting COVID-19 and Improving Well-Being in Urban Informal Settlements. J Urban Health., et al. Available from: 2020. https://doi.org/10.1007/s11524-020-00438-6. Cited 2020 May 6

  8. Lall S, Wahba S. Cities, crowding, and the coronavirus: Predicting contagion risk hotspots. World Bank Blogs. 2020. Available from: https://blogs.worldbank.org/sustainablecities/cities-crowding-and-coronavirus-predicting-contagion-risk-hotspots. Cited 2020 May 3.

  9. Bilal U, Alazraqui M, Caiaffa WT, Lopez-Olmedo N, Martinez-Folgar K, Miranda JJ, et al. Inequalities in life expectancy in six large Latin American cities from the SALURBAL study: an ecological analysis. The Lancet Planetary Health. 2019;0. Available from: https://www.thelancet.com/journals/lanplh/article/PIIS2542-5196(19)30235-9/abstract. Cited 2019 Dec 12.

  10. Bilal U, Tabb LP, Barber S, Diez Roux AV. Spatial Inequities in COVID-19 Testing, Positivity, Confirmed Cases, and Mortality in 3 U.S. Cities. Ann Intern Med. 2021; Available from: https://www.acpjournals.org/doi/https://doi.org/10.7326/M20-3936. Cited 2021 Mar 29.

  11. NYC Health. COVID-19: Data. 2020. Available from: https://www1.nyc.gov/site/doh/covid/covid-19-data.page. Cited 2020 May 3.

  12. Public Health Agency of Canada. From risk to resilience: An equity approach to COVID-19 – The Chief Public Health Officer of Canada’s Report on the State of Public Health in Canada 2020. 2020 Oct. Available from: https://www.canada.ca/en/public-health/corporate/publications/chief-public-health-officer-reports-state-public-health-canada/from-risk-resilience-equity-approach-covid-19.html.

  13. Caiaffa WT, Andrade AC de S, Sales ADF, Coelho DM. Saúde Urbana, Cidades e a Interseção de Sistemas: Panorama, Agendas, Gaps e Oportunidades Rumo à Equidade, Sustentabilidade, Resiliência e Promoção à Saúde. Rio de Janeiro: Fundação Oswaldo Cruz; 2021 Oct. Available from: https://saudeamanha.fiocruz.br/wp-content/uploads/2021/11/Caiaffa-WT-et-al_Sa%C3%BAde-urbana-cidades-e-a-interse%C3%A7%C3%A3o-de-sistemas_TD_77_final.pdf.

  14. Martins-Filho PR, de Souza Araújo AA, Quintans-Júnior LJ, Santos VS, COVID-19 fatality rates related to social inequality in Northeast Brazil: a neighbourhood-level analysis. Journal of Travel Medicine.;27. Available from: 2020. https://doi.org/10.1093/jtm/taaa128. Cited 2021 Jan 8.

  15. Sales ADF, Lopes AA dos S, Magalhães AS, Coelho DM, Oliveira ES de, Lima LL, et al. Evolução espaço temporal de Covid-19 em Belo Horizonte: um olhar para as disparidades intraurbanas nos dois primeiros anos da epidemia no município. Cidade e Pandemia: Contribuições para políticas públicas em crises sanitárias. São Paulo, SP: Instituto Pólis; 2023. Available from: https://polis.org.br/projeto/cidade-e-pandemia/.

  16. Wallace K. They’ve been called hot spots. It’s actually ‘code’ for social inequity. The Toronto Star. 2021 May 11; Available from: https://www.thestar.com/news/gta/2021/05/11/theyve-been-called-hot-spots-its-actually-code-for-social-inequity.html. Cited 2021 Dec 20.

  17. Briso CB, Phillips T. Rio’s favelas count the cost as deadly spread of Covid-19 hits city’s poor. The Guardian. 2020 Apr 25; Available from: https://www.theguardian.com/world/2020/apr/25/rio-favelas-coronavirus-brazil. Cited 2020 May 3.

  18. Fisher M, Bubola E. As Coronavirus Deepens Inequality, Inequality Worsens Its Spread. The New York Times. 2020 Mar 15; Available from: https://www.nytimes.com/2020/03/15/world/europe/coronavirus-inequality.html. Cited 2020 May 3.

  19. Office for National Statistics. Coronavirus (COVID-19) related deaths by ethnic group, England and Wales. United Kingdom; 2020 May. Available from: https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/articles/coronavirusrelateddeathsbyethnicgroupenglandandwales/2march2020to10april2020?hootPostID=b229db5cd884a4f73d5bd4fadcd8959b.

  20. Ribeiro KB, Ribeiro AF, Veras MA de SM, de Castro MC. Social inequalities and COVID-19 mortality in the city of São Paulo, Brazil. Int J Epidemiol. 2021;50:732–42.

  21. Wang Z, Tang K. Combating COVID-19: health equity matters. Nat Med. 2020;26:458–458.

    Article  CAS  PubMed  Google Scholar 

  22. Williamson EJ, Walker AJ, Bhaskaran K, Bacon S, Bates C, Morton CE, et al. Factors associated with COVID-19-related death using OpenSAFELY. Nature. 2020;584:430–6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. de Marcos I. Reimagining slums in Latin America: Lessons from the COVID-19 pandemic. Global Americans. 2021. Available from: https://globalamericans.org/reimagining-slums-in-latin-america-lessons-from-the-covid-19-pandemic/.

  24. International Labour Organization. Women and men in the informal economy: A statistical picture. Third edition. 2018 Apr. Available from: http://www.ilo.org/global/publications/books/WCMS_626831/lang--en/index.htm.

  25. Menezes-Júnior LA, Parajára MD, Silva JI, Friche AA, Andrade AC, Silva LS, et al. O trabalho terceirizado e por turnos está associado ao aumento do risco de COVID-19? Descobertas de um estudo com trabalhadores da mineração. Rev Bras Saúde Ocup. 2023;48:edepi14.

    Google Scholar 

  26. Parajára MD, Silva JI, Menezes-Júnior LA, Sales AD, Friche AA, Machado EL, et al. COVID-19: profile of Brazilian workers of a mining company and the dynamic of community cases. Rev Bras Saúde Ocup. 2023;48:edepi13.

    Google Scholar 

  27. Poirier MJP, Barraza D, Caxaj CS, Martínez AM, Hard J, Montoya F. Informality, social citizenship, and wellbeing among migrant workers in Costa Rica in the context of COVID-19. Int J Environ Res Public Health. 2022;19:6224.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Lancet T. COVID-19 in Latin America: a humanitarian crisis. The Lancet. 2020;396:1463.

    Article  Google Scholar 

  29. Bottan NL, Vera-Cossio DA, Hoffmann B. The Unequal Impact of The Coronavirus Pandemic: Evidence from Seventeen Developing Countries. Inter-American Development Bank; 2020. https://doi.org/10.18235/0002451.

  30. UN Women. The Economic Impact of COVID-19 on Women in Latin America and the Caribbean. UN Women | Americas and the Caribbean. 2020. Available from: https://lac.unwomen.org/en/noticias-y-eventos/articulos/2020/11/impacto-economico-covid-19-mujeres-america-latina-y-el-caribe.

  31. Blair A, Pan SY, Subedi R, Yang F-J, Aitken N, Steensma C. Social inequalities in COVID-19 mortality by area and individual-level characteristics in Canada, January to July/August 2020: Results from two national data integrations. CCDR. 2022;48:27–38.

    Article  PubMed  PubMed Central  Google Scholar 

  32. Mishra S, Ma H, Moloney G, Yiu KCY, Darvin D, Landsman D, et al. Increasing concentration of COVID-19 by socioeconomic determinants and geography in Toronto, Canada: an observational study. Ann Epidemiol. 2022;65:84–92.

    Article  PubMed  Google Scholar 

  33. Wyly E. Strategic Positivism. Prof Geogr. 2009;61:310–22.

    Article  Google Scholar 

  34. Krieger N. Proximal, distal, and the politics of causation: What’s level got to do with it? Am J Public Health. 2008;98:221–30.

    Article  PubMed  PubMed Central  Google Scholar 

  35. Krieger N. A glossary for social epidemiology. J Epidemiol Community Health. 2001;55:693–700.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Krieger N. Epidemiology and the People’s Health: Theory and Context. 1st ed. Oxford: Oxford University Press; 2013.

    Google Scholar 

  37. Krieger N. Methods for the scientific study of discrimination and health: an ecosocial approach. Am J Public Health. 2012;102:936–44.

    Article  PubMed  PubMed Central  Google Scholar 

  38. UN-Habitat. State of Latin American and Caribbean cities 2012: Towards a new urban transition. 2012. Available from: https://unhabitat.org/state-of-latin-american-and-caribbean-cities-2.

  39. Statistics Canada. Canada’s population estimates: Subprovincial areas, July 1, 2018. Government of Canada; 2019 Mar. Report No.: 91–214–X. Available from: https://www150.statcan.gc.ca/n1/daily-quotidien/190328/dq190328b-eng.htm.

  40. Chen JT, Krieger N. Revealing the unequal burden of COVID-19 by income, race/ethnicity, and household crowding: US county vs. ZIP code analyses. Harvard Center for Population and Development Studies; 2020 Apr. Report No.: HCPDS Working Paper Volume 19, Number 1. Available from: https://cdn1.sph.harvard.edu/wp-content/uploads/sites/1266/2020/04/HCPDS_Volume-19_No_1_20_covid19_RevealingUnequalBurden_HCPDSWorkingPaper_04212020-1.pdf.

  41. Maroko AR, Nash D, Pavilonis B. Covid-19 and Inequity: A comparative spatial analysis of New York City and Chicago hot spots. medRxiv. 2020;97(4):461–70. https://doi.org/10.1007/s11524-020-00468-0.

    Article  Google Scholar 

  42. Jay J, Bor J, Nsoesie E, Lipson SK, Jones DK, Galea S, et al. Neighborhood income and physical distancing during the COVID-19 pandemic in the U.S. medRxiv. 2020;4(12):1294–302. https://doi.org/10.1038/s41562-020-00998-2.

    Article  Google Scholar 

  43. Ransome Y, Ojikutu BO, Buchanan M, Johnston D, Neighborhood KI, Cohesion S, Inequalities in COVID-19 Diagnosis Rates by Area-Level Black, African American Racial Composition. J Urban Health. Available from: 2021. https://doi.org/10.1007/s11524-021-00532-3. Cited 2021 Mar 28.

  44. Laajaj R, Webb D, Aristizabal D, Behrentz E, Bernal R, Buitrago G, et al. Understanding how socioeconomic inequalities drive inequalities in COVID-19 infections. Sci Rep. 2022;12:8269.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  45. Poirier MJP, Grépin KA, Approaches GM, Alternatives to the Wealth Index to Measure Socioeconomic Status Using Survey Data: A Critical Interpretive Synthesis. Soc Indic Res. Available from: 2019. https://doi.org/10.1007/s11205-019-02187-9. Cited 2019 Oct 18.

  46. Dirección General de Alimentación y Desarrollo Comunitario. Índice de Vulnerabilidad Social. Unidad de Atención a Población Vulnerable; 2017. Available from: https://transparencia.info.jalisco.gob.mx/sites/default/files/%C3%8Dndice%20de%20Vulnerabilidad%20Social%202017.pdf.

  47. Prefeitura da Cidade do Rio de Janeiro. Índice de Desenvolvimento Social (IDS) por Áreas de Planejamento (AP), Regiões de Planejamento (RP), Regiões Administrativas (RA), Bairros e Favelas do Município do Rio de Janeiro. 2010. Available from: https://www.data.rio/documents/PCRJ::%C3%ADndice-de-desenvolvimento-social-ids-por-%C3%A1reas-de-planejamento-ap-regi%C3%B5es-de-planejamento-rp-regi%C3%B5es-administrativas-ra-bairros-e-favelas-do-munic%C3%ADpio-do-rio-de-janeiro-2010/about. Cited 2024 Jan 18.

  48. O’Donnell O, van Doorslaer E, Wagstaff A, Lindelow M. Analyzing health equity using household survey data; a guide to techniques and their implementation. Washington, D.C.: The World Bank; 2008.

    Google Scholar 

  49. Wagstaff A, Paci P, van Doorslaer E. On the measurement of inequalities in health. Soc Sci Med. 1991;33:545–57.

    Article  CAS  PubMed  Google Scholar 

  50. O’Donnell O, O’Neill S, Van Ourti T, Walsh B. conindex: Estimation of concentration indices. Stata J. 2016;16:112–38.

    Article  PubMed  PubMed Central  Google Scholar 

  51. StataCorp. Stata Statistical Software: Release 15. College Station, TX: StataCorp LLC; 2017.

    Google Scholar 

  52. Marconi AM, Castillo Salgado C, Sarrouf EB, Zamora RJ, Irurzun AM, Islam N. Socioeconomic inequities of COVID-19 mortality in vulnerable Comunas of the City of Buenos Aires. Sci Rep. 2023;13:13642.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Reyes-Vega MF, Soto-Cabezas MG, Cárdenas F, Martel KS, Valle A, Valverde J, et al. SARS-CoV-2 prevalence associated to low socioeconomic status and overcrowding in an LMIC megacity: A population-based seroepidemiological survey in Lima. Peru EClinicalMedicine. 2021;34:100801.

    Article  PubMed  Google Scholar 

  54. Mena GE, Martinez PP, Mahmud AS, Marquet PA, Buckee CO, Santillana M. Socioeconomic status determines COVID-19 incidence and related mortality in Santiago, Chile. Science. 2021;372:eabg5298.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  55. França EB, Ishitani LH, Abreu DMX de, Teixeira RA, Corrêa PRL, Jesus E dos S de, et al. Measuring misclassification of Covid-19 as garbage codes: Results of investigating 1,365 deaths and implications for vital statistics in Brazil. 2022; Available from: https://journals.plos.org/globalpublichealth/article?id=10.1371/journal.pgph.0000199 . Cited 2024 Sep 8.

  56. Flowerdew R, Manley DJ, Sabel CE. Neighbourhood effects on health: Does it matter where you draw the boundaries? Soc Sci Med. 2008;66:1241–55.

    Article  PubMed  Google Scholar 

  57. Laajaj R, De Los RC, Sarmiento-Barbieri I, Aristizabal D, Behrentz E, Bernal R, et al. COVID-19 spread, detection, and dynamics in Bogota. Colombia Nat Commun. 2021;12:4726.

    Article  CAS  PubMed  Google Scholar 

  58. Msemburi W, Karlinsky A, Knutson V, Aleshin-Guendel S, Chatterji S, Wakefield J. The WHO estimates of excess mortality associated with the COVID-19 pandemic. Nature. 2023;613:130–7.

    Article  CAS  PubMed  Google Scholar 

  59. Hargreaves JR, Davey C, White RG. Does the ‘ inverse equity hypothesis ’ explain how both poverty and wealth can be associated with HIV prevalence in sub-Saharan Africa ? J Epidemiol Community Health. 2013;67:526–9.

    Article  PubMed  Google Scholar 

  60. Victora CG, Vaughan JP, Barros FC, Silva AC, Tomasi E. Explaining trends in inequities: evidence from Brazilian child health studies. Lancet. 2000;356:1093–8.

    Article  CAS  PubMed  Google Scholar 

  61. Victora CG, Joseph G, Silva ICM, Maia FS, Vaughan JP, Barros FC, et al. The inverse equity hypothesis: analyses of institutional deliveries in 286 national surveys. Am J Public Health. 2018;108:464–71.

    Article  PubMed  PubMed Central  Google Scholar 

  62. Roelen K, Ackley C, Boyce P, Farina N, Ripoll S. COVID-19 in LMICs: The Need to Place Stigma Front and Centre to Its Response. Eur J Dev Res. 2020;32:1592–612.

    Article  PubMed  PubMed Central  Google Scholar 

  63. Privor-Dumm L, Excler J-L, Gilbert S, Abdool Karim SS, Hotez PJ, Thompson D, et al. Vaccine access, equity and justice: COVID-19 vaccines and vaccination. BMJ Glob Health. 2023;8:e011881.

    Article  PubMed  PubMed Central  Google Scholar 

  64. WHO Commission on Social Determinants of Health. Commission on Social Determinants of Health final report. In: World Health Organization, editor. Closing the gap in a generation: health equity through action on the social determinants of health. Geneva, Switzerland: World Health Organization, Commission on Social Determinants of Health; 2008.

    Google Scholar 

  65. Perry H, Zulliger R. How effective are community health workers? An Overview of Current Evidence with Recommendations for Strengthening Community Health Worker Programs to Accelerate Progress in Achieving the Health-related Millennium Development Goals. 2012. Available from: https://www.semanticscholar.org/paper/How-effective-are-community-health-workers-An-of-to-Perry-Zulliger/27ef205af6ea59d034b9595634e9035983ec993f. Cited 2024 Sep 15.

  66. Victora CG, Hanson K, Bryce J, Vaughan JP. Achieving universal coverage with health interventions. The Lancet. 2004;364:1541–8.

    Article  Google Scholar 

  67. Friel S, Akerman M, Hancock T, Kumaresan J, Marmot M, Melin T, et al. Addressing the social and environmental determinants of urban health equity: evidence for action and a research agenda. J Urban Health. 2011;88:860–74.

    Article  PubMed  PubMed Central  Google Scholar 

  68. Baum FE, Bégin M, Houweling TAJ, Taylor S. Changes not for the fainthearted: reorienting health care systems toward health equity through action on the social determinants of health. Am J Public Health. 2009;99:1967–74.

    Article  PubMed  PubMed Central  Google Scholar 

  69. Yamey G, Schäferhoff M, Hatchett R, Pate M, Zhao F, McDade KK. Ensuring global access to COVID-19 vaccines. The Lancet. 2020;395:1405–6.

    Article  CAS  Google Scholar 

  70. Moon S, Jambert E, Childs M, von Schoen-Angerer T. A win-win solution?: A critical analysis of tiered pricing to improve access to medicines in developing countries. Global Health. 2011;7:39.

    Article  PubMed  PubMed Central  Google Scholar 

  71. Nunn AS, Fonseca EM, Bastos FI, Gruskin S, Salomon JA. Evolution of antiretroviral drug costs in Brazil in the context of free and universal access to AIDS treatment. PLoS Med. 2007;4:e305.

    Article  PubMed  PubMed Central  Google Scholar 

  72. Marmot M. Social determinants of health inequalities. Lancet. 2005;365:1099–104.

    Article  PubMed  Google Scholar 

  73. Glanz K, Rimer BK, Viswanath K, editors. Health behavior: theory, research, and practice. 5th ed. San Francisco, Calif: Jossey-Bass; 2015. ISBN: 978-1-118-62898-0.

  74. Vélez MC. COVID-19 and vaccination in Latin America and the Caribbean: challenges, needs and opportunities. UNESCO; 2021. Report No.: MTD/SHS/2021/PI/02. Available from: https://unesdoc.unesco.org/ark:/48223/pf0000378377_eng.

  75. Macintyre S, Ellaway A, Cummins S. Place effects on health: how can we conseptualise, operationalise and measure them? Soc Sci Med. 2002;55:125–39.

    Article  PubMed  Google Scholar 

  76. Arcaya MC, Arcaya AL, Subramanian SV. Inequalities in health: definitions, concepts, and theories. Revista Panamericana de Salud Pública. 2015;1:1–12.

    Google Scholar 

Download references

Acknowledgements

We gratefully acknowledge research assistance from Kaysie Ngo and Gabriel Fezza in collecting and verifying data and conducting literature reviews in support of this study.

Funding

This work was supported by the York University Faculty of Health Junior Faculty Funds & Minor Research Grant program. The funder had no role in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the article for publication.

Author information

Authors and Affiliations

  1. Global Strategy Lab, Faculty of Health, Dahdaleh Institute for Global Health Research, York University, 4700 Keele Street, Dahdaleh Building 2120, Toronto, ON, M3J 1P3, Canada

    Mathieu JP Poirier, Andrea Morales Caceres & Tieneke E Dykstra

  2. School of Global Health, Faculty of Health, York University, 4700 Keele Street, Toronto, ON, M3J 1P3, Canada

    Mathieu JP Poirier

  3. Observatory for Urban Health in Belo Horizonte (OSUBH), Federal University of Minas Gerais, Belo Horizonte, MG, 30130‑100, Brazil

    Aline Dayrell Ferreira Sales & Waleska Teixeira Caiaffa

Authors
  1. Mathieu JP Poirier
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Andrea Morales Caceres
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Tieneke E Dykstra
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Aline Dayrell Ferreira Sales
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Waleska Teixeira Caiaffa
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Conceptualization: MJP; Data curation: MJP, AMC, TD; Formal analysis: MJP; Funding acquisition: MJP; Project administration: MJP, AMC, TD; Validation: MJP, AMC, TD, ADFS, WTC; Visualization: MJP; Roles/Writing—original draft: MJP; and Writing—review & editing: MJP, AMC, TD, ADFS, WTC.

Corresponding author

Correspondence to Mathieu JP Poirier.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Poirier, M.J., Morales Caceres, A., Dykstra, T.E. et al. Social epidemiology of urban COVID-19 inequalities in Latin America and Canada. Int J Equity Health 23, 212 (2024). https://doi.org/10.1186/s12939-024-02301-5

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12939-024-02301-5

Keywords

International Journal for Equity in Health

ISSN: 1475-9276

Contact us