Differences in the prevalence of risk factors for severe COVID-19 across regions of São Paulo City

IDepartamento de Medicina Preventiva, Escola Paulista de Medicina, Universidade Federal de São Paulo – São Paulo (SP), Brazil. IIHospital Israelita Albert Einstein – São Paulo (SP), Brazil. IIIDepartamento de Medicina Preventiva, Faculdade de Medicina, Universidade de São Paulo – São Paulo (SP), Brazil. Corresponding author: Beatriz Thomé. Departamento de Medicina Preventiva, Escola Paulista de Medicina, Universidade Federal de São Paulo. Rua Botucatu, 740, Vila Clementino, São Paulo, SP, Brazil. E-mail: biathome@gmail.com Conflict of interests: nothing to declare – Financial support: none. DOI: https://doi.org/10.1590/1980-549720200087


INTRODUCTION
São Paulo City stands as the epicenter of the outbreak spreaded by SARS-CoV-2 in Latin America. On July 2 nd , there were 134,984 confirmed COVID-19 cases and 7,370 deaths, roughly 10% of confirmed cases and deaths in Brazil 1 . São Paulo City is composed of 5 regions (Northern, Eastern, Southern, Midwestern and Southeastern), each with its unique socioeconomic and epidemiological characteristics.
The health systems of São Paulo are currently being challenged in the attempt to control the transmission of SARS-CoV-2 while providing adequate care, in particular to a subset of infected patients with severe disease. Among deaths due to COVID-19 in Brazil, 69% were 60 years or older and 63% had at least one of the identified clinical risk factors for severe disease, amongst which the most prevalent were cardiovascular disease and diabetes 2 . As part of response planning, it is of utmost importance to identify segments of the population that may be at risk for severe COVID-19, and describe their sociodemographic characteristics and how they are geographically distributed. The scientific community warns about the disproportionate impact of the pandemic among the population subgroups of lower socioeconomic status 3 .
In this study, we estimated the prevalence of risk factors for severe COVID-19 for people living in São Paulo City according to sociodemographic characteristics and city region using data from a population-based household survey. REV BRAS EPIDEMIOL 2020; 23: E200087

METHODS
We retrieved data from the most recent household-based survey conducted in São Paulo, ISA Capital 2015, which collected information of a representative sample of non-institutionalized residents. ISA-Capital 2015 collected the respondents' self-reported health conditions, weight and height, smoking habit, among other information. The survey was based on probabilistic sample. Two-stage sampling was done within census tracts (primary sampling unit) and households (second stage). A total of 4,043 respondents were interviewed. Data were collected with a structured questionnaire with mostly closed questions. The design, characteristics, and questionnaires of ISA-Capital 2015 have been described in detail in: https://www.prefeitura.sp.gov.br/cidade/secretarias/saude/ epidemiologia_e_informacao/isacapitalsp/.
In our analysis, we included risk factors for severe disease described in the literature and other reliable public health resources 4 , which were available in ISA-Capital survey. From the total of people interviewed we included data of 3,223 adults (≥ 18 years old) for whom information on chronic diseases and lifestyle risk factors was available. Criteria of risk for severe disease included people aged ≥ 65 years old or with a diagnosis of cardiovascular disease, diabetes, chronic respiratory disease, hypertension, (current) cancer, history of stroke, obesity (BMI ≥ 30 kg/m 2 ), current smoking, or moderate to severe asthma (defined as asthma that moderately/severely limits daily activities as per respondents). We estimated the prevalence of one or more risk factors for severe COVID-19 by sex, age, education, income, race/ethnicity and São Paulo City region.
All statistical analyses considered ISA complex multistage sampling design and were carried out using Stata 15.0 software (StataCorp, TX, USA).

RESULTS
Of the participants included in our study, 47% were men, 66% had at least secondary education, 51% were white, and 54% lived with less than a minimal wage per capita.
Prevalence of single risk factors for severe COVID-19 among older adults (≥ 65 years old) were as high as 58% (hypertension), whereas for younger adults (< 65 years old) obesity was the most prevalent risk factor (21%). In general, risk factors were more prevalent among older adults, with two exceptions: obesity and smoking were more prevalent in younger adults (data not shown).
The prevalence of one or more risk factors for severe COVID-19 was 56.4% (4.7 million) in São Paulo City ( Table 1). The proportion was lower in adults < 65 years old (51%) versus in older adults (80%). Among less educated adults, that is, those who had no formal education reported, 86% had at least one risk factor for severe COVID-19, as compared to 49% among those with university education initiated. Distribution of risk factors was similar according to income or race (Figure 1). REV BRAS EPIDEMIOL 2020; 23: E200087

DISCUSSION
We found that more than half of the population, and hence a large number of adults, presented with at least one risk factor for severe COVID-19 in São Paulo City, including those under 65. A similar estimate for worldwide risk for severe disease pointed out to a fifth of the population, but is likely to be underestimated, since the calculation did not include obesity, a highly prevalent risk factor globally 5 .
We also described how the prevalence of risk factors is unequally distributed across São Paulo City: in the Northern and Southeastern regions, these risk factors were more prevalent, which may partially explain the higher death rates, alongside with inequalities in resources for care, particularly in the Northern region. Population subgroups with lower education rates, a robust indicator of lower socioeconomic status, had higher prevalence of risk factors for severe disease. Data analysis of the first weeks of the pandemic in São Paulo City pointed out to a concentration of COVID-19 cases in the Midwestern and Southeastern regions, whereas COVID-19 deaths were concentrated in the Northern, Southeastern and Eastern regions 6 , probably also reflecting disparities in accessing the necessary health services.
As part of an effective COVID-19 response, describing the distribution of risk factors for severe disease in the population is important to identify vulnerabilities and tailor prevention and care strategies.
The present study has limitations. Risk factors were self-reported and prone to misclassification bias. In addition, ISA-Capital dates from 2015, and other risk factors (known and unknown) were not captured. Nevertheless, our findings contribute to a better understanding of the greater impact of COVID-19 in lower-resource settings and population subgroups. Response strategies will need to be tailored to address such vulnerabilities.