Introduction

People have been encouraged to stay in their homes during the COVID-19 pandemic. In Brazil, the initial social distancing strategies were implemented in late March 2020 ¹. Personal, social, and programmatic vulnerabilities ² quickly worsened in the Brazilian sexual and gender minorities (i.e., lesbian, gay, bisexual, transsexual, travesti, and related identities - LGBT+) and flagged the COVID-19 pandemic as a syndemic. A syndemic is a set of closely intertwined and mutual enhancing health problems that significantly affect the health of the most socially vulnerable groups ³. Thus, noxious social conditions and prior worse health epidemic conditions synergistically interact with COVID-19 exposure, comprising a mutually caused epidemic ⁴.

A syndemic might occur in historically neglected populations such as sexual and gender minorities, which are more prone to worsen their personal and social vulnerabilities, leading to higher disease susceptibility ² due to the structures that define the availability of resources in the health-disease process ³. Personal vulnerability includes cognitive and behavioral aspects linked to disease awareness and the possibility of change. Social vulnerability refers to the social aspects of personal vulnerability, such as political decisions and cultural barriers. Programmatic vulnerability constitutes the social level of the government, which includes the commitment to promote preventive and education actions to avoid diseases ². Sexual and gender minorities are crossed by a personal vulnerability related to minority prejudice events that might cause chronic stress and biological processes to compensate them, such as elevated blood pressure and proinflammatory cytokines ^5,6. Biological processes and higher rates of substance use ^7,8 to deal with stress are associated with a higher cardiometabolic risk during the sexual and gender minorities' lifetime ⁹ and worse health than cisgender heterosexuals ^7,10.

Prejudice events are mostly related to heteronormative sociality and lack of family support ¹¹, along with multilevel, psychological, and social stressors, including exposure to discrimination and violence ⁵. For example, one study conducted in 12 Brazilian capital municipalities in 2016 with men who have sex with men showed that 65% reported discrimination based on sexual orientation in the last 12 months, and 23.5% experienced physical violence ¹². Discrimination occurred mainly from classmates, family, and neighbors ¹². According to official violence records against sexual and gender minorities in 2015-2017, homes were the main place of violence, ranging from 54.6% in teenagers to 78.9% in older adults ¹³. Increased tweets related to family violence revealed a higher vulnerability among sexual and gender minorities during the pandemic ¹⁴. Moreover, adverse psychological distress ^15,16, including increased loneliness, social isolation, and reduced emotional support ^17,18, has increased during the COVID-19 pandemic. These adverse mental outcomes were also observed in Brazil ¹⁹.

Alongside these adverse health outcomes, the pandemic exposes the social vulnerabilities caused by inequality in Brazil ²⁰. For example, in Belo Horizonte (Minas Gerais State), the number of hospitalizations due to nonspecified and COVID-19 was higher among people living in the most deprived areas ²¹. A significant difference was also observed when comparing data on race/skin color: from March to July 2020, the standard mortality rate for white people was 115 deaths per 100,000 population, while for black people was 172 deaths per 100,000 population in the city of São Paulo ²².

Brazilian data on COVID-19 vulnerability lack information on sexual and gender minorities. Some LGBT+-related institutions and researchers provided some data, since official representative data does not account for gender identity. During the COVID-19 pandemic, 42.7% of the Brazilian sexual and gender minorities considered emotional problems the worst consequence, whereas nearly 11% considered loneliness and decreased family interactions ¹⁹. Moreover, about one quarter reported unattainability to adhere to social distancing, which is statistically associated with being non-white and having lower schooling level or income ²³.

This cross-sectional analysis aimed to identify and to better understand the factors associated with the highest COVID-19 vulnerability in the Brazilian sexual and gender minorities.

Methods

Study design and sample

The Brazilian LGBT+ Health Survey is a cross-sectional online study of individuals who identified themselves as sexual and gender minorities. The study sample comprised a convenience sample with all individuals who met the inclusion criteria and agreed to anonymously participate by an online link. Inclusion criteria were: individuals who self-declare in one of the sexual and gender minority categories, aged ≥ 18 years, living in Brazil, having Internet and computer, tablet, or smartphone access to answer the questionnaire, and understanding the questions.

Initially, the link to the survey was divulgated on social media (i.e., Facebook, Instagram, and WhatsApp), on the official website of the participating universities, face-to-face contact with students of the universities, and via radio and online press. Groups and associations of pro-sexual and gender minorities from different Brazilian regions were contacted and the study was divulgated in some primary health care units from Belo Horizonte and Rio de Janeiro to achieve more participants. The answering period of the survey was from August 19 to November 30, 2020, about five months after the national initial social distancing strategies. On August 19, new daily cases were 48,800, with 1,100 daily deaths, which means a decreasing tendency that achieved 639 daily deaths in late November ²⁴. Further details can be found elsewhere ²⁵.

The Brazilian LGBT+ Health Survey was approved by the Ethics Research Committee of the Minas Gerais Federal University (protocol 34123920.9.0000.5149). Only participants who agreed to participate (i.e., consent to participate after a brief description of the aims of the research and potential risks and benefits).

COVID-19 vulnerability

A vulnerability index previously created by an LGBT+ institution was used to measure sexual and gender minorities' personal and social COVID-19 vulnerability ¹⁹, which applied the same methodology as the social vulnerability index used by the Institute of Applied Economic Research (IPEA). Three vulnerability dimensions were included: income, COVID-19 exposure, and health. The income vulnerability dimension included two aspects: (1) having up to one minimum wage before the onset of the COVID-19 pandemic in Brazil (i.e., before March), including those without wage; and (2) affording yourself for less than one month even if you lose your income resource. In the second aspect, those with missing data and who reported receiving up to one minimum wage were also considered “vulnerable” (n = 102). Different from the original vulnerability index ¹⁹, the question about “being up to 24 years old without studying or working” was excluded because it refers to a specific age (i.e., up to 24 years old) and, therefore, does not reflect an individual vulnerability for the whole population.

The COVID-19 exposure vulnerability dimension included two aspects: (1) self-reported non-adherence to social distancing measures during the pandemic, including all participants who partially disagreed with the sentence “I respected the social distancing measures imposed by health authorities”; and (2) knowing close people previously or currently diagnosed with COVID-19. Finally, the health vulnerability dimension included two aspects: (1) exclusively using the public health care system (i.e., not having a private health insurance plan); and (2) having at least one diagnosis of a chronic condition, including diabetes, hypertension, heart disease, stroke, pulmonary disease, autoimmune disease, renal disease, or cancer.

The answers from the three dimensions were summed to create an individual vulnerability score, generating a score ranging from 1 to 6, divided into quartiles. Those in the highest quartile (i.e., score of ≥ 3) were considered “high vulnerability”, and those in other quartiles were considered “low vulnerability”. The three dimensions were also used separately, considering the highest vulnerability when the participants positively scored in both questions of each dimension.

Associated factors

Three categories of associated factors were included: gender-related, sociodemographic, and health-related characteristics.

Gender-related characteristics: sexual orientation (homosexual, bisexual, or heterosexual (considering only those transgender) and other scarce sexual orientations (i.e., asexual, pansexual, or queer), gender identity (cisgender women, cisgender men, or transgender, and other scarce gender identities (i.e., travesti or non-binary);

Sociodemographic characteristics: age groups (18-29, 30-39, 40-49, or ≥ 50 years), schooling level (complete high school, complete undergraduate education, or complete graduate education), race/skin color (non-white or white), living alone (yes or no), the mean number of people per room in the household (1 or > 1), Brazilian region (North, Northeast, Southeast, South, or Central-West), current work status (at home, as usual, or unemployed), and receiving government income support during the COVID-19 pandemic (yes or no);

Health-related characteristics: self-rated health (very good/good, fair, or very poor/poor), self-reported diagnosis of depression (yes or no), positive COVID-19 test during the pandemic (yes or no), proper facemask use during the pandemic, including all participants who totally agreed with the sentence “I properly used facemask outside the home” (yes or no), and perceiving worse mental health during the pandemic (yes or no).

Statistical analysis

Differences across the COVID-19 vulnerability categories were estimated using the Pearson's chi-square test. Logistic regression models were used to estimate the odds ratios (OR) and their 95% confidence intervals (95%CI) to assess factors associated with the highest COVID-19 vulnerability. Multivariate analyses were sequentially performed by adding blocks of characteristics in the following order: (1) gender-related characteristics; (2) sociodemographic characteristics; and (3) health-related characteristics. The fully adjusted model included only variables with p < 0.20 in the block analyses due to evidence of multicollinearity (variance inflation factor > 5). Hosmer-Lemeshow goodness-of-fit test was implemented to assess model fit after fitting the logistic regression final models. Post-stratification was used to estimate weights according to Brazilian regions, considering the population estimates of the general Brazilian population aged ≥ 18 years used in the 2019 Brazilian National Health Survey (PNS 2019). This procedure was used to enhance representativeness, since the participants' selection probability was unknown ²⁶ and the participants were concentrated in the Southeast Region. All analyses were performed using Stata 14.0 SE (https://www.stata.com).

Results

Out of 976 individuals who agreed to participate and met the inclusion criteria, 826 participants had complete information to classify the COVID-19 vulnerability and were included in our analysis. Details on the flow of original participants until inclusion in the Brazilian LGBT+ Health Survey were described elsewhere ²⁷. Table 1 describes the characteristics of the study population and according to the COVID-19 vulnerability (total and by the three aspects). The mean age was 31.8 years (± 11.2). The participants were mainly homosexual (75.7%), cisgender men (58.2%), and white (54.6%). Total COVID-19 vulnerability statistically varied according to sexual orientation, age groups, schooling level, mean number of people per room in the household, Brazilian region, and self-rated health. Regarding the separate vulnerability aspects, COVID-19 exposure showed more participants (n = 241), followed by income vulnerability (n = 221), and health vulnerability (n = 94). A higher proportion of heterosexual and other scarce sexual orientations showed income vulnerability (11.9%) than the non-vulnerable, whereas transgender and other scarce gender identities showed a lower proportion (8.8%). Moreover, they also showed a higher proportion of COVID-19 exposure (14.9%) than non-vulnerable.

Table 1 Gender-related, sociodemographic, and health-related characteristics and according to the high COVID-19 vulnerability. The Brazilian LGBT+ Health Survey, August-November, 2020.

We also described our sample by age groups since age significantly influences the composition of sexual orientation and gender identity in non-representative samples. Figure 1 shows that regarding sexual orientation, bisexual, heterosexual, and other scarce sexual orientations are mostly concentrated at younger ages: only 1.6% bisexual and 1.3% heterosexual and other scarce sexual orientations were aged ≥ 50 years. The same pattern did not occur with gender identity, although transgender and other scarce gender identities showed a proportion of 5.4% among individuals aged ≥ 50 years.

Figure 1 Participants' sexual orientation and gender identity according to age group. The Brazilian LGBT+ Health Survey, August-November, 2020.

Table 2 shows the results of multivariate associations between participants' characteristics and the highest COVID-19 vulnerability. The fully adjusted model showed higher odds of the COVID-19 vulnerability among heterosexual and other scarce sexual orientations (OR = 2.34; 95%CI: 1.01-9.20, compared to homosexual), cisgender men (OR = 3.52; 95%CI: 1.35-4.44, compared to cisgender women), and those aged ≥ 50 years (OR = 3.74; 95%CI: 1.24-11.25, compared to 18-29 years). Lower odds of COVID-19 vulnerability included individuals with complete graduate education (OR = 0.06; 95%CI: 0.02-0.22, compared to complete high school), being white (OR = 0.44; 95%CI: 0.23-0.83), and reporting proper facemask use (OR = 0.31; 95%CI: 0.13-0.76).

Table 2 Sequential models of the association between participant's characteristics and COVID-19 vulnerability. The Brazilian LGBT+ Health Survey, August-November, 2020.

Discussion

This study found that some sexual and gender minorities are more prone to higher COVID-19 vulnerability. They included heterosexual and other sexual orientation minorities, cisgender men, and those aged ≥ 50 years. Moreover, individuals with higher schooling level, white race/skin color, and reporting proper facemask use were less likely to have a higher COVID-19 vulnerability.

By establishing the analysis standpoint in the collective dimension as producer and reproducer of mechanisms of illness and vitality, based on historically constructed social vulnerabilities, human beings assume the character of a product of civilization and, therefore, the status of a social product ²⁸. Health is understood as the full development of the human potentialities, according to the level of progress achieved by society in a specific historical period, depending on the anatomical and functional regularity of the body and on the possibility to use what humanity has produced ^28,29. Hence, humans are not born ready, but acquire the human condition according to the access produced by society, such as food, education, health care services, stable and dignified employment conditions, and environmental safety. The relationships that are established within this dynamic determine different possibilities and restrictions to develop life and, consequently, different ways or possibilities of living, getting sick, and dying ²⁸.

Despite sexual and gender minorities being treated as a whole in our analysis, the results show a different COVID-19 vulnerability according to sexual orientation and gender identity categories. Although the higher vulnerability was not significant to transgender and other scarce gender identities, heterosexual, and other scarce sexual orientations showed a higher COVID-19 vulnerability. These categories show a constructed gender identity different from the born gender ³⁰, which increases social vulnerability due to discriminative environments. Cisgender women and men also show different gender identities, leading to different social and political constructions ³⁰. However, these sexual and gender minorities share social and environmental characteristics, leading to a higher COVID-19 vulnerability.

According to our results, factors associated with a higher COVID-19 vulnerability, except for proper facemask use, are structural determinates and suggest overlapping vulnerabilities, as described by the COVID-19 syndemic model ^3,4. Multiple historical and present-day factors have created the syndemic condition, including lower schooling levels, non-white race/skin color, worse working conditions or unemployment, and receiving income support during the COVID-19 pandemic. Although not all those factors were associated with higher vulnerability in the fully adjusted model, the descriptive analysis showed that they were worse in the income vulnerability dimension. Nearly 28% of the Brazilian people have received government income support during the COVID-19 pandemic ³¹. Although an online-based sample inherently excludes the most vulnerable individuals, 24.6% of the participants were enrolled and received government income support. Furthermore, income vulnerability reflects vulnerability in the COVID-19 exposure dimension. A similar online survey showed that 26.3% of the Brazilian sexual and gender minorities reported difficulty to maintain social distancing and other preventive measures related to COVID-19, 42.3% due to job/salary reduced or lost, and 19.4% due to transportation availability ²³. Therefore, home-office and stay-at-home are not commonly chosen by a historically neglected and discriminated population embedded in a heteronormative ^5,11 and racist ³² society, precluding them from friendly schooling environments, having better job opportunities, and economic prospects. The home can also be a discriminative environment ¹², decreasing emotional support during the pandemic ¹⁹ and affecting mental health ^15,16,17,19.

The income dimension, embedded in the social context, also affects the health dimension. Our findings did not show association in the fully adjusted model. Nevertheless, descriptive analysis evidenced a higher proportion of individuals diagnosed with depression and worse mental health during the pandemic in the higher income vulnerability group. For example, transgender individuals use fewer health care services due to disrespect to their social name ³³ which is a barrier to health care access ³⁴. Moreover, they experience harassment, trauma, and mental health disorders more frequently than cisgender individuals ³⁵, derived from higher discrimination in several life aspects ³⁶. The non-white race/skin color is an essential determinant of poor access to health care and higher job losses during the pandemic ³⁷, which partially explains the higher COVID-19 mortality rates among non-whites in Brazil ³⁸. Data from the United States show that non-white and sexual and gender minorities are worse economically affected than non-LGBT+ counterparts: 15% non-white LGBT+ individuals recently laid off work, whereas this proportion is only 11.5% among non-LGBT+ counterparts ³². Moreover, the literature reports that minority ethnic groups, minority gender-related groups, and people living in areas of higher socioeconomic deprivation generally experience long-term exposures that may cause an unequal COVID-19 vulnerability distribution ^3,39,40.

A total of 83.3% reported proper facemask use. Among the Brazilian sexual and gender minorities, proper facemask use was lower among those individuals with increased alcohol use during the pandemic ²⁷, which might derive from worse mental health during the pandemic ^15,16,17,19. Although worse mental health during the pandemic did not increase in the COVID-19 exposure vulnerability in our study, the proportion increased in the income vulnerability. Proper facemask use indirectly reflects synergic overlapping across vulnerability dimensions. However, lower facemask use might be related to lack of COVID-19 awareness, leading to lower perceived susceptibility and worry or greater self-confidence in coping with it ⁴¹. Regardless, government and health care providers must immediately implement strategies to ensure equity, such as using sexual orientation and gender identity measures in surveillance data and include equity-focused initiatives ⁴².

Study strengths and limitations

Our study strengths and limitations should be considered. Firstly, online surveys decrease the response rate, comprise a convenience sample, and only include participants with internet access. Therefore, the most vulnerable population was not included. However, considering the unavailability of nationally representative datasets and the difficulty to design a nationally representative study with the sexual and gender minorities, this study might contribute to understand the sexual and gender minorities' higher COVID-19 vulnerability. Secondly, the lack of programmatic vulnerability ² in the COVID-19 vulnerability operationalization and the statistical approach hindered a straight vulnerability overview as a syndemic model. Therefore, further analyses must consider different approaches. Thirdly, the cross-sectional design limits the establishment of causal chain, but vulnerabilities are usually bidirectional. Finally, having a private health care was used to classify a lower COVID-19 vulnerability, despite during periods of increased COVID-19 cases in Brazil, such as in July-August, at the beginning of our data collection, both private and public health care systems lacked hospital beds. Regarding the strengths of our study, we used anonymous data of the participants, which is considered the best form to increase adherence of this population. Moreover, this is the first study in Brazil with broad coverage of participants from the five geographical regions of Brazil and includes questions on a wide range of health dimensions. We used post-stratification regarding Brazilian regions to strengthen sample representativeness.

Conclusion

Our outcomes emphasize structural factors associated with the highest COVID-19 vulnerability among sexual and gender minorities, which suggests overlapping vulnerabilities, as described by a syndemic of a mutually caused epidemic. This model guides health care providers and governments' strategies to deal with the pandemic, which includes a joint approach to the common epidemic that affects sexual and gender minorities. They include broad multi-sectorial approach to decrease inequalities, promoting sexual and gender minorities' friendly environments, supporting social and economic vulnerable individuals, increasing primary health care and emergency access, and better understand care and psychosocial care network in the public health care system.

References

This is an open-access article distributed under the terms of the Creative Commons Attribution License