Articles

Food insecurity in households with persons with disabilities

in a situation of extreme vulnerability in Brazil: A secondary
cross-sectional analysis
Ursula Viana Bagni,a,* Alexia Vieira de Abreu Rodrigues,b Eloah Costa de Sant’Anna Ribeiro,b Rosana Salles-Costa,b,c and Aline Alves Ferreirab,c
a
Department of Social Nutrition, Emília de Jesus Ferreira College of Nutrition, Federal Fluminense University, Niterói, RJ, Brazil
b
Postgraduate Program in Nutrition, Josué de Castro Nutrition Institute, Federal University of Rio de Janeiro, Rio de Janeiro, RJ, Brazil
c
Department of Social and Applied Nutrition, Josué de Castro Nutrition Institute, Federal University of Rio de Janeiro, Rio de Janeiro, RJ,
Brazil

Summary The Lancet Regional

Health - Americas
Background Inequities in access to education, work and health care are striking among persons with disabilities,
2023;18: 100417
making this population more vulnerable to poverty, lack of access to basic services and violation of rights such as
Published Online
access to food. Household food insecurity (HFI), marked by precarious income, has increased among persons with
https://doi.org/10.
disabilities. In Brazil, the Continuous Cash Benefit (In Portuguese, Benefício de Prestação Continuada - BPC) is the 1016/j.lana.2022.
guarantee of a minimum wage for persons with disabilities as a measure to promote social security and access to 100417
income in a situation of extreme poverty. Thus, the objective of this study was to assess HFI among persons with
disabilities in extreme poverty in Brazil.

Methods A cross-sectional study with national representation was carried out with data from the 2017/2018 Family
Budget Survey, with moderate and severe food insecurity as the dependent variable, and the situation of food
insecurity measured using the Brazilian Food Insecurity Scale. Prevalence and odds ratio estimates were
generated with 99% confidence intervals.

Findings Approximately 25% of households experienced HFI, with a higher prevalence in the North Region (41%),
receiving up to 1 income quintile (36.6%), with a female (26.2%) and black person (31%) as a reference. The analysis
model found that region, per capita household income, and social benefits received in the household were statistically
significant factors.

Interpretation The BPC proved to be an important source of household income for persons with disabilities living in
extreme poverty in Brazil: in almost three-quarters of the households, it was the only social benefit received, and, for
most of them, it represented more than half of the total household income.

Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit
sectors.

Copyright © 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Brazil; Disabled persons; Food insecurity; Poverty; Social vulnerability

Introduction disabilities.1 In addition, they have a higher cost of living

Persons with disabilities (PWD) experience worse living associated with health care and rehabilitation, assistive
conditions, such as poor education, unequal employment devices, special diets, personal assistance, transport, and
opportunities and inadequate access to health care.1–3 Their other specific needs than non-disabled person. As a result,
required disability-related services are often neglected, this population is more likely to face poverty, including
and their everyday lives are surrounded by exclusion, poor housing, lack of access to safe water and sanitation,
violence, abuse, prejudice, or disrespect because of their and limited financial resources for food.1–3

*Corresponding author. Department of Social Nutrition, Emília de Jesus Ferreira College of Nutrition, Federal Fluminense University, Rua Mario
Santos Braga nº 30, 4º Andar, Campus Valonguinho, Centro, Niterói, Rio de Janeiro 24020-140, Brazil.
E-mail addresses: ursulaviana@gmail.com, ursulaviana@id.uff.br (U.V. Bagni).

www.thelancet.com Vol 18 February, 2023 1

 Articles

Research in context
Evidence before this study considered those with BPC beneficiaries under the age of 65,
Inequities in access to education, work and health care are as the only eligibility criterion for receiving the benefit would
striking among persons with disabilities (PWD) in Brazil, be disability.
making this population more vulnerable to poverty and
Added value of this study
household food insecurity (HFI). Previous studies have
We highlight the originality of the study, the socioeconomic
indicated that in households with PWD unable to work,
conditions and food insecurity of households with PWD in
exposure to HFI is greater, that is, income limitation implies
Brazil had not yet been investigated. In fact, studies on
food insecurity. Income transfer programs have been adopted
inequalities in health and nutrition of PWD in Brazil are scarce,
in some countries as income guarantee measures. In Brazil,
evidencing the social invisibility of this population. Evidence
the Continuous Cash Benefit (In Portuguese, Benefício de
suggests that BPC is an important source of household
Prestação Continuada - BPC) is the guarantee of a minimum
income for person with disabilities living in extreme poverty
wage for person with disabilities as a measure to promote
in Brazil and can contribute to the food security of this
social security and access to income in a situation of extreme
population.
poverty. Considering limited access to income for PWD, the
relationship between income and HFI and the importance of Implications of all the available evidence
social policies to reduce inequalities, we searched Pubmed and The vulnerability of the population with disabilities impacts
Scielo databases for articles published in English and access to food insecurity. The HFI investigation, considering
Portuguese, in the last 10 years, that have investigated HFI, the quantitative, qualitative and psychological dimensions of
the sociodemographic profile and economic conditions of access to food conditional on sufficient family income to
PWD. As search terms, “Persons with Disabilities”, “Food purchase food, may be related to the living and health
Insecurity”, “Poverty” and “Social Vulnerability” were used. conditions of this population, including access to social
There are still few studies that investigate the security through the BPC. The findings suggest the relevance
sociodemographic and economic conditions of PWD and that of generating visibility to a group under-discussed in the
relate it to exposure to HFI. For data analysis, we used public academic field, contributing to the discussion on social
data from a nationally representative survey of the Brazilian inequalities and public policies for minorities in Brazil.
population called the Family Budget Survey (Pesquisa de Highlighting the importance of the BPC for access to income,
Orçamentos Familiares - POF), one of the most important contributing to the reduction of inequalities and the most
surveys in the country. Although, the POF does not directly serious levels of HFI of the PWD.
investigate the presence of PWD in households and we

Disability is consistently related to an increase in equivalent to $238.5 Brazilian reais - BRL or $74.5
household food insecurity (HFI) across different pop- United States dollars USD, considering its average price
ulations and geographic settings,2,4,5 with a higher risk of $3.2 in January of the same year).
among younger disabled adults and those with mental Although there are other benefits guaranteed by law
disabilities.2,4 In the United States in the years 2009/ in Brazil, BPC is the only one intended for PWD living
2010, HFI was higher both in households with a in great poverty. Providing the amount of one minimum
working-age adult with a disability that prevented work monthly salary to these subjects, BPC is the only social
(33.5%) and in those with a labour force-excluded assistance benefit guaranteed by the Brazilian Federal
working-age adult (24.8%) than in households with no Constitution of 1988.9–11 It represents approximately
working-age adult with a disability (12.0%).6 In 2018, 80% of the family budget of PWD, and for almost half of
these numbers barely changed: HFI was present in them, it is the only source of family income.10
33.0% of households with a member with work-limiting Studies on the sociodemographic characteristics,
disabilities and 22.4% of those with disabled working- living conditions and health of BPC beneficiaries, that
age adults in the labour force.7 is, PWD in a situation of extreme poverty in Brazil, are
To minimize these inequities, many countries scarce.12–14 However, the impact of this benefit on
around the world, such as Chile, Bangladesh, India, and improving living conditions, reducing poverty and
South Africa, provide income support for PWD through mitigating the income inequality of beneficiary families
unconditional cash transfers.1,8 In Brazil, this income was shown to be significant.15–21
support is provided by the Continuous Cash Benefit (In The current dimension of HFI and its associated
Portuguese, Benefício de Prestação Continuada - BPC) factors in BPC beneficiaries still need to be studied to
for persons with all types of disabilities in great poverty guide decision-making in public policies and to quantify
who have a monthly per capita family income below a the care provided by health and social care professionals
quarter of the current minimum wage9–11 (in 2018, was to PWD in situations of extreme social vulnerability. The

2 www.thelancet.com Vol 18 February, 2023

 Articles

hypothesis of this study is that families composed of of gross monetary and nonmonetary earnings of all family
PWD are more exposed to HFI. members. The first quintile of income represents ⅕ of the
Thus, the aim was to investigate the prevalence of poorest population (lowest income) and the fifth quintile
HFI and the socio-demographic and economic condi- the richest 20% of the population (highest income). The
tions of PWD in great poverty in Brazil based on data degree of dependence on the BPC for household income
from the Family Budget Survey 2017/2018. [up to 30%, from 30 to 50% and above 50%]. Biological sex
defined based on self-report [male; female], racial identity
[white; black; asian/indigenous, birracial], and schooling
Methods [no schooling; elementary school I/II; high school or
This was a cross-sectional population-based observational university education] of the reference person in the
and analytical study with an analysis of public domain household were also included for analysis, defined as the
microdata from the 2017/2018 Family Budget Survey person considered to be the primarily responsible for the
developed by the Brazilian Institute of Geography and household, that is, the person responsible for the family’s
Statistics (IBGE). The STROBE reporting standard decisions, regardless of being the one who received the
(Strengthening the Reporting of Observational Studies in BPC.22–24
Epidemiology) was adopted to guide the research design. Food insecurity, defined as limited access to safe and
Based on the census sectors of the 2010 Brazilian nutritious food in sufficient quality and quantity,25 was
Demographic Census, the IBGE defined a single sample assessed in this study using the Brazilian Food Inse-
for the other population surveys, called the master curity Scale,25 which classifies HFI based on available
sample, consisting of a set of primary sampling units. household income. It is a psychometric scale that
Thus, in the household budget survey, the two-stage assesses the interviewee’s perception of access to food,
cluster sampling design, with geographic and statisti- in terms of quality and quantity, in the three months
cal stratification, defined the primary sampling units prior to the interview. Composed of 14 dichotomous
and permanent private households, from the master questions [yes/no], each affirmative answer adds one
sample. The primary sampling units of the 2017/2018 point to classify the level of HFI in the household or
Family Budget Survey were selected by simple random indicate its absence - food security, when all answers are
sampling. Likewise, private households were selected, negative.25 Food security is present when there is no
which correspond to private households, selected for concern or possible limitation in access to food, and the
each primary unit of private units with probability pro- HFI can be present in three levels of severity: mild HFI,
portional to the number of private households in the when there is concern about access to food or replacing
sector. A total of 57,920 Brazilian households were food with cheaper options; moderate HFI, when the
interviewed from July 2017 to July 2018 and made up limitation of food occurs or there is reduced access to
the sample, classifying the 2017/2018 Family Budget food, reflecting a rupture in the dietary pattern; and
Survey as one of the largest population-based surveys in severe HFI, when access to a sufficient quantity of food
Brazil.22–24 More details on the design of the household is also compromised, which may indicate hunger.25 In
budget survey can be found in the first results release this study, the outcome variable was the occurrence of
report released by the IBGE.24 the most severe forms of HFI in the household (mod-
A total of 2178 households were identified with at erate or severe HFI), treated dichotomously [yes/no].
least one resident receiving BPC. Among these house- Given the complexity of the data from the household
holds, only those with beneficiaries under 65 years of budget survey, the sampling unit and the sample weight
age (n = 1251) were included in the study. We adopted provided by IBGE were analyzed. The prevalence and
this criterion to ensure that BPC was received due to the respective 99% confidence intervals (99% CI) were
presence of a disability and not due to age since the estimated for a descriptive analysis of the sociodemo-
2017/2018 Family Budget Survey did not have any graphic characteristics of Brazilian households with at
questions to investigate directly the presence of PWD in least one PWD. To assess the association between the
the household. Thus, households in which no resident study variables and moderate/severe HFI, the chi-
was a beneficiary of the BPC or that had beneficiaries square test was used, and variables that presented
aged 65 or over were considered ineligible.10 p values <0.01 were considered statistically significant.
The features of the households of persons with dis- The logistic regression model was used to verify the
abilities assisted by the BPC were drawn in relation to the odds ratio (OR) for the occurrence of moderate/severe
region [North; Northeast; South; Southeast; Center-West], HFI. The households in which resident was a benefi-
territory [urban area; rural area], number of residents in ciary of the BPC as the exposure variable. The way of
the household [up to 3 residents; from 4 to 6 residents; 7 controlling the variables used was all study variables
or more residents], number of PWD in the household [1 were considered in the bivariate model. However, those
person; more than 1 person] and social benefits received at with p values <0.05 in the bivariate analysis were
home [only BPC; BPC and other(s)]. Household income considered in the adjusted model (region, per capita
per capita [in quintiles] was calculated considering the sum household income and social benefits at home).

www.thelancet.com Vol 18 February, 2023 3

 Articles

Verification of multilinearity was performed, evaluating United States Dollars (USD). In the reference period of
independent variables that can correlate. In the adjusted the study (January 15, 2018), this amount was equivalent
model, the variables were tested together, and a stricter to approximately one Brazilian minimum wage ($954.00
criterion was adopted to verify the associations; those BLR), considering the average exchange rate of $3.20
that presented values of p < 0.01 were considered sig- USD during this period. On average, the BPC contrib-
nificant. Results were expressed as ORs and respective uted 45.7% (99% CI 42.9–48.4) of household income,
99% CI. and in 52.9% (52.9%; 99% CI 42.1–63.4) of households,
In the 2017/2018 Family Budget Survey sample, each it contributed more than half of the household income.
household represents a group of households in the pop- In 74.1% (99% CI 69.3–78.3) of households, the BPC
ulation, associated with a sample weight or expansion was the only social benefit received (Table 1). Also,
factor.24 The expansion factor allows obtaining estimates 40.8% (99% CI 36.2–45.6) of the households with PWD
for the sample population of the research, thus the sample receiving BPC were in a situation of food security, with
data are expanded for the Brazilian population.24 The an- 34.1% (99% CI 29.5–39.1) with mild HFI, 16.2% (99%
alyses of this study were performed using STATA software CI 12.6–20.6) with moderate HFI, and 8.9% (99% CI
version 16.0 (StataCorp LP, College Station, United States) 6.6–11.7) with severe HFI.
using the Survey Data Analysis command (svy prefix), In the bivariate analysis, the prevalence of moderate/
used before any command, manages the research analysis severe HFI was higher in the North Region (41%; 99%
settings, being able to designate variables that contain in- CI 27.8–53.6) and lower in the South Region (9.0%; 99%
formation about the research sample design (sample unit CI 3.8–19.9). Moderate/severe HFI was also lower in
and sample weight, for example). The command can also households with per capita household income in the 5th
specify characteristics such as the number of sampling quintile (15.6%; 99% CI 8.8–26.0), that is, among those
stages, sampling method and analysis patterns. Missing with the higher income, compared to households in
data, expected due to the complexity of the survey, were other lower income brackets (Table 2).
treated by imputation procedures conducted by IBGE. The When investigating the strength of the association
procedure was used for the attribution of expenditure or between moderate/severe HFI and sociodemographic
income values, when total or partial non-response variables, there was statistical significance in the crude
occurred, as well as for rejected values, identified as analysis for region, per capita household income, and
response errors in the critical review stage also conducted social benefits received at the household level (Table 3).
by IBGE.24 After adjustment, the odds ratio of moderate/severe
According to Resolution No. 466 of December 12, HFI was found to be approximately five times higher in
2012 from the National Committee of Ethics in the North Region (OR = 5.5; 99% CI 1.8–16.4) and
Research (CONEP), for studies that use secondary data approximately three times higher in the Central-West
available in the public domain, as in this study that used (OR = 3.9; 99% CI 1.3–12.0), Northeast (OR = 3.0;
data available in the public domain from the Brazilian 99% CI 1.1–9.0) and Southeast Regions (OR = 3.0; 99%
Institute of Geography and Statistics, approval by a local CI 1.0–8.9) compared to the South Region of Brazil
Ethics Committee CEP-CONEP System is not required. (Table 3). In households where the per capita household
income was in the 1st quintile (OR = 2.6; 99% CI
1.1–5.8) or in the 2nd quintile (OR = 1.9; 99% CI
Role of the funding source 1.0–4.3), the odds ratio of moderate/severe HFI was
This research did not receive any specific grant from approximately twice as high as in the other income
funding agencies in the public, commercial, or not-for- brackets (Table 3).
profit sectors.

Discussion
Results In Brazil, the BPC is a social benefit guaranteed by law,
The investigated households were predominantly and its main objective is to grant the assisted population
located in urban areas (78.3%; 99% CI 73.5–82.5) and the inalienable right to social assistance. This benefit is of
located in the Northeast (43.7%; 99% CI 38.1–49.4) and paramount importance for PWD, as it guarantees a min-
Southeast Regions (32.0%; 99% CI 26.0–38.5) (Table 1). imum monthly income for individuals who can prove that
The households were mostly composed of up to they do not have the means to provide for themselves in
three residents (55.6%; 99% CI 51.0–60.3) and only one terms of their livelihood and quality of life.9–11
person with a disability (91.3%; 99% CI 88.0–93.7). The Thus, in the present study, the BPC proved to be an
reference person in the household was mostly brown important source of household income for families with
(54.9%; 99% CI 49.9–59.7), female (55%; 99% CI PWD. In almost ¾ of the households, BPC was the only
49.7–60.1) and had a low education level (Table 1). social benefit received, and for most of them, it repre-
The average per capita household income (which sented more than half of the household income. A
included money received from the BPC) was $312.73 similar situation was reported in a study carried out in

4 www.thelancet.com Vol 18 February, 2023

 Articles

Variables na %b 99% CIc

Region
North 185 8.9 6.5–12.0
North east 609 43.7 38.1–49.4
Central-West 143 7.8 5.7–10.6
South 89 7.7 5.4–10.7
Southeast 225 32.0 26.0–38.5
Location of the household
Urban area 919 78.3 73.5–82.5
Rural area 332 21.7 17.5–26.5
Household per capita income (quintile)d
1 (36.3–156.2) 251 18.6 15.1–22.7
2 (156.3–219.0) 250 18.9 15.5–22.9
3 (219.1–300.6) 250 21.2 17.0–26.0
4 (300.7–404.2) 250 20.5 16.8–24.8
5 (404,3–4695.8) 250 20.8 17.0–25.2
BPC contribution to household income
<30% 57 19.0 9.8–33.7
From 30 to 50% 109 28.1 20.8–36.8
More than 50% 194 52.9 42.1–63.4
Social benefits in household
Only BPC 916 74.1 69.3–78.3
BPC and others 335 25.9 21.6–30.7
Number of residents
Up to 3 residents 677 55.6 51.0–60.3
From 4 to 6 residents 474 36.2 31.8–40.7
7 or more residents 100 8.2 5.7–11.6
Number of persons with disabilities in the household
1 person 1143 91.3 88.0–93.7
More than 1 person 108 8.7 6.3–12.0
Race of the reference person in the household
White 364 31.4 26.8–36.3
Black 164 13.3 10.0–17.4
Asian/Indigenous 7 0.4 0.1–1.6
Biracial 714 54.9 49.9–59.7
Sex of the reference person in the household
Male 585 45.0 39.9–50.3
Female 666 55.0 49.7–60.1
Education of the reference person in the household
No schooling 262 21.3 17.4–25.8
Elementary School I/II 746 61.6 56.8–66.2
High school/University education 243 17.1 14.0–20.7
Note: Sample weights were considered for all variables studied. Brazil, 2017/2018 Family Budget Survey. aNumber of observations considering the expanded data.
b
Prevalence. c99% Confidence interval. dAmounts in United States dollars, considering the average exchange rate of $3.20 during the survey period.

Table 1: Characteristics of the households with persons with disabilities receiving Continuous Cash Benefit (BPC).

the State of Rio de Janeiro, in which 100% of persons benefits. This scenario, already reported in previous
with disabilities were completely dependent on social national studies,19 may reflect the higher percentage of
assistance for their income and this was primarily pro- persons with disabilities in these regions (9.9% and
vided by the BPC, since 80% of households did not 8.1%, respectively), pointed out by the latest National
receive any other type of benefit.13 In Brazil, the Study on Health.26
Northeast and Southeast regions have the largest pop- Another point worth mentioning is the regional
ulation,24,26 and in this study presented the highest inequality of moderate/severe HFI observed in this
prevalence of households of PWD receiving social study, which was approximately five times higher in the

www.thelancet.com Vol 18 February, 2023 5

 Articles

Variables Moderate/severe HFI

Yes No
na %b 99% CIc na % 99% CIc p-valued
Region 0.0020
North 76 40.1 27.8–53.6 109 59.9 46.4–72.1
North east 157 26.6 21.1–32.9 452 73.4 67.1–79.1
Central-West 28 25.9 15.9–39.3 115 74.1 60.6–84.1
South 9 9.0 3.8–19.9 80 91.0 80.1–96.2
Southeast 42 22.5 14.2–33.7 183 77.5 66.3–85.8
Location of the household 0.7092
Urban area 233 24.8 20.0–30.3 686 75.2 69.7–80.0
Rural area 79 26.1 19.0–34.8 253 73.9 65.2–81.0
Household per capita income (quintile)e 0.0022
1 (36.3–156.2) 97 36.6 26.7–47.9 154 63.4 52.1–73.3
2 (156.3–219.0) 74 28.3 20.0–38.3 176 71.7 61.7–80.0
3 (219.1–300.6) 60 25.4 16.8–36.5 190 74.6 63.5–83.2
4 (300.7–404.2) 45 20.9 12.7–32.4 205 79.1 67.6–87.2
5 (404,3–4695.8) 36 15.6 8.8–26.0 214 84.4 73.9–91.2
BPC contribution to household income 0.0907
<30% 10 17.7 4.9–47.5 47 82.3 52.5–95.1
From 30 to 50% 20 15.0 7.7–27.2 89 85.0 72.8–92.3
More than 50% 59 30.4 20.6–42.4 135 69.6 57.6–79.4
Social benefits of household 0.0081
Only BPC 204 22.4 17.9–27.7 712 77.5 72.3–82.1
BPC and others 108 32.6 24.1–42.3 227 67.4 57.7–75.9
Number of residents 0.6146
Up to 3 residents 159 23.8 18.3–30.3 518 76.2 69.7–81.7
From 4 to 6 residents 123 26.0 20.0–33.0 351 74.0 67.0–80.0
7 or more residents 30 29.7 14.9–50.4 70 70.3 49.6–85.0
Number of persons with disabilities in the household 0.6557
1 person 283 25.3 20.9–30.2 860 74.7 69.8–79.1
More than 1 person 29 23.0 13.1–37.2 79 77.0 62.8–86.9
Race of the reference person in the household 0.3511
White 76 23.2 16.4–31.8 288 76.8 68.2–83.6
Black 48 31.0 17.9–48.1 116 69.0 51.9–82.1
Asian/Indigenous 1 0.4 0.02–4.6 6 95.6 53.7–99.7
Biracial 187 24.9 19.9–30.8 527 75.0 69.2–80.1
Sex of the reference person in the household 0.4852
Male 135 23.7 18.1–30.5 450 76.3 69.5–81.9
Female 177 26.2 20.4–32.9 489 73.8 67.1–79.6
Education of the reference person in the household 0.4515
No schooling 77 26.7 19.2–35.8 185 73.3 64.1–80.8
Elementary School I/II 185 26.0 20.6–32.1 561 74.0 67.9–79.3
High school/University education 50 21.2 12.9–32.7 193 78.8 67.3–87.0
Note: Sample weights were considered for all variables studied. Brazil, 2017/2018 Family Budget Survey. aNumber of observations considering the expanded data.
b
Prevalence. c99% Confidence interval. dChi square test. eAmounts in United States dollars, considering the average exchange rate of $3.20 during the survey period.

Table 2: Association between sociodemographic variables and moderate/severe household food insecurity (HFI) in households with persons with
disabilities receiving Continuous Cash Benefit (BPC).

North Region, followed to a lesser extent by the North- observed in the general population, in which moderate/
east, Central-West and Southeast Regions when severe HFI was approximately five times higher in the
compared to the South Region of Brazil, the region with North and Northeast Regions than in the South Region
the highest socioeconomic indices in the country. The of Brazil.23 When evaluating severe HFI in Brazilian
results of the present study follow the national trend regions, Gubert and Pérez-Escamilla (2018) observed a

6 www.thelancet.com Vol 18 February, 2023

 Articles

Variables Moderate/severe HFI

Bivariate model Adjusted model
ORa 99% CIb OR 99% CIb
Region
North 6.8e 2.3–19.8 5.5e 1.8–16.4
e
North east 3.7 1.4–9.7 3.0e 1.1–9.0
Central-West 3.5e 1.2–10.8 3.9e 1.3–12.0
South 1.0c 1.0c
e
Southeast 2.9 1.0–9.7 3.0e 1.0–8.9
Location of the household
Urban area 1.0c
Rural area 1.1 0.6–1.8
Household per capita income (quintile)f
1 (36.3–156.2) 3.1e 1.4–7.0 2.6e 1.1–5.8
2 (156.3–219.0) 2.1e 1.0–4.6 1.9d 1.0–4.3
3 (219.1–300.6) 1.8 0.8–4.2 1.7 0.7–4.1
4 (300.7–404.2) 1.4 0.6–3.4 1.3 0.5–3.2
5 (404,3–4695.8) 1.0c 1.0c
BPC contribution to household income
<30% 1.0c
From 30 to 50% 0.8 0.1–4.3
More than 50% 2.0 0.4–9.4
Social benefits of household
Only BPC 1.0c 1.0c
BPC and others 1.7e 1.0–2.7 1.3 0.8–2.3
Number of residents
Up to 3 residents 1.0c
From 4 to 6 residents 1.1 0.7–1.8
7 or more residents 1.3 0.5–3.5
Number of persons with disabilities in the household
1 person 1.0c
More than 1 person 0.9 0.4–1.8
Race of the reference person in the household
White 1.0c
Black 1.5 0.6–3.5
Asian/Indigenous 0.1 0.07–2.9
Biracial 1.1 0.6–1.8
Sex of the reference person in the household
Male 1.0c
Female 1.1 0.7–1.8
Education of the reference person in the household
No schooling 1.3 0.6–2.8
Elementary School I/II 1.3 0.7–2.5
High school/University education 1.0c
Note: Sample weights were considered for all variables studied. Brazil, 2017/2018 Family Budget Survey. aOdds ratio. b99% Confidence interval. cReference category. dp–
value < 0.05. ep–value < 0.01. fAmounts in US dollars, considering the average exchange rate of $3.20 US dollars during the survey period.

Table 3: Odds Ratios (OR) and 99% confidence intervals (CI) of the association between sociodemographic variables and moderate/severe household
food insecurity (HFI) in households with persons with disabilities receiving Continuous Cash Benefit (BPC).

heterogeneous distribution profile, highlighting an (18.6%–0.8%) and smaller differences between munic-
increased prevalence in the Northeast (6.7%) compared ipalities in the Southeast Region (0.7%–4.9%).27
to the South Region (1.9%). They also highlighted a In Brazil, the distribution of HFI is based on the
marked difference in the distribution of intraregional inequities of society. In strata of greater economic, so-
prevalence of severe HFI, with large discrepancies cial and demographic vulnerability, the prevalence of
between municipalities in the Northeast Region HFI tends to be higher.23 In this study, for the Southeast

www.thelancet.com Vol 18 February, 2023 7

 Articles

and Central-West Regions, the prevalence of moderate/ inequities that permeate the lives of PWD, these results
severe HFI observed in PWD receiving BPC was higher highlight the fact that social benefits are the main
than that observed in the general population in these source of income for these individuals. That is, these
regions (8.7% and 12%, respectively23), when compared social benefits may be insufficient to mitigate social
to the South Region. inequalities and prevent HFI and hunger. PWD may
Other factors may contribute to sociodemographic need a higher income to overcome the material hard-
inequalities in receiving BPC benefits, such as differ- ship imposed by their disability and maintain food
ences related to the dissemination of information about security than persons without disabilities, as observed in
benefit programs to the population, little access to places a previous study by She and Livermore (2007).28 Ac-
where PWD can apply for the benefit, low availability of cording to these authors, for a PWD incapable of
professionals for medical expert evaluation, heteroge- working to reach the same level of food security as a
neity in the form of assessment of disability and in- person without disability, both on the poverty line, the
capacity for work, among others. In addition, the social annual income of the PWD would need to be more than
movements of PWD and support organizations in 2.5 times higher. Thus, the disregard of the intersec-
different regions of the country tend to be heteroge- tional barriers that permeate social vulnerability among
neous and heavily influenced by the historical and social PWD makes disability benefits often inadequate to
formation of each region, which directly impacts access prevent HFI across the population.2 In Brazil, Law
to education and health.2,16 14176 was enacted in 2021, which increases the limit
The receipt of BPC benefits by PWD occurred more value of monthly family income per capita to request the
frequently in urban regions, which accounts for 83.2% benefit from ¼ minimum wage to ½ minimum wage.
of PWD in Brazil.26 But although the majority of PWD This measure seeks to increase the breadth of BPC
live in urban areas, the proportion of benefit receivers is coverage for PWDs, and its impacts should be investi-
still likely to be related to its greater ease of access in gated in future research.29
urban areas, since these are the sites in Brazil that It is important to highlight that the 2017/2018
concentrate the best services of health, transportation, Family Budget Survey form did not contain questions
education, and others. The rural regions of the country about the presence of PWD in Brazilian households.
still have limited geographic access to social and health Therefore, an inherent limitation of this study was the
services, which are also more precarious when assessment of only BPC beneficiaries under 65 years of
compared to urban areas. In rural areas, the absence of age, since elderly person above this age without dis-
Brazilian civil registration documents is also common, abilities living in great poverty can also receive this
which creates obstacles to requesting the benefit; in benefit. By restricting the age of the beneficiaries to less
addition, difficulties related to topography, the road than 65 years of age, we assumed that the only reason
system or the absence of social security agencies within for receiving BPC was the presence of a disability. As
a reasonable distance are obstacles in accessing medical the total number of PWD in the survey was unknown, it
evaluation points that allow person to receive the was not possible to estimate the prevalence of HFI in all
benefit.16 Brazilian households with PWD and in households of
In this study, almost 60% of households of PWD PWD living in great poverty and not receiving BPC, as
receiving BPC had some degree of HFI, a prevalence well as comparing the HFI between PWD receiving and
similar to that of a household-based survey (64%) not receiving BPC. Therefore, the presence of house-
developed in 2009 involving 961 PWD receiving BPC holds with BPC beneficiaries over 65 years of age and
benefits in 543 Brazilian municipalities.14 The magni- with a disability can not be ruled out and would infer a
tude of HFI identified in this study was almost double level of underestimation to our results. Yet, the known
that observed in the Brazilian population (36.7%)23 and reduced information and access to BPC and the signif-
also exceeded the prevalence of HFI of PWD indicated icant level of PWD living in great poverty and not
in international studies (between 22.4% and 33.5%).6,7 receiving BPC was not captured in the survey and sug-
This fact highlights the extreme social vulnerability gests our estimated number would be underestimated.
that PWD dependent on social benefits endure in Brazil. Furthermore, the stratification of data by sex and race of
The high HFI values also express the changes in eating residents of the person of reference was limited by the
patterns resulting, among other factors, from political size of the sample.
crises in the country, culminating in a lack of food at Still, this study innovatively presents the magnitude
home. Hunger, then, seems to reach higher levels than of HFI for the most vulnerable PWD in the country,
those found in households that do not have PWD.27 which has never been addressed in Brazilian population
It was also observed that the lower the household studies. This directly contributes to the development
income, the greater the frequency of HFI: in the poorest and implementation of public policies aimed at
families, with a per capita household income in the 1st reducing poverty and guaranteeing human rights,
or 2nd quintile, moderate/severe HFI was approxi- especially in a population that experiences numerous
mately twice as high as in other families. Given the prejudices in society. The development of transversal

8 www.thelancet.com Vol 18 February, 2023

 Articles

policies including PWD is a challenging topic for public References

1 World Health Organization. World report on disability. Available
health policy-makers in Brazil, especially in the context in: https://apps.who.int/iris/handle/10665/44575; 2011.
of public policy dismantling since 2016. The political 2 Schwartza N, Buliungb R, Wilsonc K. Disability and food access
crises that have widened since 2018 and more recently and insecurity: a scoping review of the literature. Health Place.
2019;57:107–121.
the COVID-19 pandemic make the situation more 3 Mitra S, Yap J. The disability data report. Disability data initiative.
difficult, and the perspective is that, in the coming years, Fordham Research Consortium on Disability: New York. Available
person with extreme vulnerability will suffer even more in: https://disabilitydata.ace.fordham.edu/wp-content/uploads/
2021/05/The-Disability-Data-Report-2021.pdf; 2021.
from an increase in food insecurity, thereby increasing 4 Brucker DL, Coleman-Jensen A. Food insecurity across the adult
social inequalities in Brazil.30 life span for persons with disabilities. J Disabil Policy Stud.
2017;28(2):109–118. https://doi.org/10.1177/1044207317710701.
Households of persons with disabilities who 5 Helfin CM, Altman CE, Rodriguez LL. Food insecurity and
receive the Continuous Cash Benefit have a high disability in the United States. Disabil Health J. 2019;12(2):220–226.
prevalence of food insecurity, which is unequal in https://doi.org/10.1016/j.dhjo.2018.09.006.
6 Coleman-Jensen A, Nord M. Food insecurity among households
magnitude in different regions of the country. These with working-age adults with disabilities, ERR-144, U.S. Depart-
individuals depend heavily on this social benefit; ment of Agriculture, Economic Research Service, January 2013.
Available in: https://www.ers.usda.gov/webdocs/publications/
however, it seems to be insufficient to prevent food 45038/34589_err_144.pdf?v=7536.6.
insecurity in the households of persons with disabil- 7 Coleman-Jensen A. U.S Food insecurity and population trends with
ities in Brazil. a focus on adults with disabilities. Physiol Behav. 2020;220:112865.
https://doi.org/10.1016/j.physbeh.2020.112865.
The fragility of social policies in guaranteeing social 8 Gooding K, Marriot A. Including persons with disabilities in social
rights of this population is accentuated by the absence of cash transfer programmes in developing countries. J Int Dev.
studies and national surveys on the impact of the 2009;21:685–698. https://doi.org/10.1002/jid.1597.
9 Ministério do Desenvolvimento Social e Combate à Fome. Cader-
Continuous Cash Benefit on the health, nutrition and nos de Estudos Desenvolvimento Social em Debate. – N. 13. Bra-
food security of persons with disabilities. Interdisci- sília, DF: Ministério do Desenvolvimento Social e Combate à Fome.
plinary and intersectoral discussions in civil society and Available in: https://aplicacoes.mds.gov.br/sagirmps/ferramentas/
docs/caderno%20-%2005.pdf; 2010.
academia are essential to guide public policies aimed at 10 Brasil. Decreto nº 6.214, de 26 de setembro de 2007. Regulamenta o
population equity, reducing prejudice and fighting benefício de prestação continuada da assistência social devido à
pessoa com deficiência e ao idoso de que trata a Lei no 8.742, de 7
hunger and poverty, especially in Brazil. de dezembro de 1993, e a Lei nº 10.741, de 1º de outubro de 2003,
acresce parágrafo ao art. 162 do Decreto no 3.048, de 6 de maio de
1999, e dá outras providências. Available in: http://www.planalto.
Contributors gov.br/ccivil_03/_Ato2007-2010/2007/Decreto/D6214.htm.
U.V.B. conceived the idea and design of the study, contributed to the 11 Stopa R. O direito constitucional ao Benefício de Prestação Con-
administration and execution of the project, interpreted the statistical tinuada (BPC): o penoso caminho para o acesso. Serv Soc Soc.
data and wrote the manuscript. E.C.d.S.A.R. contributed to the 2019;135:231–248. https://doi.org/10.1590/0101-6628.176.
analysis and the interpretation of statistical data, in the execution and 12 Lobato L, Góis JB, Senna M, Reis M, Vasconcellos M. Avaliação do
visualization of the manuscript and helped in the writing of the Benefício de Prestação Continuada. In: BRASIL. Ministério do
manuscript. A.V.d.A.R. contributed to the curation and analysis of Desenvolvimento Social e Combate à Fome. Avaliação de políticas e
programas do MDS: resultados. Volume 2: Bolsa Família e Assistência
the data, applying statistical and computational techniques, inter-
Social. Brasília: Ministério do Desenvolvimento Social e Combate à
preted the data and assisted in the writing of the manuscript. R.S.-C. Fome (MDS). Secretaria de Avaliação e Gestão da Informação
contributes to the methodology and design of the study and to the (SAGI); 2007.
interpretation of statistical data. A.A.F. contributes to the supervi- 13 Sochaczevski J, Tavares LHGL. Suas, BPC e pessoas com defi-
sion, coordination and planning of the study, methodology and study ciência: análise a partir de municípios fluminenses. SER Soc.
design, interpretation of statistical data and writing of the manu- 2013;15(32):141–165.
script. All authors who assisted in writing the manuscript and read 14 BRASIL. Ministério do Desenvolvimento Social e Combate à Fome.
and approved the final manuscript. Secretaria de Avaliação e Gestão da Informação. Síntese das pes-
quisas de avaliação de programas sociais do MDS – Versão atuali-
zada e revisada 2006-2010. Cadernos de Estudos Desenvolvimento
Data sharing statement Social em Debate, Brasília, n. 13, 2010. Available in: https://goo.gl/
SWv0n1.
This study analyzed public domain microdata from the Family Budget Survey
15 Denes G, Komatsu BK, Menezes Filho N. Uma Avaliação dos
(2018), made available by the Brazilian Institute of Geography of Statistics at Impactos Macroeconômicos e Sociais de Programas de Trans-
its electronic address https://www.ibge.gov.br/estatisticas/sociais/saude/ ferência de Renda nos Municípios Brasileiros. Policy Paper nº 21.
24786-pesquisa-de-orcamentos-familiares-2.html?=&t=microdados. São Paulo: Insper Centro de Políticas Públicas, 2016. Available in:
https://www.insper.edu.br/wp-content/uploads/2018/09/Avaliacao-
Impactos-Macroeconomicos-Sociais-Transferencia-Renda-Municipios-
Funding Brasileiros.pdf.
This research did not receive any specific grant from funding agencies 16 Medeiros M, Diniz D, Squinca F. Transferências de Renda Para a
in the public, commercial, or not-for-profit sectors. População com Deficiência no Brasil: Uma Análise do Benefício de
Prestação Continuada. Texto Para Discussão no 1184. Brasília:
IPEA, 2006. Available in: https://www.ipea.gov.br/portal/index.
php?option=com_content&view=article&id=4782.
Declaration of interests 17 Peliano AM. Desafios e Perspectivas da Política Social. Texto Para
All authors declare that there is no conflict of interest. Discussão no 1248. Brasília: IPEA, 2006. Available in: https://www.
ipea.gov.br/portal/index.php?option=com_content&view=article&
id=4810.
18 Soares S, Ribas RP, Soares FV. Focalização e cobertura do Pro-
Appendix A. Supplementary data grama Bolsa- -Família: qual o significado dos 11 milhões de famí-
Supplementary data related to this article can be found at https://doi. lias? Rio de Janeiro: Ipea, 2009. (Texto para Discussão, n. 1396).
org/10.1016/j.lana.2022.100417. Available in: https://goo.gl/zJBq3N.

www.thelancet.com Vol 18 February, 2023 9

 Articles

19 Soares FV, Soares S, Medeiros M, Osório RG. Programas de trans- 25 Segall-Corrêa AM, Marin-León L, Melgar-Quiñonez H,
ferência de renda no brasil: impactos sobre a desigualdade. Texto para Pérez-Escamilla R. Refinement of the Brazilian household
discussão Nº 1228. Brasília: IPEA, 2006. Available in: http://www.ipea. food insecurity measurement scale: recommendation for a 14-item
gov.br/portal/images/stories/PDFs/TDs/td_1228.pdf. EBIA. Rev Nutr. 2014;27(2):241–251. https://doi.org/10.1590/1415-
20 Santos WR. Deficiência e BPC: o que muda na vida das pessoas 52732014000200010.
atendidas? Ciên Saúde Colet. 2011;16(Supl. 1):787–796. 26 Pesquisa nacional de saúde: 2019: ciclos de vida: Brasil/IBGE,
21 Lobato L, Senna M. Benefício de Prestação Continuada (BPC): os Coordenação de Trabalho e Rendimento. Rio de Janeiro: IBGE,
pobres na mira das políticas de austeridade. Análise de conjuntura. 2021. 139p. Available in: https://biblioteca.ibge.gov.br/index.php/
Rio de Janeiro: Centro de Estudos Estratégicos da Fiocruz, 2020. biblioteca-catalogo?view=detalhes&id=2101846.
Available in: https://cee.fiocruz.br/?q=BenefIcio-de-Prestacao- 27 Gubert MB, Perez-Escamilla R. Severe food insecurity in
Continuada-BPC-os-pobres-na-mira-das-politicas-de-austeridade. Brazilian municipalities, 2013. Ciên Saúde Colet. 2018;23(10):
Accessed November 3, 2021. 3433–3444.
22 Instituto Brasileiro de Geografia e Estatística. Pesquisa de Orça- 28 She P, Livermore GA. Material hardship, poverty, and disability
mentos Familiares 2017- 2018: Manual do agente de pesquisa. Rio among working-age adults. Soc Sci Q. 2007;88(4):970–989. https://
de Janeiro: IBGE, 2017. Available in: https://biblioteca.ibge.gov.br/ doi.org/10.1111/j.1540-6237.2007.00513.x.
visualizacao/instrumentos_de_coleta/doc5568.pdf 29 BRASIL. LEI Nº 14.176, DE 22 DE JUNHO DE 2021. Available in:
23 Instituto Brasileiro de Geografia e Estatística. Pesquisa de Orçamentos http://www.planalto.gov.br/ccivil_03/_Ato2019-2022/2021/Lei/
Familiares 2017-2018: análise da segurança alimentar no Brasil. Rio de L14176.htm#art1. Accessed February 18, 2022.
Janeiro: IBGE, 2020. Available in: https://biblioteca.ibge.gov.br/index. 30 Brazilian Research Network on Food and Nutrition Sovereignty and
php/biblioteca-catalogo?view=detalhes&id=2101749. Security (Rede Brasileira de Pesquisa em Soberania e Segurança
24 Instituto Brasileiro de Geografia e Estatística. Pesquisa de Orça- Alimentar e Nutricional - Rede PENSSAN. II VIGISAN - National
mentos Familiares 2017-2018 : primeiros resultados. Rio de Survey of Food Insecurity in the Context of the Covid-19 in Brazil,
Janeiro: IBGE, 2019. Available in: https://biblioteca.ibge.gov.br/ 2022. Available in: https://olheparaafome.com.br/. Accessed
visualizacao/livros/liv101670.pdf. December 15, 2022.

10 www.thelancet.com Vol 18 February, 2023