Women and evaluation of inequalities in the distribution of risk factors for Chronic non-communicable diseases (NCD), Vigitel 2016–2017

Objective To compare the distribution of chronic non-communicable diseases (CNCD) indicators among adult female beneficiaries and non-beneficiaries of the Bolsa Família Program (BFP) in Brazilian capitals. Methods Analysis of Vigitel telephone survey data in 2016 and 2017. Gross and adjusted prevalence ratios (PR) and their respective confidence intervals were estimated using Poisson Regression model. Results Women with BF have lower schooling, are young people, live more frequently in the Northeast and North of the country. Higher prevalence of risk factors were found in woman receiving BF. The adjusted PR of the BF women were: smokers (PR = 1.98), overweight (PR = 1.21), obesity (PR = 1.63), fruits and vegetables (PR = 0.63), consumption of soft drinks (PR = 1.68), bean consumption (PR = 1.25), physical activity at leisure (PR = 0.65), physical activity at home (PR = 1.35), time watching TV (PR = 1.37), self-assessment of poor health status (PR = 2.04), mammography (PR = 0.86), Pap smears (PR = 0.91), hypertension (PR = 1.46) and diabetes (PR = 1,66). When women were compared among strata of the same schooling, these differences were reduced. Conclusion Worst indicators among women receiving BF reflect social inequalities inherent in this most vulnerable group. The study also shows that BF is being targeted at the most vulnerable women.


INTRODUCTION
Chronic non-communicable diseases (CNCD) are considered the main causes of death and disability in the world population, responsible for high costs and financial burdens on individuals, societies and health systems 1,2 . Studies show that CNCDs affect quality of life and well-being 3,4 , as well as the importance of social determinants, particularly poverty, in the occurrence of CNCDs, with the worst indicators in the most vulnerable and socially marginalized population [5][6][7] . In addition, health inequities cause unequal possibilities related to the scientific and technological advances that have occurred in this area, changing the chances of exposure to the factors that determine health and illness, further expanding inequalities and the risks of illness and death 7 .
In this regard, several social protection strategies have been used in the world to promote social mobility and face the problem of hunger and misery. Among them, conditional cash transfer (CCTs) programs have stood out for their contractual design and governance structure, which promotes the individual's positive behavior, as a central element, and not care 8 .
In the Brazilian context, the Bolsa Família Program (BFP) was created by the federal government through Provisional Measure No. 132 9,10 . The program is a social welfare action, being one of the largest direct cash transfer programs in the world and, by legal definition, women are the preferred holders for receiving the benefit (Law No. 10,836/2004) 11,12 .
The aim of the present study was to compare the distribution of CNCD indicators among adult women beneficiaries and non-beneficiaries of the BFP in Brazilian capitals and the Federal District.

METHODS
This is a population-based epidemiological study with a cross-sectional design, with data from Vigitel, years 2016 and 2017, from the 26 Brazilian capitals and the Federal District. The surveyed population was comprised of adult women (≥ 18 years and over). Approximately two thousand interviews are carried out per capital, about 54 thousand interviews each year 14. In the current study, 3,330 women benefiting from the PBF and 63,152 non-beneficiaries were studied. Data on the sample design can be read in other publications 14,15 .
In 2016, the following question was introduced: "Do you or someone in your family who lives at home receive Bolsa Família (BF)?". This is the first study to analyze this question. Currently, 92% of women are the beneficiary of the family, regardless of the type of family arrangement to which they belong 12 . Thus, indicators from the Vigitel database for adult women were analyzed.
The prevalences were compared, with the respective 95% confidence intervals (95% CI) of the following indicators: • Risk factors: a) smokers (reporting smoking, regardless of quantity); b) body weight (overweight -body mass index ≥ 25 kg/m 2 ; obesity -body mass index ≥ 30 kg/m 2 ) -the absent values of overweight and obesity suffered imputation, according to the methodology used by Vigitel and previously described 14; c) regular consumption of soft drinks or artificial juice (five or more days a week); d) habit of watching TV for 3 hours or more a day; e) abusive consumption of alcoholic beverages (four or more doses for women) on the same occasion in the last 30 days; f ) self-assessment of poor health status; g) referred morbidities (previous medical diagnosis of hypertension and diabetes); • Protective factors: h) recommended consumption of fruits and vegetables (five or more servings daily, five or more days a week); REV BRAS EPIDEMIOL 2020; 23: E200058 i) regular consumption of beans (five or more days a week); j) physical activity (PA) in free time (PA ≥ 150 minutes in the free time domains); k) conducting tests for early detection of cancer in women (mammography for women aged 50 to 69 years, in the last two years; and Pap smear for women aged 25 to 59 years, in the last three years).
For this analysis, it was necessary to join the Vigitel 2016 and 2017 databases and select only the women who received or did not receive the BFP benefit. The choice to study two years was a methodological option to increase the sample size of women who received the benefit.
It was necessary to calculate post-stratification weights, since the sample of adults interviewed by Vigitel was extracted from the register of residential telephone lines existing in each city, which only allows, strictly, inferences for the adult population residing in households covered by fixed telephone network. Thus, the use of post-stratification weights aims to minimize the addiction resulting from the low coverage of landline registrations, especially in the North and Northeast regions 14,15 .
In order to construct the post-stratification weights of this subsample of women, the estimate of the female population with or without PBF obtained by the 2016 National Household Sample Survey (PNAD) was used as the reference population 16 . The variables age, education and region were used in the construction of the weights, as these characteristics are associated with the possession of the PBF. Weights were obtained using the rake method 15 .
The descriptive analyzes consisted of calculating the distribution of the proportions of women who receive and do not receive the benefit of the PBF, according to sociodemographic characteristics and NCD indicators. The prevalence and the prevalence ratio of the indicators were analyzed among the population with and without PBF, using the Poisson regression model with robust variance, with adjustment for age. A stratified analysis was also carried out according to the different schooling strata (zero to eight years, nine to 11 years and 12 years and more), comparing women with the same level of education, beneficiaries or not of the PBF, with adjustment for age. Data analysis was performed using the Statistical Software for Professional (Stata) statistical program, using the commands of the survey module, taking into account the weights and, therefore, the sample's representativeness. The

RESULTS
Among the 66,482 women studied in 2016 and 2017, 3,330 live in households that received BF during the study period, and these residents are concentrated mainly in the Northeast REV BRAS EPIDEMIOL 2020; 23: E200058 and North regions -1,674 (50.3%) and 922 (27.7%), respectively. There is also a greater proportion of these beneficiaries between the ages of 25 to 54 years old (2,046/61.4%), without education and/or with incomplete elementary school (926/27.8%) and complete high school and or incomplete higher education (1,538/46.19%) ( Table 1).
The distribution of CNCD risk and protection factors for women with and without BFP, in the group of 26 capitals, and the estimated prevalence ratios (PR) are shown in Table 2  . There were no statistically significant differences between the prevalence of beneficiaries and non-beneficiaries for alcohol abuse ( Table 2). Table 3 shows the prevalence, PR and age-adjusted PR of CNCD risk and protection factors among women with and without BFP from the same age-adjusted level of education (zero to eight years of study). Women who are beneficiaries of the BFP with low education have a higher prevalence of tobacco use (PR = 1.51; 95% CI 1.16 -1.97), a higher prevalence of obesity (PR = 1.24; 95% CI 1.06 -1.45), lower recommended consumption of fruits and vegetables (PR = 0.71; 95% CI 0.56 -0.89), less PA practice at leisure (PR = 0.79; 95% CI 0.63 -0 , 98) and worse self-assessment of health status (PR = 1.46; 95% CI 1.07 -1.97). As a protection indicator, there was a higher consumption of beans (PR = 1.09; 95% CI 1.01 -1.17).

DISCUSSION
This is the first study that analyzes data from Vigitel for adult women who are beneficiaries of the BFP. This population is more vulnerable and has a lower family income. Among the sociodemographic indicators, the study highlighted: less education, greater concentration in the Northeast Region and younger population, among women beneficiaries of BFP. Among the indicators analyzed here, in general, the prevalence of risk factors was higher among women who receive the BFP benefit. They have a higher prevalence of smoking, overweight and obesity, less consumption of fruits and vegetables, higher consumption of soft drinks, however, on the other hand, the consumption of beans was higher. These women also had less PA practice during leisure time, however greater PA at home, or performing heavy cleaning tasks at home, in addition to more time watching TV. They had a worse self-assessment of their health status considered poor, less coverage of mammography and Pap smears and higher prevalence of self-reported morbidities (hypertension and diabetes). The study points out worse indicators in the poorest populations.
These differences decreased when women between strata with the same education were compared. When comparing the women beneficiaries of the PBF with 12 years or more of study with the others, the prevalence of risk factors was reduced, and the differences found were: higher prevalence of overweight, higher consumption of soft drinks and higher prevalence of hypertension self-reported among women with BFP. All other variables were similar, revealing that education is a major protective factor for women's health.
These data highlight the importance of social determinants, particularly poverty, in the occurrence of CNCDs, with the worst indicators in the most vulnerable population 5,17 .
The prevalence of smoking indicators was higher among women beneficiaries of BFP. Studies indicate that populations with higher education, income and better socioeconomic conditions have lower prevalence of tobacco use 18 , which has been explained by the greater access to information on the harms of this habit. It is noteworthy that the differences disappeared among more educated women, showing that their disadvantage can be overcome through greater investment in education.
Foods such as fruits and vegetables are considered protective and prevent cardiovascular diseases and cancers 4,19 . Vigitel's recommended consumption of fruits and vegetables (FLV) is a proxy for what would be recommended by the World Health Organization, or the daily consumption of 400g/day of fruits and vegetables 14,19,20 .
Likewise, beans are a protective food, as they are rich in fiber and nutrients, resulting in greater satiety and obesity prevention 21 , and are recommended by the Food Guide for the Brazilian Population, which recognizes its role in the national culinary tradition 22 . Because it is a lower cost food and has an important participation in traditional Brazilian cuisine, its consumption is higher among lower income populations. On the contrary, soft drinks, fats, sugars and salt are foods that increase the risk of CNCD 4,20,22 .
The study pointed out that women from the BFP have less consumption of FLV and more of soft drinks; the exception was beans. The Strategic Action Plan to Combat CNCDs encourages countries to adopt measures to regulate ultra-processed foods with a high sugar, salt and fat content, as well as measures that can increase production and consumption and reduce prices of fresh food such as vegetables 4,22 . These measures would bring great benefits to the population 4,22 , especially to the poorest, such as women beneficiaries of the BFP and their families. It is also noteworthy that having more education brought important benefits to the diet of the women studied here, regardless of the economic level.
Leisure-time PA was less practiced by women benefiting from the PBF, in contrast to domestic PA, represented by heavy cleaning, as well as longer TV time or sedentary leisure. Leisure time PA is associated with populations with high education and income, due to greater access to spaces for PA practice and greater knowledge about the benefits of PA 23 . On the contrary, PA practices during work and commuting to work are associated with low-income populations. In a special way, PA in the home is more affectionate to women, due to a macho aspect of Brazilian society, which considers that domestic activities should be performed by women 24 . It was also found that in the analysis stratified by education, these differences were reduced, pointing out that it is possible to change this reality, investing in public policies of access to public spaces for PA practice, as well as improving the population's income and education, aiming to decrease these inequalities 5 . Data from the National Health Survey indicate that overweight affects more than half of women, and obesity, 24%, revealing the extent of the problem in the country 25 . Also, obesity can affect the achievement of the global goal of reducing CNCD mortality 26 . The study shows that, although it is a widespread problem, it is even more serious among women beneficiaries of PBF, especially with low education.
Studies indicate that the prevalence of at least one chronic disease increased with age and was higher among women 27 , in addition to the presence of multimorbidities (two diseases and three or more diseases) 28,29 .
Self-assessment of poor health status was more frequent among women beneficiaries of BFP, but, when stratified by schooling, in strata with 12 or more years of study, this difference disappeared, which is in accordance with the literature 29 . This indicator is classically associated with worse health and life conditions, low income and the elderly population, and is an important predictor of mortality 30,31 .
The study showed that hypertension and diabetes morbidities were more prevalent among women with BFP, or low income, which has been confirmed in other studies, both by Vigitel, pointing out that the prevalence of hypertension in the population with low education was three times as high 32 than for diabetes 33 , with frequencies up to four times higher in the population with low education. Also, in international studies 35 , an association was found between educational level and diabetes mellitus, after adjusting the variables income and occupation 34 . The explanations for this finding are schooling, as a proxy for socioeconomic level, and less access to health promotion practices, such as healthy eating, PA, access to medicines and health services 17,35 .
The coverage of preventive cancer exams was lower among women beneficiaries of PBF, with frequencies below that recommended by the Strategic Action Plan for Coping with Chronic Diseases in Brazil, which foresees reaching 75% of mammography coverage and 85% of Pap smears in 2022 36 . It is noteworthy that the prevalence of oncotic cytology (Pap smear) is higher, possibly because they are performed by the Family Health Strategy, through the Unified Health System (SUS) 35 .
Alcohol abuse did not differ between the two groups of women, but, when stratified for 12 years, it was slightly higher among women with higher income, which is consistent with other studies 14 .
The study highlights health inequities, which are the worst indicators in the population benefiting from BFP, revealing the importance of social determinants in the health-disease process. Higher income populations have easier access to health services and promotion practices 4,5 . These results reflect the importance of investments in improving living conditions and education, which can directly interfere with health indicators.
Cash transfer programs are fundamental in tackling inequalities, especially with regard to strengthening actions related to gender, in the production of equity 37 . The study showed greater vulnerability of this population, which should also be prioritized by health services, in positive discrimination with regard to access to services, educational practices and public policies for health promotion. Thus, this study may contribute to the definition of health policies aimed at the prevention and control of these diseases.
Among the limits of the study, the use of telephone interviews stands out, seeking to reduce this bias using post-stratification weights. The fact that the information is self-reported can also result in information bias, although national and international experience highlights that variables such as arterial hypertension and health status assessment can obtain good estimates using this methodology, in addition to presenting advantages such as speed of information, sensitivity and low cost 38 . Due to the reduced number of interviews with BFP, it was decided to analyze 2016 and 2017, reducing the standard error of the estimates. Another limitation is the fact that the respondent is not the beneficiary of the program, since the question is about the benefit in the family. In this case, a proxy of the results is made. In addition, the cross-sectional design of the study does not allow the establishment of a temporal cause and effect relationship.

CONCLUSION
The study showed differences in risk and protection factors among women, with worse indicators among those who receive the BFP benefit. It is noteworthy that this does not reflect a causal relationship between receiving BFP and having a worse performance, but rather the social inequalities inherent to this most vulnerable group, characterized by little access to healthy food, places to practice PA, health services and practices health promotion, in addition to differences in schooling and less information about disease prevention and health promotion practices.