AIDS Among Young ... - CiteSeerX

Autonomy and Susceptibility To HIV/AIDS Among Young Women Living In a Slum In Belo Horizonte, Brazil Alessandra Sampaio Chacham 1 Malco B. Camargos2 Mônica Bara Maia3 Marília Greco4 Ana Paula Silva5 Dirceu B. Greco6 Keywords: Sexual Health; Reproductive Health; Gender; Youth.

ABSTRACT In Brazil in the last decade there has been an increased susceptibility in young women to HIV. This study explored the link between autonomy, or control over major domains of economic, social and sexual life, and susceptibility to HIV infection among poor young women living in an urban slum area of Brazil. A cross-sectional survey was carried out of a random sample of 356 young women between 15 to 24 years old. The study found that indicators of autonomy that relate to sexuality, mobility and freedom from threat by partners were significantly correlated with practices linked to HIV prevention and with access to health services promoting prevention, especially among adolescents. Young women who talked to their partner about condom use before first intercourse and co-decided with them on condom use were more likely to use condoms. In contrast, those who had ever been victim of physical violence by a partner, or whose mobility did a partner restrict were less likely to use condoms. The study found that young women were more likely to have accessed health services for gynaecological examinations and HIV testing if they had ever been pregnant. Access to family planning services was also more common for pregnant and married young women, and less likely for those who were less mobile, or who had been victims of physical violence. The findings indicate that structural gender inequalities translate into unequal relationships and reduced autonomy, increasing young women's susceptibility to HIV. The study suggests the development of programmes in public schools to discuss gender inequality, domestic violence and sexual and reproductive health, and reproductive health policy that ensures that young women access services before they get pregnant and that reaches young males. 1

Pontificia Universidade Catolica de Minas Gerais Pontificia Universidade Catolica de Minas Gerais 3 Rede National Feminista de Saúde e Direitos Sexuais e Direitos Reprodutivos 4 Projeto Horizonte, Universidade Federal of Minas Gerais 5 Projeto Horizonte, Universidade Federal of Minas Gerais 5 Universidade Federal of Minas Gerais 2

Autonomy and Susceptibility To HIV/AIDS Among Young Women Living In a Slum In Belo Horizonte, Brazil Alessandra Sampaio Chacham 7 Malco B. Camargos8 Mônica Bara Maia9 Marília Greco10 Ana Paula Silva11 Dirceu B. Greco12

1.INTRODUCTION With an estimated 660 000 people infected by HIV in 2004 (UNAIDS, 2005), the AIDS epidemic is certainly one of the major health concerns in Brazil. Although the number of notified AIDS cases have been steadily declining in the country in the last few years, there is a worrisome trend of feminisation of the AIDS epidemic in Brazil with the incidence in women increasing to that of men in recent years. According to the Brazilian Ministry of Health (MS, 2004), the male/female ratio of notified AIDS cases that was over 6:1 in the 1980’s has dropped to less than 2:1 in 2000 and is decreasing each year. Between January and July of 2004, the ratio of men to women infected fell to 1.5. However, this ratio conceals a difference among age groups: amongst young people aged 13-24 years the rate is higher among females. AIDS numbers are also growing among the more impoverished sectors of Brazilian population: blacks, people living in rural areas and in the outskirts of big cities. AIDS incidence rates have also been concentrating among people (especially women) with lower educational levels (MS, 2004). AIDS incidence patterns in Brazil are starting to follow the deep social and economic inequalities that characterize this country. Although Brazil has a very large economy, (one of the top 10 world economies in 2005 according to the World Bank), it occupies the 72nd position in the Human Development Index Rank (UNDP, 2003). The reason for this disparity is in the fact that Brazil is one of most unequal countries in the world: in 2001 the richest 10% controlled 48% of the GDP while the poorest 10% had a participation of 0.7% (UNDP, 2003). Inequality is pervasive in the country and also marked by gender, race and social class that put women, especially poor black young women in a very disadvantaged position and the shifting on HIV/AIDS trends reflects those inequalities.

7

Pontificia Universidade Catolica de Minas Gerais Pontificia Universidade Catolica de Minas Gerais 9 Rede National Feminista de Saúde e Direitos Sexuais e Direitos Reprodutivos 10 Projeto Horizonte, Universidade Federal of Minas Gerais 11 Projeto Horizonte, Universidade Federal of Minas Gerais 12 Universidade Federal of Minas Gerais 8

1

This data indicates that although an internationally acknowledged success in HIV/AIDS control, the Brazilian STD/AIDS Program has not been able yet to prevent the expansion of the epidemic among young women. One possible reason may be a failure of the government program and of other kinds of intervention to tackle the gender inequalities that are an important factor in young women’s susceptibility to HIV. Gender inequality creates a situation where young women lack of power and autonomy and so have difficulty negotiating prevention practices (Gage 2000). Hence the need for interventions focused on HIV/AIDS prevention to empower young women in the sexual sphere as well as other dimensions of their lives, so they will have enough autonomy to make informed choices and enforce their decisions concerning their sexual lives.

2. DIFFERENT DIMENSIONS OF AUTONOMY IN WOMEN’S LIVES AND SUSCEPTIBILITY TO HIV As defined by Mason (1997:158), gender refers to the socially constructed set of relations, attributions, roles, beliefs and attitudes that define what it means to be a woman or a man in social life. In most societies gender relations are unequal, with imbalances in relation to power between men and women. Gender imbalance is reflected in law, policies and social practices, as well as in people’s identities, attitudes and behaviours. The concept of gender, when it incorporates dimensions of power, exposes the asymmetries and the hierarchies (hierarchies in which women occupy a subordinate position) within the relations between men and women. Unequal gender relations between men and women tend to make difficult, if not impossible sometimes, for women to negotiate the use of condoms and to prevent HIV infection. Numerous studies have pointed out that the negotiation of safer sex through condom use is rendered problematic not only by the negative connotations associated with the method, but also by cultural attitudes toward female sexuality: women who want to practice safer sex may not be able to do so for fear of being considered immoral and untrusting and for fear of reprisals in the form of anger and rejection (Gage 2000:195). Negotiating of condom use is even more problematic when a woman is totally or mostly economically dependent on their partner and sex is one of few bargaining tools they have. Young women and girls are in a particularly vulnerable position, given their economic vulnerability, especially when they have unplanned pregnancies and/or an early marriage. In Brazil, both teen pregnancy and early marriage are more common in poor rural areas and urban slums (BEMFAM 1996). In this context, the ability of adolescent girls to negotiate whether sex will occur and whether condoms and contraceptives will be used may be further reduced. In many instances, the threat of male violence can also contribute to the pressure on teenage girls to agree to unsafe sexual practices. In addition to that, the presence of traditional gender roles and expectations reinforce especially among females the notion of romantic love, as Giffen (1999:284) comments: “for many women, unprotected sex means trust and intimacy while the use of a condom symbolizes multiple partners, and lack of trust and intimacy.” In Brazil, adolescents both male and female tend to reproduce traditional gender roles and to present a conservative approach in relation to sexual matters. The masculine model to be emulated is based on the capacity to provide for a family and to be sexually potent, while the female model

2

is based on maternity, with sexual attractiveness and romantic love as ideals. Sexual autonomy can be further reduced in this context. Autonomy has been defined as: “The degree of women’s access to, and control over, material resources (including food, income, land and other forms of wealth) and to social resources (including knowledge, power and prestige) with the family, community and society at large” (Jejeebhoy 2000: 205). Basically the concept of autonomy relates to the extent women exercise control over their own lives within the family in which they live (Jejeebhoy 2000: 205). Based in this definition, Jejeebhoy (2000:218) created five dimensions of autonomy and selected indicators for each. They are: economic and child-related decision-making; mobility; freedom from threat from husband/partner; access to economic/social resources; and control over economic resources. In regard specifically to the reproductive and sexual sphere autonomy means, according to Sen & Batliwala (2000:29), whether a woman/girl can safely determine when and with whom she will engage in sexual relations, sexual health, regulate her fertility and safe childbirth. Lack of autonomy in the sphere of sexuality can thus be considered a risk to women’s sexual health. In our study we use some of those indicators proposed to analyse which autonomy factors are more or less linked to susceptibility to HIV infection among poor young women living in a slum urban area of Brazil. We also sought to identify whether and how social organisations in the community; social policies and community services influence women’s autonomy and capacity to confront HIV and AIDS.

3. RESEARCH DESIGN AND METHODS Our research took place in one of Belo Horizonte’s poorest neighbourhoods: Taquaril. The city of Belo Horizonte, capital of the State of Minas Gerais in Brazil, is the third largest city of the country with 2,234,000 inhabitants (2000 Census), and over 3,500,000 in its greater metropolitan area. Taquaril is located on the extreme border of the eastern region of Belo Horizonte and it is one of the poorest areas of the city. According to the Belo Horizonte Censo Social of 2001, there were 2,437 young women between 15 to 24 years old living in Taquaril. Community health agents collected the data for this local Census and the information is kept in the families’ files at the local health centres. We used these files to locate the address of young women between 15 to 24 years old living in the area. A random sample of 332 young women was drawn from the list. We calculated the sample size using the equation for a probabilistic random sample with a significance level of 5% for the total population of young women between 15 to 24 years old living in the area. During fieldwork we discovered that around half of the people listed had moved away or could not be localized (even in the official city maps), making it necessary to draw 300 further names of young women to complete our sample. The fieldwork took two months in March to May 2005. Every adolescent/woman interviewed was contacted at home and signed an informed consent form. The interviews were done at any place the volunteer felt comfortable to talk--it was her decision. The interview lasted on average one hour. At the end, we had interviewed 372 women between 15 to 24 years old living in the Taquaril area.

3

Among those, 5 interviews were discarded because these volunteers were under 15 years of age, and 11 were discarded because of unexplained missing or contradictory information in the questionnaire. The final sample comprised of 356 questionnaires, divided in 179 from adolescents (between 15 to 19 years old) and 177 from young women (between 20 to 24 years old). Our 9 interviewers were under and post graduate students of Social Sciences, trained and supervised by the researchers who checked and coded questionnaires. The interviewers received a map of the neighbourhood and each one was responsible for 1 or 2 sectors (according to the official division of Taquaril area). Normally they were well received by the women and their families. Although the interviewers were instructed never to insist, the overwhelming majority of women accepted to answer our questions right away. We had only three reported cases of refusals, and in two cases it was because a father did not allow his daughter to participate. We got authorization to do this research from the Ethical Committee of Federal University of Minas Gerais. 3.1 INSTRUMENT FOR DATA COLLECTION AND ANALYSIS: We used the following more specific indicators of the five dimensions of autonomy: Area of autonomy Economic decision making Child related decision making Mobility and access to social resources

Control over economic resources

Freedom from threat

Indicator used in the study Who bought major household goods and more expensive goods, such as a car Not tested in this study (see below) The number of places a woman can go alone: to health centres, community centres, relatives' and/or friends' houses, shopping centres or to another city; if she has some activities for leisure; if she has access to TV, radio and book if the women has a house key; curfew hours; if she can go out with friends, with any kind of clothes she wants to wear and if she can use makeup. If has paid work and to control over how her own money is spent and how household money is spent; if she does not work outside home, if she has money for personal and household expenses; if she is free to buy objects for personal use and gifts; if she has a bank accoun Whether she fears and/or is exposed to physical emotional or sexual violence and abuse from a partner or other relative; if she has ever seen he mother being victim of domestic violence; if sh feels she can avoid a sexual intercourse or interrupt if she wants; if she can safely demand the use of condoms.

4

A preliminary round of 20 in-depth interviews with young women living in the Alto Vera Cruz (a neighbourhood with social and economic conditions very similar to Taquaril) explored which questions and approaches were most useful to understand sexual behaviour and practices among local young women and to understand the meanings they attach to autonomy and community support. A questionnaire was developed from this and also pre-tested at Alto do Vera Cruz. It included questions about: the social and economic conditions of the women interviewed; their sexual, marital and reproductive history; their knowledge, attitudes and practices on sexuality in general and in the prevention of HIV/AIDS in particular. Questions were asked about their autonomy level and capacity to make and enforce their decisions, not only about their sexual lives but also on three other dimensions of autonomy defined in section 2. The results led us to focus on the groups of indicators from the three dimensions of autonomy, beside sexual autonomy (being able to chose when and where to have sex and with whom), that seemed to be particularly related to the understanding of the relationship between autonomy, vulnerability and service uptake. The impact of local public service and community institutions on young women’s susceptibility to HIV was evaluated through their perceptions of those impacts and also indirectly, through their knowledge and practices of HIV/AIDS prevention. The questionnaire explored their utilization of the social support network available to them for health generally, particularly sexual and reproductive health, with a major focus on their access to and perceptions of public and nongovernment organisation (NGO) services. The questionnaire also explored barriers encountered in access to services. Finally the questionnaire explored knowledge, attitudes and practices on sexuality in general and on the prevention of HIV in particular. The answers from both the closed and open-ended questions were codified, entered into a database and analysed using the Statistical programme for Social Sciences (SPSS 11.5). The statistic test chi-square was performed and correlations were accepted upon an approximate significance level of 0.05 or below. In the logistic regression model, the variables were selected using the method stepwise forward, what means, at each step the variable that were not statistically meaningful (p-value> 0,050) were excluded until all the variable in the model presented significance level of 0,05 or less.

4. MAIN FINDINGS 4.1 Social, Economic Profile of Adolescents and Young Women Interviewed at Taquaril Data with social, economic and demographic information on the adolescents/young women interviewed in Taquaril is presented at table 1. The median household monthly income was low amongst our respondents, as expected, around US$ 150.00. Just over half of them (51%) declared an income below the poverty level established by UN of one dollar a day per capita. Over a third of all households were headed by females. Monthly income is lower than average in those households, however, albeit poorer, adolescents and young women living with their mothers had the same levels of education as those living in father headed households, and even had higher levels than those living with their husbands or by themselves. The most vulnerable households are those headed by respondents alone. Among the households with incomes of less than US$100.00 a month, respondents themselves headed 62%.

5

Three quarters of respondents were born in city of Belo Horizonte. Eighty-four percent identified themselves as black or brown (The Brazilian Census uses colours- white, brown, black and yellowinstead of race/ethnicity), and only 16% as whites. In the last census almost 50% of Brazilians reported to be black or brown, indicating a disproportionately high number of blacks living in this area. A third of the respondents (107 women) was married or united at the time of the interview and 137 (38%) had ever been married at least once, however, among 20 to 24 years old, the proportion of those married/united at least once almost doubled. Among those respondents who were not married or in a union at the time of the interview, 47.0% were dating and only 14.1% declared never having a boyfriend. Three women declared having a female partner at the time of the interview. There was a predominance of Catholics (49%), although this number is low compared to the 70% of Catholics among Brazilian population found in the last Census (IBGE, 2000). Evangelic/Pentecostals compose 34% of the sample, as compared to 15% for the general population (IBGE, 2000). Church frequency is quite high, over 60% reported going to church at least once a week, and for a large majority religion is very important in their lives. However 12% declared they did not belong to any religion at all. We found a higher number of single women among Catholics (50%), than among Evangelic/Pentecostal women (32%). In relation to schooling, among the 15-19 years age group, as expected a much higher number (70%) are still studying when compared to the 20-24 years age group (19%). It is important to notice that among those who left school 39% had already finished high school. This number rises to 45% among 15 to 19 years old and falls to 36% for 20 to 24 years old. Our data does not allows any inference on why those 15 to 19 years old tend to be better educated than those 20 to 24 years old, but it probably follows a general trend towards increasing secondary level enrolment in Brazil in recent years. Among those who quit school, 37% did so because of pregnancy or to care for children, or to marry. A much lower number of single women left school because of pregnancy (13%) than among those married now (22%). Also, among those who are still single a much higher number left school because they finished high school (58%) than among those married/united (20%). The probability of finishing high school is directly linked to family income. Sixty-five percent of the higher income group (with a monthly household income of over 500 US dollars) had finished high school as opposed to 24% of the general sample. Employment rate is quite low, considering that this is a population with a higher level of education (9 years of schooling) than most Brazilians have (where 5 to 6 years is the average for the general population). Only 26.7% among them have regular paid work. Even among the older ones, the employment rate is low. Although the number of 20 to 24 years old working (28.8%) is higher than among those between 15 to 19 years old (24.6%), it is not as high as might be expected. Among married women the employment rate is lower (23.4%) than among single women (28.2%).

6

Table 1 Data on social, economic, demographic characteristics and reproductive and sexual behavior of 15 to 24 years old females interviewed in Taquaril. Brazil, 2005. AGE GROUPS

Data on Social, Economic and Demographic Characteristics and Reproductive and sexual Behaviour

15 to 19 (%) N=178

20 to 24 (%) N=175

Monthly Income¹ Up to 100 dollars From 100 to 300 dollars 300 dollars and up Schooling Levels¹ 2ª to 5ª grade 6ª to 8ª grade Uncompleted high school Completed high school/some college Marital Status¹ Married/united Single/separated/widow Ever been married¹ Yes No Paid work² Yes No Religion² No religion Catholic Evangelical Head of household¹ Father Mother Husband Respondent Others Sexual experience and pregnancy¹ Had sex and got pregnant Had sex and didn’t get pregnant Never had sex Age at first intercourse¹ Up to 14 year old From 15 to 18 years old 19 years old and up Never had intercourse

22.7 65.6 11.7

26.9 52.0 21.1

5.6 40.4 39.3 14.6

21.3 23.6 21.8 23.3

11.2 89.4

49.1 50.9

14.0 86.0

62.9 37.1

24.2 75.8

28.6 71.4

14.6 48.3 37.1

11.7 51.5 36.8

47.2 38.2 7.9 0.6 6.2

26.3 20.6 41.7 5.7 5.7

18.5 33.1 48.3

73.7 19.4 6.9

16.3 35.4 0.0 48.3

20.6 60.6 12.0 6.9

7

Age at first Pregnancy¹ Up to 14 year old From 15 to 18 years old 19 years old and up Sexual Practices¹ Sex vaginal, oral and anal Sex vaginal and oral Sex vaginal only No sexual experience

24.2 69.7 6.1

7.0 61.7 31.8

2.3 10.8 38.1 48.9

8.6 28.2 56.3 7.5

¹Significant at p-value =0.05 or below ²Not significant at p-value=0.05 or below

Among those employed, only 32% have access to social security and other benefits. Over 90% of those who work are concentrated in only five types of jobs: working as maids, nannies, hairdressers and/or manicurists, cashiers or salesperson, and a few as secretaries or receptionists. Our respondents earn in average less than national minimum wage (around US$ 100.00). Among those who have an occasional occupation, for example as a manicurist or babysitter, earnings are much lower, around US 20.00 per month. Only 14% have a checking account or a savings account and 16.3% have no source of income at all. Compared to their partners or husbands, while they are generally not much older (with a 4 year difference), and have the same amount of schooling on average (9 years), with even fewer having graduated high school (19%) than the respondents, 76% of them are working right now. When we consider the kinds of jobs these men hold, we find much more diverse and better paid occupations: construction workers, mechanics, plumbers, repairmen, drivers and police officers, as well as other types of jobs in the service sector. The findings suggest a gender imbalance in economic power and access to economic resources. Most respondents (72%) had sexual intercourse at least once. At the time of first intercourse they were on average 15.8 years old, a similar age reported in other recent Brazilian surveys on youth and sexuality (BEMFAM, 1996; UNICEF, 2002; Ministério da Saúde, 2000; Aquino et al., 2003). Among single women, 55% already had their first sexual experience. On average they had two sexual partners. Religion did not seem to play a role in avoiding or delaying sexual initiation, nor in avoiding pre-marital sex. Catholics and Pentecostals had the same probability of having engaged in pre-marital sex by age of 16 (data not shown). Sexual practices other than vaginal sex (such as oral and anal sex) were reported by a much lower number of respondents and tended to occur after they had their first vaginal intercourse. Non-penetrative practices were also reported in lower numbers, the most common was sex at the entrance of vagina, reported by 41% of them. Among those who ever had sex, 63% have been pregnant at least once and most of them (84%) had their first pregnancy before 20 years old. The rate of very early pregnancy was high: 11% got pregnant before they were 15 years old. The median age at first pregnancy was 17 years, so that on average respondents get pregnant one year after they start having sex. The median age they enter their first marriage/union is also 17 suggesting both phenomenon are related. The prevalence of adolescent pregnancy (from 13 to 19 years old) was 38%. Not surprisingly adolescents/women who were married/united at least once are much more likely to ever been pregnant than single women: 90% of them compared to 17%.

8

While 70% currently use some contraceptive method, 86% declared not using any contraceptive method at the time of their first pregnancy. Lack of knowledge is not an explanation for this low level of contraceptive use, with only one respondent not knowing any contraceptive method. Most respondents (53%) declared they wanted their first pregnancy, so that non use of contraceptives or condoms at the time is more likely to be related to a desire for early motherhood. The practice of abortion seems to be extremely underreported, only 2.0% admitted ever having an abortion. This number is when compared to Aquino at al. (2004) that reported a 15.3% prevalence of abortion in a sample of women in Brazil. Low reporting may be because abortion is illegal and a taboo in Brazil, or because this sample truly abort less since teenage pregnancy is widespread. A much higher number (18%) declared having miscarriages, induced abortions may also be concealed among these numbers. Among those who have ever been pregnant, 87.6% had at least one child born alive and 44.6% two or more children. About eight percent (7.8%) were pregnant at the time of the interviews (this number was about the same for both age groups). Among those with at least one child born alive, 19.3% had children with 2 or 3 different partners. Pre-natal care prevalence is high with 97% of them getting pre-natal care (and 80% having 6 or more consultations) at their last pregnancy, generally at their local health centre.

4.2. DETERMINANTS OF CONDOM USE We used condom use as a major indicator of susceptibility. Condom use prevents risk of HIV and sexually transmitted infections and also enables women to exercise choice over pregnancy. We explored condom use at both first and last intercourse. We decided to use them to be able to pinpoint a pattern of condom use, since the beginning of their sexual history, comparing with current use, rather then a more generic question about condom use, such as “do you use condoms”, more likely to incite automatic positive answers. We found that 60.2% of respondents declared having used condoms during first sexual intercourse. Most of these did it to avoid STDs and/or pregnancy, with 9.1% mentioning avoiding STDs only. Among those who did not use condoms, not knowing about the importance of condom was the most commonly cited reason for non use (33%), while only 4% reported not using condoms because their partner did not want to. Though condom use is relatively high at the beginning of their sexual lives, it seems that it is rapidly forgotten or employed infrequently thereafter. A lower number reported using condoms the last time they had sex (45.7%). Older women (38% among 20 to 24 years old to 58.7% among 15 to 19 years old) and married women (21.7% to 62.4% among single) were less likely to have used condoms at their last time. For those who did use condoms, the main reason was to avoid STDs and/or pregnancy. Condoms are more commonly used as a contraceptive (33.3%). Non-use was thus related to acquired trust in partner (41.6%), followed by the use of another contraceptive method and being currently pregnant. Only 7.3% reported being pressured by their partners to not use condoms, while 10.9% declared it was their choice not to use it. 4.2.1 Determinants of Condom Use at First Intercourse: When controlling either by age condom use at first intercourse did not correlate with income, educational level, religious affiliation or colour/race. Tables 2 shows frequency of condom use at

9

first intercourse distributed according to some characteristics and sexual autonomy indicators, controlling by age group (15 to 19 years and 20 to 24 years). Current marital status was not significantly associated to condom use at first intercourse. Table 2 Frequency of condom use at first intercourse among 15 to 24 year females living at Taquaril, according to some autonomy indicators, controlled by age. Brazil, 2005. Frequency of condom use 20-24 15-19 at first intercourse by age groups Autonomy indicators Yes % No % Yes % No % N=66

N=26

N=88

N=75

Yes No

19.7 80.3

46.2 53.8

55.7 44.3

81.3 18.7

Yes No ¹ Ever been tested to HIV Yes No Sexual experience and pregnancy¹ Had sex and got pregnant Had sex and didn’t get pregnant Talked with partner about how to avoid children before first intercourse¹ Yes No Finds difficult to propose condom use to partner² Finds very difficult Finds somewhat difficult Finds no difficult at all Partner ever refused to wear condoms² Yes No Wanted first intercourse³ Yes No

28.8 71.2

53.8 46.2

67.0 33.0

93.3 6.7

39.4 60.6

65.4 34.6

69.0 31.0

88.0 12.0

28.8

53.8

67.0

93.3

71.2

46.2

33.0

6.7

68.2 31.8

28.0 72.0

72.7 27.3

32.0 68.0

86.4 9.1 4.5

62.5 20.8 16.7

80.5 16.3 9.2

76.1 14.9 9.0

16.7 83.3

45.8 54.2

26.1 73.9

39.7 60.3

77.3 22.7

65.4 34.6

79.5 20.5

61.3 38.7

Ever been married¹

Ever been pregnant¹

10

Who decides about what method to use³ She alone Partner alone Both of them

10.0 4.0 86.0

20.0 ---80.0

26.2 ---73.8

49.0 ---56.0

¹Significant for both age groups at p-value=0.05 or below ²Significant for age group 15 to 19 at P=0.05 or below ³Significant for age group 20 to 24 at P=0.05 or below

Adolescents/young women who have ever been married were less likely to have had used a condom in their first intercourse than those who never married, even though more than 90% of respondents who ever been married engaged in pre-marital sex. Adolescents were more likely to have used condoms in their first sexual intercourse than 20 to 24 year old women, even among those who had ever been married. Age thus has an effect independently of marital status. Those who have ever been pregnant were less likely to have used condoms at their first intercourse, in both age groups and among single women. This suggests that condoms are used to prevent pregnancy more than to prevent risk of sexually transmitted infection, and also that women trust partners after marriage. To have ever been tested for HIV is also associated with non use of condoms. This probably reflects the high number (97%) of women who were tested for HIV during their last pregnancy, reflecting the correlation between pregnancy and reduced condom use. Adolescents/young women who live with their mother were also more likely to have used a condom at their first intercourse. Although not a significant association, among single women, those who live in households headed by their mothers present the highest proportion of condom use at first intercourse (75.4%) when compared to those who live in household headed by their fathers, husbands or other relatives. Some indicators of sexual autonomy were correlated with the use of condom at first sexual intercourse. Respondents who discussed with their partner about how to avoid a pregnancy before having their first sexual experience had a higher frequency of condom use at their first intercourse, even when controlling by age and marital status. This is an important variable related both to the capacity of women to negotiate condom use and men’s important role in this process. Those who wanted to have their first sexual intercourse at the time they had it were more likely to use condoms at the first intercourse, as well those who decided together with partner what contraceptive method to use. This correlation was, however, only significant among 20 to 24 years old respondents. Variables related to mobility such as: “to ever been forbidden of wearing some clothes”, “to ever been forbidden to have a friend” and “to have a time set to arrive at home”, are all related to a lower frequency of condom use at the first sexual intercourse among adolescents (Table 3). The kind of relationship (if more equal or authoritarian) that a young woman is likely to establish with her partners seems to influence in the likelihood of condom use, even before she gets married (or united), especially among younger respondents. To have ever been a victim of physical violence and to have ever been a victim of physical violence by a husband was associated with a lower frequency of condom use at the first time. We found a high level of domestic violence found in the area. Thirty-seven percent of them have seen their mothers being hit by their fathers/stepfathers. Among women who have ever

11

been married, 27% have been victims of physical violence from a partner/husband. Domestic violence significantly increases young women’s susceptibility to HIV and is directly linked to gender inequality and lack of autonomy among women. Table 3 Frequency of condom use at first intercourse among 15 to 24 year females living at Taquaril, related to some autonomy indicators, controlled by age group. Brazil, 2005. Frequency of condom use at first 15-19 20-24 intercourse by age group Autonomy indicators Yes % No % Yes % No % N=66 N=26 N=88 N=75 Ever been victim of physical violence³ Yes 40.9 57.7 25.0 56.0 No 59.1 42.3 75.0 44.0 Who hit¹ Parent 37.9 38.5 13.6 17.3 Partner 3.0 19.2 11.4 38.7 Never was hit 59.1 42.3 75.8 44.0 Ever prohibited to wear some kind of clothes³ By parents 16.7 19.2 9.1 4.0 By boyfriend 36.4 15.4 20.5 13.3 By husband/partner 6.0 19.2 17.0 38.7 Never was prohibited 40.9 46.2 53.4 44.0 Ever prohibited to have friend(s)² By parents 19.7 30.8 6.8 5.3 By boyfriend 21.2 3.8 11.4 12.0 By husband/partner 7.6 19.2 19.3 34.7 Never was prohibited 51.5 46.2 62.5 48.0 Has time limit to get home² Established by parents 56.1 30.8 11.4 13.3 Established by husband/partner 1.5 11.5 4.5 6.7 Has no time limit 42.4 57.7 84.1 80.0 ¹Significant for both age groups at p-value=0.05 or below ²Significant for age group 15 to 19 at P=0.05 or below ³Significant for age group 20 to 24 at P=0.05 or below

A multivariate analysis, using logistic regression, indicated risk factors associated with the use/non use of condoms at first sexual intercourse. The results indicate that women who talked with their partners about how avoid children before first intercourse are 5.8 times more likely to use condoms compared to those who did not talk about it. Women who have ever been tested for HIV were four times less likely to have used condoms at their first time, while partner refusal at any time to use condoms also increases the chances of not using a condom at first time when compared with those whose partners never refused. The first two variables can be considered as indicators of sexual autonomy, as they reflect women’s capacity to negotiate condom use with their partners and the importance of an open communication between partners and men’s role in this process.

12

Given the importance found in our evidence that talking to partners about preventing pregnancy before having first sexual intercourse in relation to condom use, we explored other independent variables that could be linked to the probability of adolescents/young women talking with their partner about contraception before intercourse. There was a greater likelihood of discussing condom use with partners in younger and single women, in women who were 18 years or older at first intercourse, and in those who talked with the mother about sex. Schooling levels, having had sex education in school and having received information on HIV prevention in school were significantly positively associated with talking with partners about contraception, but only for 20 to 24 year olds. These results suggest that the exposure to sexual information either at home or at school increases the likelihood of has young women engaging in preventive behaviour regarding HIV and AIDS.

4.2.2 Determinants of Condom Use at Last Intercourse: At last sexual intercourse, both older and married women were less likely to have used condoms, as were those who had ever been married, or who had ever been pregnant. Condom use at first intercourse was not associated to condom use at last intercourse. Condom use at last intercourse was higher for adolescents in a relationship where they decided together what contraceptive to use and for single women who lived in mother headed household. Some indicators women’s sexual autonomy and less traditional relationships between men and women seem thus to be positively associated with condom use at last sexual intercourse. There are some factors that mediate in this. Condom breakage during sex is associated with less condom use for adolescents and single women (maybe reflecting their lack of experience with condoms). Women’s belief that condoms interfere with sex (either by hurting or by diminishing the pleasure) also influenced the frequency of condom use at the last time among adolescents. This is a subject that is treated frequently from the male point of view however it also seems to affect how women perceive condoms and their willingness to use them. Adolescents’ complaints about condoms hurting or interfering in pleasure may also indicate that they do not have enough lubrication (related maybe to not feel enough pleasure during sex or the presence of some vaginal infection). Two variables presented an unexpected influence at on condom use at last time: single women who used alcohol or drugs were more likely to have used condoms and single and married women who declared they enjoyed sex were less likely to have used it. Women in a more stable relationship were more likely to declare enjoying sex and also to use fewer condoms. Women who used alcohol or drugs at their last intercourse were less likely to be married (91.7% of them were single) and thus more likely to wear condoms. Table 4 indicates that married women who participated in a family planning group were more likely to have used condoms, the opposite association found with first time. This is an encouraging finding about the role of health centres in prevention, after women star frequenting them.

13

Table 4 Frequency of condom use at last intercourse among single and married females living at Taquaril. Brazil. 2005. Has used condom at last Married Single/separated intercourse by marital status Autonomy indicators Yes % No % Yes % No % Ever been pregnant³ Yes 95.7 88.0 37.6 57.1 No 4.3 12.0 62.4 42.9 Head of household³ Father 4.3 4.8 38.7 55.4 Mother 4.3 3.6 53.8 32.1 78.3 83.1 --------Husband Respondent 8.7 2.4 2.2 7.1 Others 4.3 6.0 5.4 5.4 Condom has ever broken³ Does not use ---7.3 ----6 Many times 4.3 1.2 1 6 A few times 22.7 39.0 39 44 Never 73.9 52.4 60 44 4.3 ___ 7.5 5.2 Used alcohol or drugs at last 95.7 100.0 92.5 94.5 intercourse³ Enjoys sex² Yes 62.5 85.5 67.7 83.9 Sometimes 17.4 3.6 4.3 1.8 No 17.4 18.8 28.0 14.3 Ever participated in Family Planning at health centre² Yes 78.3 52.4 32.3 35.2 No 21.7 47.6 67.7 64.8 Sexual experience and pregnancy³ Had sex and got pregnant 95.7 88.0 37.6 57.1 Had sex and didn’t get pregnant 4.3 12.0 62.4 42.9 ¹Significant for both age groups at p-value=0.05 or below ²Significant for married/united women only at P=0.05 or below ³Significant for single/separated women only at P=0.05 or below

A multivariate analysis, using logistic regression, indicated risk factors associated with the use or non use of condoms at last sexual intercourse. Results indicate that condom use in the last sexual intercourse was lower in married women, in those attending church more often and in women who declared enjoying sex. These findings differ from those related to condom use at first time, indicating that the non use of condoms is a woman’s decision too, reflecting her beliefs about love, relationships and intimacy.

14

4.4 AUTONOMY AND ACCESS TO INFORMATION AND HEALTH SERVICES How HIV is transmitted and how to avoid it is common knowledge among our respondents. All respondents (100%) knew that having sex without condoms can infect them and only 3.9% declared that they did not know how to avoid the AIDS virus. AIDS is clearly not seen anymore as a disease restricted to “risk groups” such as homosexuals, sex workers or drug addicts: 42.2% of them believe that they are at risk of contracting HIV/AIDS. The disease is also not removed from their direct experience: 35.9% know someone who has the virus and 48.4% knew someone who died of AIDS. However, reasonably high numbers of them still hold several misconceptions and misunderstandings regarding the possibility of getting infected by: donating blood (53.5%), kissing (32.4%), mosquito bite (25.8%), or by sitting on a toilet (20.3%). We did not find any association of the respondents’ levels of knowledge and their attitudes about prevention and risk perception with prevalence of condom use both at first and last intercourse. These results indicate that the factors linked to condom use go much beyond information and knowledge on where to access condoms. Of course that does not mean that other barriers to access, such as shame, lack of privacy during utilization of services, lack of methods at health centres and lack of money to buy condoms at a drugstores do not play a large role in blocking youth’s access to condoms. Nonetheless, it is also clear that access and knowledge are necessary but not sufficient conditions to ensure condom use. In relation to source of information on prevention, 44.1% of respondents said they had talked with a professional at the health centre about STD/HIV/AIDS. However, among those who had at least one gynaecological consultation (63% of respondents), only 34.8% reported having discussed about HIV/STD/ AIDS at their last consultation. Our respondents declared that in the last 6 months, the main source of information for them about AIDS was the media, followed by school and then in third place the health centre. However, they still expected health services to have an active role in helping them to confront the HIV/AIDS epidemic, since the large majority of respondents declared they would seek a health professional if they needed more information about STDs/AIDS. When we analysed the impact of the different dimensions of women’s autonomy on health service uptake we used as indicators of access to health centre two variables: (1) “has ever participated in the family planning group at the health centre”; and (2) has ever received a visit from the community health agent. With those variables we expected to identify two different forms of access to public health services. One, represented by the participation on Family Planning Groups, depends on active demand from the user for contraceptive methods. The second, by receiving a visit from the community health agent does not depend of users’ demand to happen, and is a routine part of the health agent’s workload. Keeping these differences in mind, our results suggest that autonomy indicators have a differentiated impact on access; depending of the kind of access we are referring to. The more “passive” form of access, that is receiving a visit from a health agent, showed no correlation with different autonomy indicators and social economic indicators. However, in relation to the participation in Family Planning groups, a “pro-active” form of access, given that women have to go to the centres to participate in them, we found that the variable “to have an established time to arrive at home by parents” (linked to mobility and access to social

15

resources) had a negative impact on participation among adolescents. In relation to sexual autonomy indicators, the variable “enjoys having sex” was positively associated with participation in a PF group only among teenagers from 15 to 19 years old, suggesting that a positive relationship with sexuality among the young is an important motivation for the search of prevention. On the other hand, to have ever had been victim of physical violence has a negative impact in the access of 20 to 24 years old to PF groups (data not shown). Among the social and demographic indicators, to have ever been married has a strong positive correlations with the participation in a PF group (60% to 24% among singles), while having ever been pregnant has a strong correlation with having frequented the Family Planning group at the health centre at least once (57% to 19% among women who never got pregnant), even after controlling for age and marital status. Pregnancy is also associated with having ever had a gynaecological consultation (98% to 72%), also controlling for both age and marital status. These results indicate that adolescents and young women tend to be included in a more structured way inside the health services only after their first pregnancy. Services offered by the primary health care services still keep focusing their actions to mother and baby duo and not to young women needs in general.

5. FINAL COMMENTS Our findings indicate that there are associations between young women’s condom use, as an indicator of susceptibility to HIV, and some indicators of autonomy. This is particularly so for indicators linked to autonomy in the dimensions of sexuality, mobility, and freedom from threat from a husband/partner. Age and marital status, and particularly the latter, also determine condom use. Our data also suggests that information alone related to prevention practices and access to condoms is not enough to assure its use among adolescents/young women. Information and access to social resources is not a major determinant of condom use due to its wide presence across all social groups in Brazil. Young women know how to avoid pregnancy and HIV and also where to get condoms and other contraceptive methods. Levels of knowledge about how to avoid the HIV/STDs and pregnancy are high, even if several misconceptions still exist. It is also a widespread knowledge that health centres distributes condoms and contraceptives, although we may question the quality of the access to services they offer, still marked by age and pregnancy status. Our respondents’ autonomy and ability to reduce risk is clearly affected by the kind of relationships they establish with their husband/partners. Unequal relationships, seems to have an explanatory role in young women’s susceptibility to HIV/AIDS. It is also very clear in our results that men’s role is fundamental in this process of successful negotiating a healthy and enjoyable sexual life for both partners. Nevertheless, men’s role are much less studied and understood. Consequently, young men almost are a lesser focus of public health policies. On the other hand, while young women are frequently the targets of sexual and reproductive programs, this appears from our data to be after they have become pregnant, in their roles as mothers, and they appear to be ignored by the other programs, governmental or not, that currently

16

focus in empowering young men through vocational training and job placement. The few programs geared towards young women tend to reproduce traditional gender roles by teaching them for instance how to be manicurists or nannies, jobs that generally are poorly paid and without opportunities for professional growth. Hence although women tend to be better educated, they are less likely to hold a job and have far less work options that men have, so that programmes that do not challenge these structural inequalities may end up and reinforcing gender inequality. As young women graduate from high school and find themselves without real prospects, getting pregnant and marrying (or not) may seem a logical solution to what to do with their lives. We propose that giving wider economic opportunities to young women, getting them into the health services before they get pregnant and attracting more men into reproductive and sexual health programmes would be a welcome, necessary shift in the public policies that have developed until now. We further propose that social programs improve young women’s livelihood opportunities through professional training, and that preventive programs be implemented to curb domestic violence.

17

REFERENCES

1. AQUINO E. M. L. et al. 2003. Adolescência e reprodução no Brasil: a heterogeneidade dos perfis sociais. Cad. Saúde Pública, vol.19, suppl. 2, Rio de Janeiro. 2. BATLIWALA, Srilatha. 1994. “The Meaning of Women’s Empowerment: New Concepts from Action”, in Gita Sen, Adrienne Germain and Lincoln Chen (eds.), Population Policies Reconsidered: Health, Empowerment and Rights. Cambridge, Mass: Harvard University Press. 3. BEMFAM – Sociedade Civil Bem-Estar Familiar no Brasil. 1999. Adolescentes, Jovens e a Pesquisa Nacional sobre Demografia e Saúde: um estudo sobre fecundidade, comportamento sexual e saúde reprodutiva. Rio de Janeiro: BEMFAM. 4. GAGE, Anastasia. 2000. “Female Empowerment and Adolescence”. Women's Empowerment and Demographic Processes. Org. Harriet B. Presser e Gita Sen. Oxford: Oxford University Press. 5. IBGE. Instituto Brasileiro de Geografia e Estatística. CENSO 2000. Available at: www.ibge.gov.br/home/estatistica/populacao/perfil. 6. JEJEEBHOY, Shireen.2000. "Women's. Autonomy in Rural India: Its dimensions, determinants and the influence of the context". Women's Empowerment and Demographic Processes. Org. Harriet B. Presser e Gita Sen. Oxford: Oxford University Press. 7. SEN, Gita, BATLIWALA, Srilatha. "Empowering Women for Reproductive Rights". Women's Empowerment and Demographic Processes. Org. Harriet B. Presser e Gita Sen. Oxford: Oxford University Press. 2000. 8. GIFFIN, Karen.1998. “Beyond Empowerment: Heterosexualties and the Prevention of AIDS.” Social Science and Medicine, Volume 46, No. 2, pp. 151-156. 9. MASON, Karen O. 1993. The Status of Women: a Review of its Relationships to Fertility and Mortality. The Rockefeller Foundation, New York. 10. MINISTÉRIO DA SAÚDE – MS. Coordenação Nacional de DST/AIDS. 2005. Buletim Epidemiológico AIDS. Ano XVII No. 1. 01 à 52 Semanas Epidemiológicas Janeiro a Dezembro. 11. Ministério da Saúde, Secretaria de Políticas de Saúde, Coordenação Nacional de DST/HIV/AIDS, CEBRAP. 2000.Comportamento Sexual da População Brasileira e Percepções sobre HIV/Aids. Brasília: Ministério da Saúde. 12. UNICEF. 2002. A voz dos adolescentes. Brasília. Disponível em www.unicef.org/. 13. United Nation Development Programme. UNDP. 2003. Human Development Report 2003. New York: Oxford University Press. Available at http://hdr.undp.org/reports/ 14. United Nations Programme on HIV/AIDS. UNAIDS. 2005.The 2004 Report on the Global AIDS Epidemic. http://www.unaids.org/bangkok2004/report_pdf.html

18