Factors affecting behaviours that address H IV risk among Black and White South Africans

The aim of this study was to identify factors affecting HIV risk reduction among 150 Black and 150 White South Africans chosen by systematic random sampling. Main outcome measures included sexual behavior and condom use, knowledge about correct condom use, intention of condom use, behavioural norms, attitudes, normative beliefs, and subjective norms about condoms, HIV/AIDS health beliefs, and HIV risk behaviour. Bivariate analysis gave positive significant relations among being single, age, HIV/ AIDS perceived severity, HIV/AIDS prevention barriers and HIV risk behaviour. Further, bivariate analysis gave nega­ tive significant relations among age at onset of puberty, age at first vaginal intercourse, correct condom use knowledge, subjective norms, intention to use condoms and HIV risk behaviour. Regression analysis indicated that for subjective norm to use condoms, less intention for condom use, less condom use knowledge and younger age of first vaginal intercourse were predictive for HIV/AIDS risk behaviour. HIV prevention intervention programmes should include the identified factors and cultural diversity. Introduction Problem statem ent Abdool Karim and Abdool Karim (2000: 3) indicate that the annual national antenatal HIV seroprevalence rate in South Africa rose more than 21-fold from 1990 (0.76%) to 1997 (16.01 %). It is expected that the prevalence of AIDS will reach about 27% of the sexually active population in South Africa by the year 2010 (Webb 1997: 214). Condoms are an integral part of STD and HIV/AIDS preven­ tion, and their use has increased significantly over the past decade. Correct use of condoms reduces the risk of HIV trans­ mission by almost 100 percent. Therefore, condom promotion has received considerable attention in the fight against the AIDS pandemic (World Health Organisation, 1995: 3ff.). Yet condom use is among the most difficult issues to address in designing programmes to reduce the sexual transmission of HIV in Africa. Campbell (1997: 187) summarizes for sub-Saha­ ran Africa that negative attitudes toward condom use are often based on cultural factors, e.g. the desire for children and fe­ male sexual compliance as ways to achieve economic status. Among Tanzanian adults it was found that 320 (20.6%) reported having ever used a condom and of the 320 respondents who had ever used a condom, 34(10.6% ) reported having used it at the last sexual intercourse (Mnyika, Klepp, Kvale, Schreiner & Seha 1997:176). In South Africa, aversion to condom use is the dominant theme, although explanations for this vary; cultural beliefs are also a barrier to condom use in South Africa, where many people consider it essential that the sperms of the men actually enters the woman (Gould 1993: 132f.). All in a sample of 100 urban black mothers questioned in Durban stated that they had never experienced sexual intercourse where their partner had used a condom (Abdool Karim, Abdool-Karim & Nkomokazi, 1991; 340). The highest incidence of condom use was in a sample of gold m iners, w here 32.6% had used them at least once (Ijsselmuiden, Padayachee, Mashaba, Martini & Van Staden, 1990: 520). Kirigia and Muthuri (1999: 488) found that white South African women have a higher likelihood of asking their new sexual partners to use condoms than their African, Col­ oured and Indian counterparts, indicating cultural diversity in factors affecting condom use.


Introduction Problem statement
A bdool Karim and Abdool Karim (2000: 3) indicate that the annual national antenatal HIV seroprevalence rate in South A frica rose more than 21-fold from 1990 (0.76%) to 1997 (16.01 %).It is expected that the prevalence of AIDS will reach about 27% o f the sexually active population in South Africa by the year 2010 (Webb 1997: 214).Condom s are an integral part of STD and HIV/AIDS preven tion, and their use has increased significantly over the past decade.Correct use of condom s reduces the risk o f HIV trans m ission by almost 100 percent.Therefore, condom promotion has received considerable attention in the fight against the AIDS pandem ic (World Health Organisation, 1995: 3ff.).Yet condom use is among the most difficult issues to address in designing program mes to reduce the sexual transm ission of HIV in Africa.Campbell (1997: 187) summarizes for sub-Saha ran Africa that negative attitudes toward condom use are often based on cultural factors, e.g. the desire for children and fe male sexual com pliance as ways to achieve economic status.Among Tanzanian adults it was found that 320 (20.6%) reported having ever used a condom and of the 320 respondents who had ever used a condom, 34(10.6% ) reported having used it at the last sexual intercourse (Mnyika, Klepp, Kvale, Schreiner & Seha 1997:176).In South Africa, aversion to condom use is the dominant theme, although explanations for this vary; cultural beliefs are also a barrier to condom use in South Africa, where many people consider it essential that the sperms of the men actually enters the wom an (Gould 1993: 132f.).All in a sample of 100 urban black mothers questioned in Durban stated that they had never experienced sexual intercourse where their partner had used a condom (Abdool Karim, A bdool-Karim & Nkomokazi, 1991; 340).The highest incidence o f condom use was in a sample of gold m iners, w here 32.6% had used them at least once (Ijsselmuiden, Padayachee, M ashaba, Martini & Van Staden, 1990: 520).Kirigia and Muthuri (1999: 488) found that white South African women have a higher likelihood o f asking their new sexual partners to use condoms than their African, C ol oured and Indian counterparts, indicating cultural diversity in factors affecting condom use.

Rational
In order to design effective education and other interventions to reduce risk, it is not only important to know what the knowl edge, beliefs, and risk behaviours are, but also to know what factors are associated with these risk behaviours (Mathai, Ross, & Hira 1997: 563).One's risk of HIV infection is a function of the number of partners one has, the behaviours one engages in with those partners (such as sexual intercourse without use of a condom), the probability that those partners have HIV/ AIDS, and the transmissibility o f the disease (Brunham & Plummer 1990: 1339).Literature indicates that factors affecting HIV risk reduction are a range of situational, interper sonal and structural factors such as knowledge about AIDS, behavioral intention, perceived vulnerability, perceived barriers, self-efficacy, and demograhic factors (Basen-Engquist 1992:120, Bengel, Belz-Merk & Farin 1996:505).

Aim
Since there was a lack of studies in South Africa, the purpose o f this study was to identify factors affecting HIV risk reduc tion among Blacks and W hites in South Africa.

Objectives
-To identify the association between biographic factors, sexual history, health beliefs, condom use knowledge, behavioural and subjective norms to use condom s, normative beliefs to use condoms, attitudes towards condom use and condom use intention with HIV risk behaviour.-To identify independent predictors for HIV risk behaviour.

Design
The study is an interview -adm inistered telephone survey us ing a cross-sectional design.

Sample and procedure
The sample included 150 Blacks in a formerly Black area (town ship) and 150 W hites in a form erly W hite urban area in the Limpopo Province of South Africa.Participants were chosen by a l-a-30 systematic random tel ephone sampling, excluding nonresidential numbers, from the directory of one formerly Black and one formerly White area.A research assistant trained in interviewing made telephone calls until, in each sample, 150 participants-in the age range from 21 to 50 years-had been identified.Participants were assured of W hites (51.4%) (51 male and 63 female), in the age range o f 21 to 50 years (M age 32.5 yr., SD=6.7) for Blacks and in the age range o f 21 to 50 years (M age 32.3 yr., SD=8.4).

Measures
The questionnaire was pilot tested on 15 Black and 15 White adults, who did not form of the final sample.No modifications were required after pilot testing.The questionnaire consisted o f questions about: (a) Biographic data (6 items); (b) Sexual activity and condom use (12 items); (c) Knowledge o f correct male condom use (10 items) (Sharma, Dave, Sharma & Chauhan, 1997:711); Cronbach alpha as well as split-half reliability coefficients for condom knowledge scale were .62 and .59for this sample.(d) Intention o f using condoms (when having sex next time w ith a new partner) (1 item ) (from -2 = stro n g ly agree to + 2-strongly disagree); (e) Behavioral norm to use condoms (3 items), such as "Do you and your friends talk about using condoms?" (rated l=Yes, -1 =No, 0=D on't know); (f) Attitudes towards condoms (3 items), such as "It is a good  complete anonymity.They were told that their name was ran domly selected from the telephone directory and that no names were written on the filled in questionnaires.
From the 300 participants 45 (15%) opted against participation, and from the remaining 255 only those cases were analyzed who had had a (vaginal) sexual experience.These were in total 222 (87%), 108 Blacks (48.6%) (36 male and 72 female) and 114 idea for me to use condom s!" (rated from -2=strongly agree to +2=strongly disagree); important to me think I should use condoms) (1 item) (rated from -2=strongly agree to +2=strongly disagree); (i) A 16-item A ID S Health B elief Scale (AHBS), developed by Zagum ny and Brady (1998: 173), to measure the four com po nents of the Health B elief M odel (HBM): perceived suscepti bility to disease, perceived severity of a specific disease, per ceived benefits o f preventive behaviour, and barriers to pre ventive behaviour.Scaling involved the use o f a 6-point Likert type scale with responses o f "strongly agree" weighted 6 and "strongly disagree" w eighted 1.For all sub-scales, higher scores represent a greater am ount of that belief.Cronbach al pha as well as split-half reliability coefficients for the AIDS Health B elief Scale were .67 and .56 for this sample.A H IV/AID S risk index included the following items: (i) H is tory o f sexually transm itted disease, (ii) Had sex under the in fluence o f alcohol, (iii) Frequency o f vaginal sex during the last 3 months (without condom), (iv) Frequency of oral sex during the last 3 months without condom, (v) Frequency of anal sex during the last 3 months without condom, (vi) Number of sexual partners in the last 12 months, (vii) Frequency of vaginal inter course in the past 3 months with a 'casual' partner and (viii) Frequency o f vaginal intercourse in the past 3 months with a commercial sex worker.

Bivariate analysis
Bivariate analysis indicated that being single, age, HIV/AIDS perceived severity as well as HIV/AIDS prevention barriers w ere positively associated, and age at onset of puberty, age at first vaginal intercourse, correct condom use knowledge, sub jective norm s as well as intention to use condoms were nega tively associated with HIV risk behaviour.Culture, gender, HIV/ AIDS perceived susceptibility, perceived prevention benefits, behavioural norm s, norm ative beliefs and attitudes towards condom use were not associated with HIV risk behaviour.

Regression analysis
Only the variables, which were correlated with the HIV/AIDS risk index, were included in the equation.Stepwise multiple regression analysis is shown in Table 2. tive of HIV/AIDS risk behaviour.The model accounted for 74% o f the variance (adjusted R2: .528).Excluded from the equation were age, age at onset of puberty, perceived prevention barri ers and perceived H IV/AIDS severity.

Discussion and conclusion
Subjective norm to use condoms ( 'M ost people who are im portant to me think I should use condom s), condom use inten tion, correct condom use know ledge were inversely and younger age of first vaginal intercourse was predictive for HIV/ AIDS risk behaviour.This means the improvement of subjec tive norms to use condoms, condom use intentions and knowl edge of correct condom use in HIV prevention programmes could lead to the reduction of HIV risk behaviour.The finding that age at first vaginal intercourse was negatively related to high risk behavior may be explained by factors such as an increase of the number and type of sexual partners, history of a sexually transmitted disease and frequency o f sex with in creasing age.Flisher, Ziervogel, Chalton, Leger and Robertson (1993: 596) state that age at first intercourse may be a poor indicator of risk owing to the possibility of a long time interval between the first and subsequent sexual encounters and infre quency of sexual intercourse.In line with other literature it was found that factors affecting HIV risk reduction are a range of situational, interpersonal and structural factors such as knowledge about correct condom use, behavioral intention, perceived severity, perceived prevention barriers, and demograhic factors (Basen-Engquist 1992:120,Bengel, Belz-Merk & Farin 1996:505).Further investigation is needed on factors associated with HIV risk behavior such as marital status, age, HIV/AIDS perceived severity, and perceived prevention barriers, and they should be included in HIV prevention intervention programmes.

Acknowledgement
The financial assistance of the National Research Foundation (South Africa) towards this research is hereby acknowledged.Opinions expressed this publication and conclusions arrived at, are those of the author and are not necessarily to be attrib uted to the National Research Foundation.

(
Items 13, 14, 15, 16  were reverse scored), *** p<.001, ** /?<.01, * p<.05 (g) Normative beliefs about condoms (3 items referring to par ents, sex partners, close friends), such as "My partner thinks I should use con d o m s" (rated from -2 = stro n g ly agree to +2=strongly disagree); (h) Subjective norms about condoms (M ost people who are 20 Curationis August 2002 Subjective norm to use condom s, intention o f condom use, correct condom use know ledge were inversely predictive, w hereas younger age o f first vaginal intercourse was predic-

Table 1 :
Descriptive and bivariate analyses for the HIV/AIDS risk index

Table 2 :
Regression analysis on HIV/AIDS risk index