Abstracts Vol. 1 [International Conference on AIDS (11th: 1996: Vancouver, Canada)]

Tu.D.471 -Tu.D.475

Tuesday, July 9, 1996

advantages of comparisons carried out downstream from survey design arrd the impor
tance of structuring a community of researchers to improve the quality arnd comparabil ity
of surveys in the future.
Miche Hubert, Facultes univer itaires Saint-Louis, Centre d'etudes socioiogiques Bid du
Jardin botanique, 43 B 1000 BruxellesTel. ([phone] Fax [phone]
e-mail hubertafusl.ac.be
Tu.D.47 I
FIVE-YEAR TRENDS IN HIV RISK BEHAVIORS AMONG YOUTH
Collins, Janet L. Warren, W., Kann, L. Division of Adolescent and School Health Centers for
Disease Control and Prevention, Atlanta, Georgia, USA
Objective: To examine trends in sexual risk behaviors among U.S. high school students.
Methods: Data from the 1990, 1991, 1993, and 1995 national Youth Risk Behavior Surveys
are used to examine whether significant changes have occurred in HIV related risk behaviors among in-school youth. These surveys provide self report data from cormparable, independent samples of approximately 12,000 youth ages 14- 1 8 (grades 9-12) each year.
Results: From 1990 through 1995 no significant change was detected in the percent of high
school students ever having had sexual intercourse (53%), having had sexual intercourse
with four or more partners (19%), or having had sexual irntercourse during the three
months preceding the survey (38%). In contrast, overall condom use at last sexual intercourse increased significantly from 1990 to 1995 (from 46% to 53%). Subgroup analyses by
sex, grade, and race/ethnicity indicated a significant increase in condom use among female
students and black students.
Conclusions: Both the leveling of sexual behavior rates and the increase irn condom use are
important strides in halting HIV infection among youth. However, nmore muist be done to
help youth people delay initiation of sexual activity and increase condom use. Implications
of these findings for school health programs will be discussed.
Janet L Collins, Ph.D., 4770 Buford Hwy, NE, Atlanta, GA USA 30341-3714
Telephone: [phone] Fax: [phone] e-mail:jlcl(Accdrshl.er.cdc.pov
Tu.D.472
SEXUAL BEHAVIOUR AND HIV RISK: LESSONS FROM A RELATION-BASED APPROACH
Huynen, Ph.,* Marquet, J.,* Ferrand, A., I Hubert, H.,* Van Campenhoudt, L." h Centre d'otudes sociologiques, Facultes univers itaires Saint Louis, Brussels, Belg um; "Institut de soci -
ologie, Universitx de Lille, Lille, France
Objectives: To determine the structlural characteristics of the sexual relations (encounter's
circumstances, partners' expectations and satisfactions, succession or concomitance of relations, relationship's social visibility, norms, etc.) that influence the way HIV risk is taken into
account (or not) in these relations.
Methods: Dat a about the sexual relations active during the year preceding the investigation
were collected in the Belgian national survey on sexual behaviour caried out in 1993
(3,733 telephone interviews in Belgium' s resident population aged I 8-49).This data collec
tion is relation-based: for- each relation, questions were asked: I) about the way HIV risk
was handled at the start of the relation (no precaution, talking about risk, condom use, HIV
tea'sting, etc.), as well as in its course (harving other partners, abandoning co domr, etc.); and
2) about characteristics of the relation.
Results: The survey showed that two out of three individuals with severIl partners handle
HIV risk differently from one relation to another:This variation in the way of 'adapting' to
HIV risk is linked to the structural characteristics of the relations. For example, HIV screening before any sexual intercourse depends on the place partners met and is less frequent
when the partners consider the relationship as important;: condom use is hi-her in sexual
intercourses that occured just one tine in a relationship (virtually all relations with prostitutes used a condom) and is more frequent when partners have discussed HIV risk togeth
er: relationships that began at school are more protected (condom use) than those that
began at the workplacef Ego's norms about idelity are clearly more influenced by Egos perception of what 's happening in his/hier social networks than by norms handed down by
those networks.
Conclusions: This analysis confirms clearly the interest -for research and prevention- of
relation-based approach according which rione's relationships ire not equivalent: the results
lead us to consider as relevant structural characteristics of the relation that are liable to
influence way(s) of adapting to HIV risk.
Phi ippe Huynen, Facultes uners itares Saint-I otis, Centre d'etudes sociologiques, BId du
Jardin botanque, 43 - B 1000 Bruxelles Tel. [phone] Fax (+32 2) 21I 79 97
e-mail huynen(afusl.ac be
Tu.D.473
CONCURRENT PARTNERSHIPS AND HIV TRANSMISSION IN UGANDA
Martina Mor is, David SerTadda2, Mrjar KretzschrNar o3, Nelsonr Sewan kambo2, Maria
W e Pennsylvania State Uniersity, USA; 2. Makerere Univers ity, JUganda; 3 RIVM,
Objective: To identify the prevalence of concuArrent partnerships in the sexually actve populatron of Rakai disti t, Uganda and to e t blrsh therr npact enr IV tansinisir ndyrnamics.
Sample and Methods: Hlousofhold-based r-ii ana saple af I1c40 adurlts I5 'Itor ea old in
Rakai district using a local network survey module for the respondents' most recent three
sexual relationships. Concurrency is established by using the reported dates of irst sex and
last sex with each partner, and comparing intervals for overlaps. Descriptiae st tis ire
used to characterize the average length of overlap, the types of partnerships involved, and
the levels of partner-specific condom use. Microsimulation is used to establish the impact of
concurrency on HIV transmission.
Findings: The reported lifetime number of sex partners in this population is similar to that
reported in the U.S.The mean (median) number reported is 8.5 (3) for men, and 2.5 (2)
for women, with 22% of men and I% of woren reporting more than 10 partners. About
40% of the total population and more than 50% of those over 30 years old report at least
one concurrent partnership in the last three sexual partners.The rate is higher for men
(57%) than for women (25%). About 10% report that all of their last three par-tners were
concurrent at some time (19% of men and 3% of women).The great maorits of concur
rent partnerships involve relatively pon aterm relationshps spouses, consensuals and
f-iends only 5% fnvolved, short term acquaintance or stranger The mean lenth of over

lap between tvo -_on' iorent partnerships is 29 months (median 16 months), and consistent
condom use pr ies protection in only about 10% of cases. Mlicros imulation suggests that
this type of oi noun n" / would dramatically amplify the speed of spread and the size of the
epidemic a-cnpre,! t sequential monogamy.
Conclusion:,,.te a overlappin g elationships in this populaticn prov de the bas is for
a stable corteCur 'res,.rk that can dramatically amplify the speed of HIV. Concurrency,
rather than the b'. oute number of partners, may account for the rapind and pervasive
spread of HIV in this population.
Martina Morris, Dept of Sociology, University of Pennsylvania, 288 McNel Bldg,
3718 Locust Walk, Philadelphia, PA 19104. Phone [phone], fax [phone],
[email protected]
Tu.D.474
TEENAGERS' SEXUAL BEHAVIOR IN FRANCE
Lagrange, Hugues*, Lhomond, Brigitte "Levinson, Sharman ", the ACSJ group. CNRS,
EHESS, France
Objectives: The first national survey on adolescent sexual behavior in France, conducted in
1994, has two objectives: 1) a comprehensive understanding of the entry into sexual experience, taking nto account personal chronology as well as individuals soal and sexual net
works,2) providing data to inform effective prevention strategies, evaluating HIV risk behaviors and knowledge and attitudes pertaining to AIDS
Methods:The survey is based on a two-level stratified random sample of- 6183 individuals
(15-18 year-olds). Face to face interviews of one hour duration, using a questionnaire, were
administered in regular and technical high schools and in vocat ional training centers
Results: I) Entry into sexuality takes place between 15 and 18 (median ge at f-st coitus:
male: 17.3~+0. 1,female: 17.5+0.1) The first relationship lasts a month or less for 58; of boys
and 4 1% of girls.Those who have longer first relationships are more likely to start a second
relationship before ending the first one or withi one mont h her its end (25 % of the sex
ually experienced). Fis st intercourse involves both vaginal and anal intercourse,n 5% of the
cases.The prevalence of homosexual practice, among the sexually experienced is I.4 for
boys and 1.3% for girls. 15 % of the girls have been forced into sexual acts, 75% of them by
boys they know.2) Cond om use at first intercourse increased dramratcaly n the ear ly
nineties. 56% of boys and 58% of girls who had their first sexual intercourse in!989 used a
condom, and respectively 82% and 87% in 1993.This means that most sexual intercourse
during the first relationship is now protected. However; condom use decreases at last itercourse.The ma in f actor effecting this evolution is oral contraception Longer ind steadier
relationships involve more frequent intercourse and lead to using the pll. Those who gae
up condom use are two times as likely to have had an HIV screening.
Conclusion: Findings indicate an emerging protection strategy combining condomn use and
HIV testing.Tha encouraging evolution in protection is less remarkab e for ndividuals n
vocational training centers.
Sharman Levinson, ACSJ CNS, 7 rue d'Anlou, 75008 Paris, France telephone I1 40 06 00 94
fax: 33 I 40 07 01 12
Tu.D.475
NATIONAL HIV PREVENTION FOR GAY AND BISEXUAL MEN IN THE UK 1987-1995
Adam Crosier. He lth Education Authority London, UK.
Objectives: -This paper assesses the impact of direct public education tar geted it gay and
bisexual men in the UK since 1987. It considers the impact of IV prevention in teim of
evidence from epidemiololgical and behavioural research, and also examines vht,say and
bisexual men think of national targeted HIV prevention camra pgrs. Ihe coirse of the IIV
epidemic in Britain has affected gay and bisexual men dispropor tionately. Latest estimates
indicate that of 21,900 adults infected with HIV 56% are thought to have acquired infection
through male homosexual exposure.The incidence of HIV infection among ay nd isexual
men rose firom 1989 onwards, and there has been a reversal of rens ii icrtors of
unsafe sexual behaviour among this group.
Methods: All national HIV prevention campaigns for gay rand bisexual men snce 1987 have
been developed through the use of research conducted to identify nfornation nreeds.
styles, methods, maessages and media whichi are lkely to be most successful in corruncating effectively with target audiences.The responses of gay and bisexua! men, to the canpaigns have been monitored through a regular tracking survey (cross sectional) of c350 gay
men who attend 25 ay pubs and clubs in Brtan.The survey is also used, a means f
assess ng this population's reported sexu Ibehanouris and contributes to the process of
reeds assessment.
Results: The prinncipal behavioural change made by gay bar attender s sa reported reduction in the number of sexual partners in the period to 1987 (mean 1.8 partners in year -to
interview), followed ya steady increase snce (I 8.6 partners in 995). Findings sdicatea
trend towards increased condom use on the last sexual occasion (up fhrom 9% of respondents repor ting peretr at ive sex in 1987 to 52% in 1995), and a sreduction nun pr otecterd
anal sex (down fron 48% n 1987 to 21!% in 1995).The use of condoms is the only permanent behaviourarrl moddcrfration reported by the sarmple.There is clear evidence of a move to
the adoption of a range of behaviours appopnate to different persoral strtep5es for proto. ay irn r oer ptsutiye 0epo nes to the campaigns, both un te-ms of awareness
(90cc) and an relovanco to their" lives (88cc).
Conclusions: These results confirs that pay men have made sIgndicant chanes to their
sexual behaviours and lifestyles, but taken in conjunction with epdemological evidence of
increased incidence of HIV infection among gay men, indicate that the stratepies that many
raen and espcially young men - have adopted, and the choices they make about what
behaviours are shafe, with different types of partners, may be risky for HIV tr tnsrssaion.
Adam CrosieHealth Education AuthorityHamilton House,Mabledon PlaceLondon,
WCIH9TXUK Tel:[phone] Fax: [phone],
email: [email protected]

261