A, B and C in Uganda: The Roles of Abstinence, Monogamy and Condom Use in HIV Decline

Abstract

Uganda is often cited as a role model in the fight against HIV/AIDS because of its success in reducing both prevalence and incidence of HIV infection since the late 1980s. Although an increase in sexual abstinence has been highlighted as a primary cause of the declines, large increases have also been recorded in monogamy and condom use. The extent to which each of these factors actually influenced the overall decline in Uganda's HIV rates has become a highly charged political issue in the United States, leading to restrictions on how US development funding for combatting HIV is allocated. The Alan Guttmacher Institute investigated changes that occurred in abstinence, monogamy and condom use in Uganda in the 1990s, using nationally representative data from Uganda's Demographic and Health Surveys of 1988 (women only), 1995 and 2000 (women and men), and national-level findings from two surveys by the Global Programme on AIDS in 1989 and 1995 (women and men). Reduction of infection risk by lowering numbers or types of partners among people with more than one relationship was not covered. Here we reprint the chapter on “Implications” from the AGI report, a commentary on the lessons that can and cannot be taken from the data, published in a 2003 Guttmacher Report on Public Policy.

Implications of behavior changes in risk for HIV infection

The evidence from national surveys of men and women suggests that changes in all three of the factors investigated–abstinence, monogamy and condom use–contributed to reducing and sustaining HIV infection at lower levels in Uganda, at least among some women.

Increased delay in initiation of sexual activity over the period 1988–2000 contributed to a reduction in the risk of HIV infection: adolescent women 15–19 were less likely to have initiated sexual intercourse in 2000 than in 1988, as were adolescent men 18–19 in 2000 compared with 1989. Furthermore, delaying the onset of sexual intercourse may also have contributed to more stable and monogamous partnerships and higher levels of effective condom use. However, when all changes in sexual behavior are taken into consideration, the effect of increased abstinence on the overall level of sexual exposure to the risk of HIV infection is moderated. Other small-scale studies have found larger changes in abstinence, but these may not be generalizable. ¹ *

For example, abstinence did not increase among those who had ever been sexually active and therefore did not contribute to reducing exposure to HIV infection for those individuals over the time period covered by the DHS data. Those adolescents and young adults who have sexual intercourse were increasingly likely to be married between 1988–1995, but between 1995–2000, this trend was reversed and they are increasingly likely to be unmarried and they were also more likely than older adults to be exposed to multiple sexual partners. These trends to some extent counterbalance the impact of reduction in risk from later onset of sexual intercourse.

The proportion of sexually experienced people who were sexually active at the time of the survey hardly changed for women, except for a decline among those 15–17, and actually increased among adult men and women, from the late 1980s to 2000, potentially increasing adults' risk for HIV infection. The net impact of changes in initiation of sexual intercourse and of abstinence among those who had had sex is reflected in the proportion of all women and men who were sexually active in the three months before each survey period. This measure indicates a decrease in ongoing levels of exposure over the period 1988–2000 to HIV risk among all adolescent women and men, but increased exposure to HIV risk (GPA) or relatively little change (DHS) for adult women and men in most age groups.

Sharp increases in monogamy also contributed to lowering the risk for HIV infection for the periods 1989-1995 (GPA) and 1995-2000 (DHS). These increases reduced the risk for HIV infection among younger married women and among unmarried sexually active women at all ages. They understate the importance of overall reduction in the number of sexual partners. Other research has shown that partner reduction among those having multiple partners can also be highly significant in terms of epidemiological impact on HIV infection. ²

Increased monogamy protected unmarried men over the period 1989-1995 (GPA data shown here and also discussed in other studies³); the limited data available for married men for this time period also show that this factor changed in the direction of reduced risk of HIV infection for them. There was little change, however, in the more recent period, 1995–2000, in the proportions of monogamous men among unmarried sexually active men, though older married men reported declines in number of partners.

Steep increases in use of the condom among the unmarried sexually active population, both men and women, also contributed to reduction in HIV risk, based on information for the period 1988–2000 for women; and for the period 1989–2000 for men (using both GPA and DHS data). While married men also reported a small increase in current use of condoms, there was little change in the level of use reported by married women. Information on consistency and correctness of use is not available from national surveys.

This paper considers trends among both women and men, because changes in either group affect both groups. It is notable that the level of monogamy among married people has not changed much, and a substantial proportion of married men are not monogamous, and most do not use condoms. As a result, married women are at high risk of infection, particularly given that it would be much more difficult for them to refuse to have sex with their partners, compared to unmarried women.

Implications for policy and programs

This analysis shows that positive behavior change in all three areas of ABC, abstinence, being faithful (monogamy) and condom use, have contributed to the decline of HIV in Uganda to sustained lower levels.

These results are consistent with the broad diversity of interventions and of groups and organizations that have been active in implementing interventions, and suggest that the current approach in Uganda–where a large number of organizations and groups implement a range of different policies and programs that together address all three of the main factors that influence HIV infection–is likely to be effective in reducing the prevalence of HIV.

The size, and probable impacts, of changes in these factors differed by gender, age and marital status, and in many cases the changes were small. They were greatest among young women, especially those 15–17, and some groups of young men, suggesting perhaps that interventions have been differentially directed at young men and women or have been more effective among young people, as well as that they have had a bigger potential for change because of their age and unmarried status. Relatively high levels of exposure to multiple sexual partners and low levels of condom use across all age groups, however, indicate the need for continued education, service and other intervention efforts.

More research is needed, both to compare patterns of change in other countries with the Ugandan situation and to ascertain more about what social changes and interventions led to these behavioral changes; some studies are already underway. ⁴ But, anecdotal evidence points to strong government commitment, destigmatization of people with HIV/AIDS and a comprehensive approach to behavior change in the area of sexual activity, sexual relationships and condom use. And, further, stopping HIV/AIDS in Uganda and worldwide requires work even beyond changes in these behaviors. As one expert cautioned in a recent overview of the current understanding regarding HIV levels and trends in Uganda:

“Another frequent mistake encountered is the notion that the decline in prevalence rates must be due to a few specific interventions introduced by the Ugandan government… the government is but one player in the fight against HIV-1. There are hundreds of non-governmental organizations (NGOs), religious groups and community activists also working to prevent the spread of HIV/AIDS in Uganda.”⁵

Coverage of the recent visit by President Bush to Uganda also highlights the multiplicity of factors that contributed to Uganda's success in containing the HIV epidemic, and points to the dual contributions of “reduction in casual sex and an increase in condom use”.⁶ This article also points out that a key factor in the decline in prevalence of HIV in Uganda is the government's “uniquely creative and strategic policy approach to enable non-state actors in their individually targeted messages about prevention”, and specifically highlights the importance of the comprehensive approach implemented in Uganda.⁵

Acknowledgements

The chapter on “Implications” is reprinted with kind permission of The Alan Guttmacher Institute from: Singh S, Darroch JE, Bankole A. A, B and C in Uganda: The Roles of Abstinence, Monogamy and Condom Use in HIV Decline. Occasional Report No. 9, New York: AGI, 2003. The references from this section of the report (numbers 12–17) were re-numbered here.

Notes

* Other research has shown much sharper declines in sexual initiation among primary school pupils aged 13–16 in one district in Uganda (the Soroti District); however, the national data indicate that such large declines did not occur in the country as a whole. It is possible that the greater behavior change in this district resulted from interventions particular to this district; in addition, in general, adolescents attending school may be more exposed to educational interventions than the average adolescent, and more likely to change their behavior in response to such interventions.