Opio, Alex MB ChB, MSc, PhD*; Mishra, Vinod MStat, MA, MPH, PhD†; Hong, Rathavuth MD, DrPH†; Musinguzi, Joshua MB ChB, MPH*; Kirungi, Wilford MB ChB, MPH*; Cross, Anne MA†; Mermin, Jonathan MD, MPH‡; Bunnell, Rebecca PhD‡
Uganda's HIV epidemic is long-standing, with the first known AIDS case reported in 1982.1 In subsequent years, a rapid rise in HIV prevalence followed, and by the late 1980s, all parts of Uganda were affected by a severe and generalized epidemic. HIV prevalence peaked in the early 1990s,2 reaching 30% at some urban antenatal clinics (ANC). In response, a national HIV prevention and control program was started in the mid-1980s. The program initially focused on surveillance and public health education; but over the years, it expanded to include promotion of condom use, treatment of sexually transmitted infections, prophylaxis and treatment of opportunistic infections, HIV counseling and testing services, blood screening services, and general care and support of people living with or affected by HIV infection. In addition, nonhealth sector interventions such as service provision for orphans, school children, and working people were implemented.
To track progress and assess the impact of the response to the HIV/AIDS epidemic, monitoring and evaluation of program activities have been implemented in Uganda and in other countries. HIV sentinel surveillance and knowledge, attitude, behavior, and practice surveys conducted on a periodic basis are examples of such activities. On the basis of above, evidence of declining trends in HIV seroprevalence was reported, especially among young ANC attendees2-6 and in 2 subnational population-based cohort studies.7,8 Improvements in HIV knowledge and behavior indicators were also reported in parts of Uganda.2-6 Most recently, it was reported that HIV prevalence is no longer declining in a cohort that had earlier experienced declining trends.9 Furthermore, HIV prevalence is no longer declining in some ANC sentinel surveillance sites.10 These reports have raised great concern.
In view of the latest trends in HIV prevalence in Uganda, critical analysis of recent trends in HIV-related behaviors and knowledge is crucial, to partly explain possible linkages. In this article, we present results of a secondary analysis of data collected in 4 national surveys conducted between 1989 and 2005. The main objective of the analysis was to determine recent trends in HIV-related behaviors and knowledge in Uganda. These surveys are nationally representative, thus provide invaluable databases for trends analysis. The 2 existing subnational population-based cohort studies in Uganda have a limitation of not being nationally representative.7,8
Overall Study Design
We conducted a secondary analysis of data from the 2004-2005 Uganda HIV/AIDS Sero-Behavioral Survey (UHSBS), and the 2000-2001, 1995, and 1988-1989 Uganda Demographic and Health Surveys (UDHSs). The details of these surveys have been described elsewhere.11-14 The most recent survey, the 2004-2005 UHSBS, was a nationally representative, population-based survey of adults aged 15-59 years and children younger than 5 years. Details about the sample design, survey questionnaires, training, and fieldwork for this survey are available in the main survey report.11 The survey protocol was approved by the Uganda National Council of Science and Technology and the Centers for Disease Control and Prevention and by the Institutional Review Boards of the Uganda Virus Research Institute and Macro International Inc.
The UHSBS collected information from 9529 households in 417 sample enumeration areas. A total of 11,454 women and 9905 men aged 15-59 years were eligible for individual interviews and blood sample collection for serological testing. Of the eligible women and men, individual questionnaires were completed for 95% of women and 89% of men and blood specimens were collected from 90% of women and 84% of men.
Additional data came from 3 previous UDHSs. The objective of these surveys was to collect and analyze data on fertility, mortality, family planning, and health. They were based on nationally representative samples of women aged 15-49 years and men aged 15-59 years. However, unlike the 1995 and the 2000-2001 surveys, the 1988-1989 UDHS only sampled women and excluded some northern Uganda districts. Sample sizes in the 1988-1989, 1995, and 2000-2001 surveys were 4730, 9066, and 9208 adult respondents, respectively. The response rates were high in these surveys, ranging from 85% among men in the 2000-2001 survey to 97% among women in the 1988-1989 survey.
The key variables examined in our analysis included indicators of sexual behaviors and knowledge about HIV/AIDS. In the case of behaviors, we examined median age at sexual debut, premarital sex, primary and secondary abstinence, multiple sex partnership, nonspousal sex, extramarital sex, and condom use at last nonspousal sex. Median age at first sexual intercourse was defined as the age by which 50% of the respondents had initiated sex, calculated from cumulative single-year percent distribution of age at first sexual intercourse. Premarital sex was defined as the proportion of never-married men and women who ever have had sex. Primary abstinence was defined as the proportion of youth (15-24 years) who have never had sex. Secondary abstinence was defined as the proportion of youth who have ever had sex but not in the past 12 months. We defined sex with a nonmarital, noncohabitating partner as “nonspousal sex.” Extramarital sex was defined as sex by currently married people with nonspousal partners.
HIV/AIDS knowledge was measured by 2 indicators related to prevention of HIV and by 1 composite indicator of comprehensive knowledge of HIV/AIDS. The indicators of knowledge of HIV prevention were measured by the following 2 variables: (1) Can people reduce their chances of getting the AIDS virus by having sex with just 1 sex partner who is not infected and who has no other partners? and (2) Can people reduce their chances of getting the AIDS virus by using a condom every time they have sex? The indicator on comprehensive knowledge of HIV/AIDS was defined as the percentage of respondents who agreed, in response to prompted questions, that people can reduce their chances of getting the AIDS virus by having sex with only 1 partner who is not infected and who has no other partners and by using condoms consistently and knowing that a healthy-looking person can have the AIDS virus and that HIV cannot be transmitted by mosquito bites or by sharing food with a person who has AIDS.
For each indicator, we compared the figures obtained from different surveys to determine whether there were any significant changes over time. Trends were examined starting from the first survey for the indicators that were recorded using comparable questions through the most recent 2004-2005 survey. Stata version 9.0 (Stata Corp, 2005) was used to determine the statistical significance of observed trends. For an indicator measured as a proportion, we used a weighted logistic regression accounting for the sampling design, which regressed a dummy variable representing the indicator on a variable representing time for each survey, that is, 1988-1989, 1995, 2000-2001, or 2004-2005. (For ease of presentation, the surveys are subsequently referred to by the year the fieldwork was completed, ie, 1989, 1995, 2001, and 2005.) The trend was considered statistically significant if the coefficient of the time variable was significant at a P value <0.05. For the median age at first sex indicator, we similarly regressed age at first sex (as a continuous variable) on the time variable using a weighted least squares regression and recorded the statistical significance of the coefficient of the time variable. In essence, the test for trend in median age at first sex is a test for trend in the mean. All temporal changes in indicators reported below as significant were statistically significant at a P value <0.05.
Between 1995 and 2005, the median age at first sex among 20- to 49-year-old respondents increased from 16.1 to 16.8 years for women and from 17.4 to 18.4 years for men (Table 1). The median age at sexual debut increased similarly for women and men in 3 age categories, and a similar trend was observed in both urban and rural areas.
The proportion reporting premarital sex decreased significantly among women aged 15-19 years from 36% in 1989 to 29% in 2005, whereas among same-age men, it decreased from 41% in 1995 to 34% in 2001 and then rose again to 40% in 2005 (Table 1). Premarital sex was higher among the 20- to 24-year-olds than among the 15- to 19-year-olds. Premarital sex was also higher in urban areas than in rural areas.
The proportion of youth aged 15-24 years who reported never to have had sex increased significantly over time (Table 1). For women aged 15-24 years, this proportion increased from 23% in 1989 to 32% in 2005. A similar trend was seen in men, for whom the proportion increased from 32% in 1995 to 42% in 2005. When the 15- to 24-year age group was broken down into 15-19 and 20-24, a similar trend was seen in both age groups. The proportion among women aged 15-19 years who reported never to have had sex increased from 38% in 1995 to 54% in 2005. In men of the same-age group, this proportion increased from 52% in 1995 to 61% in 2001 but then declined slightly to 58% in 2005. The proportion reporting primary abstinence also increased among older youth aged 20-24 years, but the levels remained much lower than among youth aged 15-19 years. From 1995 to 2005, the proportion of youth aged 20-24 years who reported primary abstinence increased from 3% to 7% among women and from 11% to 15% among men.
The proportion of sexually experienced women aged 15-19 years who reported no sex in the 12 months preceding the surveys rose from 7% in 1989 to 18% in 2005 (Table 1). For men, this proportion decreased from 33% in 1995 to 16% in 2005. Overall, the proportion of sexually experienced women and men aged 15-24 years who reported secondary abstinence was higher in urban areas than in rural areas.
Multiple Sex Partners
Between 2001 and 2005, the proportion of respondents aged 15-49 years who reported having sex with 2 or more partners in the past year increased from 2% to 4% in women and from 24% to 29% in men (Table 1). The proportion of women and men reporting 2 or more sexual partners in the past year increased similarly in both rural and urban areas, but the levels were higher in urban than in rural areas.
Between 2001 and 2005, the proportion of men aged 15-49 years who reported having sex with a nonspousal partner increased from 28% to 37%, whereas for women, it remained stable (14% in 2001, 15% in 2005; Table 1). The proportion of men who reported nonspousal sex was however more than twice as high as that of women. When data were disaggregated by age and residence, nonspousal sex was much more common in the younger age group (15-24 years) and among urban residents than in the older age group (25-49 years) and among rural residents, respectively. Unmarried people were more likely to report this practice than married persons.
The proportion of married respondents who reported sex with someone other than a spouse in the 12 months preceding the survey were much lower among women (2% in 2001 and 3% in 2005) than among men (11% in 2001 and 18% in 2005). From 2001 to 2005, the proportion reporting extramarital sex increased slightly for women from 2% to 3% but significantly for men from 11% to 18%.
Among respondents who had nonspousal sex in the 12 months preceding the survey, the proportion who reported using a condom during the last nonspousal sex increased in women aged 15-49 years, from 39% in 2001 to 48% in 2005, whereas in same-age men, it decreased from 61% in 2001 to 54% in 2005. A similar change was seen for respondents aged 15-24 years, where among women, this proportion increased from 45% in 2001 to 55% in 2005, whereas among men, it decreased from 65% in 2001 to 55% in 2005. The changes were similar in urban and rural areas, but the proportion reporting condom use with nonspousal partner was much higher in urban areas than in rural areas for both women and men. In regard to consistent use of condom among the 15- to 49-year-olds who had sex in the last 12 months, the practice remained unchanged in men at 14% between 2001 and 2005. In women, there was a slight improvement from 7% in 2001 to 9% in 2005 (Table 1). Consistent condom use was more common among urban residents than among rural residents.
Most women and men aged 15-49 years knew that “having only 1 sexual partner who is not infected and who has no other partners” reduces the risk of HIV infection. This knowledge increased in women from 83% in 2001 to 88% in 2005 but remained unchanged in men-90% in 2001 and 89% in 2005 (Table 2). The changes were similar in both urban and rural areas.
The proportion of women who reported knowledge of condoms to prevent HIV infection increased slightly from 65% in 2001 to 68% in 2005, whereas for men, it remained constant at 77% (Table 2). During 2001-2005, condom knowledge increased among women in rural areas but decreased slightly in urban areas. Among men, it remained unchanged in both urban and rural areas. Overall, condom knowledge was higher in urban areas than in rural areas.
Comprehensive Knowledge of HIV/AIDS
Comprehensive knowledge of HIV/AIDS remained stable in women, 27% in 2001 and 28% in 2005, and declined slightly in men from 39% in 2001 to 36% in 2005 (Table 2). The level of comprehensive knowledge of HIV/AIDS was much higher in urban and male respondents than in rural and female respondents, respectively. For men, comprehensive knowledge decreased some in both urban and rural areas, but for women, it increased some in rural areas and decreased some in urban areas.
Analysis of trends in HIV-related behaviors and knowledge from 1989 to 2005 indicates prevention successes but also highlights remaining challenges. In this article, we present information that some important HIV-related behavior and knowledge indicators recently deteriorated or experienced stagnation and that there is a shift toward more risk-taking sexual behaviors. Specifically, nonspousal sex, nonuse of condoms during nonspousal sex, extramarital sex, and engagement in sex with multiple partners increased among Ugandan men. Among women, nonspousal sex has remained at about the same level, condom use in nonspousal sex rose, and extramarital sex remained unchanged, although more women are now engaging in sex with multiple partners. Additionally, in both women and men, the proportion of 15- to 24-year-old respondents who reported never have had sex increased significantly, and the median age at sexual debut increased steadily between 1995 and 2001, although it stagnated thereafter.
Our analysis showed that knowledge indicators that had improved steadily during the 1980s and the 1990s and reached a high level by 2001 have stagnated thereafter. Both the knowledge that having only 1 uninfected faithful partner and that using condoms can prevent HIV transmission, as well as comprehensive knowledge of HIV/AIDS have remained stable as during 2001-2005.
In Uganda and elsewhere, declines in HIV prevalence and incidence have been associated with risk reduction behaviors.15-18 Conversely, increases in unsafe sexual behaviors have resulted in increased risk for HIV infection and transmission.19 Therefore, our findings of stagnation or deterioration of certain HIV-related behavior and knowledge indicators warrant further investigation and may indicate, among others, complacency in the HIV/AIDS prevention and control efforts among the general population. Information on Uganda's HIV prevention successes is widespread. Furthermore, between 2000 and 2005, access to antiretroviral therapy rapidly increased in Uganda. By December 2005, more than 60,000 patients were estimated to have been on antiretroviral therapy. These factors could have changed people's perception about the HIV/AIDS epidemic. In the United States, Katz et al20 suggested that highly active antiretroviral treatment resulted in increases in rates of unsafe sexual behavior among men having sex with men. Such reversals to unsafe behaviors may threaten past HIV prevention successes.21
The timing of our findings parallels with reports that HIV prevalence is no longer declining in a Ugandan cohort that earlier experienced declining trends of HIV infection.9 It also coincides with a period when the previously evident declining trends in HIV prevalence among ANC attendees have stopped in some sentinel surveillance sites.10 Furthermore, over the same period, a decline in the intensity of information-education-communication (IEC) to promote behavior change has been observed.22 The number of actors in community sensitization and education also dwindled over the same period.22 The decline in intensity of IEC programs probably arose from a drop in funding for primary prevention programs. The current level of funding for primary prevention interventions is relatively low compared with the past one for similar activities. On the other hand, funds seem to be more available for service access messages as opposed to behavior change messages.22,23 Behavior change messages were at the center of the life skills education programs. Therefore, it is possible that a decline in intensity of IEC programs is partly responsible for the deterioration of some of the HIV-related behavior and knowledge indicators. To forestall further deterioration, more support is required for a comprehensive national IEC program. It has been noted that providing people with comprehensive information on reducing HIV risk including abstinence, reduction of number of sexual partners, and correct condom use is most effective at preventing new infections.24,25
Researchers have also argued that fear-arousing IEC messages were responsible for the Uganda's success story through instilling behavior change.24 These types of messages are no longer being disseminated. Asingwire et al have suggested that Uganda needs to reactivate these types of messages to combat the recent deterioration of behavior indicators.22
Another factor which could have contributed to the recent trends in behavior indicators is “normalization of AIDS.”22 It is possible that the perception of the population on HIV has become a normal feature. Uganda's HIV experience dates back to almost 26 years ago. To date, the epidemic has affected the lives of every citizen, either directly or indirectly. Although 10 years ago, people may have feared AIDS and taken great measures to avoid it, today people may think of AIDS as simply a normal part of life. This normalization phenomenon could have led to complacency in risk avoidance, hence, partly explaining our results.
Between 2000 and 2005, Uganda experienced negative reports on condoms. This reached a climax in 2004 when the quality of a few batches of the government-procured “Engabu” condoms was found wanting, with some condoms developing bad smell. The above could have changed the perception of some people on the effectiveness of condoms, thus contributing to the drop in its use.
Our findings underscore a need to reinvigorate and adopt additional prevention approaches in a country with long-standing HIV prevention programs and successes. More epidemiological and behavioral investigations are required to better understand the determinants of increasing high-risk behaviors, stabilizing HIV prevalence, and possible increases in HIV incidence. The negative trends seen for certain knowledge and behavior indicators should be a warning sign that declining trends in HIV prevalence may be reversed. Indeed, as alluded to earlier, there is some evidence showing that past HIV prevalence declines have not continued over the last few years,9,10 perhaps as a result of recent changes in HIV-related risk behaviors.
Furthermore, on the basis of further analysis of the 2005 UHSBS data, we recently reported a high level of unprotected sex among HIV-infected persons.26 This report is disturbing and underscores the need to design prevention strategies that target HIV-positive persons, especially in a country where HIV discordance is very high. In 2005, HIV discordance was 57% among couples in which 1 partner was HIV-infected.11
In conclusion, our study shows that over the last few years, some HIV prevention indicators in Uganda have either deteriorated or stagnated, thus, indicating a shift toward more risk-taking behaviors. These findings have raised concern as they parallel a halted decline in HIV prevalence and incidence.9 To address the concern, the government designated 2006 as the year for intensifying HIV prevention and developed a new road map for HIV prevention.27 The road map outlines a number of key areas for priority action to support the delivery of a comprehensive HIV prevention package, including acceleration of reduction of sexual transmission of HIV through intensified ABC+ (abstinence, being faithful, and condom use plus strategy) campaigns to promote risk reduction/risk avoidance, expansion of services for HIV counseling and testing with mutual disclosure of serostatus to one's partner, scaling up programs for prevention of mother-to-child-transmission, designing strategies for HIV-positive individuals, and intensification of programs for prevention and treatment of sexually transmitted infections.
The findings in this study are subject to some limitations. First, the data we used were derived from responses to questions from different surveys, the UDHSs and the UHSBS. Although the designs of the surveys were broadly similar, there was variability in the way some questions were asked. This variability could have led to differences in the responses given, possibly biasing certain comparisons. Second, interpreting behavioral information obtained from self-reporting has limitations due to possible social desirability bias. However, this potential bias may have been similar across surveys and hence little effect on trend analysis.
The authors thank Drs. Lisa Butler, Willi Macfarlane, Rand Stoneburner, and George Rutherford for their comments on earlier versions of this article and Leonard Atuhairwe for data analysis. We are also grateful to Dr. Wolfgang Hladik for editorial review of the final draft. Financial assistance was provided by the US Agency for International Development and the Centers for Disease Control and Prevention. Macro International Inc. and Institute for Global Health of the University of California, San Francisco provided technical support.
1. Goodgame RW. AIDS in Uganda-clinical and social features. N Engl J Med. 1990;323:383-389.
2. Kirungi L, Musinguzi J, Madraa E, et al. Trends in antenatal HIV prevalence in urban Uganda associated with uptake of preventive sexual behaviour. Sex Transm Infect. 2006;82(Suppl 1):36-41.
3. Asiimwe-Okiror G, Opio AA, Musinguzi J, et al. Changes in sexual behaviour and decline in HIV infection among young pregnant women in urban Uganda. AIDS. 1997;11:1757-1764.
4. Asiimwe-Okiror G, Musinguzi Opio AA, Biryahwaho B, et al. Declines in HIV prevalence in Ugandan pregnant women and its relationship to HIV incidence and risk reduction. Presented at: XI International Conference on AIDS; July 6-12, 1996; Vancouver, British Columbia. Abstract Mo.C.905.
5. Opio AA, Mulumba N, Asiimwe-Okiror G, et al. Changes in HIV/AIDS prevention indicators in a rural district of Lira in Uganda. AIDS. 2000;14:1461-1463.
6. Opio AA, Asiimwe-Okiror G, Musinguzi J, et al. Sexual behaviour change due to HIV/AIDS: results from population-based surveys conducted in five districts of Uganda. Presented at: XI International Conference on AIDS; 1996; Vancouver, British Columbia. Abstract LB.D.6069.
7. Mulder DW, Nunn AJ, Wagner HA, et al. HIV-1 incidence and HIV-1-associated mortality in a rural Ugandan population cohort. AIDS. 1994;8:87-92.
8. Wawer MJ, Serwadda D, Gray RH, et al. Trends in HIV-1 prevalence may not reflect trends in incidence in mature epidemics: data from the Rakai population-based cohort, Uganda. AIDS. 1997;11:1023-1030.
9. Shafer LA, Biraro S, Kamali A, et al. HIV prevalence and incidence are no longer falling in Uganda-a case for renewed prevention efforts: evidence from a rural population cohort 1989-2005, and from ANC surveillance. Presented at: XVI International Conference on AIDS; 2006; Toronto, Ontario. Abstract THLB0108.
10. Ministry of Health [Uganda]. Draft HIV Surveillance Report, 2006. Kampala: Ministry of Health, Government of Uganda.
11. Ministry of Health (MOH) [Uganda] and ORC Macro. Uganda HIV/AIDS Sero-Behavioural Survey 2004-2005. Calverton, MD: Ministry of Health and ORC Macro; 2006.
12. Kaijuka EM, Kaija EZA, Cross A, et al. Uganda Demographic and Health Survey 1988-1989. Columbia, MD: Ministry of Health [Uganda] and Institute for Resource Development/Macro Systems Inc; 1989.
13. Statistics Department [Uganda] and Macro International Inc. Uganda Demographic and Health Survey, 1995. Calverton, MD: Statistics Department [Uganda] and Macro International Inc; 1996.
14. Uganda Bureau of Statistics (UBOS) and ORC Macro. Uganda Demographic and Health Survey 2000-2001. Calverton, MD: UBOS and ORC Macro; 2001.
15. Stoneburner RL, Low-Beer DL. Population-level HIV declines and behavioral risk avoidance in Uganda. Science. 2004;304:714-718.
16. Killian AH, Gregson S, Ndyanabangi B, et al. Reductions in risk behaviour provide the most consistent explanation for declining HIV-1 prevalence in Uganda. AIDS. 1999;13:391-398.
17. Hallet TB, Aberle-Grasse J, Bello G, et al. Declines in HIV prevalence can be associated with changing sexual behaviour in Uganda, urban Kenya, Zimbabwe, and urban Haiti. Sex Transm Infect. 2006;82(Suppl 1):i1-i8.
18. Gregson S, Garnett GP, Ynamukapa CA, et al. HIV decline associated with behavior change in eastern Zimbabwe. Science. 2006;311:664-666.
19. Centers for Disease Control and Prevention (CDC). Increases in unsafe sex and rectal gonorrhoea among men who have sex with men-San Francisco, California, 1994-1997. MMWR Morb Mortal Wkly Rep. 1999;48:45-48.
20. Katz MH, Schwarcz SK, Kellog TA, et al. Impact of highly active antiretroviral treatment on HIV seroincidence among men who have sex with men: San Francisco. Am J Public Health. 2002;92:388-394.
21. Wolitski RJ, Valdiserri RO, Denning PH, et al. Are we headed for a resurgence of the HIV epidemic among men who have sex with men? Am J Public Health. 2001;91:883-888.
22. Uganda AIDS Commission. Rapid Assessment of Trendsand Drivers of the HIV Epidemic and Effectiveness of Prevention Interventions in Uganda: A Review of HIV/AIDS Prevention Interventions in Uganda. 2006. Kampala: Uganda AIDS Commission, Government of Uganda.
23. Uganda AIDS Commission. Rapid Assessment of Trends and Drivers of the HIV Epidemic and Effectiveness of Prevention Interventions in Uganda: A Synthesis Report. 2006. Kampala: Uganda AIDS Commission, Government of Uganda.
24. Green EC, Halperin DT, Nantulya V, et al. Uganda's HIV prevention success: the role of sexual behavior change and the national response. AIDS Behav. 2006;10:335-346.
25. Shelton JD, Halperin D, Nantulya V, et al. Partner reduction is crucial for balanced “ABC” approach to HIV prevention. BMJ. 2004;328:891-893.
26. Bunnell R, Opio A, Musinguzi J, et al. HIV transmission risk behavior among HIV-infected adults in Uganda: results of a nationally representative survey. AIDS. 2008;22:617-624.
27. Uganda AIDS Commission. Accelerating HIVPrevention in Uganda. The Road Towards Universal Access. 2007. Kampala: Uganda AIDS Commission, Government of Uganda.
© 2008 Lippincott Williams & Wilkins, Inc.