Selected demographic characteristics of women surveyed in 1990, 1993 and 1994/1995 are shown in Table 1. The selection of these factors was based on availability of comparable data in all three surveys. There was a significant association between all of these characteristics and HIV infection of the woman in each of the cohorts (with the exception of maternal age in 1993). Women who were young, with fewer pregnancies, and more education, were most likely to be HIV-infected. Statistically significant trends in prevalence across categories of age, education and parity were also observed between cohorts (as indicated by P values in the last column of Table 1, with the exception of maternal age < 20 years and illiterate male partners of the women surveyed). Much of the increases occurred between 1990 and 1993, with little or no change between 1993 and 1994/1995 (the time intervals, however, were not exactly comparable).
In the follow-up study, 1173 HIV-seronegative women were evaluated for 2302 PY (median duration, 30.9 months per person; range, 0.03–73.4 months). During this period, 97 women seroconverted at an incidence of 4.21 per 100 PY (95% CI, 3.17–5.25). The age-specific seroconversion rates are shown in Table 2. HIV incidence was highest among young women below 20 years of age (incidence rate, 5.98 per 100 PY; 95% CI, 3.91–8.05) and progressively decreased with increase in age. The annual seroconversion rate was highest during the first year of follow-up and steadily declined in the subsequent years. For example, the rates of seroconversions during the years 1990, 1991, 1992, 1993, and 1994–1995 were 21.26, 12.77, 8.15, 3.33 and 1.11 per 100 PY, respectively.
The prevalence of STD (syphilis, trichomoniasis, gonorrhea, and genital warts and ulcers) among women surveyed in 1990, 1993, 1995 and 1996 is shown in Table 3. There was a significant trend of decline over time for all STD. Although this decline occurred among both HIV-positive and negative women, it had been more consistent for all STD among HIV-positive women. For example, the decrease in prevalence of genital warts and ulcers among HIV-seronegative women was not statistically significant.
Overall, there was an increase in reported lifetime use of condoms from 5.6% (95% CI, 5.1–6.2) in 1990 to 17.5% (95% CI, 15.9–19.1) in 1993, but this declined to 4.9% (95% CI, 3.4–6.4) in 1995. HIV-positive women reported higher use of condoms (7.1% of 1502 women in 1990, 22.2% of 616 women in 1993, and 5% of 701 women in 1995) than HIV-negative women (5.2% of 5101 women in 1990, 15.6% of 1487 women in 1993, and 4.1% of 97 women in 1995).
This study shows that HIV prevalence is high in urban pregnant women in Blantyre, Malawi. Similar to the situation in many African urban cities , there has been an increasing trend over time. However, the most recent figures (e.g., 1993–1996) suggest either slowing rate of increase or a trend towards leveling off (albeit at a high level). This is consistent with reports from other sub-Saharan countries  and is in agreement with modeling work, which suggested a rapid increase in HIV prevalence followed by stabilization within about 10 years of the onset of the epidemic . It is unlikely that the observed HIV prevalence is due to changes in serologic methods for testing. We performed Western blot tests on all ELISA-seropositive specimens in the earlier surveys when HIV seroprevalence was not known, and have followed conventional testing strategies in subsequent surveys [13–16].
Analysis of the age-specific prevalence is more revealing than the overall seroprevalence rates. HIV prevalence in the younger age-groups (< 20 and 20–24 years) for women surveyed in 1996 was not significantly higher than among women of these ages surveyed in 1993 (among women < 20 years, HIV prevalence in 1996 was slightly lower than in 1993). The 1996 prevalence (32.8%) was high mainly due to higher HIV prevalence amongst older women (30–34 years) than in earlier years. The highest age-specific HIV prevalence shifted over the years from younger to older age-groups (20–24 years in 1990, 25–29 years in 1993, and 30–34 years in 1996), suggesting an aging cohort of women who became infected at a younger age.
The overall incidence of HIV (4.21 per 100 PY) was high among women of childbearing age included in this study. The incidence of HIV was highest among women aged under 20 years (5.98 per 100 PY) compared with the incidence among older women. Since young women are the most at risk of seroconversion, it is important that intervention programs to control heterosexual transmission of HIV are specifically developed to target this group (e.g., educational activities including condom promotion). In this study, the HIV incidence was also high among women aged 20–34 years (4.58 per 100 PY among women aged 20–24 years declining to 3.50 per 100 PY among women aged 30–34 years; Table 2). Therefore, HIV/STD control activities should continue throughout the entire reproductive life.
The observation that seroconversion was highest during the first year of enrollment with a declining trend in subsequent years suggests that women are at highest risk during the postpartum period. It is likely that male partners are exposed to infected non-regular partners during late pregnancy and early postpartum when sexual abstinence is common . The regular female partners of these men are therefore at greatest risk when sexual activity is resumed postpartum. This finding emphasizes the need for HIV intervention measures that women could use after delivery other than condoms. A vaginal microbicide with virucidal, bactericidal and spermicidal properties would be highly desirable. It is also important to note that in studies designed to evaluate the efficacy of HIV prevention measures in similar populations of postpartum women, most events (seroconversions) occur early during the follow-up period.
The high HIV incidence among young women could explain the increasing prevalence over time. Several studies have shown that when prevalence is stable or declining, the incidence of HIV could be high or increasing [21,22]. Although antenatal HIV serial seroprevalence data have been helpful for surveillance purposes, underestimation of seroprevalence or seroincidence is likely. For example, in a population-based study in Uganda, HIV-positive women were found to have reduced fertility due to lower rates of conception . Biologic factors limiting fertility or sociodemo-graphic variables that influence antenatal attendance of seropositive women could bias serial seroprevalence estimates.
A declining trend of STD prevalence has been observed in the last 5 years. STD prevalence was higher in HIV-positive women (Table 3), but the decline in prevalence occurred among both HIV-positive and negative women with the exception of genital warts and ulcers. Since higher rates of genital warts and ulcers amongst HIV-infected women are usually attributed to immunosuppression, it is difficult to interpret the declining trend we observed in our data for these conditions. These conspicuous genital lesions could not have been misclassified by the clinic staff. It is probable that these declines are due to a secular beneficial effect of other factors (e.g., improvements in treatment of concomitant bacterial infections such as syphilis or chancroid) or decreased sexual behavior due to a decline in health status. There was no significant decrease in rates of warts and ulcers among HIV-seronegative women. In this group of women, warts and ulcers are mainly due to chronic viral infections (e.g., human papillomaviruses and herpesviruses), and are not expected to be cured by conventional STD bacterial therapy.
A potential limitation when comparing our STD data is the time when testing was performed, since sexual practices during pregnancy and postpartum might not be the same (e.g., abstinence in the first few months postpartum) . In the 1990 and 1993 cohorts, women were tested for STD during pregnancy, whereas in the 1994 study women were tested for STD 5 months or more after delivery in 1995 (after women had resumed sexual activity). These differences in design are unlikely to have influenced our results. A decline in all STD and a continuing decline for syphilis in 1996 suggest that the decrease in STD rates amongst the 1995 women is not the result of a bias.
An important finding is that the decline in STD has occurred without substantial increases in reported condom use. The decline in condom use in the 1995 cohort of women is of concern and does not parallel increasing condom distribution in the southern region of Malawi in recent years (Population Service International, personal communication, 1996). We could not verify whether this decline is due to changes in degree of reporting of condom use among women in different cohorts (e.g., a decline in over-reporting of ever use of condoms in the 1994/1995 cohort compared with the 1993 cohort of women). We showed in an earlier analysis of the 1990 and 1993 cohorts that HIV-seropositive women over-report condom use and that reported condom use is inconsistent amongst these women . An in-depth investigation is urgently needed to verify rates and patterns (i.e., use of condoms with regular and non-regular partners) of condom use in this community. Currently, it not clear whether the large amounts of condoms distributed in the region during the study period were not used or whether women were under-reporting condom use.
A recent study from a neighboring country in East Africa found similar trends of condom use. Among male workers from a textile factory in Tanzania, significant changes in sexual behavior demonstrated by decrease in number of sexual partners was noted without substantial increases in condom use . Interestingly, condom use with casual partners in Tanzania increased and declined over a period of 2 years of follow-up, a trend similar to that observed among women in Malawi. These findings underscore the value of applying multiple interventions to prevent heterosexual transmission of HIV/STD.
The large decline in STD prevalence in this population suggests that the HIV prevention program is having an impact either through improved STD treatment or reduced risk behaviors. The National AIDS Control Programme of Malawi and other governmental and non-governmental organizations have been active in spreading health education messages and adopting a country-wide syndromic management of STD . Within Africa, reports from Uganda indicate that mass education, reduction of sex partners, control of STD and condom use have contributed to declining prevalence of HIV among young adults in the rural areas . Recent findings in Thailand are another example that behavioral change can reduce the prevalence of both HIV and STD . Sequential cross-sectional STD prevalence measures may be useful in monitoring the impact of STD and HIV prevention programs. In this population, since there was no simultaneous increase in condom use, the observed trends possibly reflect improvements in diagnosis and treatment of STD.
The authors thank the Ministry of Health of Malawi and the staff of the Johns Hopkins University, Ministry of Health Research Project for their active and dedicated collaboration.
1. Over M, Piot P: Health sector priorities review: HIV infection and sexually transmitted diseases
. In Disease Control Priorities in Developing Countries
. Edited by Jamison DT, Mosley WH. New York: Oxford University Press; 1993:455–527.
2. Wasserheit JN: Epidemiological synergy: interrelationship between human immunodeficiency virus infection and other sexually transmitted diseases
. Sex Transm Dis
3. Laga M, Diallo MO, Buvé A: Inter-relationship of sexually transmitted diseases and HIV: where are we now? AIDS
4. Grosskurth H, Mosha F, Todd J, et al.
: Impact of improved treatment of sexually transmitted diseases on HIV infection in rural Tanzania: randomized controlled trial
5. Nelson KE, Celentano DD, Eiumtrakol S, et al.
: Changes in sexual behavior and a decline in HIV infection among young men in Thailand
. N Engl J Med
6. Schwarcz SK, Kellogg TA, Kohn RP, et al.
: Temporal trends in human immunodeficiency virus seroprevalence and sexual behavior at the San Francisco municipal sexually transmitted disease clinic, 1989–1992
. Am J Epidemiol
7. Weinstock HS, Sidhu J, Gwinn M, Karon J, Petersen LR: Trends in HIV seroprevalence among persons attending sexually transmitted disease clinics in the United States, 1988–1992
. J Acquir Immune Defic Syndr
8. Strickler H, Hoover DR, Dersimonian R: Problems in interpreting HIV sentinel seroprevalence studies
. Ann Epidemiol
9. Quinn TC: Association of sexually transmitted diseases and infection with the human immunodeficiency virus: biological co-factors and markers of behavioral interventions
. Int J STD AIDS
10. Dallabetta G, Miotti P, Chiphnagwi J, et al.
: High socio-economic status is a risk factor for HIV-1 infection, but not for sexually tranmsitted diseases in women in Malawi: implications for HIV control
. J Infect Dis
11. Taha TET, Miotti P, Liomba, G, Dallabetta G, Chiphangwi J: HIV, maternal death and child survival in Africa [letter]
12. Taha TE, Dallabetta GA, Canner JK, et al.
: The effect of human immunodeficiency virus infection on birthweight and infant and child mortality in Malawi
. Int J Epidemiol
13. Van der Groen G, Van Kerckhoven I, Vercauteren G, Piot P: Simplified and less expensive confirmatory HIV testing
. Bull WHO
14. Mortimer J: An alternative approach to confirming anti-HIV reactivity: a multi-country collaborative study
. Bull WHO
15. Sato PA, Maskill WJ, Tamashiro H, Yeymann DL: Strategies for laboratory HIV testing: an examination of alternative approaches not requiring Western blot
. Bull WHO
16. Gurtler L: Difficulties and strategies of HIV diagnosis
17. US Agency for International Development: Confronting AIDS in the Developing World
. Washington, DC: US Agency for International Development; 1992.
18. Sokal DC, Buzingo T, Nitunga N, Kadende P, Standaert B: Geographic and temporal stability of HIV prevalence among pregnant women in Bujumbura, Burundi
19. Stoneburner RL, Low-Beer D, Tembo GS, Mertens TE, Asiimwe-Okiror G: Human immunodeficiency virus infection dynamics in East Africa deduced from surveillance data
. Am J Epidemiol
20. Miotti P, Canner J, Chiphangwi J, Liomba G, Saah A, Dallabetta G: Rate of new HIV infection in a cohort of women of child-bearing age in Malawi
. AIDS Res Hum Retroviruses
21. Batter V, Matela B, Nsuami M, et al.
: High HIV-1 incidence in young women masked by stable overall seroprevalence among chldbearing women in Kinshasa, Zaïre: estimating incidence from serial seroprevalence data
22. Peterman TA, Zaidi AA, Wroten J: Decreasing prevalence hides a high HIV incidence: Miami
23. Gray RH, Wawer MJ, Sewankambo N, et al.
: Reduced fertility in women with human immunodeficiency virus type 1 infection; a population-based study in Uganda
1998 (in press).
24. Taha TET, Canner JK, Chiphangwi JD, et al.
: Reported condom use is not associated with incidence of sexually transmitted diseases in Malawi
25. Ng'weshemi JZL, Boerma JT, Pool R, et al.
: Changes in male sexual behaviour in response to the AIDS epidemic: evidence from a cohort study in urban Tanzania
26. National AIDS Control Programme: STD Management Guidelines
. Blantyre, Malawi: Ministry of Health and Population, Malawi Government; 1995.
27. Mulder DW, Nunn A, Kamali A, et al.
: Decreasing HIV-1 sero-prevalence in young adults in rural Uganda
Keywords:© Lippincott-Raven Publishers.
Condoms; HIV; incidence; Malawi; women; prevalence; sexually transmitted diseases; trends