The HIV-1 pandemic in Kenya is characterized by an estimated prevalence of 7% with regional heterogeneity.1,2 Previous studies undertaken in the late 1990s at different agricultural plantations in the Rift Valley Province reported an HIV-1 prevalence from 10% to 14% in the Kericho area (unpublished data). The main objective of this study was to further characterize the existing HIV-1 epidemic, with the goal of determining the suitability of agricultural employees and their dependents for a future vaccine cohort feasibility study.
East Africa is of particular interest for evaluating HIV-1 vaccines. The US military has long maintained an infectious disease medical research facility in Nairobi, Kenya. The US Army Medical Research Unit-Kenya HIV/AIDS Program, also known as the Walter Reed Project (WRP), is 1 of the 3 established field stations for the US Military HIV Research Program in East Africa (Kenya, Uganda, and Tanzania). More important, however, is the diversity of HIV-1 subtypes present in the region. Although clade B is the predominant HIV-1 subtype in developed countries, globally, the epidemic is dominated by other subtypes.3,4 The HIV-1 pandemic in Kenya is characterized predominantly by clade A and, to a lesser extent, clades D, C, and recombinants.5-9 Vaccine trials conducted in areas of high diversity will allow assessment of the protective efficacy of candidate vaccines against differentsubtypes.10
MATERIALS AND METHODS
The WRP is located on a tea plantation neighboring Kericho, Kenya, approximately 260 km northwest of Nairobi. More than 50,000 employees and dependents reside in housing clusters (estates) over the 104-km2 plantation. Institutional review board approval was obtained in Kenya and the United States for a study of HIV-1 prevalence and risk factors among volunteers (employees and adult dependents) at 1 tea plantation in Kericho. Enrollment was open to individuals residing at 1 of 6 estates at which HIV/AIDS education was being delivered by WRP staff. Inclusion criteria included being an adult between 18 and 45 years of age, residing on the estate, ability to provide informed consent, and planning to reside on the estate for at least 18 months. Exclusion criteria included not meeting inclusion criteria or unwillingness to participate in the survey or blooddraw.
Enrollment and Data Collection Procedures
With the consent of plantation management, investigators recruited volunteers from local HIV/AIDS education sessions in June 2000. Prospective volunteers were informed of study goals and procedures, including the fact that neither individually identified survey data nor HIV test results would be shared with plantation management. After enrollment and documentation of informed consent, each volunteer was interviewed and provided a blood sample. Confidential interviews were administered in either English or Kiswahili using a structured interview instrument developed with input from local personnel.
HIV Testing Procedures
Rapid HIV-1 serologic testing was performed at the WRP laboratory on the same day as phlebotomy. Anticoagulated whole blood was used for Determine HIV-1/2 (Abbott Laboratories, Inc, Abbott Park, IL) and ImmunoGold HIV-1/2 (Orgenics, Waltham, MA) testing in parallel. Any positive or discrepant HIV-1 result noted on rapid HIV testing was then confirmed by ELISA (Ortho HIV-1/2 Ab-Cap. ELISA System, Ortho Diagnostic Systems, Rochester, NY). HIV-1 results were given to the volunteers approximately 2 weeks after testing along with posttest counseling.
χ 2 or Fisher exact tests were applied to compare HIV-1 prevalences in categorical variables (marital status, education, tribal affiliation, employment, travel, history of malaria attacks, circumcision [men], and receipt of goods in exchange for sex [women]). χ 2 test for trend was used to examine trends in HIV-1 prevalence for age and age at first sex. To evaluate the association of risk factors with HIV-1 infection, adjusted odds ratios (AORs) stratified by sex and controlled for age, education, and tribal affiliation, with 95% confidence intervals (CIs), were estimated using logistic regression analysis. All analyses were performed using JMP 5.0 (SAS Institute, Cary, NC).
A total of 845 volunteers were recruited, with HIV-1 serology results available for 97%. Quality was insufficient to determine HIV-1 serologic status for 25 samples. All analyses were based on 820 individuals.
Approximately 80% of volunteers were men. Fifty-seven percent of the volunteers were younger than 30 years. Nearly 70% of the volunteers were married. Most of the volunteers (96%) had received at least a primary school education. More than one third of the volunteers belonged to the Kalenjin tribe, with approximately 20% belonging to each of the next 3 most common tribes (Table 1).
Eighty-one (52 men and 29 women) of 820 volunteers were found to be HIV infected, yielding an overall HIV-1 prevalence of 9.9% (95% CI, 7.9%-12.1%) (Table 2). HIV-1 prevalence was twice as high among women as men (17.4% vs 8.0%, P < 0.001).
Among men, HIV-1 prevalence increased with increasing age group (P = 0.008). A trend was not seen among women; however, those 30 years or older had a 2-fold increase in HIV-1 prevalence compared with those younger than 30 years.
Married men had a statistically significantly higher HIV-1 prevalence than single men (P < 0.001). Among women, there was no statistically significant difference in HIV-1 prevalence among marital status groups, however.
Among men, HIV-1 prevalence was 8% for those who had ever attended school and those who had not. HIV-1 prevalence among women who never attended school was 37%, dropping to 20% among those who had completed a primary school education and 10% who had completed a secondary school education (P = 0.006).
HIV-1 prevalence varied by sex and tribal affiliation. Luo tribe members had the highest HIV-1 prevalence (23.5% among men and 29.4% among women). Members of the Kalenjin tribe (the largest tribal group represented) had the lowest HIV-1 prevalence among men (3.3%) and the second lowest among women (10.4%).
Employed men and women had higher HIV-1 prevalences than dependents, but this difference was only statistically significant for men (8.9% vs 3.1%, P = 0.039).
Volunteers who had traveled to Kericho in the past 2 months had a higher HIV-1 prevalence, but again, this difference was statistically significant only for men (11.0% vs 6.1%, P = 0.005).
Among men, HIV-1 prevalence increased by increasing age at first intercourse from 8% to 10%. However, HIV-1 prevalences for intervals of age at first intercourse were 2 to 3 times higher for women than for men.
Men and women who reported 4 or more malaria attacks in the previous year had higher HIV-1 prevalences than those who reported no attacks. This difference was statistically significant for men only (11.0% vs 4.8%, P = 0.006).
Overall, 73% of men reported having been circumcised. The proportion circumcised varied greatly by tribal affiliation. More than 99% of Kalenjin men were circumcised compared with 18.6% of Luo men. Men who reported not having been circumcised had an HIV-1 prevalence more than 7-fold higher than men who reported having been circumcised (29.2% vs 4.3, P < 0.001).
Fifteen percent of women reported having received goods in exchange for sex. HIV-1 prevalence in these women was 36.0% compared with 13.9% in women who reported no exchange of goods for sex (P = 0.017).
More than 97% of all volunteers indicated a willingness to participate in future HIV studies involving periodic follow-up visits. There were no statistically significant differences in this proportion by HIV-1 status, sex, age, marital status, education, or tribal affiliation.
Among male volunteers, significant risk for prevalent HIV-1 infection was associated with age older than 30 years (AOR, 2.3), marriage (AOR, 4.0), tribal affiliation (Luo: AOR, 9.3; Luhya: AOR, 2.8), and 4 or more malaria attacks in the past 12 months (AOR, 2.2; Table 3). In addition, circumcision was a statistically significant protective factor against HIV-1 infection (AOR, 0.3). Among women, significant risk for prevalent HIV-1 infection was seen in those who never attended school (AOR, 4.9) and those who received goods for sex (AOR, 2.9).
The results from this study indicate a high prevalence of HIV-1 infection among these rural agricultural residents. Our findings are consistent with the data reported by the Kenyan Demographic Health Survey of 2003.1 The sex and age distribution for HIV-1 infection in this study in Kericho reflects a heterosexually transmitted epidemic typical of sub-Saharan African populations.
In this study HIV-1 prevalence varied greatly by tribal affiliation and male circumcision status. Luo tribe men had as much as an 8-fold increase in HIV-1 prevalence compared with men from other tribes. Previous studies have reported the highest HIV-1 prevalences among the Luo.1,11 Circumcised men had a 7-fold lower HIV-1 prevalence compared with uncircumcised men in our study. Circumcision rates varied by tribe, ranging from 18.6% (Luo tribe) to 100% (Kisii tribe). This finding is consistent with previous research which has documented a lower HIV prevalence among circumcised men.11-18 A recent systematic review of 37 studies of male circumcision and HIV infection in both high-risk and general populations found that the quality and design of the studies were variable. In particular, the importance of confounding factors was not uniformly considered across the studies.19 For example, factors such as religion and confirmed sexually transmitted infections could play a significant role in confounding the relationship between circumcision and HIV infection. The findings from our study also may be potentially confounded by these issues as the study was not specifically designed to examine the association between circumcision and HIV risk. Subsequent longitudinal studies in this population will assess this relationship more fully.
HIV-1 prevalence was higher in women than in men in our study. This sex disparity may be a reflection of varying societal norms throughout the world that place women at greater risk of HIV-1 infection, such as coerced sex, prostitution, rape, and genital mutilation.20 Unequal power dynamics in some communities frequently result in increases in domestic violence directed toward women.21-23 In addition, due to economic constraints, women may be forced to turn to commodity-driven sex. With an often limited ability to negotiate male condom use with their partners, these women are at high risk for HIV infection.1 Our study was not designed to investigate these factors. Consistent with our study, marriage has been reported as a risk factor for HIV infection in other studies. In Kisumu, Kenya, among adolescent girls aged 15 to 19 years, HIV-1 prevalence was 10% higher in those who were married than in those who were single and sexually active.24
Limitations and Lessons Learned
The volunteers enrolled in this study were drawn from convenience sampling from estates at which HIV/AIDS education was ongoing. Furthermore, concern was expressed by some community members about the relationship between the study organizers and the employer. Despite messages that data on risk behavior and HIV status would never be shared with management, it is possible that some individuals were still reluctant to volunteer for this study. Although there was no evidence to suggest that volunteers were at higher or lower risk than nonvolunteers the fact that nonparticipation was not assessed leaves open the possibility of bias in some estimates.
An effort was made to hire project staff to serve as field workers who were both educated and not employed by the plantation. Thus, the field staff was made up of young dependents of plantation workers who commanded less respect within the community than older individuals might have. In addition, in some instances, project coordinators and field workers could have received additional training about the project and about HIV-1 to allow them to better address questions and concerns from the community. It was recognized that these issues needed to be addressed before cohort studies could move forward.
Women represented only 20% of the total enrollment in this study. Focus groups were later held throughout the tea plantation to evaluate sex-specific participation. Based on these findings, women-only education forums were instituted, and flexible hours for enrollment and follow-up were devised so that women could schedule visits around work or domestic responsibilities. Enrollment and empowerment of women to openly enroll in HIV-1 research trials remain a concern in the region of East Africa and will need to be continuously evaluated and addressed.
The self-reported history of malaria attacks in our study was significantly associated with HIV-1 infection. However, these reports were not laboratory-confirmed cases of malaria. Thus, they may have represented medical events other than malaria. The significance of this association remains unclear.
Vaccine Cohort Development
Research on HIV-1 and other vaccines has helped to highlight 3 critical features of potential vaccine cohorts:25 (1) moderate to high annual HIV incidence rate (minimum of 1%-2%), (2) continued HIV exposure risk even in the presence of directed counseling, and (3) community willingness to participate in a vaccine trial. The present study suggests a substantial HIV-1 infection rate and documented a high willingness for continued participation in HIV-1 research; the currently ongoing prospective cohort study begun after this study will provide further critical information.
Medical and educational infrastructures within the tea plantation have been expanded and will be used as a platform to support future vaccine trials. The Kericho community is actively involved with the establishment of a community advisory board and ongoing community level education. Finally, a mutually beneficial relationship has developed between the investigators, the tea plantation management, and the employees, suggesting that cohort studies can be successful in this area.
A major goal of this project was to determine the suitability of this site for future HIV-1 vaccine cohort development. Lessons learned from this study have facilitated current cohort activities in this area where more than 2800 volunteers have been recruited for extended follow-up. This research will characterize HIV-1 baseline prevalence, incidence, risk behaviors, and willingness to participate in future HIV-1 vaccine trials.
The authors thank Ms Lilian Langat who was program manager in Kericho during this study. Her dedication and motivation paved the way for the current cohort activities in Kericho. The study would not have been possible without the wholehearted cooperation of the tea plantation management including the medical staff and employees of James Finlay Kenya, Ltd. The authors also thank Drs Nelson Michael and Paul Scott of the US Military HIV Research Program for their review of the manuscript.
1. Kenya Central Bureau of Statistics, Ministry of Health, and ORC Macro. Kenya Demographic and Health Survey 2003
. Calverton, MD: CBS, MOH and ORC Macro; 2004.
3. McCutchan FE. Understanding the genetic diversity of HIV-1. AIDS
. 2000;14(suppl 3):S31-S44.
4. Peeters M, Toure-Kane C, Nkengasong JN. Genetic diversity of HIV in Africa: impact on diagnosis, treatment, vaccine development and trials. AIDS
5. Louwagie J, Janssens W, Mascola J, et al. Genetic diversity of the envelope glycoprotein from human immunodeficiency virus type 1 isolates of African origin. J Virol
6. Neilson JR, John GC, Carr JK, et al. Subtypes of human immunodeficiency virus type 1 and disease stage among women in Nairobi, Kenya. J Virol
7. Dowling WE, Kim B, Mason CJ, et al. Forty-one near full-length HIV-1 sequences from Kenya reveal an epidemic of subtype A and A-containing recombinants. AIDS
8. Yang C, Li M, Shi Y, et al. Genetic diversity and high proportion of intersubtype recombinants among HIV type 1-infected pregnant women in Kisumu, western Kenya. AIDS Res Hum Retroviruses
9. Songok EM, Lwembe RM, Kibaya R, et al. Active generation and selection for HIV intersubtype A/D recombinant forms in a coinfected patient in Kenya. AIDS Res Hum Retroviruses
10. Gaschen B, Taylor J, Yusim K, et al. Diversity considerations in HIV-1 vaccine selection. Science
11. Agot KE, Ndinya-Achola JO, Kreiss JK, et al. Risk of HIV-1 in rural Kenya-a comparison of circumcised and uncircumcised men. Epidemiology
12. Moses S, Bradley JE, Nagelkerke NJ, et al. Geographical patterns of male circumcision practices in Africa: association with HIV seroprevalence. Int J Epidemiol
13. Moses S, Plummer FA, Bradley JE, et al. The association between lack of male circumcision and risk for HIV infection: a review of the epidemiological data. Sex Transm Dis
14. Van Howe RS. Circumcision and HIV infection: review of the literature and meta-analysis. Int J STD AIDS
15. Gray RH, Kiwanuka N, Quinn TC, et al. Male circumcision and HIV acquisition and transmission: cohort studies in Rakai, Uganda. Rakai Project Team. AIDS
16. Weiss HA, Quigley MA, Hayes RJ. Male circumcision and risk of HIV infection in sub-Saharan Africa: a systematic review and meta-analysis. AIDS
17. Bailey RC, Plummer FA, Moses S. Male circumcision and HIV prevention: current knowledge and future research directions. Lancet Infect Dis
18. Baeten JM, Richardson BA, Lavreys L, et al. Female-to-male infectivity of HIV-1 among circumcised and uncircumcised Kenyan men. J Infect Dis
19. Siegfried N, Muller M, Deeks J, et al. HIV and male circumcision-a systematic review with assessment of the quality of studies. Lancet Infect Dis
20. Heise L, Ellsbergd M, Gottemoeller M. Ending violence against women. Population reports, series L, no. 11. Baltimore: Johns Hopkins University School of Public Health, Population Information Program; December 1999.
21. Odujinrin O. Wife battering in Nigeria. Int J Gynecol Obstet
22. Fikree FF, Bhatti LI. Domestic violence and health of Pakistani women. Int J Gynecol Obstet
23. Ilika AL, Okonkwo PI, Adogu P. Intimate partner violence among women of childbearing age in a primary health care centre in Nigeria. Afr J Reprod Health
24. Glynn JR, Carael M, Auvert B, et al. Why do women have a much higher prevalence of HIV than young men? A study in Kisumu, Kenya and Ndola, Zambia. AIDS
. 2001;15(suppl 4):S51-S60.
25. Suligoi B, Wagner TM, Ciccozzi M, et al. The epidemiological contribution to the preparation of field trials for HIV and STI vaccines: objectives and methods of feasibility studies. Vaccine