HIV-1 incidence rates and risk factors in agricultural workers and dependents in rural Kenya: 36-month follow-up of the Kericho HIV cohort study.

Background: Incidence data from prospective cohort studies using rigorous laboratory methods are important in designing and evaluating HIV vaccine and therapeutic clinical trials and health care programs. We report 36-month HIV-1 incidence rates and demographic and psychosocial risks from the Kericho cohort in rural Kenya's southern Rift Valley Province.
Methods: Thirty-six month, prospective, closed, observational cohort study of adult plantation workers and dependents followed biannually. HIV-1 incidence rates per 100 person-years (py) were calculated, and Cox regression analyses were used to estimate hazards ratios (HR) associated with seroconversion.
Results: Two thousand four hundred volunteers (mean age +/- SD = 30.1 +/- 8.5 years; 36.5% women) participated. Twenty-nine new HIV cases were identified in year 1 of follow-up, which increased to cumulative totals of 49 and 63 cases in years 2 and 3, respectively. The corresponding 1-, 2-, and 3-year incidence rates were 1.41 [95% confidence interval (CI) = 0.95-2.02], 1.16 (95% CI = 0.86-1.54), and 1.00 (95% CI = 0.77-1.28) per 100 py. Risk factors associated with HIV seroconversion included the following: of the Luo tribe (HR = 3.31; 95% CI = 1.65-6.63), marriage more than once (HR = 2.83; 95% CI = 1.20-6.69), self-reported male circumcision (HR = 0.32; 95% CI = 0.17-0.60), history of sexually transmitted infection (HR = 2.40; 95% CI = 1.09-5.26), history of substance abuse during sex (HR = 2.44; 95% CI = 1.16-5.13), and history of transactional sex (HR = 3.30; 95% CI = 1.79-6.09).
Conclusions: HIV-1 incidence rates were relatively low in adult plantation workers and dependents in rural Kenya. Cohorts including higher risk populations (eg, commercial sex workers) warrant consideration for regional HIV preventive vaccine trials. Even low incidence, well-described cohorts generate valuable epidemiological clinical trial data.