Date Published: February 6, 2015
Publisher: Public Library of Science
Author(s): A. Desiree LaBeaud, Tamara Banda, Julie Brichard, Eric M. Muchiri, Peter L. Mungai, Francis M. Mutuku, Erin Borland, Ginny Gildengorin, Sarah Pfeil, Crystal Y. Teng, Kristin Long, Mark Heise, Ann M. Powers, Uriel Kitron, Charles H. King, Matthew Kasper. http://doi.org/10.1371/journal.pntd.0003436
Abstract: BackgroundChikungunya virus (CHIKV) and o’nyong nyong virus (ONNV) are mosquito-borne alphaviruses endemic in East Africa that cause acute febrile illness and arthritis. The objectives of this study were to measure the seroprevalence of CHIKV and ONNV in coastal Kenya and link it to demographics and other risk factors.MethodologyDemographic and exposure questionnaires were administered to 1,848 participants recruited from two village clusters (Milalani-Nganja and Vuga) in 2009. Sera were tested for alphavirus exposure using standardized CHIKV IgG ELISA protocols and confirmed with plaque reduction neutralization tests (PRNT). Logistic regression models were used to determine the variables associated with seropositivity. Weighted K test for global clustering of houses with alphavirus positive participants was performed for distance ranges of 50–1,000 meters, and G* statistic and kernel density mapping were used to identify locations of higher seroprevalence.Principal Findings486 (26%) participants were seropositive by IgG ELISA. Of 443 PRNT confirmed positives, 25 samples (6%) were CHIKV+, 250 samples (56%) were ONNV+, and 168 samples (38%) had high titers for both. Age was significantly associated with seropositivity (OR 1.01 per year, 95% C.I. 1.00–1.01); however, younger adults were more likely to be seropositive than older adults. Males were less likely to be seropositive (p<0.05; OR 0.79, 95% C.I. 0.64–0.97). Adults who owned a bicycle (p<0.05; OR 1.37, 95% C.I. 1.00–1.85) or motor vehicle (p<0.05; OR 4.64, 95% C.I. 1.19–18.05) were more likely to be seropositive. Spatial analysis demonstrated hotspots of transmission within each village and clustering among local households in Milalani-Nganja, peaking at the 200–500m range.Conclusions/SignificanceAlphavirus exposure, particularly ONNV exposure, is common in coastal Kenya with ongoing interepidemic transmission of both ONNV and CHIKV. Women and adults were more likely to be seropositive. Household location may be a defining factor for the ecology of alphaviral transmission in this region.
Partial Text: Alphaviruses are endemic to many regions of Kenya; however, due to limited surveillance and documentation during and in between known outbreaks, the alphaviral burden is yet to be fully recognized. O’nyong-nyong virus (ONNV) and chikungunya virus (CHIKV) are closely related alphaviruses in the Semliki Forest antigenic complex . CHIKV was initially isolated from a febrile person in Tanzania in 1953  and small outbreaks occurred in Kenya in the late 1900s. In 2004, CHIKV re-emerged in Kenya and subsequently spread eastward around the Indian Ocean countries, causing a severe epidemic, and resulting in significant morbidity that heavily taxed the healthcare and public health infrastructure in many regions . More recently in 2013, CHIKV spread to the Americas, where it continues to cause outbreaks in many Caribbean islands with over 100,000 cases reported within the first 6 months . ONNV was initially isolated in Northern Uganda from anopheline mosquitoes and human serum during a 1959 epidemic . ONNV has been associated with relatively few but large-scale epidemics. In 1996, ONNV resurfaced in Southern Uganda, causing major isolated epidemics, but was last reported in Kenya in 1961 .
Of the 1864 samples initially tested for anti-alphavirus IgG antibodies by ELISA, 413 (22%) were positive, 134 (7%) were equivocal and 1317 (70%) were negative. Ninety percent (371/413) of ELISA positives, eighty seven percent (117/134) of ELISA equivocal and 4.5 percent (59/1317) of ELISA negatives were PRNT tested. In the 83 cases in which the PRNT result differed from the ELISA result, the PRNT result was used as the final result and the sixteen equivocal samples that were not PRNT tested were removed from the final analysis. Eighty four percent (313/371) of ELISA positive, 89% (105/117) of ELISA equivocal and 42% (25/59) of ELISA negative were confirmed positive by PRNT. In our final analysis, 486 out of 1848 (26%; 95% CI 24.3%–28.4%) participants were positive for anti-alphavirus IgG antibodies by ELISA and 443 were confirmed positive by PRNT (Fig. 1). PRNT data for all of the ELISA positive and equivocal samples are not available because of specimen volume limitations inhibiting both ONNV and CHIKV PRNT to be performed.
Alphavirus exposure, particularly ONNV exposure, was found to be common in coastal Kenya, despite little previous public health attention or research. Evidence of high transmission rates for ONNV was unexpected, given that the last known outbreak in this region was in 1961. The fact that multiple children and young adults were exposed shows that ONNV transmission has occurred in the past five years and remains undetected or poorly reported within the clinical setting. Not surprisingly, children were less likely to be seropositive than adults, likely due to fewer years of mosquito exposure. This suggests that there is endemic circulation with additive risk for exposure over time. Interestingly, younger adults were more likely to be exposed than older adults in this study; however, we did not find evidence that older adults were losing their titers over time (Fig. 3). These two findings suggest an increase in transmission risk of alphaviruses within the last 3 decades in these communities.