WNV is transmitted primarily by the bite of infected mosquitoes that acquire the virus by feeding on infected birds. The intensity of transmission to humans is dependent on abundance and feeding patterns of infected mosquitoes and on local ecology and behavior that influence human exposure to mosquitoes. Although up to 55% of affected populations became infected during epidemics in Africa, more recent outbreaks in Europe and North America have yielded much lower attack rates (1
). In the area of most intense WNV transmission in Queens, New York, in 1999, ≈2.6% of residents were infected (most of these were asymptomatic infections), and similarly low prevalence of infection has been seen in other areas of the United States (3
). WNV outbreaks in Europe and the Middle East since 1995 appear to have caused infection in <5% of affected populations (1
). These levels of infection are too low to decrease the frequency of epidemics or modulate their intensity through protective immunity.
Data on the incidence of WNV in most of the world are not readily available. WNV transmission has been reported in Europe, the Middle East, Africa, India, parts of Asia, Australia (in the form of Kunjin virus, a subtype of WNV), North America, and parts of Central America and the Caribbean (1
). In recent years human WNV disease in the Eastern Hemisphere has been reported mostly from areas in the Mediterranean Basin: in Algeria in 1994, Morocco in 1996, Tunisia in 1997 and 2003, Romania in 1996 through 2000, the Czech Republic in 1997, Israel in 1999 and 2000, Russia in 1999 through 2001, and France in 2003 (1
). Enzootics involving horses were reported in Morocco in 1996 and 2003, Italy in 1998, Israel in 2000, and southern France in 2000, 2003, and 2004 (6
In the Western Hemisphere, most human WNV disease has occurred in the United States. Since the virus was detected in New York from 1999 through 2004, 16,706 cases have been reported to the Centers for Disease Control and Prevention (CDC); 7,096 of these were classified as neuroinvasive disease, 9,268 as West Nile fever (WNF), and 342 had other or unspecified clinical presentation (reported through June 8, 2005; the proportion of total cases reported that are neuroinvasive disease is artificially higher than what is believed to occur naturally since neuroinvasive disease is more likely to be reported than WNF or asymptomatic infection) (). Transmission of WNV has spread dramatically from New York to the north, south, and west (). From 2002 to 2003, the most intense transmission shifted from the Midwest and south-central states to the western plains and Front Range of the Rocky Mountains. In 2004, most WNV disease cases were reported in California, Arizona, and western Colorado, but foci of highest incidence were scattered across the United States (). In the East, WNV transmission recurred for 6 consecutive years with the highest number of human disease cases reported in 2003, indicating that WNV disease has become seasonally endemic. In Canada, transmission of WNV to humans has been documented in Quebec, Ontario, Manitoba, Saskatchewan, and Alberta, and WNV-infected birds have also been found in New Brunswick and Nova Scotia (http://www.phac-aspc.gc.ca/wnv-vwn
). Evidence of WNV transmission has been reported from the Cayman Islands, Jamaica, Dominican Republic, Mexico, Guadeloupe, El Salvador, Belize, Puerto Rico, and Cuba, but only 1 human case has been reported from Mexico and 1 from the Cayman Islands (http://www.paho.org/English/DD/PIN/ptoday15_oct03.htm;
). The paucity of human cases thus far in Latin America and the Caribbean is surprising, considering the ecologic conditions that favor arbovirus transmission in these areas. WNV isolated from a bird in Mexico in 2003 appeared to be attenuated, but whether viral mutation accounts for the scarcity of human disease remains to be seen (9
Human West Nile virus disease cases by clinical syndrome, United States, 1999–2004*
Reported incidence of neuroinvasive West Nile virus disease by county, United States, 1999–2004. Reported to Centers for Disease Control and Prevention by states through April 21, 2005.
The incidence of WNV disease is seasonal in the temperate zones of North America, Europe, and the Mediterranean Basin, with peak activity from July through October (6
). In the United States, the transmission season has lengthened as the virus has moved south; in 2003, onset of human illness began as late as December, and in 2004 as early as April (CDC, unpub. data). Transmission of WNV in southern Africa and of Kunjin virus in Australia increases in the early months of the year after heavy spring and summer rainfall (2
In the United States, persons of all ages appear to be equally susceptible to WNV infection, but the incidence of neuroinvasive WNV disease and death increases with age, especially among those 60 to 89 years of age, and is slightly higher among male patients () (10
). During 2002, the median age among neuroinvasive disease cases was 64 years (range 1 month to 99 years), compared to a median age of 49 years (range 1–97 years) for WNF cases (10
). Of the 2,942 neuroinvasive disease cases, 276 (9%) were fatal (10
). Although severe disease occurs primarily in adults, neuroinvasive disease in children has been reported. From 2002 through 2004, 1,051 WNV disease cases among children <19 years of age were reported in the United States; 317 (30%) had neuroinvasive disease; and 106 (34%) of these were <10 years (CDC, unpub. data; reported through June 8, 2005). Two (0.6%) pediatric patients with neuroinvasive WNV disease died: an infant with underlying lissencephaly and a 14-year-old boy with immune dysfunction.
Reported incidence of neuroinvasive West Nile virus disease by age group and sex, United States, 1999–2004. Reported to the Centers for Disease Control and Prevention by states through April 14, 2005.
The most important risk factor for acquiring WNV infection is exposure to infected mosquitoes. In Romania the risk for WNV infection was higher among persons with mosquitoes in their homes and with flooded basements (12
). An analysis of the locations of WNV disease cases during the 1999 outbreak in New York found that cases were clustered in an area with higher vegetation cover, indicating favorable mosquito habitat (13
). A study of the outbreak in Chicago in 2002 indicated that human disease cases tended to occur in areas with more vegetation, older housing, lower population density, predominance of older Caucasian residents, and proximity to dead birds, but the effects of these variables were influenced by differences in mosquito abatement efforts (14
). Risk factors for infection not related to mosquito exposure include receiving blood transfusions or organ donations, maternal infection during pregnancy or breastfeeding, and occupational exposure to the virus (15
Apart from older age and immunosuppression after organ transplantation, the risk factors for the development of severe neuroinvasive WNV disease have yet to be determined (10
). Underlying hypertension, cerebrovascular disease, and diabetes have been considered as possible predisposing factors; further study may elucidate the role of these or other host factors that might modify the risk for severe disease or death (12
). Genetic predisposition for severe disease has been described in mice but has not yet been elucidated in humans (18
). The role of innate and adaptive immune responses in determining outcome deserves further study.