Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
From the Centers for Disease Control and Prevention
August 8, 2007

West Nile Virus Activity—United States, 2006

JAMA. 2007;298(6):619-621. doi:10.1001/jama.298.6.619

MMWR. 2007;56:556-559

2 figures omitted

West Nile virus (WNV) is the leading cause of arboviral encephalitis in the United States. Originally discovered in Africa in 1937, WNV was first detected in the western hemisphere in 1999 in New York City. Since then, WNV has caused seasonal epidemics of febrile illness and severe neurologic disease in the United States. This report summarizes provisional WNV surveillance data for 2006 reported to CDC as of April 3, 2007. During 2006, WNV transmission to humans or animals expanded into 52 counties that had not previously reported transmission and recurred in 1,350 counties where transmission had been reported in previous years. In addition, 1,491 cases of WNV neuroinvasive disease (WNND) were reported in the United States during this period, amounting to a 14% increase from 2005 and the largest number reported since 2003. On the basis of extrapolations from past serosurveys, an estimated 41,750 cases of non-neuroinvasive WNV disease occurred in 2006; of these cases, 2,770 were reported. These findings highlight the need for ongoing surveillance, mosquito control, promotion of personal protection from mosquito bites, and research into additional prevention strategies.

WNV data are reported to CDC through ArboNET, an Internet-based arbovirus surveillance system managed by state health departments and CDC. State and local health departments (1) collect reports from health-care providers and clinical laboratories regarding cases of WNV disease in humans; (2) collect and test dead birds, often focusing on corvids (e.g., crows, jays, and magpies), which have high mortality attributed to WNV infection; (3) collaborate with veterinarians to collect reports of WNV infection in nonhuman mammals; and (4) collect mosquitoes to test for evidence of WNV infection. Human WNV disease cases are classified as (1) WNND (i.e., meningitis, encephalitis, or acute flaccid paralysis); (2) West Nile fever (WNF), which is symptomatic WNV disease that does not affect the nervous system; (3) other clinical illness; or (4) unspecified (i.e., unknown) illness. WNF reporting is highly variable by jurisdiction, depending on the level of interest in reporting and utilization of diagnostic testing; therefore, this report focuses on WNND cases, which are thought to be more consistently identified and reported because of the severity of the illness.

Human Surveillance

During 2006, a total of 4,261 cases of WNV disease in humans were reported from 731 counties in 43 states and the District of Columbia, accounting for 23.3% of the 3,142 counties in the United States. Of these cases, 1,491 were WNND, 2,612 were WNF, and 158 were unspecified illnesses. Idaho, a state that reported four WNND cases (from a total of 17 human cases) during 2003-2005, reported 139 WNND cases in 2006, accounting for 9.3% of the national total. Idaho first reported any WNV activity in 2002; the first human case in the state was reported in 2003. Other focal outbreaks of WNND occurred in states that experienced outbreaks in previous years, including Texas (229 WNND cases), Illinois (127), Louisiana (91), and Mississippi (89). In the New York City metropolitan area, WNV disease recurred for the eighth consecutive year, with eight WNND cases reported. The counties with the highest incidence of WNND were primarily in the west-central United States. The states with the highest incidence included Idaho (9.9 cases per 100,000 residents), South Dakota (4.9), and North Dakota (3.2). The incidence of WNND peaked during the first week in August, and the overall trend was consistent with the seasonality observed in the preceding 6 years.

The median age of the 1,491 persons with WNND was 58 years (range: 3 months–99 years), and 891 (59.8%) were male. A total of 1,311 (87.9%) persons were hospitalized, and 161 (10.8%) died. A total of 101 (6.8%) persons with WNND had acute flaccid paralysis; the median age among these persons was 53 years (range: 1-87 years), and 62 (61.4%) were male. Twelve (11.9%) died; the median age of these persons was 76 years (range: 19-99 years).

Animal Surveillance

In 2006, a total of 4,106 dead WNV-infected birds were reported from 701 counties in 43 states; 404 counties from 38 states reported infected birds but no human disease. Collection of WNV-infected birds peaked during mid-August. Corvids accounted for 3,292 (80%) of the birds; the majority of states targeted corvids for surveillance. Since 1999, WNV infection has been identified in approximately 300 avian species, including 11 species in which WNV was identified for the first time during 2006.

Of 1,121 reported WNV disease cases among nonhuman mammals, 1,086 (96.9%) occurred in equine animals, and 35 (3.2%) occurred in other species (squirrels [33] and unspecified species [two]). Equine cases were reported from 414 counties in 34 states; Idaho reported 31% of all equine cases. Peak reported incidences of equine disease occurred during mid-August.

A total of 11,898 mosquito pools* from 459 counties in 38 states and the District of Columbia tested positive for WNV. Among the WNV-positive pools, 8,665 (72.8%) were made up of Culex mosquitoes thought to be the principal vectors of WNV transmission (i.e., Cx. pipiens, Cx. quinquefasciatus, Cx. restuans, Cx. salinarius, and Cx. tarsalis).1 Unidentified or other species of Culex mosquitoes made up 3,032 (25.5%) pools, and non-Culex species (i.e., Aedes spp., Anopheles spp., Coquillettidia spp., Culiseta spp., Ochlerotatus spp., and Psorophora spp.) made up 135 (1.1%) pools. Data from 2006 included the first report of WNV infection in Culexapicalis, which was collected in Arizona. The number of reported WNV-infected mosquito pools peaked during the first week in August.

Reported by:

NP Lindsey, MS, JA Lehman, EB Hayes, MD, RS Nasci, PhD, N Komar, ScD, LR Petersen, MD, Div of Vector-Borne Infectious Diseases, National Center for Zoonotic, Vector-Borne, and Enteric Diseases, CDC.

CDC Editorial Note:

In 2006, a total of 1,491 cases of WNND were reported, the highest number reported since 2003.† WNV activity was detected in all 48 contiguous states for the second consecutive year. Human WNV disease was scattered throughout the United States, but the majority of cases were reported in Idaho and in the west-central states. One state (Washington) reported human cases for the first time. The increase in reported cases since 2004 suggests that endemic transmission of WNV in the United States will continue. Although WNND case reports from Idaho (a state that reported only four WNND cases during 2003-2005) accounted for nearly 10% of all WNND cases reported in 2006, focal outbreaks also recurred in areas where seasonal transmission has occurred for several years.1

This report focuses on WNND cases because of the variability in WNF reporting by jurisdiction and by year. Reporting of WNND is thought to be more consistent and complete because of the higher likelihood of hospitalization and testing. Although the Council of State and Territorial Epidemiologists designated WNF as a notifiable disease in 2005, the true incidence and public health impact of WNF remains underestimated by national surveillance data.2,3 Population-based serologic surveys indicate that approximately 140 WNV infections occur for every case of WNND and that of all persons who become infected, approximately 20% have onset of WNF and 80% remain asymptomatic.2,4 By applying these ratios to the 1,491 reported WNND cases, an estimated 208,700 cases of WNV infection (1,491 WNND cases × 140) and 41,750 cases of WNF (208,700 × 0.20; only 2,612 cases were reported) occurred in the United States in 2006.

Although persons of all ages appear equally susceptible to WNV infection, both the incidence of WNND and the incidence of death related to WNND increase with age, especially among persons aged >60 years, and are slightly higher among males.1,5 During 2006, the median age among persons with fatal WNND was similar to that of previous years.4,6

Reports of WNV disease in equine animals have decreased annually since 2002 (CDC, unpublished data, 2007). Whether this decline represents a true decrease in disease incidence resulting from naturally acquired immunity or vaccination7 or is a result of reduced emphasis on equine WNV disease reporting is not clear. Nonetheless, the temporal and geographic distribution of equine WNV cases continues to correlate with human cases, suggesting that surveillance of equine animals can continue to help indicate areas of increased risk for human WNV disease.

Since 1999, corvids have accounted for the majority (>70%) of all WNV-infected dead birds reported to CDC. The substantial number of reported corvid deaths likely results from the size of corvids and their susceptibility to WNV disease and from surveillance programs specifically targeted at corvids. Geographically, surveillance of WNV in different bird species can vary in usefulness as indicators for WNV transmission; targeting locally relevant species can optimize efficiency of WNV surveillance.

As of December 31, 2006, WNV had been detected in 62 of the approximately 175 mosquito species found in the United States. In 2006, Culex mosquitoes (specifically Cx. pipiens, Cx. quinquefasciatus, Cx. restuans, Cx. salinarius, and Cx. tarsalis) continued to be the most prevalent in WNV-positive pools. Although 33 different WNV-infected mosquito species were identified in 2006, Culex mosquitoes are believed to account for the majority of WNV transmission in the United States.1 Therefore, Culex mosquitoes remain the primary vector target for prevention of WNV disease in the United States.

WNV surveillance is important for monitoring further spread of the virus and targeting prevention and control strategies. The ArboNET surveillance system focuses on arboviral diagnosis, testing, and reporting and is well positioned to detect increased transmission of all domestic arboviruses, to identify future introduction of foreign arboviruses, and to monitor effects of climate and other determinants of arboviral disease incidence.

In the absence of an effective human vaccine, prevention of WNV disease depends on community-level mosquito control (e.g., larviciding, adulticiding, and breeding-site reduction) and promotion of personal protection against mosquito bites, such as use of repellents and avoiding outdoor exposure when mosquitoes are most active (usually from dusk to dawn). Repellents containing DEET, picaridin, or oil of lemon eucalyptus provide protection against mosquito bites. Intact window screens or air conditioning can reduce mosquito exposure in homes. Numbers of mosquitoes can be reduced by removing or emptying water from larval habitats such as flower pots, buckets, gutters, and barrels.


This report is based, in part, on data provided by ArboNET surveillance coordinators in local and state health departments and ArboNET technical staff, Div of Vector-Borne Infectious Diseases, National Center for Infectious Diseases, CDC.

*A sample of mosquitoes (usually no more than 50) of the same species and sex, collected within a defined sampling area and period.

†Additional information available at

Hayes EB, Komar N, Nasci RS, Montgomery SP, O’Leary DR, Campbell GL. Epidemiology and transmission dynamics of West Nile virus disease.  Emerg Infect Dis. 2005;11(8):1167-1173PubMedArticle
Mostashari F, Bunning ML, Kitsutani PT.  et al.  Epidemic West Nile encephalitis, New York, 1999: results of a household-based seroepidemiological survey.  Lancet. 2001;358(9278):261-264PubMedArticle
Watson JT, Pertel PE, Jones RC.  et al.  Clinical characteristics and functional outcomes of West Nile fever.  Ann Intern Med. 2004;141(5):360-365PubMedArticle
Tsai TF, Popovici F. West Nile encephalitis epidemic in southeastern Romania.  Lancet. 1998;352(9130):767-771Cernescu C, et alPubMedArticle
O’Leary DR, Marfin AA, Montgomery SP.  et al.  The epidemic of West Nile virus in the United States, 2002.  Vector Borne Zoonotic Dis. 2004;4(1):61-70PubMedArticle
CDC.  West Nile virus activity—United States, 2001.  MMWR Morb Mortal Wkly Rep. 2002;51(23):497-501PubMed
Davidson AH, Traub-Dargatz JL, Rodeheaver RM.  et al.  Immunologic responses to West Nile virus in vaccinated and clinically affected horses.  J Am Vet Med Assoc. 2005;226(2):240-245PubMedArticle