West Nile virus (WNV) is an arthropod-borne virus of the genus Flavivirus and belongs to the Japanese encephalitis virus (JEV) antigenic complex under family Flaviviridae (1). Since the first isolation of WNV in 1937 in the West Nile district of Uganda from a febrile woman (2), the virus has been spreading and is considered an emerging zoonosis (3). WNV can cause encephalitis in severe cases and a few patients develop acute flaccid paralysis (4, 5). Although WNV is transmitted in a cycle involving different birds and mosquitoes (6), domestic birds are unlikely to play a major role in the virus transmission as they usually do not develop sufficient viraemia (7, 8). However, seroconversion in poultry in any geographical region indicates the existence of the virus and also the consequent threat to human population, and therefore poultry can be used as sentinel for disease surveillance (9).

Assam is a state in North-eastern India, situated south of the eastern Himalayas along the Brahmaputra and Barak River valleys, covering an area of 78,438 km². To its north lies Bhutan and Arunachal Pradesh with Nagaland and Manipur to the East; Meghalaya, Tripura, Mizoram and Bangladesh to the South and West Bengal to the West. The state extends from 89°42′E to 96°E longitude and 24°8′ N to 28°2′ N latitude. It is the most populated state with about 50% of the 38.8 million inhabitants of Northeast India (10). Recognition of WNV as the causative agent of acute encephalitis syndrome in Assam was first confirmed in 2006, and the study by Khan et al. (10) was the first report of this flavivirus from eastern parts of India. Although JEV has been reported to cause epidemics in Assam since 1976, studies have shown that only 53.7% of persons with acute encephalitis syndrome (AES) were infected with this virus and most cases remain undiagnosed (11). Since many patients with symptoms of neurotropic viral infection are JEV negative, there is likely circulation of some other virus, which might be WNV, as signs and symptoms produced by JEV and WNV are similar, but little is known about WNV circulation in the state, although the virus has been confirmed in some cases of AES (12).

In Assam, a total of 52 mosquito species under 11 genera have been detected, the most predominant being Anopheles followed by Culex, Aedes, Mansonia, Armigeres, Mimomyia, Ochlerostatus, Malaya, Toxorhynchites, Ficalbia and Aedeomyia (13). Mosquitoes belonging to Ma. uniformis and Cx. vishnui group comprising of Cx. tritaeniorhynchus, Cx. vishnui and Cx. pseudovishnui were reported to be highly abundant, breeding predominantly in paddy fields (14).

Assam is a state with many national reserves and wetlands, and hence a diverse avifauna (15), but in addition to all wild birds, poultry is also very common, and there has been substantial development of poultry farming in India (16). Agriculture and allied activities play a major role in the economy and the farmers are considered to be the backbone of agriculture in Assam. Poultry has become a vital component of the farm economy as it generates additional income and employment in the rural area (17). Poultry rearing is traditionally popular in this state, especially in rural and tribal areas for meat and egg production and poultry is the primary source of animal protein and supplementary income for the rural people.

In Assam, WNV has been detected by reverse transcriptase-PCR in Cx. vishnui, Cx. whitmorei, Cx. tritaeniorhynchus, Cx. quinquefasciatus, Cx. pseudovishnui and Mansonia uniformis, and the viruses detected were found to be WNV lineage 5, and 99-100% similar to a WNV strain earlier found in a human sample in Assam (18).WNV has been identified from various regions of Assam to which JEV is endemic indicating that WNV might be a substantial contributor to the acute encephalitis syndrome (AES) cases in this region.

Rice is the main crop of Assam and occupies about two-thirds of the total cultivated area of the state. However, abundance of rainfall, rice fields, animal reservoirs and diverse mosquito species makes Assam the most vulnerable state for spread of vector-borne diseases. Circulation of highly pathogenic neuro-invasive WNV has also been reported from this part of the country (12), and it is therefore urgent to identify ways to protect the Assamese population, using a one-health approach (19). Considering the above facts, the present study was conducted to better understand WNV epidemiology and to map serological evidence of WNV infection in poultry in urban and peri-urban areas of Guwahati.

This repeated cross-sectional study was conducted from April, 2018 to March, 2019.

Based on the questionnaire data, it was found that the purpose of rearing chicken in the present study was mainly for earning livelihood by selling the birds, while only a small number of farmers reared the birds for home consumption and egg production.

In the present study, 25% of the total farms were under integrated production system and all of these belonged to peri-urban areas. In integrated systems the farmers only invest by providing the sheds, equipment and the existing land and the integrator provides all other requisites including day-old chick, feed, litter (bedding), medications, vaccines, technical support etc. for raising the birds. No farm had an abattoir and the farmers sold the birds directly in the market, or the birds were collected by the company in case of integrated farms. The farmers and their families were all living in the same compound at walking distance from the poultry house.

Out of the 24 visited farms, four had indigenous, which means local breeds or non-descript birds, with between 5 and 11 birds per farm. The other 20 farms kept broilers, and reported keeping between 200 and 100,000 birds ( Table 1 ). Farmers in the peri-urban areas of Guwahati reared comparatively more birds than the urban farmers. In the present study, 75.0% of peri-urban farms had more than 1000 birds followed by 16.7% farms with 101-500 birds. In contrast to peri-urban farms, 58.3% of urban farms had 101-500 birds followed by 33.3% farms with less than 100 birds. Most farms were deemed to have moderately clean shed (15/24 farms, defined as: Bedding dry or little bit damp. Feeders and water sources may have some contamination), and similarly most farms had moderately clean birds (16/24 farms, defined as: Majority of birds having plumage color more towards brownish or dusty).

The poultry farmers of urban and peri-urban Guwahati also reared cow, pig, goat, duck, quail, pigeon, and fishes ( Table 2 ). Out of the 24 farms, 66.7% farms were found to have at least one other animal species. Household of eight farms, five in urban and three in peri-urban areas, were only kept poultry. Other animals in contact with chickens were recorded to be comparatively higher in peri-urban (55.6%) than urban (44.4%) poultry farms of Guwahati. Presences of pigs were recorded only in peri-urban poultry farms and multiple in-contact of other animals with chickens were recorded from both urban and peri-urban farms.

Out of 864 samples, 36 were positive for antibodies against WNV by ELISA (4.2%, 95% confidence interval (CI) 2.9-5.7%). Further confirmation of these 36 ELISA positive samples by HI revealed 27 to be positive for WNV with an overall apparent prevalence of 3.1% (95% CI 2.07-4.5%) ( Table 3 ). Location-wise sero-prevalence of WNV in chicken was recorded to be 0.9% and 0.7% in urban areas and 7.4% and 5.6% in peri-urban areas, as evidenced by ELISA and HI, respectively, which was comparatively higher in peri-urban than urban areas of Guwahati. Half of the peri-urban farms (50%) had at least one positive bird, compared to one (8.3%) of the urban farms (p=0.07). The sero-prevalence was recorded from Hatigaon in urban Guwahati and three peri-urban locations namely, Jugukuchi, Kamalajari and Garal. Location-wise, highest WNV sero-prevalence was recorded in Jugukuchi (11.1%), followed by Kamalajari (8.3%) with lowest in Hatigaon and Garal, showing 2.8% sero-positivity each. No WNV sero-positivity in chicken was found in the other sites ( Figures 1 and 2 ).

WNV antibodies were detected in one poultry farm of urban Guwahati compared to six sero-positive farms belonging to peri-urban areas ( Figure 1 ). Through mapping of the farm coordinates, it was possible to see that positive farms were both in densely populated as well as in more peri-urban locations. More farms with seropositive birds were observed among those that kept pigs (2/3, 66.7% (the negative farm had only one pig)) compared to farms not keeping pigs (5/21, 23.8%), and the same was observed for farms keeping any kind of livestock apart from poultry (4/8, 50%) compared to farms with only poultry (3/16, 18.8%), but the difference was not significant (p>0.1). None of the small farms with indigenous poultry had any seropositive birds.

Maximum sero-positivity of WNV was reported in July and August, 2018 with 8.3% prevalence each. Lowest sero-positivity was recorded in November (1.4%) and no positive birds were recorded from December to April ( Figure 3 ). Out of a total of 27 WNV sero-positive samples, three were from urban farms, one each in the months of July, August and September 2018. The remaining 24 sero-positive samples from May to November 2018 were from peri-urban farms.

Most seroconversions were observed in monsoon (7.3%), followed by post-monsoon (2.8%) and pre-monsoon (0.9%) with no sero-positive birds detected in winter season ( Table 4 ). All the sero-positive samples in urban areas of Guwahati were recorded during monsoon season.

In the present study, the most common disease symptoms in the poultry farmers or their house-hold members were reported to be fever (75.0%) followed by flu-like symptoms (70.8%), headache (66.7%), fatigue (50.0%), off-fed (29.2%), myalgia (25.0%), arthralgia (25.0%), nausea (16.7%) and vomiting (12.5%). No farmers reported drowsiness, neurological symptoms or motor dysfunction. Peri-urban farmers (41.7%) suffered more than urban farmers (8.3%) from myalgia (p=0.06). The only symptom with an association with serological status of the farm was myalgia (p=0.02), where 4/7 (7.1%) seropositive farms reported myalgia, while this was only reported by 2/17 (11.8%) seronegative farms.

This study demonstrates that poultry in Assam, India, are exposed to WNV and that exposure is more common in peri-urban areas, and around the monsoon season. The higher positivity (4.17%) found in our study by ELISA compared to HI (3.13%) is likely due mainly to higher sensitivity, but could also be indicative of lower specificity, and more cross-reactions, maybe with JEV. Flaviviruses are prone to serological cross-reactions (31), and it may be difficult to conclude which virus an animal has been exposed to by only using one serological method, particularly in an area with more than one flavivirus circulating, like Assam.

Higher (29.40%) as well as lower (0.60-1.70%) WNV sero-positivity in domestic birds has been reported in other countries (7, 32). Domestic chickens do not usually show clinical signs of WNV, and therefore they are the most commonly used avian model for the study of immune responses to WNV infection. However, abnormal death and paralysis of domestic chickens has been found associated with sero-positivity against WNV (32).

There is expected variation in WNV sero-prevalence over the year, as the WNV epidemiology is determined by both abiotic and biotic factors, including climate (33–35). In this study, the sero-prevalence were higher during the months with high average monthly temperature and precipitation. Higher environmental temperature increases the ability of Culex mosquitoes to transmit WNV (36, 37), and in addition, higher temperatures facilities faster larval development and this was also shown in an earlier study in Assam, which found a peak of the relative mosquito abundance during monsoon and least mosquitoes during winter (38).

Sero-prevalence may also depend on managemental and husbandry practices as well as genetic variation in disease resistance (39). Under poor housing and sanitary conditions, vectors proliferate in dirty water, filthy environment, thereby, having their highest impact on the underprivileged section of the society (40), making them especially vulnerable. Vector densities have been reported to be higher in rural areas compared to urban localities due to more favorable habitats for some mosquito species (41), which is similar to our finding of more common seroconversions in peri-urban compared to urban areas. Similarly, presence of potential factors facilitating vector-borne and zoonotic diseases such as the way of living (42), socio-economic status of the people (43), livestock dependence (44), cultural aspects of community (45), drainage system (46), may also contribute to higher sero-prevalence of WNV in peri-urban areas.

In disease epidemiology, GIS and Global Positioning System (GPS) helps in envisioning topographical distribution of disease, with respect to time and space, which is difficult to perform in any other system (47). Through mapping of the coordinates of the positive farms in this study into online tools such as Google Maps, or Google Earth, it was possible to show that seroconversions in birds occurred in densely populated as well as peri-urban areas.

WNV infections are asymptomatic in 80% of infected humans (48), while most symptomatic persons develop fever, headache, muscle pain, joint pain, rash, or gastrointestinal symptoms and very rarely (less than 1%) infected persons show neurological disorders, including encephalitis, meningitis, or acute flaccid paralysis (49). However, humans are dead-end host as there is no danger of WNV transmission from infected humans because of insufficient viremia to permit virus transmission through mosquitoes (50). None of the poultry farmers interviewed in this study reported neurological symptoms, but fever and headache were commonly reported.

Agriculture along with livestock farming was adopted as a source of livelihood by majority of the peri-urban farmers in contrary to the absence of agriculture in urban areas which might be due to lesser land holding by farmers in urban areas. Due to presence of numerous job avenues offered by different companies in urban areas, combining farming with private job was also another option for urban farmers which was non-existent in peri-urban areas of Guwahati. Agriculture plays an important role in the Indian economy and 70% of its rural households still depend primarily on agriculture for their livelihood, comprising of 82% of small and marginal farmers (51). All the poultry farmers during the present study were found to be either small or marginal and most of them were engaged in agriculture and allied sectors corroborating the above statement.

The noticeable difference in the number of birds reared in urban and peri-urban farms might be due to more conveniences of farming in peri-urban areas because of larger land holding by farmers, joint families extending help in farming, involvement of more household members in farm activities, availability of agricultural produce which can be used as feed stuffs, proximity to wholesale markets, less availability of other sources of livelihood as well as pro-activeness of different companies expanding integrated system of broiler production by involving the peri-urban farmers which was evident from the present study. Mixed farming is a commonly adopted farming practice in India involving combination of crop production with one or more type of animal species like rearing of cattle, sheep, goat, pigs and poultry as well as fishery, beekeeping, sericulture, etc. (52).

The limitations of this study include the lack of confirmation with VNT, as well as the relatively low number of birds sampled each month per farm. Given the known cross-reactions between different flaviviruses, particularly JEV and WNV, and the fact that the ELISA kit used cannot differentiate these viruses (23, 53), it is possible that some of these sero-conversions could be against JEV instead of WNV. The best method for distinguishing the causative virus would be VNT, but this would require a biosafety level 3 laboratory, which was not available. However, the study shows the use of serological testing of birds to understand the pattern of flavivirus infections in the city area, thereby indicating risks for humans.

After completion of the present study, all the farmers under study were included in an awareness program about mosquito-borne diseases and which precautions that can be done, including to remove the possible mosquito breeding sites like stagnant-standing water, empty containers, broken flower pots, tip over buckets, tyres, etc. The farmers of the sero-positive farms were informed about the presence of West Nile virus in their respective areas. The farmers were also advised to seek medical help if they developed fever, headache, muscle pain, joint pain, neurological disorder or any other symptoms common to WNV infection.

In conclusion, this study found circulation of WNV in urban and peri-urban Guwahati, indicating that this virus may be an important contributor to the undiagnosed cases of acute encephalitis that occurs every year. Understanding the distribution of WNV sero-prevalence in chicken can support future strategic planning of vector-borne disease control approaches by identifying the underlying environmental factors as well as demography of urban and peri-urban farmers in a city such as Guwahati. For improved understanding of the risks for WNV as well as other mosquito-borne flaviviruses, it would be recommended to set up surveillance programmes with a One Health approach, where sampling in animals can advise preventive measures in humans.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The animal study was reviewed and approved by Institutional Animal Ethics Committee, Assam Agricultural University, Khanapara (Approval Registration No. 770/ac/CPCSEA/FVSc/AAU/IAEC/17-18/584 dated 09.08.2017).

Conceptualization, RH and JL. Methodology, JL, RH, and AT. Formal analysis, JL and AT. Investigation, AT, RH, JL, ND, NK, PT, and SM. Resources, JL. Writing—original draft preparation, AT, JL, and RH. Writing—review and editing, all authors. Supervision, JL, RH, SP, and DB. Project administration, RH and JL. Funding acquisition, RH and JL. All authors have read and agreed to the published version of the manuscript.

The authors are thankful to Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (Formas, grant number 2016-00364) for funding the research. The authors are also grateful to Assam Agricultural University, Khanapara, Guwahati, NRC on Pig, Rani, Guwahati, Assam, the CGIAR system and Uppsala University, Sweden for the supports provided during the study.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.