One Health Resource Center

One Health Resource Center

September 23, 2019
11 min read

Nipah virus: Going from bat to worse

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

Thomas M. Yuill, PhD
Thomas M.
Donald Kaye

Nipah virus (NiV), an RNA virus, and Nipah virus infection (NiVI) are relatively recently discovered entities. NiVI is a particularly lethal infection, with a mortality rate usually higher than 50%. Its reservoir is the flying fruit bat, but it can infect a wide range of mammals, including humans. NiV has been cited as a potential agent of bioterrorism because of many suitable physical attributes of the virus and the high mortality caused by the infection in people. Although there is no evidence of any country or group attempting to weaponize NiV, the potential certainly exists.

History and discovery

Suddenly, in September 1998, pigs in major production areas in Peninsular Malaysia began falling ill with febrile encephalitis, and piglets began dying of a disease not previously known. Most adult pigs survived. That outbreak persisted until February 1999. During that time, other outbreaks popped up in a different state in two different locations. Initially, the disease was attributed to Japanese encephalitis (JE) virus infections because pigs were known to be amplifying hosts of that virus. However, death is not characteristic of JE virus infections in adult pigs. At the same time, people were also becoming ill with fever, myalgia and encephalitis in the affected area, and more than 200 individuals were admitted to the hospital. Human cases did not fit the usual pattern for JE outbreaks. Clinical human encephalitis cases in JE occur more frequently in children than in adults. In Malaysia, most affected individuals were adult men, many of whom had physical contact with pigs. Many human cases were in clusters within the same household. These observations cast doubts on JE as the etiological agent responsible for the human and pig cases. Interestingly, no cases were reported in Malay villages that are predominantly Muslim and do not maintain pigs, even though they were located close to the affected ethnic Chinese pig farms. Coincidentally, several abattoir workers in Singapore also became infected after handling pigs imported from Malaysia.

A scientist from Australia’s animal health system and another from the CDC worked with the Malaysian Department of Veterinary Services to conduct animal postmortems and collect blood samples from pigs and other animals for testing at the Australian Animal Health Laboratory in Geelong. Testing confirmed that humans, pigs, dogs, cats, horses, goats and bats had been infected by NiV. That was critical in assisting the Malaysian veterinary services in successfully containing the virus and preventing further deaths. More than 1 million pigs in Malaysia were culled in the affected areas to halt the outbreak in humans. The outbreak in humans ended with the last case on May 27, 1999. Altogether, there were 265 human cases and 105 deaths — a case fatality rate of about 40%. No more cases have been reported there since then.


Virologists from the University of Malaysia isolated a virus from the cerebrospinal fluid (CSF) of several hospitalized patients that on electron microscopy looked like those of the Paramyxoviridae family. This observation definitively excluded JE virus, a completely different flavivirus. Serologically, the isolated virus reacted with Hendra virus antibodies. Also, subsequent genetic analysis showed its sequences to be about 80% similar to Hendra virus found in Australia but significantly different enough to classify it as a new and different virus, unlike any other known paramyxoviruses. The new virus was designated Nipah virus after the river that runs through the affected area.

Epidemiology and lethality

The question of the origin of the outbreak prompted field studies to determine if other animals were involved. In addition to humans and pigs, evidence of NiVI was found in dogs, cats and horses with flying fox fruit bats (genus Pteropus) as the most likely animals involved. Subsequent studies showed these bats to be the reservoirs of NiV. The outbreak was believed to be due to NiV contamination of pig food, perhaps from contaminated fallen fruit.

In the winter (cool season December through May) of 2001, NiV was recognized as the cause of human encephalitis cases in Bangladesh, 2,500-km distant from the Malaysia outbreak. NiV-caused encephalitis has occurred there nearly every winter season since, but the epidemiologic picture is far different from that of Malaysia. Because Bangladesh is a predominantly Muslim country, pigs are rarely kept. Pteropus fruit bats were the main suspected source of the virus, but the question was: What was the route of transmission to people? Consumption of virus-contaminated fallen fruit by children was one likely route but did not account for encephalitis cases in adults. Consumption of contaminated date palm sap emerged as a more common route of virus exposure. People climb palm trees, cut the bark and place pots below the cuts to collect the sap, which is consumed untreated. Studies have shown that the bats visit the pots and cut areas, contaminating the sap with saliva and urine. The sap is a delicacy for local people, who resist heating or pasteurization because it alters the flavor. Placing a skirt made of locally available materials around the pots and cut bark has been shown to provide a physical barrier for bats that prevents contamination of the sap. However, this is not the whole transmission story. In 2004, during a significant outbreak, person-to-person transmission was also documented.


Although palm sap is consumed across Bangladesh, human cases occur in a “Nipah belt” in densely populated western Bangladesh in districts near the Ganges River. The reason for this geographic restriction is not clear. A study showed that there were 23 introductions of the virus into Bangladesh in 2001 to 2017. Of these, 10 affected multiple people. During this period, there were 122 human cases, with a mean age of 27 years. Eighty-seven of these infected people died (71% case-fatality rate [CFR]), most likely among those with respiratory complications caused by infection. Nine individuals transmitted the virus to others, with an incubation period of 5 to 15 days after close contact, accounting for more than half of the positive cases.

Human NiVI has not been restricted to Malaysia and Bangladesh. In January and February of 2001, an NiV outbreak occurred in Siliguri, a city in India’s West Bengal state located across the border from Bangladesh. In that outbreak, there were 66 probable cases and 45 deaths (68% CFR). Epidemiological linkages between cases implicated person-to-person transmission and an incubation period of around 10 days. There was another outbreak in Kerala state in India in 2018.

A second NiV outbreak occurred in two districts of Kerala state in May 2019 — Kozhikode and Malappuram. The index case was hospitalized but not laboratory confirmed as a NiVI. However, an infection was confirmed by PCR of virus RNA in throat swabs, urine or blood samples from 18 patients who were in contact with the index case or other infected individuals. Of these, 16 patients died (89% CFR). Of the 18 laboratory-confirmed cases, 17 developed acute neurological or respiratory symptoms. One patient survived after treatment with ribavirin and supportive therapy. Around 2,600 contacts were monitored for 21 days after possible exposure to infected individuals. A serum survey of 155 health care workers and 124 household and community members found three positives with subclinical infections.

Pteropus fruit bats have been implicated as the source of Nipah virus outbreaks.
Pteropus fruit bats have been implicated as the source of Nipah virus outbreaks.
Source: Adobe Stock

Initially, the source of infection of the index cases was blamed on bats roosting in a well, but none of the bats were Pteropus fruit bats, which roost in trees. However, NiV sequences were later obtained from four humans and three Pteropus fruit bats from this outbreak. This phylogenetic analysis of the human strains demonstrated a 96% similarity to Bangladesh strains of the virus and a 99.7% to 100% similarity to strains from fruit bats there. The highest similarity was between human complete NiV sequences from Kerala and gene sequences from Pteropus species fruit bats (99.7%-100%), compared with NiV sequences reported from Malaysia, Cambodia and Bangladesh (85.14%-96.15%). These results indicate that Pteropus fruit bats were most likely the source for human infection in the 2019 Kerala outbreak.

Human cases can occur without direct evidence of bat involvement, however. On April 2, 2014, the National Epidemiology Center in the Philippines received a report of human deaths in two villages — Tinalon and Midtungok, located in the municipality of Senator Ninoy Aquino, province of Sultan Kudarat, island of Mindanao. The villages are approximately 15 km apart. An investigation identified additional human deaths and nonfatal infections and concurrent neurologic disease and sudden deaths in several horses. The horses were subsequently consumed by villagers. There were 17 individuals, 11 with acute encephalitis syndrome, five with an influenza-like illness (ILI) and one with meningitis. The CFR among those with acute encephalitis syndrome was 82%; no patient with ILI or meningitis died. Of acute encephalitis syndrome survivors, one had residual severe cognitive impairment, motor weakness and ataxia, and the other experienced persistent ophthalmoplegia. The median incubation period for case-patients with known exposure was 8 days. Three of the survivors had NiV antibodies. One serum sample and one of CSF sample was positive for a NiV RNA sequence with 99% identity with NiV isolates from Malaysia and 94% to 96% identity with NiV isolates from Bangladesh. Epidemiologic data suggest that the most common route of virus transmission to humans was direct exposure to infected horses, contact with contaminated body fluids during slaughtering of sick horses and/or consumption of undercooked meat from infected horses. However, for at least five cases, clinical and epidemiologic evidence suggest direct human-to-human virus transmission.


Before the human cases, on March 3 to May 11, 2014, 10 horses were infected in the two villages. Of these, two were found dead and all but one of the others showed neurologic signs. Among other domestic animals, 4 cats that had eaten horse meat eventually died within 5 days of their probable exposure date; three were found dead and the fourth also died after exhibiting terminal bleeding from the nose and/or mouth. A dog was also found dead after eating horsemeat. Blood was collected from surviving suspected domestic animals, including horses, cats, buffalo, dogs, pigs and goats. Only samples from dogs were serologically positive. Tests were done in the Geelong Laboratory in Australia, and positive results were only obtained for henipavirus agents, which includes NiV. It is presumed that the horses were exposed to the virus via contaminated feed, forage or water from infected Pteropus fruit bats. The humans, dogs and cats were infected from eating horsemeat, followed by human-to-human transmission.

A very recent paper reviewed 248 cases of NiV infection that occurred in Bangladesh in 2001 to 2019. One-third of these (82 cases) were found to be caused by person-to-person transmission. Nosocomial transmission was a significant cause of infection.

Clinical Syndrome

Following an incubation period of 5 to 14 days, 3 to 14 days of fever, vomiting and headache may occur, followed by drowsiness, disorientation, behavioral changes, spasms, uncoordinated gate, myoclonus and mental confusion that may progress to coma within 24 to 48 hours. Death may ensue, probably from brainstem involvement. About half of the patients with neurological signs may develop respiratory illness with cough during the early part of their infections. Myocarditis occurred in some cases in patients in the Kerala outbreak. Treatment is limited to supportive care. The effectiveness of ribavirin is uncertain. In survivors, relapses may occur months or years after recovery; up to an 8% relapse rate has been reported.

It is extremely important to use appropriate isolation techniques for patients hospitalized with NiVI because person-to-person spread can occur via blood, other fluids and the respiratory route.


A preliminary diagnosis can be made based on signs and symptoms of the disease. None is pathognomonic for NiVI. Because of the similarity of symptoms of NiPI to other virus diseases such as JE, laboratory confirmation is essential.

The CDC indicates that laboratory diagnosis can be made using a combination of tests. Virus isolation and demonstration of virus RNA segments with real-time PCR can be made from throat and nasal swab samples, CSF, urine, and blood early in the course of disease. Serological diagnosis can be made later by detection of IgG and IgM antibody by enzyme-linked immunosorbent assay. The presence of virus can be determined by immunohistochemistry in tissues collected at autopsy. Attempts at isolation of the virus should be limited to level 4 biocontainment facilities.


Prevention of infection is the most effective measure to avoid disease. The appearance of disease in pigs in areas where Pteropus fruit bats are found is a clue that NiV may be present. In those areas where palm sap is consumed, it is important to deny bat’s access to collection pots and cut bark. Boiling the sap for 10 minutes would inactivate the virus, but local consumers are resistant to this advice, as mentioned earlier. Avoidance of virus infection requires continuous public education — a challenge because human cases are sporadic and specific localities may change from year to year. Prevention of human-to-human transmission is difficult in households, particularly early in the course of disease, but frequent hand-washing can be helpful. Health care personnel dealing with suspected cases should use personal protective equipment to avoid becoming infected.

A vaccine is under development. A simian adenovirus-based vaccine encoding NiV glycoprotein (G) Bangladesh was tested in Syrian hamsters. Prime-only as well as prime-boost vaccination resulted in uniform protection against a lethal challenge with NiV Bangladesh, and all animals survived challenge. One hopes that the vaccine will be similarly safe and effective in primate tests and then human trials.


Is there a risk of spread of NiV virus to other countries? Only those countries where Pteropus fruit bats are found are at potential or actual risk of human infection spilling over from them. The virus is present in some of these countries but without reported human cases. In Cambodia in 2000, 1,072 bat serum samples from several species were tested for antibodies. Only Pteropus lylei fruit bats were positive, and six other bat species were negative. One NiV-like isolate was obtained from 769 urine samples collected at roosts of P. lylei specimens. A survey of virus strains in P. lylei, and seasonal occurrence of these viruses was completed in seven provinces of Central Thailand between May 2005 and June 2007. NiV RNA sequences, which belonged to those of the Malaysian and Bangladesh strains, were detected in bat urine, the Bangladesh strain being dominant.

But is there risk of NiVI in humans or domestic animals in the absence of Pteropus fruit bats? The single outbreak in the Philippines provided evidence that NiV is present there, but that Pteropus bats were not directly responsible for the infection in humans. There is clear evidence of person-to-person transmission. With an incubation period of 5 to 14 days, it is certainly possible for an infected individual to travel considerable distances before becoming ill. That happened to an individual infected with Ebola virus infection who traveled from Nigeria to Dallas, Texas. That kind of international spread would be more likely if the virus was first spread to a large city, then transmitted to many individuals there and subsequently onward internationally. Should that occur, it might take some time for an accurate diagnosis to be made in Europe or the Americas, but we need to be prepared.

Disclosures: Kaye and Yuill report no relevant financial disclosures.