Hand, Foot & Mouth Disease: Coxsackievirus Explained

Hand, foot, and mouth disease (HFMD) sounds almost whimsical, and for most children it is a miserable week and nothing more. Fever, painful mouth sores, and a spotty rash on the palms and soles arrive together, then fade. Yet the same disease name covers a spread of viruses with very different personalities. Coxsackievirus A16 usually behaves. Enterovirus A71 occasionally attacks the brainstem, and in China it accounts for most of the children who die. This article explains what causes HFMD, how it moves between children, and why one virus is so much more dangerous than another.

Educational content, not medical advice. Talk to your doctor before changing any medication, treatment, or routine. Information here reflects current research but every person's situation differs.

What causes hand, foot, and mouth disease?

HFMD is caused by enteroviruses, a genus in the Picornaviridae family. Enteroviruses are small RNA viruses that have no outer fatty envelope. That detail matters in practice, and it is explained further below. For the general mechanics of how a virus enters a cell and copies itself, see how viruses work from infection to replication.

The genus is a large one. It includes poliovirus, the echoviruses, and the many coxsackieviruses, and most of its members cause nothing worse than a cold or a stomach upset. The respiratory relatives are covered in rhinovirus and the common cold, and the one with a reputation for paralysis is described in how poliovirus works. HFMD sits between these extremes: usually trivial, occasionally neurological.

Several serotypes (strains that the immune system recognizes as distinct) can cause the same clinical picture:

• Coxsackievirus A16 (CVA16), long the classic cause of mild HFMD outbreaks.
• Enterovirus A71 (EV-A71), the serotype linked to most severe neurological disease.
• Coxsackievirus A6 (CVA6) and A10 (CVA10), along with assorted other enteroviruses.

Think of HFMD as a diagnosis by costume. The rash, the ulcers, and the fever are the costume, and several different viruses can wear it. Only a laboratory test can say which one is underneath, and that is the question that determines how risky the illness is.

Why the serotype matters more than the diagnosis

A national surveillance study in China, covering 2008 to 2015, shows how different the serotypes are in practice. Among laboratory-confirmed cases, the share attributable to each virus changed sharply depending on how severe the illness was.

Group of confirmed cases (China, 2008 to 2015)	EV-A71	CV-A16	Other enteroviruses
All cases	43.6%	24.9%	31.6%
Mild cases	40.0%	27.1%	32.9%
Severe cases	73.8%	5.7%	20.5%
Fatal cases	92.5%	1.9%	5.6%

Source: Yang et al., Euro Surveillance, 2017.

EV-A71 was less than half of mild cases but more than nine in ten of the deaths. CV-A16, the virus many people associate with HFMD, was almost absent from fatal cases. This asymmetry is the central fact of the whole topic.

How HFMD spreads between children

The viruses leave the body by several routes at once, which is why HFMD travels so efficiently through households and childcare settings. The US Centers for Disease Control and Prevention (CDC) lists four: respiratory droplets from sneezing, coughing, or talking; contact with contaminated objects and surfaces; fluid from blisters; and feces.

The fecal-oral route deserves particular attention. Enteroviruses multiply in the gut as well as the throat, so stool can carry virus long after the child looks well. A parent changing a diaper, a toddler putting a toy in their mouth, and a shared cup can all complete the loop.

Respiratory spread works the way it does for a cold. A child who coughs near a sibling sprays virus-laden droplets that land on a mucous membrane or a surface that someone else then touches.

Blister fluid is the most distinctive route. The lesions on the hands and feet contain virus, so a child who pops a blister and then rubs their eyes or sticks a hand in their mouth has delivered the virus to a vulnerable surface.

Why daycares are amplifiers

CDC notes that HFMD is most common in infants and children under 5, though anyone can catch it. Young children have usually met few enteroviruses, so they lack antibodies to most serotypes. They also put objects in their mouths, share toys, and cannot be relied on to wash their hands.

The surveillance data in China show how sharply the burden falls on the youngest children. The median age across all cases was 2.1 years, and the median age among fatal cases was 1.6 years. Incidence peaked at age 1. It was 31.9 per 1,000 person-years in children aged 6 to 11 months, compared with 2.6 per 1,000 in children under 5 months, which fits the idea that antibodies from the mother protect newborns for a few months before fading.

Why adults usually escape

Adults can get HFMD, but they usually do not. The most likely explanation is accumulated immunity: by adulthood, most people have been infected with several of the common serotypes and carry antibodies that neutralize them. Immunity is serotype-specific, though, which is why a child who has recovered from one virus can get HFMD again from another. The general principles are covered in how the human immune system fights viruses.

An analogy helps here. Each serotype is a different brand of lock. An antibody is a key cut for one brand, so a child who has had CVA16 has a key for CVA16 and nothing for EV-A71. Childhood is the long process of collecting keys.

What HFMD looks like

CDC describes the typical picture as fever, sore throat, painful blistering mouth sores, and a rash, usually on the hands and feet. Most of the details below are standard clinical descriptions rather than numbers drawn from a single study, so treat the timing as approximate.

Fever and malaise often come first. A child may be fussy, go off their food, and complain of a sore throat before any spots appear.

Mouth sores follow. They appear as small red spots that blister and break into shallow ulcers, on the tongue, gums, and inside the cheeks. They hurt, and the pain is the main reason young children stop eating and drinking.

Skin lesions appear on the palms and soles, and they can also show up on the buttocks and elsewhere. They are usually flat or slightly raised red spots, some with a small blister. They are often less painful than the mouth sores.

Because the pattern of hands, feet, and mouth is so characteristic, clinicians usually diagnose HFMD by looking. Laboratory confirmation by PCR (a test that detects viral genetic material) of throat swabs or blister fluid is reserved for outbreaks, severe cases, and surveillance. In their 2010 review in The Lancet Neurology, Ooi and colleagues single out PCR of throat swabs and vesicle fluid as among the most efficient tests for EV-A71.

Conditions that can look similar

Herpangina, another enteroviral illness, causes ulcers confined to the back of the throat without the skin lesions. Herpes simplex gingivostomatitis can cause severe mouth sores. Chickenpox produces a rash with a different distribution and a pattern of lesions at different stages. These distinctions are for a clinician to make, not a parent. A child with an unexplained rash and fever deserves a medical assessment.

The usual course: a week or so of misery

CDC says most people recover on their own in 7 to 10 days, and that treatment is usually not needed. The main risk in an uncomplicated case is dehydration, because mouth sores make swallowing painful.

The care is supportive. CDC's home-care advice is over-the-counter medicine for fever and pain (never aspirin for children) and plenty of fluids. This article does not recommend specific doses or products, and parents should follow the advice of their child's doctor.

Nails that fall off

Some children lose a fingernail or toenail a few weeks after HFMD. The medical term is onychomadesis, meaning the nail separates from its base and sheds. It became widely discussed after a 2008 outbreak in Finland, where researchers reported nail shedding as a common feature of HFMD caused by coxsackievirus A6. In one figure from that report, a 10-year-old boy lost a nail 2 months after infection (Österback and colleagues, Emerging Infectious Diseases, 2009).

The leading explanation is that the infection briefly disrupts the nail matrix, the tissue that makes the nail. A nail that was being made during the illness is therefore weakly attached, and it falls off when the healthy nail grows out beneath it. The shedding looks dramatic but is not considered dangerous.

Coxsackievirus A6: the strain that changed the picture

For years, textbook HFMD meant CVA16 in the West and EV-A71 in Asia. That tidy split no longer holds. Before the 2008 outbreak in Finland, CVA6 had been seldom detected and mostly in association with herpangina; in Finland it was identified on only four occasions from 2000 to 2007. In the outbreak itself, 31 of 47 acute-phase specimens from 43 patients tested positive for CVA6, and the virus was also found in a shed nail fragment. After 2008, CVA6 appeared repeatedly as a cause of HFMD epidemics, and the Finnish authors suggested that it may be emerging as a new and major cause.

Surveillance elsewhere has followed a similar pattern. Regional studies in China have reported that EV-A71 and CV-A16 have in places given way to CV-A6 and other enteroviruses as the leading serotypes. This matters for two reasons. First, any vaccine aimed at EV-A71 alone cannot protect against CVA6. Second, the dominant virus in a given city can change within a few years, so a pattern that held in one decade may not hold in the next.

EV-A71: when the same illness becomes dangerous

Most EV-A71 infections are still mild. What sets the virus apart is a small fraction of cases in which it reaches the nervous system, and what happens then can be fast and severe.

How EV-A71 gets into cells

A virus enters a cell by attaching to a molecule on the surface, the way a key goes into a lock. For EV-A71, reviews describe two main receptors. SCARB2 is regarded as the primary receptor that recognizes all EV-A71 strains and helps with entry and uncoating, the step in which the virus sheds its protein shell and releases its genome. PSGL-1, a molecule found on white blood cells, is commonly described as a second receptor.

The picture is still being revised. A 2024 study in PLoS Pathogens by Nishimura and colleagues reported that SCARB2 was not detectable on the surface of susceptible cell lines and sat instead in late endosomes and lysosomes, which suggests another molecule, such as PSGL-1 in T cells, handles the first attachment while SCARB2 triggers uncoating later. Research also suggests that receptor binding is only the first step in determining which tissues a virus infects, and that viral and host factors both shape the outcome. Severe disease, including neurogenic pulmonary edema, appears to depend partly on the host's own immune response. In plain terms, scientists know that EV-A71 can reach the nervous system but have not worked out precisely why it does so in only some children.

The neurological complications

Ooi and colleagues describe the range of EV-A71 disease as including HFMD, aseptic meningitis, polio-like flaccid paralysis, brainstem encephalitis, and severe systemic complications, especially pulmonary edema and cardiorespiratory collapse.

• Aseptic meningitis means inflammation of the membranes around the brain without a bacterial cause. It produces headache, neck stiffness, and sensitivity to light, and many children recover fully.
• Brainstem encephalitis is inflammation of the part of the brain that controls breathing, heart rate, and consciousness. On MRI, the changes tend to appear in the dorsal pons and medulla, and sometimes in the spinal cord.
• Neurogenic pulmonary edema is fluid in the lungs that results from brainstem injury rather than from a failing heart or a lung infection.

The last of these explains why the disease can turn so quickly. Imagine the brainstem as the control room of a power plant. When it is inflamed, it can send out erratic signals that push the heart and blood vessels into overdrive, and fluid then leaks into the lungs. The lungs themselves were not infected. They were the casualties of a failing controller.

Clinical markers of severe disease reported in the literature include high temperature and lethargy. Parents should know that an HFMD patient who becomes unusually drowsy, floppy, or difficult to wake needs urgent medical attention.

What the Taiwan epidemic revealed

EV-A71 itself is not new. It was first isolated in California in 1969. Lin and colleagues, writing in Emerging Infectious Diseases in 2003, list earlier outbreaks with deaths: Bulgaria in 1975 (44 deaths), Hungary in 1978 (45 deaths), and Malaysia in 1997 (30 deaths). Then came Taiwan in 1998.

Year (Taiwan)	What was reported
1998	129,106 cases of HFMD or herpangina reported; 405 severe EV71 cases; 78 deaths
1999	35 severe cases
2000	291 severe cases; 25 EV71 deaths
2001	389 severe cases; 26 EV71 deaths

Source: Lin et al., Emerging Infectious Diseases, 2003.

In the 1998 epidemic, 91% of the fatal cases were in children under 5. Of 96 severe cases reviewed, 67 (70%) had encephalitis. Survivors of severe disease often had lasting problems such as central hypoventilation (weak automatic breathing), cranial nerve palsies, and weakness or wasting in the shoulder.

Two features of the Taiwanese data stand out. First, the authors cautioned that the reported figures may be only the tip of the iceberg, because few laboratories could diagnose EV71. Second, serology suggested that EV71 had been circulating in Taiwan for at least 16 years before 1998. A virus does not have to be new to produce a catastrophe. A population of susceptible children can build up quietly until an outbreak finds them.

Lin and colleagues also noted that most 1998 strains were genotype C, while genotype B was the most common in 1999 and 2000, and that genotype did not seem to affect clinical outcome. This is part of why the severity is not easily explained by a single "bad strain."

The 2008 outbreak in China

In early 2008, children in Fuyang City, in Anhui Province, began falling ill. The World Health Organization's outbreak report, with data as of May 5, 2008, listed 4,496 cases of HFMD due to EV71 in Fuyang, including 22 deaths. WHO noted that 1,391 of those cases were receiving hospital care and that no additional deaths had occurred for five consecutive days. HFMD became a notifiable disease in China on May 2, 2008, so case counts were expected to rise as reporting began.

Fatality figures from the early days of an outbreak tend to overstate the risk. The earliest recognized cases are the sickest, and as awareness grows, many more mild cases are counted, so the apparent fatality rate drops.

What the national data showed afterward

Once HFMD was notifiable, China had a surveillance system that could count cases across the country. Yang and colleagues analyzed the first eight years of data.

Measure (China, 2008 to 2015)	Figure
Notified cases	13,686,985
Severe cases	123,261 (0.90%)
Fatal cases	3,322 (0.02%)
Estimated overall case-fatality risk	0.024%
Estimated EV-A71 case-fatality risk	0.055%
Highest annual incidence	about 1,980 per million person-years (2014)
Male-to-female ratio	1.6 to 1

Source: Yang et al., Euro Surveillance, 2017.

The numbers carry two messages. HFMD is overwhelmingly mild: more than 99% of notified cases were not classed as severe. But with 13.7 million cases, even a tiny fatality rate produces thousands of deaths. A disease can be safe for the individual child and still be a public health problem for a country.

The vaccines that exist, and the ones that do not

China began licensing inactivated EV-A71 vaccines in 2015.

These are inactivated vaccines, meaning they contain killed whole virus that can train the immune system without causing infection. For the underlying biology, see how vaccines work. In the two phase 3 trials below, children received two intramuscular doses about four weeks apart.

What the trials found

Two large trials were published in the New England Journal of Medicine in 2014.

Trial	Participants	Result
Zhu et al.	10,007 children aged 6 to 35 months	94.8% efficacy against EV71-associated HFMD or herpangina over 12 months; 100% against hospitalization and against neurologic complications
Li et al. (Guangxi)	12,000 children aged 6 to 71 months	97.4% efficacy against EV71-associated HFMD over two epidemic seasons

In the Zhu trial, the 100% figures for hospitalization and neurologic complications rest on small counts (0 versus 24 hospitalizations and 0 versus 8 neurologic cases in vaccine versus placebo groups), so they are less certain than the headline number. Serious adverse events were similar between the two groups. In the Li trial, all vaccinated children whose antibodies were measured seroconverted, and fever and other adverse events were more common in the week after vaccination than with placebo.

Limits of the vaccines

The limit of a single-target vaccine is built into its design. The Chinese surveillance paper notes that the licensed vaccines offered no cross-protection against CV-A16 or other serotypes. Because EV-A71 caused 92.5% of fatal cases, the vaccine addresses the deadliest part of the problem, but it will not stop HFMD overall. The same paper noted that a CV-A16 vaccine and a combined EV-A71 and CV-A16 vaccine were still under development.

In the United States, CDC states plainly that there is no vaccine in the United States to protect against the viruses that cause HFMD. The decision about whether any vaccine should be recommended in a particular country rests with that country's health authorities.

CVA16 and EV-A71 side by side

Feature	Coxsackievirus A16	Enterovirus A71
Typical illness	Mild HFMD	Mild HFMD in most children
Share of fatal cases in China, 2008 to 2015	1.9%	92.5%
Share of severe cases in China, 2008 to 2015	5.7%	73.8%
Neurological disease	Rare	Aseptic meningitis, brainstem encephalitis, paralysis
Pulmonary edema from brainstem injury	Not a recognized pattern	Described in severe cases
Licensed vaccine	None	Inactivated vaccines licensed in China
Approved antiviral	None	None

Why outbreaks come in waves

HFMD does not arrive at a steady rate. Taiwan's 1998 epidemic had two waves, a large one that peaked in the week of June 7 and a smaller one in the week of October 4, mainly in southern Taiwan. After that, severe cases dropped sharply in 1999 and returned in 2000 and 2001. China's national data show the same unevenness: severe illness and mortality rates were highest in 2010, while the highest incidence of notified cases came in 2014, at about 1,980 per million person-years.

Lin and colleagues offered two broad explanations for the 1998 Taiwan outbreak. The virus may have mutated toward higher virulence, or host factors may have been responsible: an accumulation of susceptible people, or individual genetic susceptibility. The accumulation of susceptible children is the easiest to picture, as dry brush in a forest. Each year of low transmission leaves more unexposed children, and eventually any spark finds enough fuel.

What this means for surveillance

Because a single serotype can dominate for several years and then give way to another, counting cases without identifying the virus tells public health officials very little. China's decision to make HFMD notifiable on May 2, 2008, together with laboratory typing of a sample of cases, produced the dataset that allowed researchers to see that EV-A71 was behind most of the deaths while CV-A16 and other viruses made up much of the mild illness. The same system showed that the mix of viruses kept shifting from year to year. In the 2008 to 2015 data, EV-A71 accounted for between 88.5% and 95.4% of fatal cases in every year, but its share of severe cases ranged from 50.7% in 2015 to 82.3% in 2011.

That range is a reminder that a vaccine program has to be judged against moving targets. A drop in EV-A71 severe disease after vaccination would be good news, but it would also leave room for other serotypes to fill the gap, and only continued typing of cases can show whether that is happening.

How HFMD is treated

No antiviral drug is established for HFMD. Care is supportive. Ooi and colleagues note that intravenous immunoglobulin (IVIG, a preparation of antibodies pooled from donors) seems to be beneficial in severe EV-A71 disease, possibly through general anti-inflammatory effects, but that it has not been tested in formal trials. They also say milrinone, a heart medication, might help patients with cardiac dysfunction,. These are decisions for hospital specialists, and they apply only to the rare severe cases.

For the ordinary case, the aim is comfort and fluids. CDC advises over-the-counter medicine for fever and pain (never aspirin for children), and fluids to prevent dehydration. Warning signs that call for urgent medical assessment include inability to drink, marked drowsiness, neck stiffness, abnormal movements, or labored breathing.

Hygiene, schools, and the problem with sanitizer

CDC's prevention advice is straightforward: wash hands frequently with soap and water for at least 20 seconds, clean and disinfect shared surfaces and items, avoid touching the face, and avoid close contact with infected people. Children can return to school once they have no fever, feel well enough to take part, and have no uncontrolled drooling with mouth sores.

The emphasis on soap and water over hand sanitizer has a reason. Many enteroviruses lack a lipid envelope, and alcohol works best against viruses that have one, because it dissolves that coat. Without an envelope, a virus is harder to knock out. Washing physically lifts it off the skin, which is why handwashing is the foundation of prevention.

Why shedding outlasts symptoms

A child can stop feeling ill while still carrying virus in the gut, so the stool can remain a source of infection after the child returns to normal activity. Good hand hygiene after diaper changes and toileting remains sensible for days after recovery. The same logic applies to other gut viruses, as in norovirus versus rotavirus.

What HFMD says about viruses in general

HFMD is a useful case study in several broader ideas. First, an illness name is not a pathogen name, and the specific virus determines the stakes. Second, a virus that is mild in most hosts can still be dangerous in a small proportion of them, and when millions are infected the small proportion becomes a large number. Third, a vaccine aimed at one serotype leaves the others untouched, and the dominant serotype can shift over time, as CVA6 did.

The pattern repeats with other enteroviruses. Enterovirus D68, for instance, usually causes respiratory illness but has been linked to a paralytic condition in some children. The question of how viruses move between hosts and change their behavior is explored in how viruses jump from animals to humans.

Frequently asked questions

Q: Can adults get hand, foot, and mouth disease? A: Yes. CDC says HFMD is most common in children under 5 but anyone can get it. Adults are less often affected, probably because they have met many enteroviruses before and carry antibodies, but immunity is specific to each serotype, so a new virus can still cause illness.

Q: Is HFMD the same as foot-and-mouth disease in cattle? A: No. They are unrelated illnesses caused by different viruses. Human HFMD is caused by enteroviruses such as coxsackievirus A16 and EV-A71, while foot-and-mouth disease in livestock is caused by a virus from a different genus. The two diseases do not spread to each other's hosts in any ordinary way, and pets and livestock are not a source of human HFMD.

Q: How long does HFMD last? A: CDC says most people recover on their own in 7 to 10 days. Fever and mouth pain are usually the hardest part. Nail shedding can follow weeks later, as in the Finnish CVA6 outbreak, where one child lost a nail about 2 months after infection.

Q: Why is EV-A71 more dangerous than coxsackievirus A16? A: EV-A71 more often reaches the nervous system, especially the brainstem. In China from 2008 to 2015, it caused 92.5% of laboratory-confirmed fatal cases, compared with 1.9% for CV-A16. Researchers have identified receptors such as SCARB2 and PSGL-1, but the full explanation is not settled and involves the host's immune response.

Q: Is there a vaccine for hand, foot, and mouth disease? A: Not in the United States, according to CDC. China has licensed inactivated EV-A71 vaccines since 2015, with phase 3 trial efficacy of 94.8% and 97.4% against EV71-associated disease. They do not protect against CV-A16 or other serotypes, so they prevent only a portion of HFMD.

Q: Why does washing with soap work better than hand sanitizer? A: Many enteroviruses lack an outer lipid envelope, and alcohol is most effective against enveloped viruses. Soap and running water physically remove virus particles from the skin. CDC recommends washing with soap and water for at least 20 seconds.

Q: When should a parent seek urgent care? A: Signs of dehydration, unusual drowsiness, neck stiffness, abnormal movements, or trouble breathing call for prompt medical assessment, since these can signal neurological involvement. Severe EV-A71 illness can worsen within a day or two, which is why a child who seems to be getting sicker rather than better should be examined.

The takeaway

The key point about hand, foot, and mouth disease is that the name describes symptoms rather than a single germ. Most cases are a short, unpleasant illness that clears within 7 to 10 days. A minority, driven largely by EV-A71, involve the brainstem, and in a country with 13.7 million notified cases over eight years, that minority amounts to thousands of deaths. The vaccines that now exist in China are a real achievement, yet they cover one virus among several, and the rise of CVA6 shows that the dominant strain can change. Handwashing, rather than any clever technology, remains the most reliable tool.

Sources

• Centers for Disease Control and Prevention. About Hand, Foot, and Mouth Disease.
• Yang B, et al. (2017). Epidemiology of hand, foot and mouth disease in China, 2008 to 2015 prior to the introduction of EV-A71 vaccine. Euro Surveillance 22(50).
• Lin TY, Twu SJ, Ho MS, Chang LY, Lee CY (2003). Enterovirus 71 Outbreaks, Taiwan: Occurrence and Recognition. Emerging Infectious Diseases 9(3):291-293.
• World Health Organization (May 7, 2008). Disease Outbreak News: Hand, foot and mouth disease in China.
• Ooi MH, et al. (2010). Clinical features, diagnosis, and management of enterovirus 71. The Lancet Neurology 9(11):1097-1105.
• Zhu F, et al. (2014). Efficacy, safety, and immunology of an inactivated alum-adjuvant enterovirus 71 vaccine in children in China. New England Journal of Medicine 370:818-828.
• Li R, et al. (2014). An inactivated enterovirus 71 vaccine in healthy children. New England Journal of Medicine 370:829-837.
• Österback R, et al. (2009). Coxsackievirus A6 and Hand, Foot, and Mouth Disease, Finland. Emerging Infectious Diseases 15(9).
• Nishimura Y, et al. (2024). Enterovirus A71 does not meet the uncoating receptor SCARB2 at the cell surface. PLoS Pathogens.

---

Read next:

• Adenovirus Explained: The Hardy Virus Behind Many Diseases
• BK Virus, JC Virus & Polyomaviruses Explained
• The Immune System: How It Works Explained