The recent surge in cases of sudden and severe hepatitis in children has been widely reported around the world. Recently, several media have highlighted a possible link between these cases and contact with pet dogs. However, the data suggesting such an approximation is extremely limited – in fact, probably much more limited than most of the other hypotheses that have been proposed.
The outbreak of hepatitis cases in children was first seen in the UK, but has now been reported in Europe, Asia and the Americas. Although the numbers are still very low globally, the disease is proving serious and some children have had to undergo liver transplants. At least 11 children have died, and it looks like the phenomenon is likely to continue for some time.
Hepatitis in humans is normally caused either by a toxic substance, such as alcohol, or by infection with one of several different types of virus. However, none of the usual viruses were detected in these children.
The UK Health Security Agency (UKHSA), the body responsible for protecting public health in the UK, works to find the cause of the disease so that it can be effectively controlled and treated.
In a recent briefing, the agency reported a high number of “dog exposures” in these cases of severe childhood hepatitis. However, before parents prevent their children from approaching the family dog, the results should be examined in detail.
The UKHSA found that 70% of patients (64 out of 92, where data was available) came from dog-owning families or had been “exposed to other dogs”. However, 33% of UK households own a dog, and many more children from non-owning households are exposed to them when visiting friends or playing with them. A 70% exposure to dogs can be completely normal.
To suggest a link, it is important to show not only that exposure to dogs in patients is high, but also that it is higher than in unaffected children. Until this has been verified in what is called a case-control study, any link is no more than a suggestion.
A second problematic point with these data is that by asking enough questions, there is a high probability that the answers to one or more of these questions will appear to be related to the cases.
When we retrospectively collect very large amounts of data, this type of spurious association can easily occur. There is also a website dedicated to collecting these statistics. Here’s an example: Maine’s divorce rate between 2000 and 2009 seems strongly related to per capita margarine consumption.
The point to remember about the links identified by historical data is that they are assumptions. They should always be verified by collecting additional information on new cases. If the link is real, it will continue to appear in new data. If it’s wrong, we won’t see it again.
One of the associations of the dummy correlations website reveals another important problem. Between 2000 and 2009, per capita cheese consumption in the United States may be linked to deaths from entanglement in bed sheets.
One can easily conceive that this could be the result of cheese-induced nightmares. The fact that we can think of a mechanism underlying the link reinforces our idea that it could be true, even if said mechanism is quite far-fetched. We tend to give more weight to associations for which we can envisage an explanation, even if the evidence is weak.
So what are the possible causes of the resurgence of hepatitis cases in children? Could any of these be related to dogs? One virus in particular, an adenovirus, was detected in the blood of 72% of patients tested (for comparison, SARS-CoV-2 was only detected in 18% of cases).
In cases where it was possible to identify the type of virus, it was human-like adenovirus serotype 41 (Ad41), which normally causes diarrhea in children. Although dogs can harbor their own adenoviruses that cause respiratory disease or hepatitis, these are not known to infect humans. Additionally, Ad41 has no known connection to dogs.
The cases seen in children do not suggest that the infection is transmitted from one individual to another – the number of cases is too low and the distribution too wide for that. Similarly, the distribution of cases does not suggest that this is a new virus transmitted from dogs to children. In other countries, cases have appeared much faster than a canine virus spreads between dogs.
Are there other possible causes? It has been suggested that the severity of hepatitis results from a malfunction of the immune system – either too strong or not strong enough. Social distancing during the pandemic has reduced transmission of a range of diseases, and lack of exposure to these conditions may have left some children unprepared for infections that would normally not be a problem.
Similarly, lack of exposure to dirt resulting from handwashing, sterilization of surfaces, and other hygiene measures may have predisposed children to overreactive immune responses (as has been suggested for allergic diseases ). Hepatitis could therefore be caused by the immune response rather than by a virus. Finally, and unsurprisingly, the possibility has been raised that previous Covid-19 infections have predisposed children to hepatitis.
All of these are only theories at this time, and there is insufficient data available to prioritize them or use them to propose control measures. Fortunately, the incidence remains extremely low, and until better data are available, parents should probably focus more on observing any symptoms in their children than on reducing their exposure to dogs.
The original version of this article was published on The Conversation, a nonprofit news site dedicated to sharing ideas between academic experts and the general public.