50% heritability for key COVID-19 symptoms

A popular UK symptom tracking app has revealed that half of the variation between people is genetic...
14 May 2020

Interview with 

Frances Williams, King's College London


Using an iPhone


The evidence is starting to come in on whether host genetics - in this case, the genetics of human hosts - changes the response to COVID-19. Here in the UK, Kings College London rolled out a popular COVID-19 Symptom Tracker app, and they’ve recently released their heritability findings: using twins to figure out how big a role genes play in key symptoms. Katie Haylor spoke to one of the team, Frances Williams, about what they’ve found...

Frances - We've done what's called a classical twin study in which you can compare symptoms among the twin pairs. Because we know that identical twins share a hundred percent of their genetic material, while non-identical twins share on average 50%, we can compare between the identicals and non-identicals and work out roughly what the contribution is of genetic factors.

Katie - And what symptoms specifically are you talking about?

Frances - Cough, fever, delirium, loss of taste and smell, shortness of breath, chest pain, abdominal pain, diarrhea, that sort of thing, all the commonist symptoms associated with many different viral illnesses. We could combine the symptoms in a mathematical model to determine the best combination of symptoms that predicts being positive on a test for the virus. So it's a combination of those symptoms, such as the loss of taste and smell and fever with persistent cough and fatigue, and age and sex are also in the model. And interestingly, that model has the highest heritability of all the symptoms that we looked at: 50%. Which means that roughly 50% of the difference in expression of those symptoms is accounted for by genetic factors.

Katie - What does that mean for an individual?

Frances - Well, it's difficult to extrapolate to an individual because heritability is all about the group differences in the population if you like. But I think what this is telling us is that while many people consider infections to be an entirely random event, in fact it's not entirely random. It is to some extent influenced by the individual, but we know that many symptoms of viral illnesses actually occur because of the host immune system reacting to the presence of that virus rather than the damage that the virus itself causes directly.

Katie - Do you think this could go any way to explain why some people seem to be relatively mildly affected whereas other people are severely affected?

Frances - We know that the people that end up with a major illness requiring hospitalization often developed high levels of inflammatory proteins in the blood. And it may well be that what we're seeing in our study is reflecting the fact that somebody's genetic makeup can impact how their immune system responds to viral infection and whether or not it responds by making very high levels of those inflammatory markers.

Katie - What's the value in understanding to what extent the symptoms or risks are heritable?

Frances - Firstly, it's a better understanding of how genetic variation influences people's susceptibility to the disease. So it might be possible if we developed, for example, a test which would advise people better about their risk of developing (FULL?) Covid-19. So rather than just have a blanket rule about by age or by sex, everybody has to stay indoors, you could make it more personalized. And the second big value I think for this app is being able to collect large volumes of data real time about how this infection is spreading across the country.


Add a comment