dcyphr | Self-reported symptoms of covid-19 including symptoms most predictive of SARS-CoV-2 infection, are heritable


Genetics might influence the chance for COVID-19 infection. 2633 volunteers from TwinsUK completed a survey. This allowed a twin study of COVID-19 symptoms.


There might be a heritable factor to infectious diseases. The likelihood to get the disease, an immune response, and the severity of the response might be partially genetic. Understanding how symptoms of COVID-19 pass through the population can help allocate hospital resources. The authors developed the "C-19 COVID Symptom Tracker app" to collect real time data during the COVID-19 pandemic. Ethics committee approval was obtained. The aim of the study was to estimate heritability of COVID-19 symptoms. 


The app asks the user daily about common COVID-19 symptoms. This includes cough, fever, chest pain, delirium, loss of smell. The app was downloaded by over 2 million people. Heritability of symptoms and 'Predicted COVID-19' was investigated. Their model considered genetic, common environmental and unique environmental factors. 'Predicted COVID-19' was determined based on a combination of age, gender and reported symptoms. The authors have shown in another study that his model is effective in predicting COVID-19 compared to swab tests. 

The authors got further data regarding the participants housing arrangement. The authors factored the likelihood of infection within households and compared this to twins not living together.


2633 adult same-sex twins provided data between 25 March and 3 April 2020. The sample had 728 pairs with 86.9%% being female. Prevalence of symptoms in TwinsUK was similar to a larger dataset of 1.85 million people.

The authors found that there may have been genetic influences for delirium, fever, fatigue, loss of smell, shortness of breath and diarrhoea. There was no influence for hoarse voice, cough, skipped meals, chest pain and abdominal pain which were based on the environment. 

Heritability estimates for twins that were living apart were similar.


The authors report that 50%% of the variance of 'Predicted COVID-19' is due to genetic reasons. Currently, 2.9%% of the population has 'Predicted COVID-19'. Symptoms related to immune response have a heritability over 35%%. These symptoms include fever, delirium and fatigue. Loss of smell, which is a good predictor of COVID-19, is 48%% heritable. 

Infection of COVID-19 is influenced by a range of genetic factors. This may be because of variation in the host's immune response. The findings can help identify therapies for COVID-19 and identify at risk groups.

This study has several strengths. TwinsUK and the symptom recording is representative of the UK population. Predicting infection based on comparing symptoms to a large training set is a practical solution for widespread testing. 

One limitation is the likelihood of healthy volunteer bias. Another limitation is that symptoms are non-specific and are prevalent during spring time due to allergies/flu season. However, the sample size was sufficiently large. The results could be biased by identical twins being likelier to live together than fraternal twins. This was accounted for bu excluding cohabiting pairs through the real time data collection. Finally, the sample is mostly female which do not represent the population fully.

The genetic influence on COVID-19 symptoms may relate to genes such as those for the ACE2 receptor. Further research is required to determine whether the ACE2R gene can predict symptoms. Identifying those with a higher risk to COVID-19 can help with policymaking.