Skip to content


"What is not generally appreciated is that there is a potentially large benefit to masking even if infections are not prevented."

During the COVID-19 pandemic, wearing face masks became an important public health measure to help reduce disease transmission. A new Journal of the Royal Society Interface study models mask-induced “variolation”, whereby someone wearing a mask is likely to receive lower doses of the virus than someone who isn’t masking. This can in turn reduce the severity of infections. We spoke to Professor David Earn and Mr Zachary Levine at McMaster University to find out more. 

Can you explain the historic context of variolation?

Variolation was a precursor of vaccination. In the eighteenth century, it was common to variolate children for smallpox, which meant intentionally administering a small amount of smallpox virus taken from an infected person. Children who were variolated often experienced mild illness (much milder than if they had acquired smallpox naturally), and nevertheless were found to be immune to subsequent infection.

Unfortunately, variolation still entailed a substantial risk of full-blown disease and death. In contrast, with his smallpox vaccine, Edward Jenner succeeded in inducing immunity with very small risk of serious illness. Jenner’s approach was to use a related virus (cowpox), which causes mild illness in humans and yet yields immunity to the more serious smallpox virus. Modern vaccines use different approaches – dead or “attenuated” versions of the virus, or synthetically designed chemicals in the case of mRNA vaccines – that induce immunity with extremely small risk of serious illness.

Can you provide an overview of the background to your study?

In the summer of 2020, Ghandi and Rutherford hypothesized that “variolation” could be an unintended (positive) side-effect of face masking during the COVID-19 pandemic.  They argued that, compared with unmasked individuals, people who are infected while wearing a mask receive a smaller dose of virus, that smaller viral doses tend to yield infections with milder symptoms, and that mild infections should still provide long-lasting natural immunity to COVID-19.

We developed and analyzed a mathematical model that captures the mechanism proposed by Ghandi and Rutherford, and allows us to explore the effects of facemask-induced variolation on the patterns of spread of disease in the population as a whole.

What are the challenges and opportunities for public health messaging in relation to mask-wearing?

The “variolation hypothesis” suggested that masking could potentially be more powerful as a disease-control measure than had been realised.  During the pandemic, public health authorities have promoted the use of masks both to protect individuals from becoming infected and to prevent infected individuals from infecting others.  What is not generally appreciated is that there is a potentially large benefit to masking even if infections are not prevented.  If masking reduces the severity of infections that do occur, then it can have enormous benefits for individuals and for the healthcare system.  Getting this message across is challenging because people tend to be skeptical of the value of masks if they are aware of infections having occurred when masks were worn.

What are the future directions for this line of research?

We don’t actually know that the effect of variolation as proposed by Ghandi and Rutherford is large enough to make a big difference.  Experiments that quantify the magnitude of the effect of facemask-induced variolation in humans are needed.  If the relationships between size of dose, severity of illness, and degree of resulting immunity can be quantified, we can then build more detailed mathematical models that may help to evaluate the benefits of masking.

For more research that applies mathematical approaches to important biological questions, check out Journal of the Royal Society Interface. We also have a collection of our COVID research published across all our journals. 

Image credit: iStock image, Jull1491:

View all