Previous Covid infection may not offer long-term protection, study finds


Previous infection with coronavirus does not necessarily protect against Covid in the longer term, especially when caused by new variants of concern, a study on healthcare workers suggests.

Researchers at Oxford University found marked differences in the immune responses of medical staff who contracted Covid, with some appearing far better equipped than others to combat the disease six months later.

Scientists on the study, conducted with the UK Coronavirus Immunology Consortium, said the findings reinforced the importance of everyone getting vaccinated regardless of whether they had been infected with the virus earlier in the pandemic.

“If you look at the trajectory of the immune response after infection, mostly it is still detectable six months later, but it’s highly variable between people,” said Eleanor Barnes, a professor of hepatology and experimental medicine at Oxford and a senior author on the study.

“That is quite different to vaccination. If you vaccinate you get a really robust response, but with natural infection there’s much more diversity in responses.”

The researchers analysed blood samples from 78 healthcare workers who had Covid, with or without symptoms, between April and June last year. The blood was checked monthly for up to six months post-infection for a range of immune responses. These included different types of antibody that target the virus, B cells that make antibodies and retain a memory of the disease, and T cells, which reduce the severity of disease by killing off infected cells.

Writing in a preprint, which has yet to be peer-reviewed, the authors describe how they used a machine learning system called Simon, for Sequential Iterative Modeling Over Night, to see whether a person’s early immune response and the severity of their infection could predict their longer-term immunity. Dr Adriana Tomic, the first author on the study, said a signature in the antibody and T-cell response at one month predicted how robust the antibody response would be at six months.

The majority of people who produced a weak immune response at one month had no detectable antibodies that could neutralise the Alpha variant, first seen in Kent, at six months. None mounted neutralising antibodies against the Beta variant first spotted in South Africa. The researchers have yet to analyse data for the Delta variant now dominant in the UK.

While most of the healthcare workers who developed symptomatic disease had a measurable immune response six months later, more than a quarter did not. More than 90% of those who had asymptomatic infections had no measurable immune response six months later, the researchers found. The work is part of the protective immunity from T cells to Covid-19 in health workers (Pitch) study, funded by the Department of Health.

“In our view, previous infection does not necessarily protect you long-term from Sars-Cov-2, particularly variants of concern,” said Barnes. “You shouldn’t depend on it to protect you from subsequent disease, you should be vaccinated.”

The wide variability in immunity triggered by natural infection in part reflects the radically different exposures people can have to the virus while going about their lives. Immunity from vaccination is more reliable because people are given a standard dose in a standard way.

Danny Altmann, a professor of immunology at Imperial College London, who was not involved in the study, said the findings cautioned against simple assumptions around how immunity waned with time. “People show rather diverse trajectories after infection, but immunity often seems to hold up well at six months,” he said. “Most of all, studies such as this remind us that policy decisions on ‘boosting’ need to be evidence based in the context of a strong programme of immune monitoring.”