Scientists have backed proposals for Covid boosters in the autumn after blood tests on hundreds of people revealed that protective antibodies can wane substantially within weeks of second vaccine shots being given.

Falls in antibodies after vaccination are expected and do not necessarily mean people are more vulnerable to disease, but the researchers are concerned that if the declines persist the effectiveness of the vaccines may diminish.

The UCL Virus Watch study found that antibodies generated by two doses of the Oxford/AstraZeneca and Pfizer/BioNTech vaccines started to wane as early as six weeks after the second shot, in some cases falling more than 50% over 10 weeks.

The researchers stress that both vaccines are extremely effective against Covid, but say the findings support plans for a booster campaign this autumn, particularly for those who were vaccinated early and with the Oxford/AstraZeneca shot.

“We know levels of antibodies start high and drop substantially,” said Prof Rob Aldridge, an infectious disease epidemiologist at University College London. “We’re concerned that if they carry on dropping at the rate we’ve seen, the protective effects of the vaccines will start to drop too, and the big question is, when is that going to happen?”

Interim advice from the Joint Committee on Vaccination and Immunisation (JCVI) last month encouraged the NHS to prepare for an autumn booster programme, but a final decision on whether to proceed has not been made. It is unclear whether protection from vaccines has weakened enough to warrant boosters and many experts argue the doses are needed more urgently in other countries.

The UCL team analysed blood from 605 vaccinated people mostly in their 50s and 60s. They found that antibody levels varied widely between patients, but a double dose of Pfizer/BioNTech tended to produce far more antibodies against the coronavirus than two shots of the Oxford/AstraZeneca vaccine.

Three to six weeks after full vaccination with Pfizer, antibody levels typically stood at about 7,500 units per millilitre (ml), but more than halved to 3,320 units per ml after 10 weeks. For AstraZeneca, antibody levels peaked at about 1,200 units per ml and typically fell to 190 units per ml after 10 weeks. Since publishing the results in a letter to the Lancet, the researchers have seen the same trend in a further 4,500 participants in the study.

While antibody levels are important for protection, the immune system has other defences that are built up after infection or vaccination. It is normal for antibody levels to wane over time and for the immune system to “remember” the infection with memory B cells. Should the virus invade, these cells rapidly churn out antibodies targeted at the virus. Further protection comes from T cells, which destroy infected cells and limit the severity of disease.

“Antibodies are not the perfect measure of risk; we don’t know if there’s a magic number, as it were, where the risk of infection or hospitalisation becomes important,” said Aldridge. “But we think these data support the JCVI case for boosters, with priority for the clinically vulnerable, the over-70s, and all people living in residential care homes for older adults.”

The findings have been considered by the JCVI but are unlikely to have a major impact on discussions over boosters. A loss of antibodies is a warning sign that vaccines may wear off over time but does not say when that point comes. Public health authorities will not know for sure until people who had their second shots early in the rollout start turning up in hospitals. A decision on the booster programme is expected before that happens.