In The Lancet, Zhenyu He and colleagues report their cross-sectional study of serological responses of more
than 9500 individuals from 3600 households in Wuhan, the early epicentre for the COVID-19 outbreak. The study was initiated shortly after lockdown in Wuhan ceased in April, 2020, with follow-up over two timepoints (June and October–December, 2020).

In this cross-sectional, longitudinal serosurvey of Wuhan residents, the authors estimated the penetration of the virus into the community who were selected using a population-stratified, random sampling approach. They also examined the development and durability of SARS-CoV-2 neutralising antibodies, the putative correlate of protection in COVID-19.

In the Wuhan population sampled by He and colleagues, 532 (5·6%) of 9542 participants were positive for pan-immunoglobulins against SARS-CoV-2 at baseline, giving an adjusted seroprevalence of 6·92% (higher than
an earlier report, which estimated a seroprevalence of 3·2–3·8%). More than 80% of those surveyed who were
seropositive were asymptomatic.

As a city of 11 million people with 9 million residents as estimated during lockdown, quick extrapolation of these estimates suggests that approximately 622800 individuals in Wuhan would have been infected as of April, 2020, far exceeding the cumulative number of confirmed cases in Wuhan (50333 as of April 17, 20207).

If the seroconversion rate is an accurate reflection of exposure to SARS-CoV-2, the apparent disparity between low case numbers and high seroconversion rate seems to suggest that most seroconverted individuals produced antibodies to SARS-CoV-2 after asymptomatic infection. Accordingly, by occupation group, the highest seroconversion rate was seen in health workers and in those with known COVID-19 contacts in the past 5 months.

He and colleagues detected neutralising antibodies in 39·8% of the seropositive subgroup; encouragingly, the
proportion was sustained over the 9-month study period, indicating that putative protection was durable when
it occurred. Moreover, titres of neutralising antibodies were lower in individuals who had asymptomatic infections, compared with those with symptomatic infection and confirmed cases.

The reason why those with asymptomatic disease developed lower neutralising antibody titres than those with symptomatic disease might be simply explained by different levels of exposure to SARS-CoV-2 antigens.