Antibodies are the large, Y-shaped proteins our immune system produces to neutralize viruses and other pathogens. While testing for antibodies is less suited for diagnosing COVID-19 in individuals, as it might miss acute or old infections, it is extremely useful for population surveys.
Relative to the size of Iceland, this is the largest country survey to date. The study looked at a group of 30,576 people, almost 9% of the island’s population. The researchers used two highly specific assays to determine seroprevalence. The antibody tests were coupled with the available results of PCR (virus) testing. Thanks to its thoroughness, the study has revealed new insights.
First of all, it looked at the antibody evolution over a longer time period than previous studies. The total level of antibodies rapidly peaked, then fell off, but stabilized between 2 to 4 months at an intermediate plateau.
Broadly speaking, this seems to be the sum of 2 different kinetic mechanisms. One group of proteins forms in the early stages of the infection and shows a rapid decay. The second group takes much longer to build up. As a consequence, decay will occur later.
Recent observations of a rapid decay of antibodies had stirred up fears that immunity against SARS-CoV-2 would be short-lived. The findings here seem to be at odds with those observations, but are not necessarily. Earlier work had looked over a one-month period. It saw the rapid decay, but not the stable plateau still observed after 4 months.
The authors have calculated an infection fatality risk (IFR) of 0.3% (95% CI, 0.2%-0.6%). This value is in line with IFR’s found in studies concerning the US state of Indiana (0.26%, 95% CI: 0.21%-0.35%) and the German community of Heinsberg, fallen victim to an outbreak after carnival festivities (0.36%, 95% CI: 0.29%-0.45%. However, these values are noticeably lower than other studies, with values ranging from 0.5 – 1%.
All in all, two positive, encouraging conclusions: a longer immunity and a lower mortality rate. But they came along with a third finding, which could be a warning. The authors showed that 56% of all infections had also been found by PCR. Only 44% of infections went therefore unnoticed. Compared with estimates for other, larger countries this is very low. Not surprising though, given the large number of tests carried out in Iceland (231 tests per 1000 inhabitants on 10 August 2020). Reasoning along these lines, the authors have estimated that 0.9% of the population of Iceland has been infected with SARS-CoV-2. Again, this is lower than elsewhere: for Spain, for example, this number was estimated at 5%. Herd immunity is light years away. The Icelandic population remains very vulnerable to new waves of infection.