A new study sponsored by the CDC looked at the presence of antibodies to CV-19 in blood donor samples from late 2019 and early 2020. (CDC Paper) the researchers looked at samples collected from December 13, 2019, through January 17, 2020, in nine states. The first identified case in the US was January 19. The authors found a fairly high percent of the samples, 106 out of 7389, or 1.4%, had such antibodies and there was at least one positive in each of the nine states, which were geographically scattered and included Iowa and Wisconsin. Almost all the samples actually showed neutralizing antibodies. Cross-reactivity from seasonal coronavirus infections seems an unlikely explanation, given the confirmatory testing that was done, but if there was extensive cross-reactivity, it would suggest that a large portion of the population may have had pre-existing protection, which could affect the course of the epidemic. Blood donors aren’t a representative sample of the population, obviously tending to be healthier, younger and with more contacts, although many of these donors were middle-aged, so this may overstate the presence in the entire population, but it is still a lot of infections. If confirmed, this clearly indicates that not only was CV-19 in the US by December at the latest, and almost certainly earlier, but it was somewhat widespread. This is not surprising in light of recent studies concluding that the virus was in China and Europe as soon as early fall.
The implications are fascinating. Somehow CV-19 may have been in as much as 1% of the population and we didn’t notice it clinically. There was no big uptick in influenza-like cases, no surge of hospitalizations or LTC deaths. I have speculated on the course of the epidemic as far back as the spring. As I hypothesized then, it is possible that early spread occurred largely among the young and healthy, who were traveling a lot and had a lot of contacts, but didn’t get seriously ill, maybe thought they had a cold. Then it gradually spread to those who had contacts with nursing home residents and the elderly in late February and March and we began to see more severe disease and deaths. I don’t know how else to account for the fact that the virus may have been causing a large number of unnoticeable infections. And instead of front-loading, makes it seem like there is mid-loading in a wave. In other words, the early part of a surge is among high-contact individuals, who definitely tend to be younger and healthier; then it moves to the vulnerable populations and creates substantial damage in that relatively small proportion of the overall population, while continuing to spread among other groups, but transmission has naturally slowed, and we see the sharp declines. The wave duration in pretty interesting and looks a lot like an unmitigated epidemic curve, not a lot of evidence of flattening or truncation, although I suspect the height is lower. Coupled with the revelations that China definitely hid both what was happening and the extent of the epidemic, we can assume that there is more to learn about the exact case shape and geographic spread.