As fall approaches, we continue to see discussion of a “second wave” of coronavirus infections, at least in more temperate northern regions. In the United States alone, we have seen regional or even relatively state specific, case surges at different times. Europe seems to be experiencing a similar phenomenon and even parts of Asia. Now it could be that differences in mitigation measures and timing of their implementation and lessening or removal have something to do with that, but at best it would seem to be a very limited explanation. It seems more likely that weather is playing a role and I have reported before on studies, primarily in regard to influenza, but some coronavirus ones, that look at a wide range of meteorological measures, such as temperature, humidity and amount and intensity of sunshine. The prior human coronaviruses have been highly “seasonal”, but most of those studies were done in the temperate northern climate, which for the US would be about the northern half to two-thirds. You would be inclined to think there will be a similar pattern with this strain, which would mean that we could see an increase in cases in the late fall/early winter at least in the northern half of the US.
Interestingly, when I eyeball the 2018/19 flu season in the US by region, there don’t appear to be dramatic differences, and flu is highly seasonal. So coronavirus is perhaps more susceptible to certain weather variables and will therefore be more variable within the United States? Back in mid-March there was an article speculating and modeling the potential role of seasonality in this epidemic. (Seasonality Study) The authors are from Switzerland and Sweden and use Swedish samples to examine seasonality of the pre-existing coronavirus strains from the years 2010 through 2019. The pattern is very strong and obvious, with a peak in December falling off to a low level in summer and picking back up in late October to November. The authors considered a variety of scenarios with lower or higher seasonality influence and lower or higher mixing of coronavirus infections from global travel. Their simulations with a strong seasonality component predicted a potential of a main peak in the spring, or a main peak in the coming winter or two similarly sized peaks, for northern temperate areas. If the seasonal component is weaker, only a single peak is projected. The authors also caution that other factors can be more important than any seasonal variation and may account for different epidemic shapes even in the same basic climate regime.
Interestingly, the study also recapped the history of some recent influenza epidemics and noted a variable seasonality pattern around the world. Flu is not as predictable as we might like to think. I continue to be left with a sense that we don’t have the necessary information to predict the future course of the epidemic. Everything I see suggests to me that transmission slows dramatically when somewhere around 20% of a population is infected, although we are just guessing about total infection rates. I suspect the shape of the epidemic in any region is an interplay between local socio-demographic factors, prevalence of pre-existing coronavirus immune responses, weather variables and behavioral changes. That is a complex mix, which would make model-derived forecasts inherently unreliable. So caution is warranted both in guessing what might happen and in shaping public policy.
I conjecture the Wu-flu would have been seasonal, too, except for the asinine “flatten the curve” madness which has morphed into “we can beat Covid if everyone will wear a mask an social distance.”
Let’s see how long we can prolong the case count while encouraging some more, deadlier mutations.
It seems flattening the curve makes the seasonal variations more confusing. On the 20% infection – there are different ways of measuring – if you’re measuring antibodies (produced by B-cells) there are different parts of the virus that can produce antibodies and tests use different measurements. There is also innate T-Cell immunity (T Killer & helper cells) that are not routinely measured. We don’t know how much of that there is in the population. It is my understanding that the 20% measures antibodies only, but I may be mistaken.