Predicting the seasonality of Covid

(also on Spotify)

      • Seasonality of respiratory illness is not driven mainly by changes in the survival rate of the virus outside the body.  If it were, respiratory illness would be rarer in the Tropics than in the summer in temperate regions.
      • We need an explanation of seasonality that fits all the facts.
      • Seasonality seems to be a side-effect of the natural temperature-sensitivity of virtually all respiratory viruses
      • CoV-2 may be less temperature-sensitive than typical respiratory viruses, therefore less seasonal
      • Experiments are needed!

This has become a very important issue in the last few weeks because Governments and health services want to know what to expect from Covid-19 over the summer, and whether the virus is likely to return in the autumn.

I’ve heard so much misleading and wrong information from scientists and commentators in the last few weeks that I’m not sure where to begin.  For example, I heard an interview with an expert yesterday who said that viruses have very varied seasonality, with some such as flu being much more common in winter than summer, while others such as polio are most common in summer.  This is true, but it’s ignoring a remarkable – and clearly very important – fact:  every respiratory illness, with one minor exception, is more common in winter than summer – often much more common.  Polio is not a respiratory virus, it’s spread by fecal contamination.  (The one exception is that Edgar Hope-Simpson, an English doctor who focused on viral diseases, reported in 1981 that parainfluenza type 3, which is not a very common virus, was more active in the warmer six months of the year**.  By the way, I’ll give references in this note by referring to the list at the end of my 2016 review.  For example Hope-Simpson’s paper is [13] in that publication).

This expert that I heard yesterday also said that “temperature and humidity certainly correlate with influenza”.  What he’s referring to is the observation that influenza is more common in e.g. the United States when the weather is cold and dry [26].  This agrees with some lab studies, which found that influenza virus particles were more stable in cold dry air [27].  Another study, however, found the opposite effect [28].  In any case, it’s only part of the story.  Although we take it for granted that in e.g. the UK colds and flu become much rarer in the summer, many of us are less aware that influenza – and almost all other common respiratory illnesses – are present all-the-year-round in the tropics at intermediate levels.  For example influenza (and colds in general) becomes much more widespread in Fortaleza, Northern Brazil, in the rainy season [1] – the exact opposite what you would expect from the explanation given by this expert and many others.  Moreover, some other viruses (including the most common “common cold” virus, rhinovirus) have been found to be less stable in dry conditions than in humid [4].

This inconsistency and ambiguity is crying out for an explanation.  And there are other things that we need to explain:

    1. When viruses such as Covid jump between species they tend to be extra-virulent, in spite of not yet being well-adapted to their new hosts.  This implies that, over time, natural selection moderates the virulence of viruses.
    2. Levels of colds and flu usually jump dramatically during and after cold-weather snaps [5, 14, 21, 22].
    3. Many respiratory viruses (including some strains of coronavirus) are easier to isolate and grow better in the lab at around 33°C than at body temperature [74, 76-80].
    4. Respiratory viruses in tissue cultures where conditions favor less viral activity tend to become temperature-sensitive spontaneously (i.e. they become less active at higher temperatures) [58, 69-72], while those grown in conditions favoring rapid growth tend to lose temperature-sensitivity spontaneously [59, 74].
    5. Several respiratory viruses have been found to possess biochemical “switches” that are operated by changes in temperature [60, 65, 68].  For example many strains of influenza have been shown to switch from generating their own proteins to generating new copies of their genetic material when the temperature fell from 39°C to 33°C [61-64, 66, 67].

I suggested in a review in Medical Hypotheses (2016) [reference and link below] that all these observations can be explained by a simple idea: that most respiratory viruses use temperature to moderate their activity, meaning that they are most active below normal body temperature.  This gives the virus two important advantages: firstly, since the nose and throat are some of the coldest parts of the body, it is less likely to spread down into the warmer lungs, and then to the heart, brain etc.  This means that the host keeps moving around, thus spreading the virus.  Secondly, the virus has a built-in transmission mechanism: by staying in the nose and throat, and irritating the cells there, it causes coughing, sneezing, and runny noses.  All three of these symptoms can help it to spread to other hosts.

I wrote an explanation of this for the layperson and explained some of the problems with other explanations of seasonality a few years ago in this blog: see

Every winter, colds and flu increase

What can we conclude about Covid?  Since it is caused by a respiratory virus, we can guess that it has at least some temperature-sensitivity (with more viral activity below normal body temperature).  We can also see that it quite often infects only the nose and throat, which also suggests temperature-sensitivity.  On the other hand it may not have been a respiratory virus for very long, and it often infects the lungs as well or instead of the throat.  This suggests less temperature-sensitivity than is typical for respiratory viruses.  So we might expect Covid to become a bit rarer in the summer, but probably not to disappear almost completely like influenza and other illnesses, and to make a comeback in the fall when the temperature drops.

If my suggestion that viral temperature-sensitivity drives seasonality is correct, it follows that viruses adapt (often in a few months) to their local environment, because we see almost the same set of common respiratory viruses all over the world.  We know that influenza strains move quite freely around the world~~, and the same is probably true of other viruses.  One conclusion is that it’s essential that we don’t allow virulent strains – which might arise within one patient^^ – to be released from institutions such as hospitals into the general population.  See the link below about the Trade-Off Model for more information.

What we need here are some simple lab experiments – if any virologist can be persuaded to look at these issues.  For example, we can look at the effect of growing the virus in the lab at different temperatures, and of temperature changes, both upwards (e.g. 33°C to 37°C) and downwards (e.g. 37°C to 33°C).  This should eventually include looking at all the steps in the life-cycle of the virus, including binding to the cell surface, entering the cell, moving around the cell, making protein, making genetic material, assembly, moving to the surface of the cell, and leaving the cell.

So we don’t yet now whether Covid-19 will show strong seasonality, but we can have a very good idea of where to look.

Patrick Shaw Stewart, Saturday, 28 March.


To find the references indicated by numbers in [square brackets] and for detailed analysis, please refer to my 2016 review:  Shaw Stewart, P.D. “Seasonality and selective trends in viral acute respiratory tract infections.” Medical hypotheses 86 (2016): 104-119.

**Remarkably, in 2019 Price and his colleagues looked at 9 common viruses and found exactly the same thing.  See Price et al. Scientific reports, 9(1), pp.1-11.

~~ See Bedford, et al. “Global circulation patterns of seasonal influenza viruses vary with antigenic drift.” Nature 523.7559 (2015): 217-220.

^^ See Adaptability of respiratory viruses

The Hypothesis

For a general discussion of the seasonality of respiratory viruses, written for the layperson, please see

Every winter, colds and flu increase

For a discussion of the strange timing and duration of influenza epidemics, please see

The strange arrivals – and departures – of influenza epidemics in the UK, 1946-1974

Applications to Covid-19

For comments about the epidemiology of Covid and other respiratory illnesses, please see

Epidemiology of respiratory illness

For discussion of how the trade-off model can be applied to the Covid epidemic see

Covid 19 and the trade-off model

For a simple model of the transmission of viruses such as CoV-2, please see

A simple model of CoV-2 transmission

For comments about how quickly we can expect viruses to adapt to new environments, please see

Adaptability or respiratory viruses

For more detailed scientific points about CoV-2, see

Technical notes on CoV-2 for scientists

For practical tips on avoiding respiratory illness see

Suggestions for avoiding colds and flu – and Covid-19

2 thoughts on “Predicting the seasonality of Covid”

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: