By Foss Tighe
There is a lot of talk about a possible “Second Wave” of coronavirus infections. Why do infectious disease outbreaks come in “waves”? If infections reach some sort of peak and then begin to decline, isn’t that a sign that the outbreak is ending?
The major pandemics recorded in history books appear to have come in waves. These include:
- The plague in Athens (430 – 426 BCE)
- Bubonic Plague in Medieval Europe (1347 -1770)
- Spanish Flu (1918-1919)
Plague In Athens:
The plague in Athens first appeared in that city during the second year of the Peloponnesian War. The great historian Thucydidies wrote about the plague in great detail. For the 2500 years since the plague doctors and historians have struggled to identify the cause of the plague. See my Peloponnesian War Blog post on the plague. The plague that struck Athens swept through the Ancient Greek world in 3 waves, the first in May of 430 BCE, the second wave came in the summer of 428 BCE and the final wave occurred in winter of 427/426 BCE. The Peloponessian War was a 30 year conflict between the ancient Greek cities of Athens and Sparta, and their respective allies. The Spartans had an army that could not be beat, and Athens had a navy that could not beat. These asymmetrical strengths led an Athenian strategy of abandoning the countryside and moving the entire population into the safety of the walls protecting Athens and her port whenever the Spartan army invaded. As any good public health official knows, increased population density, increases the risk of infections. Based on data in Thucydides history, it is estimated that as much as 25% of the Athenian population died of the plague by the end of the third wave.
The bubonic plague caused by the bacteria Yersinia pestis, swept across Europe and large sections of the world in multiple waves between 1346 and 1770. It has been estimated that between 30 and 50% of the population of Europe may have been killed by the initial wave (1347-1351). One of the final waves occurred in France in 1770. Stat, a health news web site published an interesting piece on the response similarities between the officials of Marseilles in 1770 and the US official response to the coronavirus outbreaks.
The origin of the name “Spanish Flu” is interesting. It was very unlikely that Spain was the location of the original outbreak. But during WWI, all the combatant nations had some form of censorship in place. Eager to hide bad news during the war, news of the deadly flu was initially suppressed in most nations. Spain was one of the few European nations neutral during the war and therefor had a more free press. So all the early stories about the outbreak came from Spain.
The world had changed a great deal between the last wave of the Bubonic Plague and 1918. In particular the US participation in the war resulted in large numbers of young men travelling between the North America and Europe increasing the speed at which the disease could spread.
A page on the CDC website outlines the three waves: the first, starting in March 1918 and subsiding by summer of 1919, followed by the most deadly, in the US, second wave in the fall of 1918 and a third wave in the winter/spring of 1919.
There are many medical research publications on pandemic waves. Most attribute the pandemic wave patterns to some combination of 5 phenomenon1. These include:
- Variable Transmission rates
- Virus Mutation
- Waning Immunity
- Sub-Group Isolation
- Human Behavior
Variable Transmission Rates:
This is perhaps the most familiar to most people. The annual flu cycle is a form of wave infection which is driven in large part by a variable transmission rate based on seasons. The flu is much more easily transmitted during the colder months of the year. The CDC graph below shows the size of annual flu outbreaks for 8 recent flu seasons.
Seasons are not the only type of variable transmission rates. For example, one of the primary modes of transmission of HIV was sexual contact. Homosexual contact had a much higher transmission rate than heterosexual contact. With coronavirus, we know there is a much higher rate of death in older people and that younger people have milder infections, but because of the high rate of asymptomatic cases and the continued paucity of good data, we don’t know if younger people are just as likely to be infected with the virus as older people.
Another version of variable transmission rate is referred to as viral interference. One viral infection, can interfere with the ability of another virus to infect a population. This mechanism of interference is related to the today’s medication known as Interferon. Researchers have speculated that the delayed initial outbreak of the HIN1 influenza in some European countries in 2009 may have been due to an usually high rate of rhinovirus infections in those countries that spring/summer.
For various reasons the effective rate of infectious transmission can vary. Reasons for this include, but are not limited to: seasonal changes, different rates of susceptibility based on sub-groups of a community, or the presence of other viral infections
Unlike humans, virus evolve quite quickly. There is a very cool site on the web that is tracking the evolution of COVID-19 in close to real time. This is the kind of stuff that allows scientists to conclude that the initial west coast infections of coronavirus likely came from China, while the east coast infections likely came via Europe, because the Eurpopean version had already evolved some unique aspects in its viral RNA. Viral mutation can contribute to waves of infection, when the virus mutates sufficiently to cause the immunity triggered by a first infection to no longer “recognize” the mutated version of the virus. Like variable transmission rates, this is a common part of the annual flu cycle. Influenza mutates sufficiently in a 12 month period to cause antibodies to last year’s version to be less effective against this year’s version of the flu. This means a population that was substantially protected from an earlier version of a virus could see the infection disappear, only to re-emerge again at a later point.
Viral mutation is an issue for vaccine development. An effective vaccine needs to be based on some aspect of the virus that hopefully does not mutate. Vaccine developers try to target vaccines at something so fundamental to the virus, that if that changed, the virus would not longer be a threat to humans.
In its effect, this is very similar to Virus Mutation. If the immunity generated by recovering from an infection is of limited duration, it means that a “recovered” population that was initially immune to re-infection can now get re-infected. So like viral mutation, it makes it possible for a second (or 3rd, 4th etc) wave to return to an area and cause a new outbreak. Of course the effective immunity generated by coronavirus infections is one of the big unknowns at this point. Does it generate immunity? If so, how long does that immunity last.
The waves of bubonic plague in Europe were so far apart that many of the people who had been previously infected may have died. People born since the last wave were susceptible to infection, while their older relatives may have been immune.
This also has big implications for vaccine development. If the immunity generated by infection wanes, it is possible the immunity generated by vaccine will also wain. Will we need an annual vaccine for coronavirus like we have for influenza?
This is related to the idea of social networks. Social networks are of course vital to how viral infections like Covid-19 are spread. My wife works in a medical setting where she might get infected, she comes home and gives the infection to our kids. My daughter visits her friend and infects her etc. In fact a big part of how social distancing works is that we temporarily shrink the size of our social network that we physically interact with to reduce the rate of viral transmission. The theory of how this can contribute to pandemic waves, is the notion that though many of use live in the same geographic areas, there may be social networks that are largely independent of each other. So an infection could be spreading in one social network and infecting a very high percentage of that network, while other networks could be virtually untouched. This can lead to a situation where the virus may naturally decline, having burned through a sub-set of networks within a geographic area, while a substantial part of the population, representing other social networks are untouched. Sometime later, when members of the untouched networks come into contact with an infected person, perhaps someone travelling from another area, a new outbreak can start fresh in these untouched networks. My wife and I were recently having a discussion with a neighbor (social distancing of course). We recounted a handful of people who we know who died of the virus. He indicated that he knows no one who has died nor anyone who has been infected. Despite living only 40 ft apart we are living in very different social networks. You can also imagine this kind of division occurring across other social divisions as well. Health care workers in nursing homes have a very high probability of coming into contact with infected people, while technology workers, working from home probably have a much lower risk. Once most health care workers have had the illness we may see a decline in infections. Then when we have opened up, and all the techies go to some sort of in-person technology conference we may see a new outbreak in a previously isolated sub-group.
Lastly we have this all important category. Human behavior includes a wide range on things. For example, it as been suggested that our academic school year starting in the fall may have been the trigger for the second wave of the 1918 Spanish Flu.
Over the summer many families travel. Upon returning home and starting school, there is a gathering of large number of students in small locations. This presents a great opportunity to spark a new spike in infections. Schools, like the walled City of Athens in 432 BCE present viral infections great opportunities. The 1770 outbreak in Marseilles France was centered around an annual festival that drew merchants from around the world to a densely populated event. The cancellation of the annual pilgrimage to the hajj in Saudi Arabia, was based on the understanding that such events can trigger new outbreaks by bringing people from many different locations into close contact. In the United States we should be very concerned about the resumption of in-person Church services.
Social Distancing itself is a form of human behavior that could help trigger new waves on infection. Having used social distancing to “flatten the curve” and in many places caused a decline in new infections many people are asking what next? There is a growing movement to reduce the social distancing to help get the economy going again. Doing so will likely result in an increase to people being exposed to the virus. Whether that increase is sufficient to trigger a significant new wave of infections awaits to be seen.
Potential impact of seasonal forcing on a SARS-CoV-2 pandemic
Publication Date: 16.03.2020
Swiss Med Wkly. 2020;150:w20224
Three waves of 1918 Flu
A Perspective on Multiple Waves of Influenza Pandemics
Anna Mummert,Howard Weiss,Li-Ping Long,José M. Amigó,Xiu-Feng Wan
Published: April 23, 2013https://doi.org/10.1371/journal.pone.0060343
Ånestad et al., J Anc Dis Prev Rem 2015, 3:1
Three waves on black death
Med Hist Suppl. 2008; (27): 74–100.
Samuel K Cohn JR
Mt Sinai J Med
. 2009 Oct;76(5):456-67. doi: 10.1002/msj.20137.
The Plague of Athens: Epidemiology and Paleopathology
Robert J Littman 1
When bubonic plague hit France in 1720, officials dithered. Sound familiar?
By CINDY ERMUSMAY 25, 2020