Stresses and strains: the evolution of Covid is not random

In the genetic diaspora of an epidemic, there is ferocious competition between strains of virus to get to the next victim first. That leads to apparently purposeful outcomes, as if the virus had a mind. One of the things people find hardest to grasp about evolution is that it appears purposeful but the mutations on which it feeds are random. How come dolphins evolved to swim if all they had to work with were random changes in genes? Viruses also mutate at random — but most people talk as though the rise and fall of these mutant versions is mainly down to chance or luck. It’s not.

Mutations occur all the time in RNA viruses; what matters is which ones find favour in natural selection. The champions of ‘Darwinian medicine’ have been calling for their colleagues to take evolution and adaptation more into account for years, and one of them, Paul Ewald of the University of Louisville, has something very relevant to say about this pandemic. Years ago, Ewald came up with a theory of why some diseases are lethal and others are mild. He argues it is all about the mode of transmission. Infections that you catch from coughs and sneezes are mostly mild; we get more than 200 different kinds of common cold virus and on the whole none of them puts you in bed, let alone kills you. Yet insect-borne diseases such as malaria, plague and yellow fever, and water-borne diseases such as cholera and typhoid, seem quite content to kill you.

The reason, says Ewald, is that in direct-contact diseases such as colds, mild strains will do better than nasty ones, because they send you out to work and to parties, coughing and sneezing. Insect-borne or water-borne diseases, meanwhile, may actually spread better if they confine you delirious to a deathbed with a high pathogen load, the better to attract mosquitoes or (sorry) maximise your contribution to local sewers.

There are two other categories of transmission: sexually transmitted diseases, such as syphilis, herpes and Aids, which may or may not kill you in the end but are good at lurking hidden so you have a chance to move to a new partner; and durable, sit-and-wait diseases such as tuberculosis and to some extent smallpox, caught from surfaces, which can afford to be lethal because they can be passed on after you are dead.

Every virus uses mutation and selection to find a compromise between maximising its offspring while killing the host, or moderating its effect and keeping the host active. Yet there was always one epidemic that didn’t seem to fit Ewald’s theory: the 1918 flu, which grew more deadly in the second wave, despite being spread by coughs and sneezes. In 2011, Ewald had a crack at explaining this exception in a way that neatly tests the rule. See if you find it convincing.

The 1918 flu was first noticed in US army training camps in Kansas in early March. Throughout the spring and summer it was about as lethal as most flus: dangerous to the very young and very old but mild in everybody else. It was in August, on the Western Front, that army doctors started noticing that the flu was turning deadly, regularly killing fit young adults. ‘Influenza increasing and becoming more fatal,’ wrote a senior US army surgeon in his diary on 17 August.

Ewald thinks that this was because the lethal strains were spreading better than mild ones. Imagine that a mild case would

be sent to a dugout to sleep it off, while a severe case was put on a stretcher and taken down the line to a series of crowded field stations, trains and hospitals. The sicker a soldier was, the more he was moved and the more nurses and doctors he saw. The peculiar conditions of the trenches allowed ‘individuals immobilised by illness to be transported repeatedly from one cluster of susceptible hosts to another, in trenches, tents, hospitals, and trains’, Ewald argued. The flu was essentially being spread by attendants, a bit like malaria is spread by mosquitoes. The deadly strains were now at an advantage. This explains why subsequent flu epidemics have never been as nasty — and, predicted Ewald, never will be.

There is a worrying parallel with Covid-19. In the early wave a lot of cases were spread by attendants in hospitals and care homes. One South African hospital traced how a single outpatient seeded an epidemic that spread from ward to ward, infecting 39 patients and 80 staff. The virus had a means to get from victim to victim even if they stayed put: it was attendant-borne, like the 1918 flu. Did that encourage the virus to be more lethal? An estimate published this week by Public Health England finds that the B117 (Kent) mutant is roughly 65 per cent more fatal than previous strains.

By contrast, because of lockdown, a mild case of Covid kept you isolated at home. Last week the Financial Times carried an article about the huge but surprisingly mild epidemic of Covid that India is suffering. It quoted one doctor as saying that ‘we are seeing a lot less severe disease than the rest of the world, and a lot more asymptomatic infections’ and another that ‘it’s pretty generally accepted that in India, we have a very mild form of the virus’. There are lots of possible explanations, but because lockdowns have been mostly ineffective in India, could it be that mild variants have done well and an attendant-borne evolution to greater virulence has not happened?

Yet for every Mumbai, there is a Manaus — a city in Brazil that had a huge first wave with little or no lockdown and saw a lot of people die. Some thought Manaus had reached the herd immunity threshold, but it is now seeing a bad second wave. Remember, however, that the deadly strain of 1918 flu started in the trenches, but soon spread everywhere. It’s the global average strategy that we use against the virus that counts, not the local one: Manaus’s new strains seem to have arrived in the city from elsewhere.

You might conclude from this logic that we have made a mistake by locking down, ensuring that the virus remains deadly or becomes more so. I hesitate to agree with that, because I have been wrong about a lot during this pandemic. And there is one crucial way in which Covid-19 differs from flu: it spares the young and clobbers the old. That might be enough to ensure that nasty strains remain competitive with mild ones even in the absence of lockdowns. A strain that causes only very mild symptoms in most people, so they go out spreading it, but occasionally kills the vulnerable, might thrive.

I don’t know if Britain would have seen more than 100,000 deaths or fewer if we had pursued a less draconian strategy like India, Sweden or Florida. But I do know that evolution is about more than mutation.