The coronavirus in winter may be worse than scientists thought

People wearing face masks walk on a street on a cold winter day, following the coronavirus disease (COVID-19) outbreak, in Beijing, China, December 21, 2020.…
People wearing face masks walk on a street on a cold winter day, following the coronavirus disease (COVID-19) outbreak, in Beijing, China, December 21, 2020.…
Image: REUTERS/Tingshu Wang
We may earn a commission from links on this page.

A year after the emergence of the disease that has killed 2 million people, humans keep underestimating the SARS-CoV-2 virus.

This winter, Covid-19 is turning out to be even more dangerous than epidemiologists and public health officials had feared—and not just because of the more contagious variants now making their way around the globe. As recently as October, Nature reported it was “too soon to say whether Covid is seasonal like the flu.” Evidence hinted that winter weather could increase transmission of the virus: In the lab, the virus persisted under cold, dry conditions and was inactivated by the ultraviolet rays in sunlight.

There was reason to hope that wasn’t the case. Coronaviruses, which generally show less seasonal variation than the influenza virus, tend to have a weak response to changing temperatures. Outbreaks such as the severe acute respiratory syndrome coronavirus (SARS-CoV) in 2002 were not thought to be seasonal at all, although the outbreak ended too quickly for scientists to definitively test that idea.

Yet SARS-CoV-2 appears to be different, and the world may face a bigger fight this winter than expected. “People are thinking they’re looking at something like influenza, and it’s a lot worse,” says professor Richard Carson, who published a preprint in November suggesting Covid’s reaction to changing temperatures, known as its temperature response curve, may be even more pronounced than influenza, the ultimate seasonal virus. “A lot of things people were doing in the summer we thought were working was actually hotter temperatures giving the appearance these things were working.”

Carson and his co-authors at the Massachusetts Institute of Technology and Wake Forest University arrived at this conclusion by excavating a unique data set: states’ death certificate records. Statistics on Covid-19 case counts and deaths in the US are notoriously noisy. Local and state governments do not report standardized data. Labs overwhelmed with tests may delay reporting positive cases long after administering tests. Deaths reported on any given day may have occurred weeks before. All these inaccurate counts stymie attempts to measure temperature’s role in transmission rates.

But Carson’s team, with experience applying econometric techniques to modeling and forecasting environmental impacts, found a solution in Massachusetts’ Covid-19 statistics reporting last year. “Buried deep in the report was an alternative death count by the date on the death certificate,” he said. “Once we found it, took us two months to get that data from most of the big states.” By pulling dates from death certificates, the team could match trends in Covid-19 deaths against maximum daily temperatures over a three-month period between Apr. 16 and Jul. 15.

According to the research (pdf) now undergoing peer review, the data shows the virulence of the virus increases below 31°C (88°F). Its “sweet spot,” says Carson, is around 4.4°C (40°F), but that’s only its preferred condition. Temperatures between 5°C and 10°C (41°F to 50°F) favor transmission and infection.

The study’s model revealed a strong correlation between changing temperatures and the number of Covid-19 cases and deaths from a baseline of 31°C, the US mid-summer average. By the time the weather drops to a chilly 5°C, the model shows deaths rising by 160% due only to the influence of cold weather, even after controlling for state shelter-in-place orders.

Covid-19 transmissions showed an even stronger effect. Four times more new Covid-19 positive cases are expected when temperatures drop to 5°C, assuming no other interventions such as masks or social distancing. Only when temperatures drop a few degrees below freezing, the point when water droplets quickly freeze in the air, do Carson and his team project transmission to slow.

“It’s a real frightening paper,” says Carson. “We know the temperature response curve for influenza. This one is much steeper.”

The temperature response curve for the SARS-CoV-2 virus.
The temperature response curve for the SARS-CoV-2 virus.
Image: Courtesy of Carson et. al.

For most of the US, sitting in the northern hemisphere, a temperature response curve like this makes for a treacherous winter. Any delay in responding to elevated virus activity, the data implies, will lead to a rapid escalation in cases because of the feedback loop between cold temperature and the virus’s exponential growth curve. “The patterns of the outbreak in mid-December almost exactly match what the temperature response would predict,” says Carson.

Adam Kaplin, a physician and public health researcher at the John Hopkins University School of Medicine, was not surprised. “It’s really hard to see it any other way,” he said, having reached a similar conclusion in his recently submitted research (pdf) to the journal PLOS ONE. “Virus transmission is going to up because of the temperature. That’s clear.”

Kaplin analyzed SARS-CoV-2 transmission rates and daily temperature records in 50 countries between January and April 2020, a period before most mask orders took effect, allowing his international team of researchers to isolate the influence of temperature. They found for each degree temperatures fell between 30°F and 100°F, the rate of transmission rose 3.7%

If this “strong and robust” association is correct, the paper states, countries must spend their spring and summer months containing the virus to have any hope of containing winter outbreaks given the effects of decreasing temperatures. “This is a race we should have been running a lot earlier,” says Kaplin. “We should have been way ahead of it. We blew this and many people died who didn’t need to die.”

What’s causing this?

Viruses kill millions of people each year, yet we know shockingly little about why their virulence fluctuates from season to season. The flu’s winter spread is commonly attributed to human behavior patterns—staying inside, where germs spread more easily—but that explanation has been questioned. ”I don’t think you’ll find a uniform answer to this question,” says Joe Eisenberg,  chair and professor of epidemiology at the University of Michigan’s School of Public Health. ”This research was not prioritized in the past.”

Carson was not able to tease out the relative influence of viral biology and human behavior on SARS-CoV-2’s response to cold temperatures. His research could only eliminate humidity as a likely factor in transmissibility; temperature and UV light (which fluctuates in lockstep with temperature) were much better at explaining Covid-19 transmission patterns.

But Kaplin thinks the evidence points to a very strong candidate. “It’s the biology of the virus,” he argues. “Yes, people are going inside more but it’s playing much less of a role than biology.” Nothing in the scientific literature, he notes, supports the idea that this seasonality is driven by human behavior. In the journal Medical Hypotheses, a 2016 paper could find no evidence that winter crowding drives seasonal viral transmission, noting that time spent indoors in the US changes less than 10% between summer and winter. Evidence in the tropics, where the flu circulates year-round, contradicts this hypothesis, too.

Policy implications

Reducing the transmission rate will mean redoubling mask and social distancing efforts. Now preventative measures must escalate merely to keep the epidemic’s rate of spread in check.

It’s a grim outlook. “The big policy notion is every week the temperature drops, you have to reduce the effective contact rate to keep the virus in check and prevent exponential growth,” says Carson. “People aren’t doing that, and that’s why you’re seeing those outbreaks.”

The new variant adds a wild card to the mix. “This new variant is more transmissible, but it may still have a strong seasonal signal,” says Eisenberg. “We just don’t know.”

So far, policymakers have waited too long to respond to Covid-19’s winter surge. In the UK, cases began piling up again this fall after a brief summer respite. Hospitals warned of a crush with the arrival of colder weather. But national restrictions weren’t reimposed until Jan. 9, when the healthcare system was already in crisis battling the more contagious variant. Today, the UK’s per capita death rate from Covid-19 is among the highest in the world and more than 40,000 Covid-19 patients are hospitalized, roughly double last year’s peak.

The US is in even worse shape. The country never got the pandemic under control. Although transmission slowed over the summer, the virus has come roaring back. The US now accounts for 20% of the world’s Covid-19 deaths—despite having less than 5% of its population—and is on track to reach 600,000 deaths in 2021. The only immediate solution is to establish herd immunity before next winter through a mass vaccination campaign that reaches more than 70% of the population. If the pandemic continues to spiral out of control, far more deaths will be unavoidable.

But we can prevent this winter from getting even worse, say Eisenberg. The public health measures we’re already taken need to ramp up if we’re going to keep SARS-CoV-2 in check. “It intensifies our recommendations,” he says, “but doesn’t change our recommendations.”