The coronavirus (C19), or SARS-CoV-2, has a strong seasonal component that appears to closely match that of earlier SARS viruses as well as seasonal influenza. This includes the two distinct caseloads we’ve experienced in the U.S. 1) in the late winter/early spring; and 2) the smaller bump we witnessed this summer in some southern states and tropics. 

COVID Seasonal Patterns and Latitude

The Ethical Skeptic on Twitter recently featured the chart below. It shows the new case count of C19 in the U.S. in the upper panel, and the 2003 SARS virus in the lower panel. Both viruses had an initial phase at higher latitudes and a summer rebound at lower latitudes.

I particularly like the following visualizations from Justin Hart demonstrating the pandemic’s pattern at different latitudes (shown in the leftmost column). The first table shows total cases by week of 2020. The second shows deaths per 100,000 of population by week. Again, notice that lower latitudes have had a crest in the contagion this summer, while higher latitudes suffered the worst of their contagion in the spring. Based on deaths in the second table, the infections at lower latitudes have been less severe.

Viral Patterns in the South

Many expected the pandemic to abate this summer, including me, as it is well known that viruses don’t thrive in higher temperatures and humidity levels, and in more direct sunlight. So it is a puzzle that southern latitudes experienced a surge in the virus during the warmest months of the year. True, the cases were less severe on average, and sunlight and humidity likely played a role in that, along with the marked reduction in the age distribution of cases. However, the SARS pandemic of 2003 followed the same pattern, and the summer surge of C19 at southern latitudes was quite typical of viruses historically.

A classic study of the seasonality of viruses was published in 1981 by Robert Edgar Hope-Simpson. The next chart summarized his findings on influenza, seasonality, and latitude based on four groups of latitudes. Northern and southern latitudes above 30° are shown in the top and bottom panels, respectively. Both show wintertime contagions with few infections during the summer months. Tropical regions are different, however. The second and third panels of the chart show flu infections at latitudes less than 30°. Influenza seems to lurk at relatively low levels through most of the year in the tropics, but the respective patterns above and below the equator look almost like very muted versions of activity further to the north and south. However, some researchers describe the tropical pattern as bimodal, meaning that there are two peaks over the course of a year.   

So the “puzzle” of the summer surge at low latitudes appears to be more of an empirical regularity. But what gives rise to this pattern in the tropics, given that direct sunlight, temperature, and humidity subdue viral activity?

There are several possible explanations. One is that the summer rainy season in the tropics leads to less sunlight as well as changes in behavior: more time spent indoors and even less exposure to sunlight. In fact, today, in tropical areas where air conditioning is more widespread, it doesn’t have to be rainy to bring people indoors, just hot. Unfortunately, air conditioning dries the air and creates a more hospitable environment for viruses. Moreover, low latitudes are populated by a larger share of dark-skinned peoples, who generally are more deficient in vitamin D. That might magnify the virulence associated with the flight indoors brought on by hot and or rainy weather.   

Mutations and Seasonal Patterns

What makes the seasonal patterns noted above so reliable in the face of successful immune responses by recovered individuals? And shouldn’t herd immunity end these seasonal repetitions? The problem is the flu is highly prone to viral mutation, having segments of genes that are highly interchangeable (prompting so-called “antigenic drift“). That’s why flu vaccines are usually different each year: they are customized to prompt an immune response to the latest strains of the virus. Still, the power of these new viral strains are sufficient to propagate the kinds of annual flu cycles documented by Hope-Simpson.

With C19, we know there have been up to 100 mutations, mostly quite minor. Two major strains have been dominant. The first was more common in Southeast Asia near the beginning of the pandemic. It was less virulent and deadly than the strain that hit much of Europe and the U.S. Of course, in July the media misrepresented this strain as “new”, when in fact it had become the most dominant strain back in March and April.

What Lies Ahead

By now, it’s possible that the herd immunity threshold has been surpassed in many areas, which means that a surge this coming fall or winter would be limited to a smaller subset of still-susceptible individuals. The key question is whether C19 will be prone to mutations that pose new danger. If so, it’s possible that the fall and winter will bring an upsurge in cases in northern latitudes both among those still susceptible to existing strains, and to the larger population without immune defenses against new strains.

Fortunately, less dangerous variants are more more likely to be in the interest of the virus’ survival. And thus far, despite the number of minor mutations, it appears that C19 is relatively stable as viruses go. This article quotes Dr. Heidi J. Zapata, an infectious disease specialist and immunologist at Yale, who says that C19:

“… has shown to be a bit slow when it comes to accumulating mutations … Coronaviruses are interesting in that they carry a protein that ‘proofreads’ [their] genetic code, thus making mutations less likely compared to viruses that do not carry these proofreading proteins.”

The flu, however, does not have such a proofreading enzyme, so there is little to check its prodigious tendency to mutate. Ironically, C19’s greater reliability in producing faithful copies of itself should help ensure more durable immunity among those already having acquired defenses against C19.

This means that C19 might not have a strong seasonal resurgence in the fall and winter. Exceptions could include: 1) the remaining susceptible population, should they be exposed to a sufficient viral load; 2) regions that have not yet reached the herd immunity threshold; and 3) the advent of a dangerous new mutation, though existing T-cell immunity may effectively cross-react to defend against such a mutation in many individuals.