Eastern promises: Are Asians relatively immune to COVID-19?

Eastern promises

Are Asians relatively immune to COVID-19?

Believe me, I really did not want to write another post on COVID. Why would I wish to put off my already limited readership?  But with so much going on, the temptation was simply too much to resist. And for what it’s worth, I do hope you’ll share the joy I got out of this minor insight and data dredge.

From the beginning of the pandemic, we’ve heard how the South East Asian countries have efficiently curbed their infection rates. Some of the writing was clearly politically motivated, contrasting them to the supposedly poor initial response of China. We were told how South Korea and Taiwan tested a large number of people and were able to control transmission. Though the testing figures don’t hold up (at <0.01 tests/1000 population, Taiwan has perhaps the lowest testing figure among well-off countries, and South Korea fell behind most of Europe in testing by the beginning of March), some of these countries did do efficient case and contact tracing. But is this sufficient explanation for the low number of cases and deaths seen in these countries? And most people forget the poor countries just across from China’s borders which have also done spectacularly, without any help from big data analytics and other flourishes that their rich neighbours used. As an example, Vietnam, a country of over 95 million people, is yet to record a single death due to COVID-19. Neither have Laos and Cambodia. None of these countries are into aggressive testing either. This is not all. If you zoom out of this region, you will find other adjacent countries with deaths in the low 2 digits (Thailand, 58; Singapore, 27). This is wildly in contrast to what we find in Europe and the Americas. The number of deaths per million population is an order of magnitude higher. (Figure1) Could we attribute all this difference to undercounting of deaths? Or is there something else at play?

Figure 1: Deaths due to COVID-19 per million population in different regions of the world

Death data as on the 27^th of July 2020, from https://ourworldindata.org/covid-deaths. Bubble size corresponds to population size. Africa may be at an earlier stage in the pandemic, but is included here for completion.

Cousins of SARS

Bat biologists (wonder what drives people to such a delightful day job!) contend that bat diversity is maximal in the South East Asian region, including southern China (https://www.economist.com/science-and-technology/2020/07/22/the-hunt-for-the-origins-of-sars-cov-2-will-look-beyond-china). Since most coronaviruses which make the jump to humans, originate in bats, they believe that the diversity of the coronaviruses must also be greatest in this region, allowing for mutations that make this jump possible. Some scientists believe that many (as yet undocumented) coronaviruses may have already become infectious to humans, what with people preferring their fruit-bat with coconut milk (being non-judgmental while experiencing overpowering nausea is a challenge, but I am trying, sincerely!). So it is plausible that some less malevolent cousins of the current SARS Cov-2 may have already been in circulation in these populations for some time now. Perhaps this partly explains the low infection and death rates from COVID-19. But, this is all just conjecture, and there is no easy way to prove the hypothesis.

Degrees of separation

While the bat biologists and infectious disease epidemiologists fret over the question, I thought that it would be interesting to see how per-capita death rates due to COVID-19 changed as a function of distance from the least affected regions. Because of the wide variability in testing rates, I used deaths rather than the case numbers. I measured distances to country capitals from the capital city (Hanoi) of the largest country with zero deaths. I included only the Asian countries; per-capita deaths in Europe and Americas are clearly well above the highest figures from Asia (figure 1). I excluded countries (such as Syria and Yemen) which are troubled by civil unrest or war. The resulting plot is an analyst’s delight. It seems that per-capita deaths due to COVID-19 show an increasing trend as we move away from Vietnam. (Figure 2) What’s more, there are at least three distinct clusters of countries sharing similar per-capita death rates (and at similar distances from Hanoi). Thailand, Singapore, and Malaysia (coded blue in figure 2) have the lowest rates, followed by the South Asian countries (India and Pakistan, names in purple). The countries in the outer reaches of Asia (Turkey, Israel, Iraq, names in red) have the highest per-capita death rates due to COVID-19.

Figure 2: Per-capita deaths due to COVID-19 in Asian countries plotted against distance from Hanoi

Death data as on the 27^th of July 2020, from https://ourworldindata.org/covid-deaths. There appear to be at least 3 distinct clusters of countries with differing per-capita deaths due to COVID-19, defined by their distance from Hanoi (colour coded blue purple and red). I haven’t shown Iran (186 deaths per million, 4500 km from Hanoi) in the graph as it is an outlier (and would need an unsightly y-axis). And excluding it does not alter the overall relationship. If we assume a linear relationship, distance from Hanoi appears to account for 60% of the difference in per-capita deaths from COVID-19.

Now, before you start making relocation plans in preparation for the next pandemic, I have to point out the caveats, as a sacred scientific duty. The most important one is of course that counting deaths is not uniform across these countries. And then there are the other differences (such as lockdown intensity and health system preparedness) which could influence the number of deaths. But even accounting for all these factors, I find it difficult to shake-off the feeling that there is something to this relationship. I do hope that the bat biologists come up with something more concrete. In any case, the most sobering take away for me is that, we know so little about a disease that has nearly overwhelmed us.