COVID-19, Air Pollution and One Health at the Climate Change Turning Point

Abstract

COVID-19 escalated into a pandemic posing humanitarians and scientific challenges. We explored the geographical feature of the first wave infection and correlated it with annual satellite and ground indexes of air quality in eight countries: China, U.S.A, Italy, Iran, France, Spain, Germany, and U.K. Controlling for population size, we found more viral infections in those areas which were afflicted by high PM 2.5 and nitrogen dioxide values. Higher mortality was also correlated with relatively poor air quality. This phenomenon also occurs in China when removing, the city of Wuhan and its province from the dataset. For long recognised to be a high-risk factor for several respiratory-related diseases and conditions, air pollution seems to be a risk factor for COVID-19 too. This finding suggests the detrimental impact climate change will have on the trajectory of future respiratory epidemics. Previous Asian epidemics and the Ebola have brought forward evidence of the natural causes of zoonoses which have become more threatening due to land-use change, ensued lack of a buffer zone between the cities and the forests, and our closer proximity to wild pathogens. Together with air pollution, these elements illustrate the need to stick to the UN targets limiting biodiversity loss and climate change.

1. Introduction

In early 2020, the world was hit by a pandemic of epochal dimensions as never seen before since a century earlier when the 1918 Spanish influenza hit worldwide. Too much of the public concern, such a global, new disease brought about several complot theories. Typical to difficult to explain phenomena, humans’ resort to an intelligent hand as the causing factor behind them. In fact, during these last 20 years, the SARS outbreak from Southern China and Hong Kong, the Zika one from Central and South America, and the Ebola one from West Africa have shown how very large epidemics can originate in every continent with similar patterns.

Like the other SARS coronaviruses, scientists have provided the best evidence that SARS-CoV-2 transferred hosts from the originating reservoir of bats to humans [1] without any laboratory modification. Despite some warnings that SARS-CoV-2 might have occurred before its allegedly first outbreak in Wuhan, the most likely place where the spillover took place remains that wet market in Wuhan. To environmental scientists, though, this is an unimportant, episodic event that will always tell just one piece of a bigger story composed by the many accounts of cross-species transfers. In the last few years, these accounts of new viral zoonotic diseases have repeated and what they brought have been harmful consequences, such as the other of the three most relevant coronaviruses, the MERS from the middle-east in which dromedaries were involved [2].

Regardless of the specific epidemics, we look at, we know that in these last epidemics both the causes of the spill overs and later the fast spreads across different habitats and cities were all anthropogenic. Land-use change and habitat loss together with the biodiversity loss causing the sixth mass extinctions are the climate change reasons behind epidemics [3, 4]. As the WWF neatly summarises (Figure 1) [5], as long as the forest ecosystems are healthy and there exists a proper countryside buffer (without intensive farming) to the cities, spill overs to humans are extremely unlikely [6].

Figure 1.

WWF pictorial.

Intensive livestock farming is another key element contributing to zoonotic diseases [7] that coupled with human overpopulation, played a critical role [8]. Evidence flourished showing that a higher frequency of contagions occurred in German [9] and French abattoirs [10]. A multidisciplinary ecological approach to study the coronavirus and other zoonotic infections is necessary as well as combining medical factors causing co-morbidity before and after the disease in patients more at risk.

COVID-19 appeared in a Chinese area affected by some of the highest air pollutions in the world, and it showed a relatively high virulence there. The areas in the world that were particularly hit by the pandemic were those with high commercial trade [11] and those hit by air pollution. These are risk factors that, together with medical issues including smoking, get associated with higher morbidity and mortality of COVID-19 [12]. In particular, smoking affects hyperactivates the ACE20 receptors in our alveolar respiratory system which are also attacked by the coronavirus [13]. Chronic exposure to air pollution causes a chronic lung and bronchial infection similar in features as that found in smokers because of the ACE20 receptors [14, 15].

2. Air pollution and COVID-19: the case of 8 studied countries

In late 2020 winter, alarmed by the spreading of the virus from China to Iran and Italy, we were struck by how it was hitting the most polluted region of the European continent, the Italian Po′ river plane. Of course, air pollution was just one of the reasons of the virus spread, mostly the travelling of passengers from China to Mecca, which brought it to Iran, and later of Chinese merchants that went to trading exhibitions in Italy were the first causes of the spread of the disease out of China. Yet, in China itself, SARS-CoV-2 was first recorded in one of the most polluted areas in the world. Conversely and paradoxically instead, the ecologically lush Chinese Yunnan Province, the one from where most of the coronavirus bat hosts have ever been sampled [16], has been one of the areas in the world’s least affected by the pandemic. A zoonotic disease was seemingly hitting again in areas of the world where unsustainable human practices were present likely because of wild animals trading in a wet market.

We, therefore, started working at the theory that chronic exposure to air pollutants was a cofactor for higher cases and deaths from COVID-19 [17]. This was the safer long-term air pollution hypothesis, while confronting some experts in the field we understood others had the same intuition and, amongst those, bold physicists and environmental doctors were working on the more alarming short-term hypothesis that the virus is also carried by the polluting particulate matter [18]. Colleagues from America showed the same day of our preprint that other individuals and environmental risk co-factors play a minor role than pollution [19]. We were glad because the same idea was picked up by scientists with more experience than ourselves, and we enlarged the analysis to include the other countries of Spain, the U.K., the U.S.A., France, and Germany [20]. Later, dozens of other studies showed a similar trend occurring for many pollutants in many regions of the world.

What was important for our analyses to work was to find territories small enough to split the countries from which we could compare COVID-19 data and 2019 averaged pollution data. We repeated the analysis throughout the months of 2020 and it was frustrating not to be able to obtain, for example, deaths at a small provincial level from Italy; the country was not disclosing this important piece of information to the scientific community! We were glad to be able to use very important and early infection data published from the Iranian state in Farsi; those data stopped being published just after a few months and the country at a larger regional level could not be followed in its progression.

Relying on the expertise of the satellite geographer and co-author Davide Fornacca, the data on air pollution posed less trouble. Not totally trusting the accuracy of ground measures collected from the various devices around the world, we relied on data sampled from a European open-access satellite that estimates ground pollution around the globe at a fine scale.

The comparison between cases and deaths from COVID-19 and 2019 air pollution had to be controlled for the most important confounding factor within those comparably small territories in the 8 countries: population density. The higher the densities of the new hosts, the higher the contagions were going to be regardless of the type of parasite and ecological conditions around; it is a simple transmission paradigm we were not interested to look at.

A plethora of other confounding risk cofactors could have been added, and many other colleagues did so at a smaller scale than eight countries. From our side, we were aiming at a simple and global illustrative variable that represents a degraded environment and is involved in respiratory disease; a variable that could have played an even more important role once a disease becomes endemic and infections from distant places play less of a role.

The maps (Figure 2 reported from our main publication [21]) show how the most polluted regions overlap with the areas where COVID-19 infects or causes most deaths. The trend is most evident in Italy, where we find a marked gradient between a highly polluted and the majority of cleaner areas. Germany and Spain provided no clear evidence of this correlation because of the character of the air pollution there. In Germany, pollution is evenly spread and the lack of a gradient does not allow for a Kendall correlation to output significant results. In Spain the pollution is negligible and the statistics also do not work out at a country level. Yet, later accounts could find some evidence also for these countries (as reviewed by [22] and later reviewed by more recent papers referenced at the end of this paper).

Figure 2.

COVID/Pollution correlation maps of 8 countries.

The pollutants we analysed were several, and amongst the most influential for finding a trend there are PM 2.5 (of the previous Figure 2) and PM 10 and, reported in the following Figure 3 (reported from the other publication [23]) for China, NO₂, CO, and HCHO).

Figure 3.

COVID/Pollution correlation maps for three pollutants in China.

The China ‘case’ drew much of our attention right because it was where the epidemic started and the country where it was contained most effectively. The pollution hypothesis should not have stood much because the onset of the virus should have been attributed to the location of the spill over. If it had happened somewhere with a high population density, that is in a megalopolis like Wuhan, pollution should have played a role possibly only later on from the initial abrupt explosion. Instead, the trend was there, even after removing Wuhan and its Hubei Province from the analysis [23].

Was the publication of these results easy in a high-impact journal? The earliest authors reflected on some limitations in a journal with high-impact factors [24]. We were unable to have such an impact for supposedly two or three reasons. Firstly, a correlational study is still an observational one missing an artificial intervention as such to be able with a control, non-modified group of patients, to compare results. Clear causation is, therefore, lacking. It is obvious that we cannot spread the virus ourselves in polluted and non-polluted places, as well as we cannot start polluting clean places for months to check whether, there, there will be more COVID cases and deaths. To overcome this ethical impossibility to turn the observation into an experiment, with years’ time longitudinal screenings performed on patients from retrospective cohorts will help cast doubts out [25]. A first example has been attempted in Spain [26] with results which agree with out hypothesis [27]. The other limitation may be given by the truly interdisciplinary nature of the study which finds it hard to gather environmental medical doctors in general science journals who can also assess the geographical methods used here. The third, more important reason is given by the apparent nature of the problem; although air pollution like smoking is a problem causing millions of deaths every year, it is not big news… Nevertheless, the editors of journals and journalists with a large impact on the general public missed the chance to communicate the message at the right time of the epidemics, that is during the earliest months, when the scariest new deaths during the lockdowns could have allowed us to ponder on the true nature of these recent new diseases, including their ecological characteristics linked to climate change [28].

Several reviews collated our studies together with well over a hundred during almost 2 years. A few recent ones are listed in refs. [29, 30, 31, 32].

3. Conclusions

The fossil fuel economy causing air pollution carried on unabated once we resumed activities after the lockdowns. At an individual level, we are fast at not wanting to tackle the emission problem because the institutions have not endorsed reforms effectively [33]. However, the basics of game theory tell us that we need to have a proper behavioural strategy at an individual level before expecting to see groups of individuals or institutions act likewise. As well as solving the Public Good Game of the pandemic [34] by paying a small individual cost and getting vaccinated for the benefit of the collectively, we must also more strongly clean our environment and strongly limit our impact for the good of the collectively and ourselves. The fact that nature is going to profit and that we need to pay a cost seems to cause us an extra layer of difficulty in the swift employment of these green strategies, when in fact, we do not understand that we belong to that layer too.

What saves the planet will save us too.

Acknowledgments

Davide Fornacca contributed to accomplishing our preprints and the two main referenced papers published in the journals Atmosphere and Frontiers in Public Health [21, 23]. This chapter as well as these two previous works received no funding for their completion.

Conflict of interest

The authors declare no conflict of interest.