Role of a Habitat's Air Humidity in Covid-19 Mortality
RRole of a Habitat’s Air Humidity in Covid-19 Mortality.
Irina V. Biktasheva , , ∗ Department of Computer Science,University of Liverpool,Liverpool L69 3BX,UK CEMPS, University of Exeter,Exeter EX4 4QF,UK (Dated: April 17, 2020)
Transient local over-dry environment might be a contributor and an explanation for theobserved asynchronous local rises in Covid-19 mortality. We propose that a habitat’sair humidity negatively correlate with Covid-19 morbidity and mortality, and supportthis hypothesis on the example of publicly available data from German federal states.
Keywords: COVID-19 Mortality; Habitat; Air Humidity; Negative Correlation
Introduction.
Covid-19 virus (Zhou et al. , 2020) is trans-mitted through droplets which last longer in humid air.Therefore, humidity is believed to be pro-Covid-19 infec-tion and mortality. There are, however, data that con-tradict this belief. For instance, Wuhan, where Covid-19was first identified and studied, is in humid subtropicalclimate zone (Wiki-Wuhan, 2020), but December, whenmortality sharply raised, is the driest month of the yearthere. The purpose of this communication is to presentand substantiate a viewpoint that air humidity negativelycorrelate with Covid-19 morbidity and mortality.
The main hypothesis consists of two parts, of differentdegree of plausibility. First, mucous membranes of theupper respiratory tract present the first and essentialbarrier against Covid-19 virus entering human organism.Hence the state of the mucous membranes is a correlateto organism’s resistivity. Second, a dry season normallycause respiratory mucosa to become over-dry. In pres-ence of Covid-19 virus the latter might become a factorof massive fatality.
Direct evidence in support of the hypothesis.
Correlationof Covid-19 with age, both in terms of registered casesand of mortality is the first and best known fact aboutthis strain (Xu, Mao, and Chen, 2020). The state ofrespiratory mucosa also correlates with age (Beule, 2010).Correlation of Covid-19 mortality with low air humid-ity is less obvious. By way of anecdotal evidence, inaddition to the coincidence of the beginning of the epi-demic with the dry season in Wuhan mentioned above, ∗ [email protected]
60 65 70 75 80 85 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 a v e r age a i r hu m i d i t y o f M a r c h , % mortality rate, %data Ham Ber Mai Saa Erf LeiKieHalFraCol Pot Mun StuBre least-squareTheil-Sen
FIG. 1 Air humidity in March vs Covid-19 mortality in Ger-man federal states (circles). The size of the circles is propor-tional to the population of the federal state; the labels areabbreviations for the largest cities of the federal lands. Redsolid line shows the linear regression. Theil-Sen regression isshown with the blue dashed line. note that Seoul and especially Tokyo, where the incidenceand mortality have been lower, have on average wetterclimate (Wiki-Seoul, 2020; Wiki-Tokyo, 2020). Corre-late of sharp raise in Covid-19 mortality with local dryperiod may also be seen on the example of Lombardy,where February is the driest month in Milan (Wiki-Milan,2020), as opposed to a wetter beginning of the springin Rome (Wiki-Rome, 2020). The much dryer Marchin Spain as opposed to e.g. more humid neighbouringPortugal (Wiki-Lisbon, 2020) seem to point in the samedirection of raise in Covid-19 mortality correlate withtransient over-dry local environment.As an illustrative and preliminary example of a moresystematic evidence, we have considered Covid-19 mor- a r X i v : . [ q - b i o . O T ] A p r tality rate, defined as the number of deaths per numberof confirmed infections, in German federal states (Wiki-Pandemic-Germany, 2020) where the majority of deathshappened last March 2020. FIG.1 shows the Covid-19mortality rate in the federal states vs the average localair humidity in March. The local air humidity was de-fined as proxy recorded in the largest city of each fed-eral state (Weather and Climate, 2020). The choice ofthe German federal states data sets is motivated by thedata availability and reliability, and, in particular, by thepresumed uniformity of the data collection protocols inGermany. Mecklenburg-Vorpommern has been excludedas the resource (Weather and Climate, 2020) gives no hu-midity data neither for Rostock nor for Schwerin. Thelinear least squares fit (red solid line), weighted by themost recent population size of the federal states as givenby Wikipedia, gives the slope of − .
09 with a standarddeviation of ± .
32, i.e. reliably negative. We have alsoapplied the Theil-Sen estimator, also weighing the datapoints proportionally to population sizes, which gives theslope of − .
10. The corresponding fits are also shown inFIG.1 (blue dashed line). The discrepancy between thelinear and Theil-Sen estimates are not surprising as theproblem is clearly multi-factorial and we are looking atonly one of the factors. Still, the two estimates concurthat the slope of the dependence is negative, that is mor-tality is on average higher in a drier air and lowers withrise of air humidity, which confirms our hypothesis onnegative correlate of Covid-19 mortality with local airhumidity.Note that (Klein et al. , 2020) presents evidence of neg-ative correlation of air humidity with Covid-19 transmis-sion rate based on all-China data. A similar study (Ma et al. , 2020) based on Wuhan data indicates negative cor-relation of air humidity with Covid-19 mortality. Both ofthe studies did not seem to take into account the effectof non-causal correlation between the seasonal increaseof humidity and decrease in transmission rate due to thetaken administrative measures. Still, both studies are in-teresting as they present an approach complementary tothe one we used in FIG.1, namely, both studies do nottake into account regional variations of air humidity andCovid-19 statistics; instead, (Ma et al. , 2020) is exploit-ing their temporal variations. In any case, the overallconclusion from those studies concurs with our hypothe-sis.The above is “direct” evidence as it allows one to hy-pothesise direct causal relationship: dry air causing over-dry respiratory mucosa in older and vulnerable popula-tion causing increase of Covid-19 mortality.
Indirect evidence in support of the hypothesis.
Dry nasalmucosa correlates with loss of smell and taste (Beule,2010), and loss of smell and taste correlates with Covid-19 statistics (Bagheri et al. , 2020). Dry air is a known risk factor for dry eyes (NHS UK,2018), with dry eyes being a risk factor for conjunctivi-tis (Brazier, 2018). Recent reports show correlates ofCovid-19 statistics with conjunctivitis (Chen et al. , 2020;The Royal College of Ophthalmologists and College ofOptometrists, 2020).Some groups of patients identified as particularly vul-nerable to Covid-19, e.g. diabetes (Xu, Mao, and Chen,2020), also correlate with the diminished function of res-piratory mucosa (Beule, 2010).For these above evidence, the direction of causal links,if any, is less clear.
Verifiable predictions of the hypothesis.
A direct verifica-tion of the proposed hypothesis would be analysis of theinstant local air humidity in the statistics of Covid-19incidence and mortality. A statistically significant corre-lation would confirm the hypothesis.A more sophisticated way of checking the hypothesiswould be via spatiotemporal modelling of the pandemic.Such modelling, which no doubt will be attempted bymany research groups, will be most successful if and whenit takes into account as many relevant factors as possi-ble. Hence if the proposed hypothesis is true, taking intoaccount the air humidity of the habitat would improvethe quality and predictive power of the models.
Discussion and practical consequences.
Air humidity de-pends on multiple parameters: the local instant and an-nual rate of precipitation, diurnal and annual tempera-ture range, altitude, etc. That is why it is so difficult toestimate air humidity based on e.g. local precipitationonly.If the negative correlate of Covid-19 mortality withair humidity is verified, it might suggest certain practi-cal steps in addition to the medical and administrativemeasures already in place, and those yet to be proposedbased on other considerations.Iceland’s Covid-19 screening showed people of all agesequally susceptible for the infection (Iceland Review,2020), with about 50% of those infected having no symp-toms at the time of testing ( ´Ciri´c, 2020). In the viewof the latter, the world wide correlation of Covid-19with age might appear to be skewed towards those pa-tients showing more symptoms and therefore more tested.Therefore, it might be useful to distinguish between thespread of Covid-19 infection and the asynchronous localrises of Covid-19 mortality. For instance, a prolonged dryweather may be taken as an indication of likely local el-evation of Covid-19 mortality. In Madrid, August is thedriest month of the year (Wiki-Madrid, 2020), so preven-tative measures might be indicated to forestall or flattenthe second wave of Covid-19 there. On a general point, itmight take at least an annual cycle of the global data tofully appreciate the spatiotemporal pattern of Covid-19pandemic, and build the data based model.In presence of Covid-19 virus, patients with tendencyto dry respiratory mucosa might be particularly vulner-able to the exposure to dry air. Indoor environmentmight become over-dry due to the winter central heat-ing, domestic devices producing heat, especially if theonly source of the indoor humidity are people themselveswhich might be not enough. It might be not possibleto alter local microclimate, not to mention an instantchange of the global one. However, control of indoor en-vironment is feasible and might mitigate patients’ expo-sure to Covid-19. Balance of exposure to Covid-19 virusin dry air against the well known exposure to bacterialinfection in a humid environment must be taken into ac-count when developing a healthy indoor technology. Tothe author’s knowledge, this aspect has not yet been dis- cussed, while, if true, this could immediately start savinglives, which is the reason for this publication.
Availability of supporting data.
The datasets used and/oranalysed in this paper are available from the websites.
Declaration of interests.
The author declares that she hasno known competing financial interests or personal rela-tionships that could have appeared to influence the workreported in this paper.
Acknowledgement
This paper is entirely the author’s ini-tiative. It does not result from any funded researchproject. The author is grateful to Professor Vadim Bik-tashev, Dr Sergey Blagodatsky, and Dr Evgenia Blago-datskaya, for much valued discussion of this paper.
REFERENCES
Bagheri, S. H. R., Asghari, A. M., Farhadi, M., Shamshiri, A. R., Kabir, A., Kamrava, S. K., Jalessi, M., Mohebbi, A., Alizadeh,R., Honarmand, A. A., Ghalehbaghi, B., and Salimi, A., “Coincidence of COVID-19 epidemic and olfactory dysfunctionoutbreak,” https://doi.org/10.1101/2020.03.23.20041889 (2020).Beule, A. G., GMS Current Topics in Otorhinolaryngology - Head and Neck Surgery , Doc07 (2010).Brazier, Y., “What is dry eye and how can I get rid of it?” , accessedApril 6, 2020 (2018).Chen, L., Deng, C., Chen, X., Zhang, X., Chen, B., Yu, H., Qin, Y., Xiao, K., Zhang, H., and Sun, X., “Ocular manifestationsand clinical characteristics of 534 cases of COVID-19 in China: A cross-sectional study,” https://doi.org/10.1101/2020.03.12.20034678 (2020).´Ciri´c, J., “Is Icelands coronavirus testing showing that 50% of cases have no symptoms?” , accessed April 12, 2020(April 2, 2020).Iceland Review, “Whats the status of COVID-19 in Iceland?” , accessed April 12, 2020 (April 2, 2020).Klein, E., Lin, G., Tseng, K., Schueller, E., Kapoor, G., and Laxminarayan, R., “COVID-19 for India updates,” https://cddep.org/wp-content/uploads/2020/03/covid19.indiasim.March23-2-4.pdf (March 20, 2020).Ma, Y., Zhao, Y., Liu, J., He, X., Wang, B., Fu, S., Yan, J., Niu, J., Zhou, J., and Luo, B., Science of The Total Environment , 138226 (2020).NHS UK, “Dry eyes,” , accessed April 6, 2020 (2018).The Royal College of Ophthalmologists, and College of Optometrists,, “Viral conjunctivitis and COVID-19,” ,accessed April 6, 2020 (March 19, 2020).Weather and Climate, https://weather-and-climate.com , accessed April 6, 2020 (2020).Wiki-Lisbon, https://en.wikipedia.org/wiki/Lisbon , accessed April 6, 2020 (2020).Wiki-Madrid, https://en.wikipedia.org/wiki/Madrid , accessed April 6, 2020 (2020).Wiki-Milan, https://en.wikipedia.org/wiki/Milan , accessed April 6, 2020 (2020).Wiki-Pandemic-Germany, https://en.wikipedia.org/wiki/2020_coronavirus_pandemic_in_Germany , accessed April 6, 2020(2020).Wiki-Rome, https://en.wikipedia.org/wiki/Rome , accessed April 6, 2020 (2020).Wiki-Seoul, https://en.wikipedia.org/wiki/Seoul , accessed April 6, 2020 (2020).Wiki-Tokyo, https://en.wikipedia.org/wiki/Tokyo , accessed April 6, 2020 (2020).Wiki-Wuhan, https://en.wikipedia.org/wiki/Wuhan , accessed April 6, 2020 (2020).Xu, L., Mao, Y., and Chen, G., “Risk factors for severe corona virus disease 2019 (COVID-19) patients : a systematic reviewand meta analysis,” https://doi.org/10.1101/2020.03.30.20047415 (2020).Zhou, P., Yang, X.-L., Wang, X.-G., Hu, B., Zhang, L., Zhang, W., Si, H.-R., Zhu, Y., Li, B., Huang, C.-L., Chen, H.-D.,Chen, J., Luo, Y., Guo, H., Jiang, R.-D., Liu, M.-Q., Chen, Y., Shen, X.-R., Wang, X., Zheng, X.-S., Zhao, K., Chen, Q.-J.,Deng, F., Liu, L.-L., Yan, B., Zhan, F.-X., Wang, Y.-Y., Xiao, G.-F., and Shi, Z.-L., Nature579