The latest research shows that seasonal temperature changes may affect the propagation trajectory of COVID-19 in different parts of the world.

On April 12, local time, the medical preprint platform medRxiv published a new regression analysis study online, pointing out that from the data in the middle and late March, 22.5 degrees Celsius or an important “breakpoint”, when the regional average maximum temperature reaches 22.5 degrees Celsius or more, the rate of confirmed cases of new coronary pneumonia will be greatly reduced.

Due to the difference in temperature and season, there is a correlation between the more serious COVID-19 epidemic and areas with a latitude of 30 degrees or more between north and south latitudes.

The study is titled “Evidence that higher temperatures are associated with lower incidence of COVID-19 in pandemic state, cumulative cases reported up to March 27, 2020”, the author of the article Michael Triplett, an engineer from American medical device manufacturer Terumo BCT and the University of Colorado Denver.

Regression analysis (regression analysis) refers to a statistical analysis method that determines the quantitative relationship between two or more variables. According to the type of relationship between independent and dependent variables, it can be divided into linear regression analysis and nonlinear regression analysis.

According to the cumulative data of patients diagnosed with New Coronary Pneumonia (COVID-19) from March 14 to March 27 of the World Health Organization, through multiple linear regression analysis and nonlinear In regression analysis, Triplett found an association between the incidence of COVID-19 and areas with latitudes of 30 degrees or more. Climate data also shows that the average maximum temperature exceeds about 22.5 degrees Celsius, the incidence of new crown cases will be greatly reduced.

Triplett analyzed the World Health Organization ’s daily report on the diagnosis of new coronary pneumonia in countries from March 14 to March 27, 2020, and according to the United Nations Statistics Division ’s The population data calculates the diagnosis rate of new coronary pneumonia in various countries.

By taking the average of the latitude and longitude of the northernmost, easternmost, southernmost and westernmost points of each country(Middle value), he assigned the population value according to latitude.

Subsequently, Triplett recorded the daily average temperature distribution of the latitude of the countries with new cases of new crowns worldwide.


Global average maximum surface temperature data on March 14, March 21, and March 27, distributed by latitude span>

For the data reported on March 27th, the researchers used the above estimated population and average maximum temperature latitude distribution as predictors for confirmed cases in different countries The data were analyzed by multiple linear regression. The researchers also analyzed the situation where the latitude is higher or lower than 30 degrees under the condition of the same predictor.

Through a series of analyses, the researchers found that 30 degrees north-south latitude seems to be a dividing line. The number of confirmed cases and the confirmed rate are higher in the areas above 30 degrees.


changes in the number of confirmed cases (left) and the rate of diagnosis (right) according to latitude from March 14 to March 27 Viral diagnosis of latitudes above and below 30 degrees is clearly visible

It is worth noting that the same can be seen from the above picture as of March On the 27th, the incidence of cases seemed to increase faster south of 30 degrees south latitude, possibly because the temperature there was falling and entering the autumn. This increase initially indicates that the correlation between the epidemic situation and latitude is likely to be affected by temperature.

Through multiple linear regression analysis of population, temperature and latitude (above / below 30 degree latitude), the researchers believe that there is a significant relationship between the diagnosed cases and the above factors. In the regression model, the transformed R2 (usually used to assess the degree of agreement between the predicted value and the actual value, the closer the R2 value is to 1, the better the independent variable interpretation of the dependent variable in the regression analysis) is 84.61%, the latitude is The credibility of the categorical variable in the case of less than 30 degrees is about 84.7%.


between the average global maximum temperature on March 14, March 21, and March 27 and the number of confirmed cases and diagnoses As the temperature increases, the diagnosis rate shows a downward trend. When the average maximum temperature reaches 22.5 degrees Celsius, the diagnosis rate can be kept at a very low level

< / div> It can be seen that when the average maximum temperature is below 22.5 degrees Celsius, the number of confirmed cases and the diagnosis rate of new crown cases have increased, but as the temperature increases to 22.5 degrees Celsius, the diagnosis rate drops sharply. When the temperature is higher than 22.5 degrees Celsius, the diagnosis rate remains very low.

The study pointed out that it should also be noted that the high-growth areas of new crown cases warmed and the temperature increased between March 14 and 27, but only The average maximum temperature is above 22.5 degrees Celsius, and the rate of confirmed cases of new coronary pneumonia can reach a low level. Above 22.5 degrees Celsius, the outbreak has developed slightly, but the growth rate of confirmed cases has slowed significantly.

In addition, the researchers found that the nonlinear regression of temperature and diagnosis rate can get the best fitting effect of all models,

The R2 value is as high as 93.78%, indicating that this factor is highly correlated with the development of the epidemic.

Triplett said that according to temperature predictions, there is a strong regional correlation between the incidence of COVID-19 cases and the average maximum surface temperature below 22.5 degrees Celsius.The case diagnosis rate peaked when the average maximum surface temperature was about 7.5 degrees Celsius.

Finally, Triplett pointed out that although he determined the relationship between several variables and the diagnosis rate of new coronary pneumonia, the geographical location and temperature could not explain the local level well Differences in infection rates. These are some preliminary findings, which will support further countermeasure research against the new coronavirus and explore the relationship between the virus transmission rate and temperature and humidity.

People usually think that certain respiratory diseases (such as flu) have obvious seasonality. The outbreak of SARS (Severe Acute Respiratory Syndrome) in 2003 also showed its dependence on specific temperature and humidity.

With the development of the new coronavirus epidemic, the relationship between virus and climate has received more and more attention. Previously, on March 9th, researchers from the Institute of Human Viruses of the University of Maryland School of Medicine and the Center for Excellence in the Global Virus Network (GVN) published a study on the preprint platform SSRN. Distribution, roughly along the 30 ° -50 ° north latitude corridor, this area has a similar temperature: 5 ℃ -11 ℃, similar relative humidity (RH): 47% -79%.

However, Marc Lipsitch, a professor of epidemiology at Harvard University, said recently that it is clear that warming will not stop the spread of the new coronavirus, and believes that the coronavirus can undoubtedly be able to Continue to spread in hotter, humid environments.

The study warns that as the southern hemisphere moves from summer to autumn and winter, the number of new crown cases in the southern hemisphere will also increase. A number of studies have so far indicated that even in warmer and humid areas, effective public health interventions should be taken to reduce the spread of new coronavirus and protect vulnerable groups from infection.