Table A (reported in Appendix) lists, for the 185 countries reported by Dorsey et al. (2018) for which temperature data were available, the two climate indices (see Methods for their definition) and the 1990-2016 variations of the PD patients’ epidemiological data relative to deaths, prevalence and DALYs numbers.
In terms of the 1990-2016 warming index, values vary between -0.3°C and 1.8°C, with a median of 0.7°C, while in terms of 2016 average temperatures (T2016), values vary between -15.9°C and 29.1°C, with a median of 23.2°C (Tab. 1). It is worth noting that warming is, in general, more pronounced in colder countries, as can be seen by the scatter diagram of the warming index versus T2016 (Fig.1).
Tab. 1 Distribution of warming indices and T2016. Minimum and maximum, 25th and 75th percentile, and median values of the distribution of the 1990-2016 warming indices and the T2016 of the 185 countries.
|
1990-2016 Warming Index
|
T2016
|
Min
|
-0.3
|
-15.9
|
25th
|
0.4
|
11.8
|
Median
|
0.7
|
23.2
|
75th
|
1.2
|
26.3
|
Max
|
1.8
|
29.1
|
By considering the 1990-2016 warming index and the T2016, the 185 countries (App. A) have been grouped in 4 categories (Fig. 1):
- the high-temperature and high-warming (HT-HW) category, which includes the 25 ones with above median climate indices (the ones represented by the dots in the top-right quadrant of the scatter plot shown in Fig. 1);
- the high-temperature and low-warming (HT-LW) category, which includes 68 countries (the ones in the bottom-right quadrant of Fig. 1);
- the low-temperature and high-warming (LT-HW category, which includes 68 countries (the ones in the top-left quadrant of Fig. 1);
- the low-temperature and low-warming (LT-LW category, which includes 24 countries (the ones in the bottom-left quadrant of Fig. 1).
To assess whether there is any relationship between climate warming and variations in the PD indices, we have computed the correlation between the 1990-2016 variations of the PD indices and the 1990-2016 climate warming index firstly for all countries, and then for the countries grouped in the 4 categories defined above. Results (Tab. 2) indicate that the correlation is higher, and reaches 26%, for the countries in the HT-HW category, while for the other countries the correlation is close to zero.
Tab. 2 Correlation coefficients distribution. Correlation coefficients between the 1990-2016 variations of the PD indices and the climate warming index, computed for PD patients’ deaths, prevalence and DALYs for countries in the 4 categories (HT-HW, HT-LW, LT-HW and LT-LW) and for all countries.
Category
|
Correlation of the 1990-2016 variation of the PD index and the climate warming index
|
Deaths
|
Prevalence
|
DALYs
|
HT-HW
|
26.3%
|
25.6%
|
26.1%
|
HT-LW
|
2.0%
|
1.6%
|
0.4%
|
LT-HW
|
0.5%
|
0.8%
|
0.4%
|
LT-LW
|
0.1%
|
4.1%
|
0.1%
|
ALL countries
|
0.2%
|
0.8%
|
0.02%
|
Tab. 3 T-test percent. T-test computed between the HT-HW distribution and the distributions of the HT-LW, LT-HW and LT-LW, for PD patients’ deaths, prevalence and DALYs.
T-TEST
|
HT-HW
|
Deaths
|
Prevalence
|
DALYs
|
HT-LW
|
0.4%
|
0.0%
|
0.1%
|
LT-HW
|
0.2%
|
2.0%
|
0.5%
|
LT-LW
|
4.8%
|
0.1%
|
1.6%
|
Figure 2 shows, for the HT-HW countries, the scatter plot of the 1990-2016 variations of the PD indices as a function of the warming index. Note the positive slope for all 3 PD indices, indicating that for countries characterized by a warm climate (warmer T2016 than the median of all 185 countries), the variations of PD indices are higher in the countries that have been subjected to the largest warming. Although the correlation is small, around 26%, this result can be seen as an indication that environmental factors linked to climate change could have contributed to the largest variations.
As a further indication that, for the HT-HW countries, climate warming played a role in inducing an increase in the PD patients’ epidemiological indices, Fig. 3 shows the median, 25th and 75th percentile of the distribution of the PD indices for the 4 categories. Note that, for all 3 PD indices, there is a clear difference between the HT-HW distribution, and the distributions of the other 3 categories. To assess how the similarity between the distributions of the 1990-2016 variations of the PD indices of the HT-HW countries, and the distributions of the other categories, the t-test has been computed. Results (Tab. 3) show that for all indices but deaths, the probability that the two distributions are similar is less than 2%, while for deaths it is 4.8%. In other words, at the 95% level we can reject the null hypothesis that there is not any difference between the distribution of the HT-HW countries, and the countries of the other 3 categories.